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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
80d6c247-e5ab-4603-b035-1f4972b5fca8 | collaborative-and-ai-aided-exam-question | 2211.08361 | null | https://arxiv.org/abs/2211.08361v1 | https://arxiv.org/pdf/2211.08361v1.pdf | Collaborative and AI-aided Exam Question Generation using Wikidata in Education | Since the COVID-19 outbreak, the use of digital learning or education platforms has significantly increased. Teachers now digitally distribute homework and provide exercise questions. In both cases, teachers need to continuously develop novel and individual questions. This process can be very time-consuming and should be facilitated and accelerated both through exchange with other teachers and by using Artificial Intelligence (AI) capabilities. To address this need, we propose a multilingual Wikimedia framework that allows for collaborative worldwide teacher knowledge engineering and subsequent AI-aided question generation, test, and correction. As a proof of concept, we present >>PhysWikiQuiz<<, a physics question generation and test engine. Our system (hosted by Wikimedia at https://physwikiquiz.wmflabs.org) retrieves physics knowledge from the open community-curated database Wikidata. It can generate questions in different variations and verify answer values and units using a Computer Algebra System (CAS). We evaluate the performance on a public benchmark dataset at each stage of the system workflow. For an average formula with three variables, the system can generate and correct up to 300 questions for individual students based on a single formula concept name as input by the teacher. | ['Bela Gipp', 'Andre Greiner-Petter', 'Andreas Spitz', 'Moritz Schubotz', 'Philipp Scharpf'] | 2022-11-15 | null | null | null | null | ['question-generation'] | ['natural-language-processing'] | [-2.63114095e-01 3.26071054e-01 3.81803840e-01 -2.82833040e-01
-6.40642464e-01 -8.65458965e-01 3.35554034e-01 7.78940678e-01
-2.59564459e-01 8.26013982e-01 -9.37645361e-02 -7.05041111e-01
-7.79115319e-01 -1.43993151e+00 -8.39063168e-01 -2.81365246e-01
5.91159463e-01 9.47479606e-01 5.66009879e-01 -4.45197135e-01
2.13679358e-01 4.96630341e-01 -2.30240536e+00 3.10435712e-01
1.73143709e+00 4.64429080e-01 4.38144565e-01 8.50337088e-01
-6.94504738e-01 1.10565019e+00 -7.50713348e-01 -6.66276157e-01
-2.50535369e-01 -3.81232649e-01 -1.15185058e+00 -5.94371915e-01
8.47510040e-01 -1.60159752e-01 -9.73081663e-02 1.04847443e+00
6.64504349e-01 1.97329193e-01 2.06450224e-01 -1.24247921e+00
-7.76372135e-01 9.12066519e-01 1.39401719e-01 4.33126204e-02
8.80628407e-01 1.68205172e-01 6.63064003e-01 -5.60233772e-01
8.08302104e-01 9.79209960e-01 4.73603070e-01 3.71092618e-01
-8.42284918e-01 -9.40257072e-01 -5.22570968e-01 1.10625589e+00
-1.16802847e+00 1.97204590e-01 4.12070543e-01 -6.51986718e-01
6.59422636e-01 4.93809849e-01 1.22048295e+00 5.87228596e-01
-3.41206901e-02 4.38592792e-01 1.10852945e+00 -6.12603903e-01
3.16037804e-01 4.49312538e-01 2.23686755e-01 9.04632032e-01
2.74499357e-01 -4.85021174e-01 -7.36486495e-01 1.92760810e-01
1.93450555e-01 -4.50778633e-01 -2.31805220e-01 -8.18952732e-03
-8.56969595e-01 5.35537779e-01 5.54777905e-02 3.15004379e-01
-1.82814285e-01 -1.97235256e-01 -3.99810448e-03 7.52271593e-01
1.32346958e-01 8.36526752e-01 -7.48461545e-01 -6.46993518e-01
-7.42694199e-01 6.38183475e-01 1.29583275e+00 9.19834793e-01
7.67196000e-01 -3.21897775e-01 -1.04669727e-01 8.04994404e-01
1.84407741e-01 4.38463867e-01 3.80560577e-01 -1.14781070e+00
1.06664654e-02 1.07720900e+00 -1.82644472e-01 -7.55337596e-01
-1.21050775e-01 -1.62036315e-01 -1.84320867e-01 3.00177783e-01
8.15907001e-01 -4.14348394e-02 -6.00214601e-01 1.28880060e+00
1.27645791e+00 4.41963643e-01 -2.31612161e-01 5.20673633e-01
1.74694586e+00 4.15037662e-01 2.96663661e-02 -7.92235881e-03
1.63636756e+00 -8.07398498e-01 -9.24645245e-01 8.71797204e-01
7.43919909e-01 -1.26337624e+00 1.01658893e+00 9.08843875e-01
-1.52384102e+00 -5.84657907e-01 -7.27096975e-01 -3.56471956e-01
-9.65430677e-01 -2.12819561e-01 3.42966646e-01 7.61548817e-01
-1.15907621e+00 5.31439006e-01 -4.43524659e-01 -3.10498387e-01
1.86045706e-01 1.54328272e-01 -2.04550967e-01 -1.73917919e-01
-1.24843264e+00 9.49703991e-01 2.62585312e-01 -5.68842530e-01
-5.53226292e-01 -1.88296497e+00 -5.75983882e-01 2.22310334e-01
4.14178759e-01 -6.98736727e-01 1.37537968e+00 -1.15646429e-01
-1.78565407e+00 7.05026150e-01 3.98320317e-01 4.70269434e-02
5.05656302e-01 -1.33153066e-01 -3.32433224e-01 3.17614883e-01
6.71897233e-02 5.19223690e-01 7.76917562e-02 -5.24865687e-01
-6.14036441e-01 -1.81072026e-01 2.94701248e-01 -4.44242619e-02
-3.88141543e-01 2.87167728e-01 -3.24929863e-01 -2.02808768e-01
-1.04874454e-01 -5.43788135e-01 1.69325456e-01 1.05507076e-01
1.25878796e-01 -9.79078233e-01 7.19164610e-01 -8.27704608e-01
1.31633103e+00 -1.50219274e+00 -1.54911235e-01 2.83933789e-01
2.41669148e-01 5.30957222e-01 7.80678540e-02 5.57755649e-01
-1.49812281e-01 -1.84138492e-03 3.00380528e-01 4.71600622e-01
2.79001057e-01 1.63103968e-01 8.74861702e-03 -1.52243540e-01
-9.34026241e-02 6.33764684e-01 -1.36694455e+00 -8.15014958e-01
4.21065807e-01 4.98667508e-01 -9.50564921e-01 5.22148192e-01
-7.49800563e-01 3.75172615e-01 -2.74816275e-01 3.95356745e-01
7.76851296e-01 -3.09265912e-01 2.99612194e-01 2.26060376e-01
-6.00716829e-01 7.01540112e-01 -1.55485666e+00 1.87253010e+00
-6.44257545e-01 5.62092841e-01 8.24152529e-02 -9.41121221e-01
7.64327347e-01 4.99511302e-01 5.53288162e-01 -4.98882085e-01
-1.27497464e-02 3.03520858e-01 -4.60664481e-02 -9.98664200e-01
4.71118361e-01 6.31970644e-01 4.43680674e-01 8.25005114e-01
4.88143593e-01 -9.28966641e-01 7.80385494e-01 6.02429271e-01
1.24246931e+00 4.24693793e-01 -8.66744295e-02 -4.24230367e-01
7.54174352e-01 2.44571745e-01 1.25303954e-01 6.17471635e-01
1.53918698e-01 -3.20469146e-03 1.80850580e-01 -2.29481727e-01
-7.71123588e-01 -1.28248680e+00 -4.63963002e-01 1.31280446e+00
-4.48802114e-01 -8.35683763e-01 -9.08604026e-01 -3.10730875e-01
-3.44747081e-02 7.22542048e-01 -1.07067034e-01 2.88768619e-01
-3.14276129e-01 5.30381948e-02 4.03754264e-01 -4.37321626e-02
4.25479800e-01 -1.29891276e+00 -4.73552436e-01 3.62220138e-01
-1.37076557e-01 -8.84475768e-01 9.96269658e-02 -3.90331149e-01
-2.56225049e-01 -1.27727544e+00 -2.18451008e-01 -8.03919613e-01
6.03880167e-01 -7.05000609e-02 1.30444872e+00 5.76352298e-01
-7.88877547e-01 9.40024614e-01 -3.89642030e-01 -7.73844361e-01
-6.01884425e-01 -8.56958330e-03 -7.47100487e-02 -7.49932587e-01
4.82841700e-01 -8.67380202e-01 -5.38566887e-01 -8.26955661e-02
-1.00896132e+00 2.84757942e-01 -1.50166243e-01 4.23833549e-01
3.96814406e-01 1.14096604e-01 6.55067623e-01 -8.60139489e-01
6.62992597e-01 -6.41601801e-01 -1.24060559e+00 5.93204200e-01
-7.76632547e-01 -1.03380330e-01 5.72521329e-01 -1.82629466e-01
-1.21077490e+00 -3.03668588e-01 -4.67949510e-01 -1.15504868e-01
-5.64288676e-01 6.54526770e-01 -2.11350881e-02 -3.14568251e-01
6.05086863e-01 -1.29989954e-03 -5.07614315e-02 -6.59530759e-01
5.59036434e-01 7.54780054e-01 5.50571740e-01 -1.01508439e+00
9.06847537e-01 -1.75874516e-01 -2.62729824e-02 -9.45043743e-01
-8.79831433e-01 -2.79710740e-01 -5.50750732e-01 -8.42694700e-01
7.47599483e-01 -8.55065107e-01 -1.60112834e+00 3.30766797e-01
-1.15429568e+00 -3.31436038e-01 -4.34209675e-01 6.94700718e-01
-1.29542440e-01 1.14898756e-01 -4.55602467e-01 -3.05470198e-01
-2.56468415e-01 -1.14915276e+00 2.68897921e-01 8.57596874e-01
-6.79825619e-02 -1.00754762e+00 3.35467607e-01 1.14922488e+00
5.54018259e-01 -1.46563977e-01 1.01926446e+00 -6.86449587e-01
-8.64681542e-01 2.11366266e-01 -2.18482129e-02 2.16973811e-01
-2.86290646e-01 5.37321329e-01 -8.48432183e-01 3.16261917e-01
-5.03105521e-01 -6.70690536e-01 2.18248591e-02 -2.96601921e-01
1.56375396e+00 -2.55691975e-01 2.49551594e-01 2.07702577e-01
1.20414138e+00 -2.07472429e-01 4.45342124e-01 2.98005879e-01
4.45415467e-01 7.88748741e-01 3.78469676e-01 4.01997149e-01
9.45675194e-01 4.83813733e-01 4.99738827e-02 5.30216157e-01
-5.31922638e-01 -1.83804035e-01 1.44534916e-01 1.69636440e+00
-1.01429828e-01 3.27934831e-01 -1.19985199e+00 8.17064106e-01
-1.45845628e+00 -1.01173663e+00 -9.06202435e-01 1.79348159e+00
1.61542165e+00 -4.80596006e-01 -2.41513431e-01 3.35164033e-02
2.57101059e-01 -6.97087824e-01 -1.06575586e-01 -7.45800614e-01
4.32156116e-01 1.39554834e+00 -2.50939906e-01 7.56788731e-01
-1.17985524e-01 8.88881862e-01 5.09048939e+00 1.13947833e+00
-8.24000180e-01 3.32678705e-01 -2.05719054e-01 -1.62059590e-01
-7.23331213e-01 -9.46708694e-02 -8.39520156e-01 4.15781796e-01
1.22145176e+00 -6.76941574e-01 5.94499528e-01 6.69187903e-01
1.52202860e-01 -1.76467225e-01 -6.49535179e-01 5.12366474e-01
-9.36277583e-02 -1.84696746e+00 -1.84373155e-01 -2.02275738e-01
1.10247469e+00 5.14800772e-02 -3.45339566e-01 5.52586973e-01
9.54724669e-01 -7.90755391e-01 3.39675695e-01 7.41718113e-01
5.75553060e-01 -7.42804825e-01 1.55517727e-01 3.18803877e-01
-1.04618883e+00 2.09554285e-01 -1.93354294e-01 -9.40917283e-02
-4.16656792e-01 7.11926341e-01 -1.11922169e+00 8.62543166e-01
1.08084261e+00 1.02812424e-01 -7.80563235e-01 1.48338079e+00
-6.92632973e-01 9.56655383e-01 -5.33266306e-01 -2.74048179e-01
-2.73323923e-01 -4.70641851e-01 2.23748550e-01 8.69616628e-01
4.86014575e-01 5.67068577e-01 6.78910166e-02 1.05100250e+00
-9.14519355e-02 4.82411981e-01 -2.17700914e-01 7.65350312e-02
8.36290777e-01 1.56508470e+00 -2.41633847e-01 -4.11235124e-01
-2.70676106e-01 3.22960526e-01 4.20471400e-01 5.34795783e-02
-7.44910181e-01 -7.33007491e-01 6.62122786e-01 2.24762276e-01
1.08983383e-01 -4.77141440e-02 -4.67534848e-02 -9.66307044e-01
-2.18616828e-01 -1.18333757e+00 4.46110249e-01 -1.12709951e+00
-1.09385705e+00 -5.39546795e-02 -1.30415909e-04 -5.38457513e-01
-2.83798516e-01 -5.39207876e-01 -7.95171678e-01 7.59105563e-01
-1.35091984e+00 -8.29761744e-01 -7.07644880e-01 8.10737252e-01
1.19468182e-01 4.37933728e-02 9.12573636e-01 7.89038420e-01
-4.56932217e-01 5.94273269e-01 -8.78154766e-03 -2.26331726e-01
7.65831769e-01 -1.50039816e+00 -2.84759074e-01 3.36337179e-01
5.37502430e-02 3.22527647e-01 8.07142019e-01 -6.36784732e-01
-1.67481983e+00 -8.47324312e-01 1.26388490e+00 -5.64899206e-01
1.07431102e+00 -8.37815329e-02 -1.11015999e+00 1.73985302e-01
6.86082482e-01 -4.11588997e-01 9.78288114e-01 1.41278142e-02
-1.26912475e-01 -5.72443604e-02 -1.13047302e+00 4.49261546e-01
8.67075026e-01 -4.87141728e-01 -7.33686030e-01 8.55286121e-01
8.72518539e-01 -6.13491774e-01 -1.66190469e+00 7.97062889e-02
3.73913199e-01 -8.29349577e-01 8.08590651e-01 -6.71261132e-01
6.41074538e-01 -4.31792408e-01 4.31070924e-01 -1.15048265e+00
2.54010230e-01 -6.30944669e-01 1.78991780e-02 1.41272187e+00
2.85591006e-01 -5.20204663e-01 6.49812937e-01 7.58136511e-01
-2.24799603e-01 -6.45681381e-01 -8.55730176e-01 -4.45313692e-01
2.58476943e-01 -8.60607862e-01 1.06354308e+00 1.40871429e+00
3.49555641e-01 -8.84694159e-02 4.83658522e-01 2.27375552e-01
4.98309374e-01 1.82507277e-01 9.59770739e-01 -1.32696462e+00
-3.26623857e-01 -5.35128355e-01 -2.86491007e-01 -3.40799212e-01
2.05963142e-02 -1.30981493e+00 -3.61458033e-01 -1.71248579e+00
-2.59553343e-01 -4.16306376e-01 1.58753335e-01 5.29719293e-01
-9.02822018e-02 4.40219343e-02 1.02579981e-01 -6.49239898e-01
-6.37682080e-01 1.96305409e-01 1.55599427e+00 1.46756873e-01
1.36642039e-01 -3.27118099e-01 -3.23565841e-01 6.11130238e-01
7.40196884e-01 -5.15573144e-01 -2.05597907e-01 -3.04809868e-01
9.17691171e-01 -1.22307122e-01 2.67760098e-01 -1.10597372e+00
7.02031195e-01 -3.96815956e-01 9.88573059e-02 -7.81566024e-01
-1.16011322e-01 -6.31490707e-01 1.77805156e-01 4.56680357e-01
-1.54634908e-01 -8.71191844e-02 3.01451087e-01 -2.65005440e-01
-1.86472088e-02 -7.39303052e-01 4.57920402e-01 5.70658594e-03
-2.10892856e-01 3.26898438e-03 -4.56320763e-01 2.16157481e-01
1.08818197e+00 2.43093610e-01 -7.42029190e-01 -2.36794844e-01
-6.70614004e-01 6.65367365e-01 -3.20219547e-02 2.22932920e-01
4.21700358e-01 -1.08330131e+00 -7.33427942e-01 2.47363731e-01
-4.64507677e-02 1.05969094e-01 4.60654736e-01 7.15616047e-01
-9.76870298e-01 3.71283501e-01 -2.32555881e-01 -4.65248078e-01
-1.40769267e+00 1.05977997e-01 1.95026889e-01 -2.15332478e-01
-2.84039587e-01 9.58808422e-01 -6.98557138e-01 -1.20166838e+00
2.45449260e-01 -2.95070291e-01 -5.55507362e-01 2.33496994e-01
8.49641681e-01 7.07337916e-01 3.37645441e-01 1.43871248e-01
2.64647424e-01 3.46755922e-01 1.18966252e-01 1.19408116e-01
1.49375582e+00 4.14747089e-01 -4.45628911e-01 -8.51749163e-03
5.39704680e-01 3.43053967e-01 -3.67810786e-01 1.34734940e-02
-1.02707241e-02 -4.02117103e-01 -8.11377764e-02 -1.30229628e+00
-8.56216311e-01 6.65634811e-01 3.73421818e-01 3.87654901e-01
8.60510170e-01 6.75119311e-02 5.94306588e-01 5.22640169e-01
3.28988545e-02 -1.26691604e+00 6.53347373e-02 6.67185545e-01
7.43218303e-01 -1.00105453e+00 7.02993199e-02 -4.50553983e-01
1.63334474e-01 1.23505163e+00 1.05978334e+00 4.53779489e-01
8.66370022e-01 1.05813704e-01 -6.24046486e-04 -5.33135951e-01
-1.12082028e+00 -3.04502666e-01 3.14626902e-01 3.86976629e-01
8.11310410e-01 9.83589217e-02 -7.03490198e-01 5.16989350e-01
-8.60140979e-01 3.97705615e-01 8.68742228e-01 9.19079185e-01
-6.30304217e-01 -1.49660659e+00 -6.61223650e-01 3.93618554e-01
-3.92757863e-01 -1.05352357e-01 -2.53824055e-01 4.53982323e-01
5.16365111e-01 1.08403420e+00 -2.77919136e-02 6.56818226e-02
3.03467810e-01 3.16610873e-01 8.64990950e-01 -7.42242873e-01
-1.24259567e+00 -8.57421517e-01 1.11381575e-01 -3.36644948e-01
-6.57744259e-02 -4.06607926e-01 -1.50775075e+00 -7.97909439e-01
-4.13216680e-01 6.57017946e-01 1.24868858e+00 8.51922631e-01
4.63893116e-01 7.45145440e-01 4.68665808e-02 2.30687745e-02
-4.69903767e-01 -8.10031712e-01 -3.74104343e-02 2.04383731e-01
-4.83621448e-01 -4.51322824e-01 -1.71277866e-01 1.05239190e-01] | [10.863378524780273, 7.9770708084106445] |
d2e7815e-5c87-4bad-8ef6-48ec6d48de8f | ji-yu-shen-jing-wang-luo-de-ban-jian-du | null | null | https://aclanthology.org/2022.ccl-1.58 | https://aclanthology.org/2022.ccl-1.58.pdf | 基于神经网络的半监督CRF中文分词(Semi-supervised CRF Chinese Word Segmentation based on Neural Network) | “分词是中文信息处理的基础任务之一。目前全监督中文分词技术已相对成熟并在通用领域取得较好效果,但全监督方法存在依赖大规模标注语料且领域迁移能力差的问题,特别是跨领域未登录词识别性能不佳。为缓解上述问题,本文提出了一种充分利用相对易得的目标领域无标注文本、实现跨领域迁移的半监督中文分词框架;并设计实现了基于词记忆网络和序列条件熵的半监督权杒杆中文分词模型。实验结果表明本该模型在多个领域数据集上杆札值和杒杏杏杖值分别取得最高朲.朳朵朥和朱朲.朱朲朥的提升,并在多个数据集上成为当前好结果。” | ['Zhilin Zhao', 'Yujiao Han', 'Mingming Zhang', 'Zhiyong Luo'] | null | null | null | null | ccl-2022-10 | ['chinese-word-segmentation'] | ['natural-language-processing'] | [-6.66921973e-01 -5.64109385e-01 6.71120346e-01 3.03466797e-01
2.83869430e-02 -7.85524726e-01 -1.05289675e-01 1.19038069e+00
3.38487849e-02 4.03807819e-01 6.26807153e-01 2.05462188e-01
-2.91888267e-01 -1.26404476e+00 -7.43031681e-01 -9.41599965e-01
-5.31877100e-01 1.94262600e+00 5.60809851e-01 -5.97569168e-01
2.71460414e-01 7.88913608e-01 -9.36058342e-01 1.11608602e-01
8.27473760e-01 7.92626679e-01 9.39814866e-01 3.25526208e-01
1.69849038e-01 4.62559253e-01 -1.20901489e+00 5.96362174e-01
-4.15982813e-01 4.60292131e-01 -9.97261927e-02 1.01550150e+00
-7.82410651e-02 -1.03970397e+00 -1.45087063e+00 9.30323541e-01
2.70124316e-01 -4.68610935e-02 1.36757421e+00 -6.38497770e-01
-6.32767081e-01 3.59987259e-01 5.00898585e-02 5.26196659e-01
-5.86440682e-01 3.61072749e-01 4.61833030e-01 -8.18457603e-01
2.64215708e-01 1.45456541e+00 1.63338518e+00 4.75344986e-01
-1.49351764e+00 -9.82028186e-01 6.75677419e-01 -4.59223032e-01
-1.57365453e+00 6.03207171e-01 3.79603028e-01 -3.36478382e-01
1.30237544e+00 7.04334736e-01 1.23837745e+00 1.67980361e+00
4.94368523e-01 1.46468616e+00 1.42816114e+00 7.42826104e-01
4.04222071e-01 1.71445382e+00 -2.28789017e-01 4.77633625e-02
9.64120030e-01 1.70672610e-01 1.63174286e-01 4.50138599e-01
2.59978682e-01 3.72504115e-01 5.72255924e-02 1.01697159e+00
-1.52758896e+00 1.40815938e+00 1.88207358e-01 1.71178651e+00
-8.30387417e-03 -1.18431759e+00 -8.74134243e-01 3.33921462e-01
1.04608670e-01 -8.25488642e-02 4.18505460e-01 9.14168835e-01
-1.45917805e-02 3.16186137e-02 7.81932056e-01 1.43113291e+00
1.33702636e+00 2.59724021e-01 -3.91546071e-01 6.89982295e-01
1.52322102e+00 7.52717197e-01 1.06310618e+00 -7.50271320e-01
-1.05287099e+00 7.22825944e-01 -4.89012361e-01 -1.94546139e+00
-1.26705110e+00 -1.58814049e+00 -5.10174036e-01 -9.17886019e-01
-1.25909358e-01 -5.74812114e-01 -5.85974574e-01 1.16790330e+00
4.06973153e-01 -5.99573553e-01 4.33362693e-01 8.26421022e-01
6.45185411e-01 1.68286383e+00 -1.59153819e-01 -2.27537975e-01
1.47872198e+00 -2.69531995e-01 -3.99842083e-01 -5.77515364e-01
-1.31758976e+00 -1.84768951e+00 1.38838470e+00 -1.19244874e-01
-6.55742526e-01 -1.78996474e-01 -1.94816661e+00 4.93299305e-01
-2.02572599e-01 -3.88301700e-01 7.02328563e-01 1.17230272e+00
-1.62834585e+00 4.63578254e-01 -9.27193880e-01 -1.48877227e+00
6.74021170e-02 4.17166084e-01 2.40657791e-01 3.96409295e-02
-9.85701978e-01 8.55417132e-01 -5.60560107e-01 8.19134533e-01
-2.03017998e+00 -4.17571157e-01 -3.17130774e-01 -3.25608820e-01
1.07703090e+00 -1.79083392e-01 1.47959542e+00 -1.12530625e+00
-7.64682651e-01 7.53292918e-01 -8.00758421e-01 -1.01646207e-01
6.89379752e-01 7.94367671e-01 4.36850302e-02 1.87571898e-01
1.98776424e-02 6.59577370e-01 7.62413263e-01 -2.40919566e+00
-3.15257013e-01 -1.66152453e+00 1.77511096e-01 -7.20884353e-02
1.85007602e-01 9.07687902e-01 -3.32885943e-02 -5.60977519e-01
2.00537056e-01 -1.93417203e+00 -2.40729436e-01 -1.56440377e+00
-4.99085784e-01 4.34801370e-01 1.17068863e+00 3.11430156e-01
1.07862628e+00 -1.99890709e+00 -1.07574010e+00 9.47644770e-01
3.28963786e-01 8.38044763e-01 9.94116887e-02 2.54321599e+00
6.17389202e-01 1.01487756e+00 -1.11998248e+00 9.17922914e-01
-2.83796728e-01 1.35467514e-01 1.23464680e+00 -4.30461653e-02
5.21941632e-02 -4.63568747e-01 -4.68609929e-02 1.70137569e-01
4.00092691e-01 6.07045352e-01 7.28479207e-01 3.10869366e-01
1.31856605e-01 2.02071756e-01 -6.87923506e-02 1.19844139e+00
1.17839634e+00 -3.72886717e-01 9.17447627e-01 -1.14118330e-01
-5.87582231e-01 -5.22032082e-01 -2.49375916e+00 1.03189301e+00
9.21714380e-02 6.00086808e-01 6.19929790e-01 -6.29379809e-01
1.41166329e+00 1.40502906e+00 9.05130386e-01 6.06434867e-02
3.36413413e-01 8.85271311e-01 2.89139450e-01 -1.09970585e-01
1.04285932e+00 -1.17355156e+00 7.50588834e-01 5.42860985e-01
-6.86147153e-01 3.43441993e-01 4.81228352e-01 -7.96454906e-01
1.50160146e+00 -8.76129091e-01 -3.81254256e-01 -7.69225359e-01
5.67771494e-01 -2.85270035e-01 2.00178936e-01 4.83146459e-02
-9.03528154e-01 3.80799204e-01 -8.32887948e-01 4.98658895e-01
-1.39668450e-01 -1.87221539e+00 -3.89700860e-01 9.58606184e-01
-7.41143152e-02 -9.97158706e-01 -6.79862499e-01 -3.82665217e-01
-2.93940425e-01 4.74011242e-01 1.49813265e-01 4.95810658e-02
-5.97934127e-01 -1.11581993e+00 -3.43630999e-01 -2.60681272e-01
1.47706258e+00 -1.44730842e+00 -2.03940511e-01 -1.48743063e-01
1.15750395e-01 -3.98336262e-01 -5.43308318e-01 -3.17662716e-01
-1.81568861e+00 -1.16860723e+00 4.46498245e-01 -1.61681306e+00
1.23463631e+00 1.46201506e-01 7.49307215e-01 6.58871055e-01
8.87571927e-03 7.11112320e-01 2.51310319e-01 -1.25285411e+00
-3.18165310e-02 2.16031224e-01 2.89187938e-01 1.40148878e+00
1.09553432e+00 -3.60353917e-01 -1.28203011e+00 5.79333246e-01
-1.38496971e+00 -2.14330405e-01 1.94966823e-01 -5.14601246e-02
1.10675442e+00 3.22016478e-01 8.39952081e-02 -7.24815130e-01
1.23006451e+00 4.46737617e-01 -1.28666568e+00 1.35829166e-01
-1.66554284e+00 -9.10547256e-01 1.18823504e+00 -4.48725998e-01
-8.25806260e-01 -8.50737572e-01 -2.75468946e-01 -1.78374633e-01
4.26511258e-01 -9.57258418e-02 -5.10153949e-01 7.67352760e-01
-5.90322495e-01 9.20067579e-02 7.04398096e-01 -7.96582043e-01
-3.90669614e-01 1.19751596e+00 2.52074957e-01 -8.02389205e-01
5.23098111e-01 4.41858172e-01 7.54504427e-02 -2.51326591e-01
3.80705535e-01 -3.84771414e-02 -6.93185866e-01 7.33191490e-01
9.24189746e-01 -1.22578883e+00 -2.51734406e-01 6.23262882e-01
-4.23859745e-01 7.54569769e-01 -5.92272222e-01 1.50903177e+00
-1.53918892e-01 -3.76293242e-01 -7.50391185e-01 -6.44934356e-01
-5.93931153e-02 -2.78446651e+00 -7.25381151e-02 2.34103227e+00
-1.99298784e-01 -8.69154930e-01 -5.36869920e-04 3.67698133e-01
8.38191986e-01 -2.00821742e-01 9.23204899e-01 -9.28927660e-01
-1.16344941e+00 -4.82054919e-01 1.45997643e-01 1.44483757e+00
7.96844482e-01 6.74670160e-01 1.96450427e-01 -3.68925899e-01
-2.65858382e-01 7.76774704e-01 5.88565588e-01 8.17993224e-01
-2.09696561e-01 -7.67627180e-01 -1.01786271e-01 4.65832144e-01
2.25514150e+00 1.28381944e+00 6.04656100e-01 2.16214299e+00
-7.89411724e-01 -5.90961128e-02 -1.65068414e-02 5.38983583e-01
8.05900842e-02 7.45593727e-01 -3.78434695e-02 1.33884394e+00
-1.25505936e+00 -2.55698144e-01 8.21831107e-01 1.33347034e+00
4.80116338e-01 2.96581835e-01 -7.75495619e-02 8.25362682e-01
-2.44493771e+00 -1.96631598e+00 5.98842621e-01 2.52459145e+00
2.76990622e-01 -5.51666737e-01 5.39700761e-02 9.64256153e-02
1.73670244e+00 -2.51326829e-01 -1.71149939e-01 -2.19663858e-01
5.05614996e-01 1.26558995e+00 5.74158192e-01 3.82070243e-01
-8.28557312e-01 4.62995209e-02 4.96690571e-01 4.96116310e-01
-2.13127589e+00 -1.87845081e-01 3.01058441e-01 -1.32596517e+00
-8.83826137e-01 7.84951150e-01 -5.24683893e-01 2.09266424e+00
8.07539344e-01 -1.11425102e+00 -1.63292438e-01 -4.73258644e-03
9.87900734e-01 1.78498313e-01 -8.56812418e-01 1.36144674e+00
2.20367998e-01 -1.70390522e+00 -1.17233492e-01 4.73707259e-01
8.07003915e-01 -3.89035612e-01 -9.94812250e-02 4.71317861e-03
1.39478117e-01 -1.27576435e+00 1.05650091e+00 -1.69836625e-01
-3.38699639e-01 -9.35219467e-01 1.45289302e+00 1.10802790e-02
-1.39285910e+00 -7.51808882e-01 1.97742090e-01 -5.55992961e-01
5.03215373e-01 2.01979682e-01 -2.06206784e-01 2.06497163e-01
-2.35130265e-01 -2.21477851e-01 -9.68966722e-01 2.30649781e+00
-8.26295912e-01 -2.86789387e-01 -4.06615973e-01 -1.81922841e+00
9.44997787e-01 -3.69712561e-01 1.01070309e+00 1.35790610e+00
3.22766542e-01 8.07398617e-01 3.85308176e-01 5.77072620e-01
-6.17385864e-01 3.34750146e-01 -1.12160549e-01 5.26470244e-01
1.06349170e+00 1.31933868e+00 -1.42544496e+00 -6.99916184e-01
-2.31322702e-02 -9.10849571e-01 -1.23767294e-01 -8.31768215e-01
-2.74845660e-02 -6.92826569e-01 -1.76223949e-01 5.62417448e-01
-5.85912108e-01 -9.72250283e-01 8.68278623e-01 -2.40151688e-01
-3.63960236e-01 -1.17876518e+00 1.34702146e-01 -6.69389009e-01
-1.40998769e+00 1.28870830e-01 3.07323694e-01 -1.15205359e+00
6.08718336e-01 -1.56472492e+00 1.29195705e-01 4.99368757e-01
3.50923061e-01 -9.33584273e-01 -6.06170185e-02 -2.01580495e-01
1.32594132e+00 -1.58173367e-01 8.20745528e-01 8.23127031e-01
-1.13255763e+00 3.11063975e-01 1.71692753e+00 8.01372051e-01
2.93763697e-01 -1.97010720e+00 -1.40742528e+00 4.87579912e-01
1.09379780e+00 1.01628087e-01 2.37095416e-01 -4.19302166e-01
-1.39609265e+00 -7.69075036e-01 5.82083583e-01 2.88112491e-01
4.46861774e-01 -2.21989065e-01 -1.04896948e-01 1.46381533e+00
6.13287807e-01 2.06986070e-01 6.58484340e-01 -4.58387911e-01
3.44213873e-01 -1.41020322e+00 -1.99509811e+00 1.26843584e+00
3.15110236e-01 2.17738450e-02 -8.06298912e-01 2.35394061e-01
2.66023353e-02 -7.85708427e-01 -1.35387754e+00 4.36297238e-01
1.76992691e+00 -9.25020635e-01 -3.59458596e-01 -4.12494600e-01
-2.90873617e-01 -2.32912794e-01 -6.79338396e-01 -1.25027502e+00
6.24153614e-01 -8.52108061e-01 1.62342966e+00 1.49629045e+00
5.51749110e-01 -1.04387999e+00 8.28756452e-01 1.74635410e+00
-7.19467580e-01 -8.19648921e-01 -1.99819267e+00 -1.11547780e+00
2.52421260e-01 -5.27705662e-02 2.51014233e-01 -2.81590134e-01
-4.13490742e-01 3.35193515e-01 4.63318557e-01 -1.86361343e-01
4.74512517e-01 -7.10597873e-01 1.10170461e-01 -1.07713461e+00
8.23283970e-01 -5.98457217e-01 -4.99801964e-01 2.17343971e-01
-9.66726780e-01 -3.12992245e-01 -2.04005575e+00 -2.09737873e+00
8.01783442e-01 -1.58185825e-01 -2.71330494e-02 -5.74555576e-01
1.71302831e+00 1.49571359e-01 1.10086322e+00 8.59715283e-01
3.87797713e-01 -6.65541351e-01 1.22469091e+00 -7.59458601e-01
-3.75081301e-01 1.41967162e-01 -5.26583135e-01 1.24019682e+00
2.92061627e-01 -1.43402234e-01 -1.35754466e-01 -7.28829205e-01
-1.02288648e-01 -1.36336255e+00 -1.12511575e+00 -1.19613779e+00
-9.40794647e-02 1.89141676e-01 1.55759916e-01 5.97325027e-01
7.09545434e-01 -6.76864386e-01 8.01690400e-01 3.23862761e-01
1.09029993e-01 1.20208764e+00 2.24825572e-02 -7.22009182e-01
3.17279696e-02 -1.61745417e+00 9.09542918e-01 -6.82113647e-01
-3.83288413e-01 -3.83111686e-01 -3.39988321e-01 -9.23732817e-01
1.42848873e+00 -1.20787509e-01 -9.57259715e-01 4.11829650e-01
-1.12093830e+00 -2.29056314e-01 6.55219078e-01 6.54738784e-01
1.46679604e+00 -1.34496641e+00 -2.06046963e+00 1.84185043e-01
-6.41668260e-01 2.91084141e-01 4.17186081e-01 8.02469969e-01
-5.71576774e-01 5.25105238e-01 -7.76173830e-01 -6.96230352e-01
-2.69351959e+00 9.24915373e-01 -3.53187084e-01 -1.69978231e-01
6.31764457e-02 6.16715431e-01 -7.29680002e-01 5.89762926e-01
3.13241899e-01 -1.86113760e-01 -3.78476739e-01 4.89742190e-01
3.63348216e-01 3.07393789e-01 2.23060790e-02 -1.42076027e+00
-8.11212361e-01 1.47114038e+00 -8.76883090e-01 -6.16187453e-01
2.90119916e-01 -2.73780465e-01 -4.97551352e-01 -1.45406835e-02
1.13150799e+00 1.18375397e+00 -3.62729698e-01 6.76825285e-01
-8.50302279e-01 -4.20793235e-01 -7.07982123e-01 -1.96661758e+00
-5.08590519e-01 6.20147586e-01 6.38416529e-01 -8.63987744e-01
6.06478155e-01 -5.18633068e-01 1.21328378e+00 3.32539469e-01
1.86757350e+00 -2.33735728e+00 -9.44912076e-01 6.84455454e-01
1.09092867e+00 -1.80577660e+00 4.53533947e-01 -2.14690194e-01
3.74560952e-01 1.60368514e+00 9.15304840e-01 -7.95554817e-01
1.21957266e+00 1.00805320e-01 -4.26700175e-01 -2.12941423e-01
-2.49119103e-01 -8.59741509e-01 -1.51615873e-01 9.10135984e-01
1.02963173e+00 1.55068621e-01 -1.49857962e+00 5.06949246e-01
-8.72032940e-01 5.95784247e-01 1.57438612e+00 1.09673738e+00
-1.01478708e+00 -5.26776969e-01 -1.69334865e+00 1.88283086e-01
-1.02269208e+00 1.22569072e+00 4.64330576e-02 1.08596373e+00
6.39153421e-01 3.02008533e+00 -4.63438332e-01 -1.12374961e+00
2.01647592e+00 -2.00860679e-01 1.06604648e+00 -1.36807159e-01
-1.62141776e+00 -4.89228487e-01 2.01142162e-01 7.24321425e-01
-4.41852719e-01 -9.17231813e-02 -8.31419230e-01 -1.16912401e+00
4.38279845e-02 2.14506125e+00 1.20928478e+00 1.74620673e-01
-1.83072239e-01 8.60717893e-01 1.09637713e+00 -7.33356297e-01
-1.36916292e+00 -8.40334654e-01 -1.03849041e+00 -3.92925233e-01
5.64914308e-02 -6.68387413e-01 -1.57715440e+00 -9.53689754e-01] | [-3.3159680366516113, 6.90767765045166] |
533f3c9d-09d2-478e-917e-05845f555c53 | graph-based-network-with-contextualized | 2109.04008 | null | https://arxiv.org/abs/2109.04008v1 | https://arxiv.org/pdf/2109.04008v1.pdf | Graph Based Network with Contextualized Representations of Turns in Dialogue | Dialogue-based relation extraction (RE) aims to extract relation(s) between two arguments that appear in a dialogue. Because dialogues have the characteristics of high personal pronoun occurrences and low information density, and since most relational facts in dialogues are not supported by any single sentence, dialogue-based relation extraction requires a comprehensive understanding of dialogue. In this paper, we propose the TUrn COntext awaRE Graph Convolutional Network (TUCORE-GCN) modeled by paying attention to the way people understand dialogues. In addition, we propose a novel approach which treats the task of emotion recognition in conversations (ERC) as a dialogue-based RE. Experiments on a dialogue-based RE dataset and three ERC datasets demonstrate that our model is very effective in various dialogue-based natural language understanding tasks. In these experiments, TUCORE-GCN outperforms the state-of-the-art models on most of the benchmark datasets. Our code is available at https://github.com/BlackNoodle/TUCORE-GCN. | ['Yong Suk Choi', 'Bongseok Lee'] | 2021-09-09 | null | https://aclanthology.org/2021.emnlp-main.36 | https://aclanthology.org/2021.emnlp-main.36.pdf | emnlp-2021-11 | ['dialog-relation-extraction', 'emotion-recognition-in-conversation'] | ['natural-language-processing', 'natural-language-processing'] | [-4.87781726e-02 5.84744632e-01 -1.94821674e-02 -6.52584851e-01
-4.20213580e-01 -3.70750666e-01 8.80168498e-01 2.41892010e-01
-2.27874249e-01 8.65550637e-01 6.40430033e-01 -4.92006153e-01
2.79778957e-01 -9.60422158e-01 -1.80163756e-01 -4.43467721e-02
1.21164396e-01 6.67758226e-01 -1.29432157e-01 -8.85583699e-01
7.33691677e-02 8.40766653e-02 -8.39501917e-01 6.60652161e-01
9.65546072e-01 8.73067021e-01 -3.71643245e-01 9.59749699e-01
-6.51740611e-01 1.45132959e+00 -7.32968748e-01 -8.65705073e-01
-4.20266569e-01 -8.50136518e-01 -1.61943543e+00 -1.58440784e-01
-2.09721252e-01 -2.75498003e-01 -4.25688297e-01 7.58961797e-01
3.83698046e-01 3.22088927e-01 6.64623022e-01 -1.30091286e+00
-5.21744370e-01 8.83268654e-01 -1.20787606e-01 1.80250272e-01
8.44094634e-01 -2.02235237e-01 1.34941518e+00 -6.83806360e-01
6.60754085e-01 1.65032494e+00 3.51084143e-01 8.88142407e-01
-9.66928422e-01 -4.74781185e-01 3.75601143e-01 3.38835090e-01
-8.48038137e-01 -4.94194090e-01 1.05157197e+00 -7.43989870e-02
1.60071087e+00 5.03378034e-01 7.33568013e-01 1.43735588e+00
7.96128586e-02 1.03294158e+00 9.71739292e-01 -4.27430362e-01
-7.06551746e-02 -1.25805857e-02 6.29753113e-01 7.26780593e-01
-3.57733846e-01 -4.89287108e-01 -6.86372101e-01 -2.05195159e-01
2.44521111e-01 -3.74008417e-01 -5.14334381e-01 3.01073968e-01
-8.96283686e-01 1.11339939e+00 2.53235698e-01 3.36457670e-01
-1.33260116e-01 -2.06051126e-01 8.65166724e-01 5.61172128e-01
9.01495516e-01 5.19573271e-01 -5.27972639e-01 -5.38520336e-01
5.60108572e-02 2.18046248e-01 1.70737016e+00 8.82316649e-01
4.01614845e-01 -4.21316177e-01 -2.91115701e-01 1.18790686e+00
1.66996107e-01 1.62213504e-01 2.42946699e-01 -8.42627466e-01
7.77945459e-01 1.06939662e+00 -2.36326247e-01 -1.18873584e+00
-7.01566935e-01 6.39008284e-02 -1.11499107e+00 -5.96159756e-01
2.13499963e-01 -6.04595542e-01 -1.32594705e-01 1.65239596e+00
4.62774038e-01 -8.19186643e-02 5.70906997e-01 5.94089627e-01
1.82126248e+00 7.98850954e-01 1.81902796e-02 -4.04296845e-01
1.56772101e+00 -1.18482161e+00 -1.35012770e+00 -3.20985556e-01
7.57538319e-01 -4.96132016e-01 9.52490628e-01 9.84537229e-02
-5.93875527e-01 -1.26054823e-01 -7.19768465e-01 -5.32178104e-01
-4.53751415e-01 -5.30169308e-02 1.00347459e+00 2.83052325e-01
-7.50527382e-01 2.72654116e-01 -4.11161125e-01 -4.11779255e-01
2.63218910e-01 1.97249413e-01 -3.50322068e-01 2.63242573e-01
-1.66486168e+00 1.08220768e+00 3.30031931e-01 4.76754248e-01
-4.54587698e-01 -5.67901097e-02 -1.14019561e+00 1.42160401e-01
6.99380577e-01 -6.20643198e-01 1.44522882e+00 -6.41433597e-01
-1.99976337e+00 9.24622655e-01 -3.47613096e-01 -5.18141270e-01
2.54925549e-01 -4.09122318e-01 -4.08626348e-01 8.55184793e-02
-2.48595074e-01 2.68450171e-01 2.22614393e-01 -1.00224090e+00
-3.93647909e-01 -1.23644918e-01 6.15847528e-01 4.47573543e-01
-1.56649396e-01 2.00109750e-01 -2.65544683e-01 -4.21451703e-02
-2.77426094e-01 -7.64891267e-01 -3.34320888e-02 -7.56923616e-01
-1.01431179e+00 -1.08770013e+00 7.45520175e-01 -6.11104310e-01
1.36896265e+00 -1.73253608e+00 2.40209043e-01 -2.62657732e-01
5.00789523e-01 3.31974953e-01 -1.83890402e-01 9.11107600e-01
-3.79772335e-02 1.94846854e-01 -4.91590947e-02 -3.82524461e-01
1.78120196e-01 3.91243786e-01 -2.55138695e-01 -7.31553556e-03
3.33648801e-01 1.11684191e+00 -1.00852752e+00 -5.47066092e-01
1.39841855e-01 4.08131570e-01 -2.56202310e-01 6.82591736e-01
-4.64953005e-01 4.92854238e-01 -6.90161228e-01 2.99551934e-01
3.43200684e-01 -4.77505714e-01 6.80483997e-01 -1.46904632e-01
2.52699375e-01 9.78476167e-01 -4.32406098e-01 1.53676677e+00
-7.78112471e-01 8.31802011e-01 1.49481585e-02 -1.08594251e+00
1.12211823e+00 5.99042416e-01 6.10499224e-03 -5.49911559e-01
4.50640529e-01 -5.56440875e-02 1.49456009e-01 -6.53998613e-01
5.86905956e-01 2.38362122e-02 -2.82688588e-01 6.40464067e-01
2.62564927e-01 -2.18578443e-01 2.96201855e-01 7.78241456e-01
9.85237837e-01 -3.01874548e-01 6.91586494e-01 4.88601997e-03
9.08942223e-01 -2.55105555e-01 6.42157197e-01 4.69706625e-01
-2.26683617e-01 3.67972590e-02 1.28746438e+00 -4.67929095e-01
-3.02869827e-01 -2.84655839e-01 2.17973366e-01 1.04606366e+00
6.44664019e-02 -9.61288393e-01 -6.88027680e-01 -1.20387137e+00
-3.40412706e-01 8.83082330e-01 -7.43953049e-01 4.89648664e-03
-5.46374261e-01 -5.29822588e-01 5.28515756e-01 2.73175955e-01
9.30880487e-01 -1.34338593e+00 -1.34506106e-01 1.59340307e-01
-9.15361345e-01 -1.39486206e+00 -3.43820870e-01 9.33347195e-02
-4.28496242e-01 -1.37171566e+00 -1.44241214e-01 -6.39035046e-01
3.45359415e-01 -5.51292896e-02 1.60586143e+00 3.71696800e-01
4.29708790e-03 5.08151472e-01 -6.67657554e-01 -5.20181715e-01
-5.23625672e-01 4.07819629e-01 -4.73264188e-01 -6.33272305e-02
8.73404741e-01 -5.38451672e-01 -3.74054611e-01 -4.61122952e-02
-3.11320424e-01 3.25141162e-01 -1.33319974e-01 1.10062277e+00
-6.95881471e-02 -2.21846417e-01 8.52673411e-01 -1.40924740e+00
1.43499351e+00 -4.49423879e-01 -9.29688364e-02 4.71162736e-01
-3.23395997e-01 -1.67416245e-01 6.15930974e-01 -6.57732189e-02
-1.33989882e+00 -3.59648585e-01 -3.95402879e-01 3.50706309e-01
-3.22326720e-01 6.72503710e-01 -2.00494081e-01 5.27951121e-01
3.58560711e-01 -7.97811896e-02 -4.06067967e-02 -2.45949566e-01
4.51655686e-01 9.54182923e-01 1.69009581e-01 -6.93232238e-01
1.33066252e-01 1.28678288e-02 -2.35981166e-01 -9.78636563e-01
-1.47428632e+00 -4.59240258e-01 -5.80025673e-01 -3.65392506e-01
1.13633287e+00 -7.44208157e-01 -1.42778027e+00 3.66743147e-01
-1.68616974e+00 -5.79895616e-01 -2.12644152e-02 1.70607120e-01
-3.01134825e-01 3.37024152e-01 -1.06395090e+00 -1.28674507e+00
-8.21243882e-01 -7.68497765e-01 8.50486457e-01 5.32180071e-01
-5.53839624e-01 -1.37661421e+00 1.09870538e-01 7.35419869e-01
1.23102657e-01 2.90361196e-01 9.06155407e-01 -1.27867007e+00
-4.18903828e-02 5.05347922e-02 -2.49181136e-01 3.75547558e-01
3.44964296e-01 -1.04652099e-01 -1.04292941e+00 1.41749650e-01
-4.79556918e-02 -9.09917295e-01 7.87917376e-01 -1.33445919e-01
9.47951138e-01 -5.22182286e-01 -1.51340127e-01 7.99179897e-02
7.17945039e-01 2.22727433e-01 5.63049018e-01 1.10502504e-01
6.51817620e-01 9.03902888e-01 5.51618874e-01 3.78002167e-01
9.67621148e-01 4.61553037e-01 2.84264684e-01 -2.06140310e-01
1.30948499e-01 -1.34188935e-01 2.67714798e-01 1.19144082e+00
-2.02152893e-01 -6.86391711e-01 -9.12831008e-01 4.61589515e-01
-2.18226290e+00 -9.50260937e-01 -4.36187178e-01 1.40392780e+00
1.26011717e+00 5.51672466e-02 1.78638715e-02 -1.66836202e-01
6.70537651e-01 5.23380518e-01 -3.84753019e-01 -1.03402162e+00
-7.64642954e-02 1.33132070e-01 -3.56592983e-01 9.11885619e-01
-1.20771813e+00 1.09222412e+00 4.96375704e+00 5.89368999e-01
-5.79168618e-01 -1.12148769e-01 9.40026164e-01 3.57861727e-01
-1.52343139e-01 -6.39335364e-02 -7.61618257e-01 -2.96461340e-02
9.53013897e-01 -3.62264305e-01 3.72154742e-01 6.08697414e-01
-4.44052666e-02 -1.44364446e-01 -1.22974765e+00 8.21265817e-01
1.36915326e-01 -1.18534613e+00 -1.12931859e-02 -1.62249506e-01
2.80233324e-01 -2.30622485e-01 -4.50969547e-01 8.09516490e-01
4.94674653e-01 -1.18936253e+00 -1.97369814e-01 4.35862988e-01
5.48688293e-01 -8.82269382e-01 1.18883216e+00 3.49121124e-01
-1.05160773e+00 3.26729685e-01 -1.74757585e-01 -3.78032118e-01
1.09888211e-01 4.96891856e-01 -9.59376991e-01 8.58551860e-01
4.58033592e-01 8.90580952e-01 -2.62770951e-01 2.69957576e-02
-7.79200137e-01 6.84061646e-01 -1.55639037e-01 -5.50406992e-01
1.81258991e-01 -3.69753569e-01 4.25637364e-01 1.60655975e+00
-3.93580884e-01 7.73728728e-01 2.56929487e-01 6.95584714e-01
-5.07749140e-01 3.09903979e-01 -6.04394495e-01 -3.78406107e-01
4.04902428e-01 1.64911139e+00 -4.13055599e-01 -3.93851548e-01
-5.35549104e-01 1.04298210e+00 7.47033417e-01 2.44416982e-01
-5.36680877e-01 -5.44887066e-01 5.05999506e-01 -6.75178945e-01
-1.04158923e-01 3.90833132e-02 1.24009565e-01 -1.33700240e+00
-9.61415172e-02 -1.12788308e+00 5.37815511e-01 -4.31201518e-01
-1.62856734e+00 1.03888202e+00 -2.79961735e-01 -6.30284429e-01
-4.79489058e-01 -6.16756737e-01 -8.98149192e-01 8.22987556e-01
-1.53565359e+00 -1.09525120e+00 -4.25958782e-01 7.07121968e-01
7.74874568e-01 -1.42729864e-01 1.44780636e+00 -5.81705868e-02
-8.08423519e-01 4.61665958e-01 -5.74318588e-01 8.42971027e-01
6.29654944e-01 -1.47106028e+00 5.08258045e-01 2.95850992e-01
1.06302567e-01 6.66734755e-01 4.67063755e-01 -5.92808545e-01
-1.20120633e+00 -7.56440282e-01 1.41831625e+00 -3.16420823e-01
7.53270268e-01 -6.24261975e-01 -1.01309192e+00 8.17599595e-01
9.35074568e-01 -3.20840150e-01 1.12485731e+00 9.15240288e-01
-3.49022746e-01 1.56270012e-01 -7.57849336e-01 5.84103942e-01
9.56360221e-01 -8.40840459e-01 -9.41052854e-01 7.04890609e-01
8.58510435e-01 -6.77382708e-01 -1.03982329e+00 2.58006185e-01
2.63747364e-01 -8.78418684e-01 6.38384998e-01 -9.38542485e-01
8.55442286e-01 3.64029050e-01 1.61468044e-01 -1.47798645e+00
1.96853444e-01 -9.46458876e-01 -4.42365944e-01 1.60452533e+00
6.29597962e-01 -7.08813667e-01 2.75270700e-01 8.20428014e-01
4.81323600e-02 -9.58307803e-01 -7.26735651e-01 -3.36047202e-01
2.80807689e-02 -1.06029212e-01 4.56393808e-01 1.25392687e+00
6.78269029e-01 1.46049595e+00 -3.95915270e-01 -2.29954690e-01
7.13961571e-02 4.41079378e-01 9.29775178e-01 -1.28597677e+00
-9.30023938e-02 -3.64413708e-01 1.75710559e-01 -1.18352997e+00
8.22069585e-01 -7.97923803e-01 6.31284192e-02 -1.95113766e+00
3.10938865e-01 -4.93596680e-02 6.99371919e-02 4.68809158e-01
-5.21401167e-01 -3.97417903e-01 7.71792233e-02 -2.23972902e-01
-9.47378278e-01 9.95155931e-01 1.36927509e+00 -2.89133817e-01
-1.17305331e-01 3.64725315e-03 -6.89873815e-01 7.74460793e-01
1.08113599e+00 -1.70675918e-01 -5.22994399e-01 -2.84602549e-02
3.80862892e-01 3.93431872e-01 1.23681836e-01 -1.61898300e-01
2.69811779e-01 -2.43793339e-01 -1.87054366e-01 -5.77009082e-01
5.12801290e-01 -3.27441692e-01 -6.06234431e-01 -2.48425407e-04
-8.42740476e-01 -1.16416283e-01 1.04928158e-01 4.73328203e-01
-4.95396316e-01 -1.56183988e-01 2.45758742e-01 -2.25427970e-01
-4.35397595e-01 6.65377220e-03 -4.28463489e-01 6.22673154e-01
5.15824795e-01 5.94737291e-01 -7.54058123e-01 -1.02672410e+00
-5.56984305e-01 6.04053617e-01 -2.82226801e-01 4.69484180e-01
6.87148213e-01 -1.01124179e+00 -9.27573025e-01 -3.98166001e-01
9.44518000e-02 2.40683377e-01 2.28738800e-01 7.04335570e-01
-2.81056315e-01 5.56231201e-01 2.76402414e-01 -1.73245758e-01
-1.71772254e+00 1.95126474e-01 4.96268898e-01 -7.76130199e-01
-7.27373600e-01 1.12708533e+00 7.75382817e-02 -9.02249515e-01
3.30417454e-01 -2.14267388e-01 -8.35139990e-01 6.57295734e-02
4.56759423e-01 -1.12497173e-02 -2.19834358e-01 -4.89726216e-01
-4.06947941e-01 -2.56104693e-02 -3.50097626e-01 1.13712341e-01
1.43667114e+00 -1.84014902e-01 -5.95711529e-01 7.56330788e-01
1.06500506e+00 -2.84340419e-02 -7.22628593e-01 -3.50361586e-01
1.20915093e-01 -4.23367281e-04 -1.80171534e-01 -9.64836955e-01
-9.22778606e-01 9.20910358e-01 -2.34750986e-01 6.54452443e-01
8.90781641e-01 8.68238136e-02 9.90091264e-01 9.66220915e-01
-2.73223687e-02 -1.26372063e+00 6.90710694e-02 1.19622111e+00
1.23058367e+00 -1.47441328e+00 -1.01122097e-03 -8.89649153e-01
-1.09453416e+00 1.27948964e+00 9.50950921e-01 1.35588825e-01
4.84375060e-01 2.60769039e-01 2.97963351e-01 -6.23815000e-01
-1.35808229e+00 -2.51109779e-01 2.29203671e-01 3.94758075e-01
1.10561347e+00 1.40274107e-01 -5.54958463e-01 7.93082535e-01
-3.52898985e-01 -3.35583746e-01 6.05386436e-01 6.89631343e-01
2.68897340e-02 -1.30484354e+00 3.77329916e-01 3.11169833e-01
-3.68918717e-01 -3.33983958e-01 -1.31344390e+00 8.24012756e-01
-5.36042869e-01 1.67185426e+00 -1.02738544e-01 -4.57828164e-01
3.39862674e-01 3.69275749e-01 1.52555555e-01 -7.63701558e-01
-1.02752936e+00 -2.44441643e-01 1.30129659e+00 -6.40963674e-01
-7.27147400e-01 -3.00431103e-01 -1.49194539e+00 -4.19542551e-01
-4.80020195e-01 6.19991660e-01 2.38542020e-01 1.12006843e+00
2.77749002e-01 9.16717410e-01 6.77499175e-01 -1.27248317e-01
-1.69915289e-01 -1.36092091e+00 -3.66738975e-01 4.67776179e-01
1.82088569e-01 -5.19058228e-01 -2.47920781e-01 -3.69669169e-01] | [12.497941970825195, 7.953983306884766] |
6bff801f-6a76-497c-8a83-fe77b585a3b9 | general-partial-label-learning-via-dual | 2001.01290 | null | https://arxiv.org/abs/2001.01290v2 | https://arxiv.org/pdf/2001.01290v2.pdf | General Partial Label Learning via Dual Bipartite Graph Autoencoder | We formulate a practical yet challenging problem: General Partial Label Learning (GPLL). Compared to the traditional Partial Label Learning (PLL) problem, GPLL relaxes the supervision assumption from instance-level -- a label set partially labels an instance -- to group-level: 1) a label set partially labels a group of instances, where the within-group instance-label link annotations are missing, and 2) cross-group links are allowed -- instances in a group may be partially linked to the label set from another group. Such ambiguous group-level supervision is more practical in real-world scenarios as additional annotation on the instance-level is no longer required, e.g., face-naming in videos where the group consists of faces in a frame, labeled by a name set in the corresponding caption. In this paper, we propose a novel graph convolutional network (GCN) called Dual Bipartite Graph Autoencoder (DB-GAE) to tackle the label ambiguity challenge of GPLL. First, we exploit the cross-group correlations to represent the instance groups as dual bipartite graphs: within-group and cross-group, which reciprocally complements each other to resolve the linking ambiguities. Second, we design a GCN autoencoder to encode and decode them, where the decodings are considered as the refined results. It is worth noting that DB-GAE is self-supervised and transductive, as it only uses the group-level supervision without a separate offline training stage. Extensive experiments on two real-world datasets demonstrate that DB-GAE significantly outperforms the best baseline over absolute 0.159 F1-score and 24.8% accuracy. We further offer analysis on various levels of label ambiguities. | ['Shih-Fu Chang', 'Brian Chen', 'Hanwang Zhang', 'Bo Wu', 'Alireza Zareian'] | 2020-01-05 | null | null | null | null | ['partial-label-learning'] | ['methodology'] | [ 3.71961027e-01 6.88132644e-01 -4.15076673e-01 -5.31091511e-01
-5.08421421e-01 -5.74813128e-01 3.80759269e-01 1.47637399e-02
8.39400664e-02 7.25088537e-01 -7.43618384e-02 -1.94926456e-01
-8.86899605e-02 -7.13460267e-01 -1.08342624e+00 -7.51962125e-01
-9.06218141e-02 6.34735703e-01 -5.43666184e-02 1.98837128e-02
-3.92939329e-01 2.50079632e-01 -1.47110355e+00 4.49137628e-01
6.20541215e-01 1.08631599e+00 -6.96471310e-04 1.05512731e-01
-2.22282231e-01 9.17990685e-01 -4.68937993e-01 -6.02750659e-01
3.62246811e-01 -6.10790551e-01 -8.88517499e-01 5.30321360e-01
7.57607460e-01 -1.07356757e-01 -3.39366198e-01 1.26369727e+00
2.67976195e-01 -1.94574744e-01 5.47347367e-01 -1.82122719e+00
-6.94712460e-01 6.92008674e-01 -7.07355499e-01 -2.65797555e-01
2.65177190e-01 -1.27179116e-01 1.37417150e+00 -6.48146451e-01
7.32573509e-01 1.26960981e+00 6.51077807e-01 6.66805387e-01
-1.22155905e+00 -1.00489986e+00 4.48173642e-01 1.55225635e-01
-1.41875207e+00 -1.32188633e-01 9.21814501e-01 -5.30359983e-01
5.92436671e-01 1.71827134e-02 3.95792425e-01 9.58977818e-01
-2.25765452e-01 5.44343352e-01 1.19224834e+00 -2.67859429e-01
4.93381061e-02 -2.75143057e-01 3.62580776e-01 1.20024109e+00
2.26947173e-01 -1.11980354e-02 -6.34388924e-01 -4.86303344e-02
7.39027619e-01 -8.04740414e-02 -3.67834717e-01 -4.31234807e-01
-1.14128458e+00 7.05884635e-01 6.91013396e-01 3.00319254e-01
-5.87415174e-02 1.63112238e-01 3.12825441e-01 3.36884767e-01
4.02709305e-01 5.83430156e-02 -5.03786922e-01 5.66904247e-01
-7.40331888e-01 -1.89858332e-01 7.86430359e-01 1.43503797e+00
1.22960436e+00 -1.43237278e-01 -8.47399309e-02 9.14090872e-01
4.28468794e-01 2.07787573e-01 1.38722926e-01 -9.56408262e-01
5.42891979e-01 8.04549158e-01 -3.19096893e-01 -1.17403638e+00
-4.33235526e-01 -6.60874546e-01 -9.68418479e-01 -2.40285099e-01
3.84355664e-01 -6.42895252e-02 -9.71473038e-01 2.25832534e+00
4.73987341e-01 7.56120443e-01 -5.29512726e-02 9.06071305e-01
1.30757427e+00 5.44321120e-01 1.39546439e-01 -4.80386972e-01
1.55208445e+00 -1.22115695e+00 -9.41225052e-01 -3.43138933e-01
7.65789092e-01 -3.70385975e-01 6.16508424e-01 1.21045895e-02
-6.09037876e-01 -5.61704636e-01 -1.00775135e+00 1.39054833e-02
-2.51426578e-01 2.13065669e-01 6.84423447e-01 3.14390093e-01
-1.15390539e+00 3.74283582e-01 -4.00128722e-01 -2.60547817e-01
2.65039623e-01 7.00196564e-01 -8.97653282e-01 -3.53789508e-01
-1.32287014e+00 2.77640402e-01 6.60599589e-01 2.34188333e-01
-6.94553792e-01 -4.47797835e-01 -1.21472597e+00 1.06861830e-01
8.22718263e-01 -4.18215305e-01 8.34872305e-01 -1.26115537e+00
-1.00106704e+00 1.28056633e+00 -1.46683887e-01 -9.81263742e-02
2.90772542e-02 4.06767637e-01 -7.88112223e-01 1.96835563e-01
4.19467211e-01 1.01084208e+00 7.20430136e-01 -1.62283146e+00
-5.85427284e-01 -3.03880900e-01 4.88828659e-01 2.59859115e-02
-1.41229391e-01 -1.90743163e-01 -7.42689371e-01 -7.14539766e-01
4.49906826e-01 -1.10191464e+00 4.01387423e-01 -2.14627966e-01
-5.33106029e-01 -4.88520026e-01 8.74305964e-01 -6.33752942e-01
1.27399719e+00 -2.14775348e+00 1.30985126e-01 9.33772922e-02
4.03911859e-01 2.09320560e-01 -3.99279445e-01 2.95983106e-01
-5.76435626e-01 1.64479762e-01 -3.44771713e-01 -3.78047258e-01
4.15715426e-02 7.62305439e-01 -8.65462720e-02 2.68848538e-01
3.76121849e-01 9.20990050e-01 -1.01225471e+00 -6.19387627e-01
-2.27069527e-01 2.31888428e-01 -5.29461265e-01 2.89021641e-01
-3.65712553e-01 5.28584003e-01 -1.72927722e-01 9.32780921e-01
8.75214577e-01 -7.24857211e-01 7.03989029e-01 -6.06095254e-01
4.27034318e-01 7.32318908e-02 -1.27929842e+00 1.51018989e+00
-9.66334343e-02 2.65672773e-01 2.16971233e-01 -1.16671503e+00
7.09474027e-01 3.39217007e-01 5.58638513e-01 -5.44403791e-01
2.46820804e-02 3.05634230e-01 -1.62504613e-01 -3.43761593e-01
4.43428149e-03 -9.89282429e-02 -1.38910919e-01 3.69933009e-01
6.63613558e-01 5.31632721e-01 3.81765217e-01 3.26398343e-01
1.13640213e+00 2.20115557e-01 3.21664810e-01 -1.03769414e-01
6.63807452e-01 -4.18534398e-01 1.15893519e+00 4.54984397e-01
-2.43839428e-01 5.37674189e-01 9.06254828e-01 -2.89960444e-01
-5.35331488e-01 -7.14817822e-01 4.22514938e-02 1.18961000e+00
3.77096504e-01 -5.46209216e-01 -9.07971978e-01 -1.31689489e+00
1.06276333e-01 1.98163509e-01 -6.84853852e-01 -5.37834018e-02
-5.84131539e-01 -5.48696160e-01 4.18810546e-01 2.38525823e-01
6.14294529e-01 -1.12394404e+00 4.40562040e-01 1.23511806e-01
-4.27599430e-01 -1.46847093e+00 -6.63340390e-01 2.42677703e-01
-4.38468337e-01 -1.47004223e+00 -2.08265826e-01 -1.20174456e+00
1.05320144e+00 1.07260145e-01 1.31367314e+00 3.43748868e-01
1.21194296e-01 4.01852697e-01 -4.94444937e-01 2.03349426e-01
-2.47641414e-01 -9.06235129e-02 -2.59805713e-02 4.44081038e-01
3.76212329e-01 -4.57650065e-01 -2.95598894e-01 5.67385674e-01
-9.38346446e-01 1.08553790e-01 5.20544410e-01 1.16239274e+00
8.26382995e-01 1.99348498e-02 6.21256888e-01 -1.16730106e+00
1.50695816e-01 -6.25088871e-01 -5.57564974e-01 5.69671631e-01
-5.29721498e-01 -2.12656651e-02 3.76713991e-01 -3.99220526e-01
-6.40433252e-01 1.46227285e-01 2.23912280e-02 -5.43575108e-01
-2.53311753e-01 5.33881783e-01 -7.76931345e-01 -2.40478963e-01
1.50313079e-01 -1.56031236e-01 -1.17798977e-01 -2.96349257e-01
2.08701238e-01 5.66276252e-01 5.46938121e-01 -7.87084460e-01
5.18390834e-01 2.06168219e-01 3.12786251e-01 -2.80795515e-01
-1.26608694e+00 -4.29203749e-01 -7.26794124e-01 -4.01423126e-01
8.95651519e-01 -1.12486804e+00 -8.57886553e-01 3.74480695e-01
-1.11601424e+00 -3.46151114e-01 -5.04819080e-02 2.30440289e-01
-3.43872964e-01 4.55618501e-01 -6.90973818e-01 -1.06787607e-01
9.81894210e-02 -1.34632683e+00 1.25006258e+00 1.33785814e-01
2.79774755e-01 -1.02718639e+00 -2.71511227e-01 6.98241830e-01
-3.82473737e-01 3.23925257e-01 9.61405039e-01 -8.37782383e-01
-5.76593697e-01 1.12485081e-01 -5.29722512e-01 3.76293331e-01
2.16943979e-01 -3.74504834e-01 -8.89380276e-01 -5.67551494e-01
-3.12492520e-01 -4.01421577e-01 5.84021568e-01 -1.30784372e-02
1.10128522e+00 -4.06825870e-01 -4.37410563e-01 5.67397237e-01
1.38724494e+00 5.97842829e-03 1.65858448e-01 -5.12168221e-02
1.25025129e+00 7.44005740e-01 5.59632361e-01 1.29518867e-01
6.65993035e-01 8.12675774e-01 6.38452232e-01 -6.86605498e-02
-4.90003645e-01 -4.05937731e-01 3.28779370e-01 1.12671483e+00
1.88509285e-01 -5.66042781e-01 -7.55318165e-01 3.25248212e-01
-2.05160928e+00 -5.94601631e-01 -3.16551030e-01 1.87530446e+00
8.10887337e-01 -2.85101384e-01 -6.58688992e-02 -7.42854550e-02
1.38561404e+00 2.15231836e-01 -3.53929639e-01 1.00276634e-01
-1.63829535e-01 -6.73015937e-02 4.14987355e-01 5.74678421e-01
-1.28932655e+00 1.13250780e+00 5.21904898e+00 9.76409674e-01
-7.98493803e-01 3.41517061e-01 5.74520826e-01 3.11353117e-01
-2.16360033e-01 1.97415799e-01 -9.89175498e-01 6.73759401e-01
5.67397416e-01 4.74355996e-01 5.56469083e-01 4.99181986e-01
-4.91250008e-01 2.47688055e-01 -1.38598251e+00 1.08821309e+00
3.08274627e-01 -8.70525599e-01 1.91831425e-01 2.95326293e-01
7.44110882e-01 -2.61257887e-01 -2.02554703e-01 5.34635663e-01
4.15305346e-01 -8.19239020e-01 6.34909332e-01 2.55228221e-01
1.10015476e+00 -7.32079506e-01 7.80048430e-01 3.79640877e-01
-1.44698298e+00 -1.57741401e-02 -2.53159225e-01 1.57305047e-01
1.16716184e-01 7.79046535e-01 -6.39122546e-01 9.94273961e-01
3.59147340e-01 7.74265528e-01 -4.46334064e-01 5.23435533e-01
-6.83154404e-01 7.88077056e-01 -1.43054590e-01 6.11567438e-01
2.72292525e-01 -3.79150897e-01 3.84427816e-01 1.04403710e+00
2.24489078e-01 2.45254949e-01 5.96880019e-01 7.19730437e-01
-6.29886985e-01 -1.29483357e-01 -5.72297513e-01 -1.16375193e-01
5.09094477e-01 1.32039976e+00 -6.88227236e-01 -2.00333327e-01
-6.43038511e-01 1.00169861e+00 6.79523528e-01 5.32077074e-01
-9.54666793e-01 -1.27090156e-01 4.32856411e-01 -9.73869115e-02
2.74252951e-01 4.34413515e-02 4.46620941e-01 -1.22848177e+00
3.56384702e-02 -9.13184643e-01 8.10565114e-01 -8.01918983e-01
-1.51814699e+00 5.33948541e-01 -1.91398337e-01 -1.16196632e+00
-1.79764345e-01 -6.32054985e-01 8.73004347e-02 4.78080720e-01
-1.56221890e+00 -1.64597702e+00 -3.02729726e-01 6.77933395e-01
9.91780683e-02 -1.52495891e-01 8.21407139e-01 6.71451151e-01
-6.53368473e-01 9.49577510e-01 -4.15133655e-01 5.78647494e-01
8.44099879e-01 -1.17160141e+00 -1.80237144e-01 8.22135091e-01
5.03786445e-01 5.35851717e-01 2.32037425e-01 -8.36474895e-01
-1.07897246e+00 -1.40844667e+00 1.00246668e+00 -5.95119363e-03
7.06541240e-01 -6.38701677e-01 -9.06785905e-01 1.09670246e+00
1.82620734e-01 6.61596954e-01 9.10729945e-01 1.81539610e-01
-7.77127087e-01 -2.26483613e-01 -1.09084845e+00 2.31688350e-01
1.54505193e+00 -9.27077711e-01 -3.66592795e-01 5.72833359e-01
9.89078701e-01 -4.79570657e-01 -9.84646976e-01 7.83254504e-01
8.43351930e-02 -8.28682244e-01 5.94130576e-01 -4.41712081e-01
2.70729095e-01 -7.38839567e-01 -2.16931820e-01 -1.17439747e+00
-4.79302436e-01 -3.75054628e-01 -3.09198916e-01 1.64597464e+00
1.70708656e-01 -7.05544114e-01 6.38859749e-01 2.82508969e-01
-2.69488990e-01 -6.86530709e-01 -1.02225721e+00 -8.00155699e-01
-4.47126091e-01 -1.98398769e-01 8.40778708e-01 1.60844100e+00
-1.66571289e-01 4.25010264e-01 -5.48944294e-01 5.31115115e-01
7.72885561e-01 2.55894572e-01 3.77062917e-01 -1.38719213e+00
-4.61570323e-01 -6.24624491e-02 -5.48927724e-01 -1.06467831e+00
8.50849628e-01 -1.39518166e+00 -1.14009604e-02 -1.43930006e+00
2.51528263e-01 -5.82056463e-01 -5.47452033e-01 1.22828889e+00
-1.63403988e-01 4.81456816e-01 2.17477813e-01 1.91603079e-01
-9.51767385e-01 3.44822049e-01 1.31691420e+00 -4.50610936e-01
2.45622039e-01 -3.27618092e-01 -6.84322119e-01 5.12516141e-01
4.02531445e-01 -5.63569665e-01 -4.66670394e-01 -4.02515143e-01
4.08593357e-01 2.47587472e-01 2.91386545e-01 -8.13540518e-01
2.94974893e-01 1.27314046e-01 -1.85125142e-01 -3.25026959e-01
8.03668201e-02 -1.14204073e+00 4.30035174e-01 -1.48073109e-02
-2.36406282e-01 -3.16338092e-01 -1.74469560e-01 8.84807825e-01
-5.22546589e-01 -2.84368664e-01 4.43555117e-01 1.35564348e-02
-8.56505930e-01 6.03413463e-01 3.31143916e-01 -1.82721429e-02
1.15768278e+00 -5.77406958e-02 -5.05950034e-01 -2.64861852e-01
-1.11174428e+00 4.68987644e-01 4.08713371e-01 3.79937321e-01
2.13121146e-01 -1.70250738e+00 -4.63447809e-01 2.41026476e-01
4.14514035e-01 2.29125321e-01 2.58721828e-01 7.16123044e-01
-5.97907938e-02 1.95553362e-01 -6.26886413e-02 -6.24360323e-01
-1.35202289e+00 7.06633925e-01 2.70505399e-01 -4.22924101e-01
-2.76539534e-01 9.79139447e-01 5.83247602e-01 -8.62538993e-01
2.11174697e-01 5.83909228e-02 -3.61303717e-01 2.92683363e-01
7.07348064e-02 -1.60757955e-02 6.78883940e-02 -1.20874763e+00
-3.25335711e-01 7.05344856e-01 8.74132216e-02 3.32102180e-01
1.05925667e+00 -3.79758552e-02 -7.29645729e-01 3.38894248e-01
1.36974740e+00 -1.44120499e-01 -1.17874527e+00 -5.03546894e-01
-5.45264669e-02 -1.89657405e-01 -1.67207152e-01 -8.81112933e-01
-1.44469965e+00 6.06798828e-01 3.36525381e-01 9.05249119e-02
1.15671480e+00 2.32403606e-01 5.71113765e-01 1.44431353e-01
6.13739729e-01 -8.85037363e-01 4.40418981e-02 5.24949431e-01
7.86449909e-01 -1.18255925e+00 -2.37596765e-01 -1.01644111e+00
-5.05836129e-01 8.36190641e-01 8.58495772e-01 2.46797636e-01
6.22694731e-01 6.05520420e-02 -6.06330037e-02 -4.87190664e-01
-5.45590997e-01 -3.39088231e-01 3.19711536e-01 4.43450451e-01
3.56257528e-01 3.08184564e-01 -2.77475208e-01 7.31210828e-01
1.63412347e-01 -1.84167400e-01 9.72712338e-02 6.53580427e-01
-3.29386368e-02 -1.42104578e+00 -2.03216612e-01 3.15341413e-01
-3.71839583e-01 1.99675839e-02 -3.45072120e-01 6.72012389e-01
7.67327905e-01 1.09333122e+00 7.51523823e-02 -5.11618793e-01
9.72031727e-02 2.74308324e-01 4.55960900e-01 -7.94961393e-01
-4.31241810e-01 2.11187795e-01 1.98999360e-01 -5.57900906e-01
-8.81484449e-01 -3.24301541e-01 -1.49018419e+00 -4.82194647e-02
-5.13583541e-01 2.69125193e-01 3.55034947e-01 9.84027565e-01
5.50033152e-01 7.05559731e-01 5.32507956e-01 -4.91044521e-01
5.34735024e-02 -8.44091773e-01 -9.01205182e-01 6.66946352e-01
2.26919591e-01 -1.02482772e+00 -5.82302451e-01 2.26166889e-01] | [9.698308944702148, 3.956148624420166] |
76c98f83-9399-462f-8316-9551fccb03cd | ldc-lightweight-dense-cnn-for-edge-detection | null | null | https://ieeexplore.ieee.org/document/9807316 | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=9807316 | LDC: Lightweight Dense CNN for Edge Detection | This paper presents a Lightweight Dense Convolutional (LDC) neural network for edge detection. The proposed model is an adaptation of two state-of-the-art approaches, but it requires less than 4% of parameters in comparison with these approaches. The proposed architecture generates thin edge maps and reaches the highest score (i.e., ODS) when compared with lightweight models (models with less than 1 million parameters), and reaches a similar performance when compare with heavy architectures (models with about 35 million parameters). Both quantitative and qualitative results and comparisons with state-of-the-art models, using different edge detection datasets, are provided. The proposed LDC does not use pre-trained weights and requires straightforward hyper-parameter settings. The source code is released at https://github.com/xavysp/LDC. | ['Angel Domingo Sappa', 'Gonzalo Pomboza-Junez', 'Xavier Soria Poma'] | 2022-06-27 | null | null | null | ieee-access-2022-6 | ['edge-detection'] | ['computer-vision'] | [-3.24640095e-01 2.13702053e-01 6.76464438e-02 2.46592332e-02
-3.72784704e-01 -1.33908048e-01 5.96008718e-01 1.28835171e-01
-8.41707885e-01 3.23563337e-01 8.75133649e-02 -5.00397801e-01
3.07258397e-01 -7.93380380e-01 -8.39384079e-01 -3.34363550e-01
-3.33607107e-01 2.34026790e-01 5.52589595e-01 -1.89878643e-01
9.68833342e-02 3.34431469e-01 -1.40125275e+00 1.61032215e-01
7.72893429e-01 1.12246549e+00 1.44559249e-01 6.94451928e-01
-2.16204941e-01 4.41355556e-01 -3.38197827e-01 -5.00441253e-01
1.58462882e-01 -1.82513341e-01 -5.73385417e-01 -3.47225338e-01
6.31955028e-01 -3.23667735e-01 -2.42681026e-01 9.53749180e-01
7.04631329e-01 -1.77397057e-01 6.79861724e-01 -1.04645324e+00
-9.08174276e-01 4.93351012e-01 -6.77384079e-01 3.60683471e-01
-1.21046245e-01 -8.00019726e-02 7.95228183e-01 -1.20433211e+00
6.32215917e-01 9.08558130e-01 1.20050240e+00 5.79624116e-01
-9.10818696e-01 -6.24354303e-01 2.08856210e-01 2.68421173e-01
-1.52695894e+00 -3.55968952e-01 4.82572436e-01 -3.34076852e-01
1.21412110e+00 -2.42470801e-01 5.79428554e-01 1.14778388e+00
7.50627518e-02 8.25584173e-01 6.37057364e-01 -5.81496298e-01
2.76805192e-01 9.14754197e-02 2.38177732e-01 1.09320116e+00
7.99544513e-01 -7.67416060e-02 -2.86869019e-01 3.26319411e-02
6.48976266e-01 -1.55558780e-01 -2.56784201e-01 -2.22125396e-01
-7.95532584e-01 6.87996864e-01 6.34186685e-01 4.23152834e-01
-6.87039316e-01 4.81766880e-01 4.87547606e-01 -1.55825049e-01
5.44174850e-01 -2.64638867e-02 -3.21093887e-01 -1.70370832e-01
-1.01915956e+00 -8.79862253e-03 6.70941591e-01 1.24446929e+00
6.55843675e-01 3.11233699e-01 -8.70528966e-02 5.99642456e-01
2.28202134e-01 2.51965314e-01 3.44383210e-01 -6.97646201e-01
3.31582159e-01 7.01534152e-01 8.38751644e-02 -8.66175175e-01
-6.81934237e-01 -6.75809860e-01 -1.03202236e+00 4.64822084e-01
3.05553138e-01 -5.37322938e-01 -1.07133126e+00 1.45162439e+00
3.41697112e-02 3.74771029e-01 -2.65275855e-02 7.95925975e-01
1.25429940e+00 6.91252112e-01 1.21076163e-02 4.52561915e-01
1.39170623e+00 -1.20119119e+00 -5.28824866e-01 -4.73775327e-01
6.91442192e-01 -8.69531631e-01 1.02594292e+00 2.01304123e-01
-1.30984640e+00 -4.99196887e-01 -1.17514944e+00 -1.41865656e-01
-5.92584133e-01 5.32437742e-01 7.52913356e-01 8.54507685e-01
-1.62391555e+00 4.78442967e-01 -8.11420679e-01 -6.07997000e-01
4.68869001e-01 2.67530292e-01 -2.42166236e-01 1.94439992e-01
-9.51404095e-01 7.50548244e-01 3.96408528e-01 1.54766470e-01
-7.81621814e-01 -6.33676469e-01 -7.71557689e-01 3.92464727e-01
2.13710412e-01 -6.99325323e-01 1.22226799e+00 -6.03810132e-01
-1.37177765e+00 7.66920447e-01 1.52999917e-02 -6.02076292e-01
6.08467340e-01 -6.57933712e-01 -3.63426566e-01 1.97140500e-01
-2.54381716e-01 1.08304811e+00 5.49980283e-01 -1.14699841e+00
-7.10057318e-01 2.15567797e-01 2.82428205e-01 -1.46785781e-01
-5.19434750e-01 1.41601339e-01 -1.02204120e+00 -6.14735425e-01
-2.56307989e-01 -9.89441454e-01 -1.17677711e-01 1.70120075e-01
-2.53891349e-01 -1.28912985e-01 7.55286217e-01 -3.17632943e-01
1.38935137e+00 -1.94121504e+00 -2.76394725e-01 -4.42694314e-02
2.12101340e-01 6.47209108e-01 -2.34380245e-01 4.23347831e-01
-5.40172532e-02 3.59665900e-01 -1.29896253e-01 -7.38482237e-01
-9.22643859e-03 -2.48702392e-01 4.25199896e-01 3.83211344e-01
9.75043327e-02 8.48761618e-01 -5.70219398e-01 -5.05774915e-01
2.75136828e-01 7.13395476e-01 -5.73665679e-01 -1.83763042e-01
2.07561135e-01 -3.06583405e-01 -1.25573412e-01 7.04218626e-01
9.46755886e-01 -5.66609025e-01 4.27508764e-02 -4.04936880e-01
-2.23337442e-01 -6.27339929e-02 -1.28639221e+00 1.65265739e+00
-5.32784641e-01 8.50202024e-01 -3.22837345e-02 -7.54165709e-01
9.67851818e-01 4.11674052e-01 1.62021220e-01 -6.14141822e-01
4.61897135e-01 3.08395177e-01 8.80202428e-02 -2.50007153e-01
4.46719825e-01 4.95723754e-01 1.08693242e-01 3.51079106e-01
3.23276609e-01 6.04829848e-01 4.73864317e-01 4.65106934e-01
1.14226806e+00 -8.91123805e-03 1.81346744e-01 -4.75470096e-01
4.22509283e-01 -3.23274136e-01 7.47674704e-01 7.41636097e-01
-2.38045529e-01 6.48753107e-01 4.79103625e-01 -7.61790991e-01
-1.16663718e+00 -8.06953669e-01 -6.95878491e-02 6.61840796e-01
1.79569960e-01 -6.04990602e-01 -1.25831437e+00 -5.00677466e-01
-9.56153423e-02 4.81781095e-01 -6.97338760e-01 9.73493210e-04
-4.78319496e-01 -7.94550121e-01 7.79092371e-01 9.99520302e-01
9.60819125e-01 -1.16853023e+00 -7.80656993e-01 1.24255478e-01
-2.32168313e-04 -1.25433910e+00 -3.32867295e-01 -6.06098175e-02
-8.41492653e-01 -1.01966870e+00 -1.04850495e+00 -1.06094885e+00
7.33025014e-01 2.49975622e-01 1.43725073e+00 3.23613763e-01
-4.32896942e-01 2.96250544e-02 -5.80900073e-01 -4.76044565e-01
1.05798384e-02 3.22865069e-01 -1.20047994e-01 -2.10588604e-01
5.97618282e-01 -4.40120220e-01 -1.09558690e+00 1.57639787e-01
-6.99004948e-01 3.02062392e-01 9.08105850e-01 6.76150620e-01
3.85046244e-01 -2.65556306e-01 6.19405389e-01 -1.04799390e+00
5.55828094e-01 -2.12947980e-01 -7.09521472e-01 4.00988311e-02
-8.77857625e-01 4.14896160e-02 5.16934633e-01 -3.19796443e-01
-1.03830731e+00 5.89495786e-02 -1.83570266e-01 -3.70911002e-01
-3.37476671e-01 2.96343416e-01 2.46161580e-01 -1.73725843e-01
6.92385375e-01 -2.77590513e-01 -2.52839088e-01 -7.44019091e-01
2.03216687e-01 6.13588631e-01 3.90395582e-01 -2.25502536e-01
4.57712114e-01 5.15414417e-01 -2.31392041e-01 -6.81546807e-01
-4.22467798e-01 -5.61815083e-01 -5.53356051e-01 -2.73361087e-01
8.44411016e-01 -9.71814930e-01 -5.34987152e-01 9.08655167e-01
-1.18309724e+00 -6.70743644e-01 1.55563742e-01 5.29375553e-01
-2.35660598e-01 1.79961786e-01 -9.81793284e-01 -6.27610505e-01
-7.62343466e-01 -7.91891336e-01 9.04736161e-01 4.79288608e-01
-5.21974228e-02 -1.02242446e+00 -1.93812713e-01 -1.11160234e-01
6.90030456e-01 4.30268049e-01 4.47396189e-01 -2.67053008e-01
-4.57224607e-01 -5.09009659e-01 -4.88613784e-01 1.59454256e-01
-5.81043772e-02 2.22934652e-02 -9.12785351e-01 -4.82622594e-01
-6.84178889e-01 -9.98440012e-02 1.13317466e+00 6.48606777e-01
1.13136590e+00 -8.53864923e-02 -4.24681038e-01 8.32872391e-01
1.70597720e+00 9.11215395e-02 1.01592338e+00 6.43618166e-01
5.02945721e-01 1.12891100e-01 2.05558464e-01 4.66941923e-01
3.85849088e-01 3.95541310e-01 5.49961865e-01 -7.44546294e-01
-5.09418428e-01 -1.23505136e-02 8.21677372e-02 6.50612950e-01
-4.64658439e-01 -5.26740789e-01 -1.09047520e+00 7.43720114e-01
-1.95102310e+00 -8.05163324e-01 -4.39297199e-01 1.79690933e+00
2.87875682e-01 6.21764481e-01 1.45886481e-01 5.95776029e-02
8.98990154e-01 1.16933525e-01 -4.83014256e-01 -6.04415298e-01
-1.20957242e-02 3.08051050e-01 8.51420581e-01 3.42774957e-01
-1.28117776e+00 9.68491197e-01 6.63546801e+00 5.43323457e-01
-9.67756450e-01 2.02961057e-01 6.33800566e-01 -1.54757723e-01
-2.52949316e-02 -1.34497985e-01 -9.13284302e-01 3.89235973e-01
9.71212029e-01 9.98235308e-03 3.89080085e-02 9.73666668e-01
8.85697752e-02 -1.40817329e-01 -7.10038126e-01 9.23179448e-01
1.53354824e-01 -1.63788021e+00 -1.45167798e-01 5.60932532e-02
8.74049306e-01 5.14865458e-01 -7.26709515e-02 3.15356463e-01
4.05383527e-01 -7.07782090e-01 8.18582833e-01 5.17291129e-01
8.18292916e-01 -7.84372985e-01 1.25730038e+00 3.90405692e-02
-1.51628363e+00 -2.21537016e-02 -5.22531331e-01 -4.44418453e-02
3.43073696e-01 6.77892923e-01 -2.09591299e-01 2.95963675e-01
1.28840423e+00 4.92907643e-01 -8.68902802e-01 1.33526027e+00
-1.11256607e-01 6.93270743e-01 -3.36537212e-01 -7.34368563e-02
5.22251964e-01 -1.75052080e-02 9.59077924e-02 1.68001473e+00
5.54758966e-01 -2.25366041e-01 -5.36116585e-02 7.03384697e-01
-3.74027640e-01 1.18398964e-01 -3.98357362e-01 2.42172524e-01
5.17522693e-01 1.57454884e+00 -1.02211881e+00 -5.58111727e-01
-9.20187593e-01 9.33179319e-01 5.52031159e-01 4.38855678e-01
-1.05777192e+00 -8.28839242e-01 3.34631532e-01 2.19244897e-01
5.11792243e-01 6.62290826e-02 -2.00226530e-01 -8.00673544e-01
1.12204693e-01 -4.85263377e-01 3.98085684e-01 -6.87571168e-01
-1.00702393e+00 8.95382583e-01 -2.55783468e-01 -1.18397605e+00
-4.10760865e-02 -9.11450267e-01 -1.10502839e+00 8.63536894e-01
-1.73407233e+00 -1.11221027e+00 -7.91904211e-01 3.32917720e-01
3.06865901e-01 -9.70600471e-02 8.97308111e-01 5.88667512e-01
-6.96585178e-01 9.41659153e-01 1.03523061e-01 4.43054289e-01
6.19512975e-01 -1.29402041e+00 1.17101598e+00 1.01045406e+00
-2.17046037e-01 4.39951479e-01 4.54619706e-01 -4.61342722e-01
-1.00419509e+00 -1.11716795e+00 9.03737009e-01 2.47407168e-01
4.96466935e-01 -3.97717983e-01 -8.32730591e-01 5.51176965e-01
6.12220347e-01 3.20059240e-01 5.93248546e-01 2.01612577e-01
-2.06288129e-01 -7.45659769e-02 -1.00675082e+00 6.06595516e-01
1.25072432e+00 -1.43904164e-01 3.43472115e-04 -9.94972810e-02
4.30932850e-01 -2.03667909e-01 -6.36793911e-01 3.87622088e-01
5.85808218e-01 -1.25155914e+00 8.80021214e-01 -1.84388936e-01
1.58095002e-01 -3.04519773e-01 2.23280191e-01 -1.13627923e+00
-8.41879845e-01 -4.40324247e-01 -3.35214645e-01 1.09432232e+00
8.22022021e-01 -6.88966751e-01 9.99497592e-01 2.37559438e-01
-5.85384965e-01 -9.77247059e-01 -4.89355892e-01 -1.03461397e+00
6.09384663e-02 -2.16136754e-01 3.66188586e-01 6.91484094e-01
-2.57491976e-01 1.20277494e-01 -2.60209173e-01 2.92295575e-01
5.75260460e-01 -1.30400673e-01 5.98991156e-01 -1.33734882e+00
-5.91384396e-02 -6.00833952e-01 -6.69463217e-01 -6.88800037e-01
-1.84759513e-01 -8.01050842e-01 -1.84893206e-01 -2.07793260e+00
4.35869209e-02 -3.61598879e-01 -3.60269368e-01 7.90398777e-01
-8.67574513e-02 7.61693120e-01 1.71177596e-01 -6.63458481e-02
-5.90067267e-01 3.69024128e-01 7.90055394e-01 -6.50090948e-02
-9.84934866e-02 -2.81388700e-01 -6.51538014e-01 1.05003548e+00
1.36134470e+00 -3.83623451e-01 -1.74015030e-01 -6.52118921e-01
2.40384117e-01 -5.90737939e-01 3.79903734e-01 -1.50223589e+00
3.61875504e-01 4.55299020e-01 5.55661678e-01 -5.75581551e-01
1.37555718e-01 -6.59418523e-01 9.32640284e-02 7.61856973e-01
-1.38240149e-02 4.74740565e-01 5.78149736e-01 2.45141700e-01
-5.26822917e-02 -4.71215874e-01 8.54029417e-01 -5.46955355e-02
-1.03100574e+00 2.85796911e-01 -4.31420773e-01 -1.32906213e-02
1.02205741e+00 -3.72891128e-01 -4.87994760e-01 -3.11330080e-01
-6.13886118e-01 2.06608474e-01 4.83578831e-01 3.87592793e-01
5.95636070e-01 -1.38535726e+00 -7.96619117e-01 -2.05692761e-02
2.21229404e-01 -6.96642026e-02 3.09729397e-01 6.98569536e-01
-1.07140279e+00 3.91513795e-01 -1.43298045e-01 -3.29315484e-01
-1.13093185e+00 4.77250516e-01 4.03593093e-01 -1.97032779e-01
-1.15277243e+00 7.23387361e-01 1.76549569e-01 -1.60502970e-01
4.38118696e-01 -2.47659579e-01 -1.00080252e-01 -1.72681078e-01
6.34898126e-01 6.25551224e-01 3.62088948e-01 -4.08036023e-01
-4.69564050e-01 6.99642539e-01 -1.90646246e-01 1.13716796e-01
1.26724470e+00 5.41844405e-02 4.11776930e-01 -3.55752371e-02
1.01235414e+00 -3.02649826e-01 -1.29955351e+00 -6.23711571e-02
-3.00235171e-02 -1.99553296e-01 4.53339547e-01 -8.67891848e-01
-1.35398316e+00 8.55422735e-01 1.10245717e+00 -6.90739043e-03
1.17479658e+00 -5.39408401e-02 8.64256740e-01 3.51699561e-01
2.29633510e-01 -1.29155815e+00 1.02938704e-01 5.17104268e-01
5.85181296e-01 -1.21860814e+00 -1.58408031e-01 -3.48738194e-01
-5.82909048e-01 9.67575312e-01 8.32480788e-01 -4.06373918e-01
8.46525073e-01 6.78278983e-01 1.81826845e-01 -2.63000846e-01
-6.76888764e-01 -5.47478974e-01 1.99404329e-01 5.10179698e-01
7.16343582e-01 -2.38705073e-02 -4.44115043e-01 2.43396938e-01
1.61608472e-01 9.05764624e-02 5.01303852e-01 8.33895802e-01
-4.59066808e-01 -9.00652647e-01 1.01537660e-01 3.33155662e-01
-7.02887774e-01 -4.16745067e-01 -3.26802045e-01 1.06449068e+00
1.33861035e-01 7.53026247e-01 4.17977422e-01 -1.99755475e-01
3.44240159e-01 5.49241863e-02 4.06715721e-02 -2.29537457e-01
-5.71864367e-01 2.94725206e-02 1.93042338e-01 -6.75380588e-01
-3.89918834e-01 -2.91934669e-01 -1.32353699e+00 -6.15837216e-01
-5.43901980e-01 -7.44972751e-02 7.26666987e-01 2.54445523e-01
8.08777750e-01 7.35988736e-01 1.56167626e-01 -1.03239691e+00
2.16121405e-01 -1.21143329e+00 -3.80588323e-01 2.87346750e-01
-4.68161702e-03 -5.95084786e-01 -3.78774345e-01 1.90512333e-02] | [9.182480812072754, 0.9318009614944458] |
5d12bbb8-c420-49a4-b3c7-a4ad411abe88 | deep-cnn-denoiser-and-multi-layer-neighbor | 1806.10726 | null | http://arxiv.org/abs/1806.10726v1 | http://arxiv.org/pdf/1806.10726v1.pdf | Deep CNN Denoiser and Multi-layer Neighbor Component Embedding for Face Hallucination | Most of the current face hallucination methods, whether they are shallow
learning-based or deep learning-based, all try to learn a relationship model
between Low-Resolution (LR) and High-Resolution (HR) spaces with the help of a
training set. They mainly focus on modeling image prior through either
model-based optimization or discriminative inference learning. However, when
the input LR face is tiny, the learned prior knowledge is no longer effective
and their performance will drop sharply. To solve this problem, in this paper
we propose a general face hallucination method that can integrate model-based
optimization and discriminative inference. In particular, to exploit the model
based prior, the Deep Convolutional Neural Networks (CNN) denoiser prior is
plugged into the super-resolution optimization model with the aid of
image-adaptive Laplacian regularization. Additionally, we further develop a
high-frequency details compensation method by dividing the face image to facial
components and performing face hallucination in a multi-layer neighbor
embedding manner. Experiments demonstrate that the proposed method can achieve
promising super-resolution results for tiny input LR faces. | ['Yi Yu', 'Junjun Jiang', 'Jinhui Hu', 'Jiayi Ma', 'Suhua Tang'] | 2018-06-28 | null | null | null | null | ['face-hallucination'] | ['computer-vision'] | [ 1.41287774e-01 2.05087826e-01 -8.95116404e-02 -5.50558507e-01
-7.82493651e-01 1.29216433e-01 4.23128963e-01 -6.46610439e-01
4.76470925e-02 5.88820696e-01 5.39114714e-01 4.65655893e-01
1.01533450e-01 -8.70367110e-01 -7.85867751e-01 -7.21030295e-01
5.55874944e-01 -1.05432719e-02 -1.36585802e-01 -2.66796350e-01
-8.63595493e-03 4.41410720e-01 -1.32110214e+00 2.78781414e-01
8.86610389e-01 8.47076356e-01 2.61245131e-01 7.87576810e-02
-6.08672127e-02 6.80389822e-01 -1.53449640e-01 -2.73076832e-01
2.08835483e-01 -4.69591826e-01 -4.08346534e-01 1.44477665e-01
3.65644038e-01 -4.04417217e-01 -6.28570914e-01 1.17660296e+00
5.94352484e-01 1.11748263e-01 4.25891697e-01 -7.00434089e-01
-1.14501143e+00 2.25489557e-01 -9.33905840e-01 4.69278023e-02
2.97189564e-01 -7.97660798e-02 5.01081645e-01 -1.42582572e+00
5.04294515e-01 1.67104125e+00 7.10500062e-01 6.63470685e-01
-1.47930479e+00 -7.28787839e-01 7.27719441e-02 2.42728591e-01
-1.60067892e+00 -7.01379359e-01 1.20769763e+00 -2.74498343e-01
4.96190131e-01 -4.11839932e-02 2.95729250e-01 9.71518993e-01
-7.13188872e-02 4.20436621e-01 1.14273143e+00 -2.54072487e-01
2.25705840e-03 3.09497148e-01 -3.51453125e-01 8.13805938e-01
1.79224946e-02 2.45448612e-02 -5.25648892e-01 -3.16611454e-02
1.26228428e+00 2.80074388e-01 -5.69692552e-01 -1.67049140e-01
-7.04690516e-01 7.73002684e-01 6.85355186e-01 4.26361203e-01
-3.97746384e-01 -2.41768777e-01 -5.06464131e-02 -3.24272886e-02
7.07240343e-01 1.71467528e-01 9.05732345e-03 6.93933904e-01
-1.21334934e+00 -8.23699683e-02 4.41196710e-01 5.88059783e-01
1.06872809e+00 4.06547755e-01 -2.27233678e-01 1.23948109e+00
5.14671206e-01 1.07307278e-01 4.87721860e-01 -1.04418898e+00
4.82785478e-02 4.05106276e-01 3.72170880e-02 -1.18072510e+00
3.15269828e-02 -5.23324907e-01 -1.19737113e+00 2.96278119e-01
-9.54507887e-02 -2.54850034e-02 -8.46410155e-01 1.92416549e+00
3.69187236e-01 5.34334362e-01 1.43377438e-01 1.11894476e+00
9.57236171e-01 6.70859635e-01 -1.46835014e-01 -5.29401660e-01
1.16316855e+00 -1.04793453e+00 -9.75015283e-01 -7.20369741e-02
-2.16763109e-01 -7.17312157e-01 1.04031062e+00 1.53480187e-01
-1.26785612e+00 -9.28546965e-01 -1.06529856e+00 -4.34861362e-01
9.18230042e-03 2.05199718e-01 1.83774933e-01 2.35133037e-01
-1.04351199e+00 5.48874974e-01 -7.37582028e-01 -1.81351408e-01
6.19571507e-01 2.18938947e-01 -4.86144185e-01 -4.19380128e-01
-1.03707314e+00 8.26001287e-01 -8.06380808e-02 3.32364917e-01
-1.09056139e+00 -5.39520025e-01 -9.83222485e-01 1.75721183e-01
1.68487087e-01 -7.18174815e-01 6.41907692e-01 -9.97683346e-01
-1.63273156e+00 5.96148789e-01 -4.16619748e-01 1.07038550e-01
1.93112329e-01 -3.03293467e-01 -4.43118572e-01 1.69621423e-01
-2.37097219e-03 4.35003370e-01 1.28874445e+00 -1.65022683e+00
-1.16488174e-01 -6.22235835e-01 -1.00133330e-01 2.92077601e-01
-5.85827887e-01 -6.58326074e-02 -5.25769711e-01 -6.52202070e-01
2.36046568e-01 -3.49909931e-01 -1.60555273e-01 2.35103086e-01
-2.27854982e-01 8.82292911e-02 1.08405232e+00 -1.00919759e+00
1.06343853e+00 -2.29460812e+00 4.60241020e-01 -4.63839918e-02
3.60010892e-01 2.89737940e-01 -2.94868648e-01 6.53451234e-02
-1.56972587e-01 -2.74430551e-02 -1.90326408e-01 -6.34713769e-01
-3.26522261e-01 2.99293935e-01 -2.77855605e-01 4.54681247e-01
3.79846275e-01 8.40110302e-01 -5.88853180e-01 -4.76454914e-01
2.51086920e-01 1.34454346e+00 -6.96863174e-01 5.22662282e-01
1.21179923e-01 7.86383092e-01 -3.89152110e-01 5.32315433e-01
1.07912397e+00 -2.17237443e-01 1.24955237e-01 -7.21019983e-01
-1.79535002e-01 -2.34008789e-01 -9.53865588e-01 1.79942560e+00
-6.68517292e-01 2.55640805e-01 5.69380462e-01 -1.04812920e+00
1.05610991e+00 2.71456033e-01 3.13059509e-01 -5.15860677e-01
-1.75791755e-01 -5.38598523e-02 -4.72404808e-01 -5.26039064e-01
1.26006812e-01 -4.49225128e-01 5.95548987e-01 1.73382480e-02
5.53490594e-02 1.30946949e-01 -4.76375848e-01 -1.33665100e-01
6.72947168e-01 3.18369538e-01 5.61747029e-02 2.48731691e-02
7.39733994e-01 -6.86170518e-01 7.42915750e-01 1.03297018e-01
2.38732293e-01 1.00600350e+00 2.08738044e-01 -2.03000560e-01
-1.07022047e+00 -9.92707491e-01 -3.33029151e-01 8.50254297e-01
1.47238776e-01 -2.00914472e-01 -8.78935099e-01 -1.95503533e-01
-2.17840984e-01 2.83334494e-01 -5.81078708e-01 -3.53004307e-01
-6.34407699e-01 -6.44324601e-01 1.59813359e-01 4.03733552e-01
9.56856608e-01 -8.88159096e-01 -9.25874785e-02 6.81734681e-02
-2.71167338e-01 -1.03923011e+00 -5.91325879e-01 -3.44726562e-01
-7.99100280e-01 -7.49427438e-01 -9.87775981e-01 -9.69172180e-01
8.31892669e-01 3.41239989e-01 6.55705571e-01 -2.88100671e-02
-3.06986690e-01 2.86375165e-01 -4.26873304e-02 2.42266640e-01
-1.72708556e-03 -3.43725294e-01 1.35864858e-02 5.09605348e-01
1.32040203e-01 -9.80566859e-01 -6.73104525e-01 1.72316819e-01
-8.46857548e-01 6.53472841e-02 7.88982689e-01 1.06771874e+00
6.99266672e-01 1.79935575e-01 5.52051306e-01 -5.96566856e-01
4.95543659e-01 -5.41305304e-01 -3.10246736e-01 2.17935190e-01
-4.72799420e-01 1.46319732e-01 6.60297692e-01 -5.01471579e-01
-1.49860096e+00 3.79497185e-02 -2.81238407e-01 -1.00345635e+00
-1.84487260e-03 4.18027610e-01 -7.35797346e-01 -2.98402131e-01
4.49874043e-01 4.38471943e-01 2.14244857e-01 -8.85371208e-01
3.32776845e-01 5.18324554e-01 6.40368879e-01 -3.60165924e-01
9.65377808e-01 6.59624696e-01 -1.20722190e-01 -9.18070316e-01
-9.03477550e-01 -4.72459868e-02 -5.46547413e-01 -8.70242268e-02
1.12298000e+00 -1.16314924e+00 -4.32064444e-01 2.42625907e-01
-1.00309372e+00 -9.96659473e-02 -2.04330757e-01 4.81105775e-01
-3.62891227e-01 5.25948226e-01 -7.83017337e-01 -7.04632938e-01
-1.71246201e-01 -1.09892583e+00 9.74784613e-01 4.03538883e-01
4.13147539e-01 -8.12931180e-01 -3.05951294e-02 4.85554695e-01
7.19419420e-01 3.26679111e-01 7.79116333e-01 3.08692418e-02
-6.66634083e-01 6.03014193e-02 -5.92282236e-01 5.67581594e-01
3.85119319e-01 -3.21101367e-01 -1.05672634e+00 -3.96193415e-01
4.66152847e-01 -3.63678038e-01 1.02583706e+00 3.12435776e-01
1.37066400e+00 -6.32004619e-01 -1.67591840e-01 1.00351501e+00
1.58551347e+00 -2.88436890e-01 8.12873483e-01 -9.97663885e-02
9.28068221e-01 6.41275585e-01 3.38847041e-01 4.38821167e-01
4.30627108e-01 8.06970239e-01 4.45319265e-01 -4.07835573e-01
-3.53291839e-01 -4.96151716e-01 4.32001263e-01 6.49871469e-01
-4.40252066e-01 3.62507045e-01 -3.36984277e-01 2.54673809e-01
-1.80117786e+00 -1.07744086e+00 3.59480649e-01 2.11034822e+00
9.08364534e-01 -2.73434848e-01 -2.58331060e-01 -1.49785161e-01
8.30875158e-01 3.05669814e-01 -6.61934137e-01 3.94777060e-02
-1.39411405e-01 2.00158283e-01 -1.09062985e-01 6.51385665e-01
-7.60263205e-01 8.83852482e-01 5.56904411e+00 8.74647141e-01
-1.24587905e+00 4.18538421e-01 6.95664763e-01 6.32931367e-02
-5.12457550e-01 -1.59295633e-01 -7.04569757e-01 2.44054571e-01
5.94896793e-01 1.07490398e-01 6.41411304e-01 7.92012095e-01
3.51655602e-01 3.07319373e-01 -8.71091366e-01 1.32411289e+00
2.84621507e-01 -1.10555661e+00 2.36905456e-01 5.96370101e-02
6.17909968e-01 -4.98536706e-01 3.84183407e-01 3.40277374e-01
-2.16048494e-01 -1.46573341e+00 3.33819389e-01 9.67568219e-01
9.86663043e-01 -7.66260087e-01 4.85699236e-01 2.64676303e-01
-1.22603643e+00 -1.07766934e-01 -8.04723978e-01 7.24651366e-02
1.77834764e-01 5.83408535e-01 -1.99628323e-01 5.47824264e-01
7.69100368e-01 9.34182882e-01 -3.25136870e-01 5.50328672e-01
-1.62901640e-01 3.19621503e-01 -3.00369877e-02 7.88023055e-01
-2.78644592e-01 -3.83290976e-01 5.30000150e-01 9.06584024e-01
3.39680076e-01 5.11954248e-01 1.56176776e-01 1.32923114e+00
-2.48440921e-01 -1.53642902e-02 -5.02848446e-01 2.15660095e-01
1.82404906e-01 1.49172604e+00 -1.36572525e-01 -1.09007873e-01
-6.62797868e-01 1.15726388e+00 5.27230203e-01 7.08024740e-01
-7.41911888e-01 -7.34126866e-02 6.75913513e-01 4.50747162e-01
3.50018620e-01 -2.28865877e-01 -8.08215439e-02 -1.43195879e+00
-1.60373794e-03 -7.39531636e-01 -1.43701449e-01 -8.23520184e-01
-1.24037993e+00 6.28443956e-01 -1.26179472e-01 -9.21613634e-01
-6.78117648e-02 -2.37356246e-01 -6.89498365e-01 1.12374210e+00
-1.88504767e+00 -1.37733316e+00 -5.76141119e-01 9.70216215e-01
5.72109163e-01 -1.50941238e-01 7.20539987e-01 5.36284268e-01
-7.69444466e-01 6.35173321e-01 -2.68098954e-02 1.09678216e-01
7.87592828e-01 -7.96220720e-01 -3.33389729e-01 6.59111679e-01
-2.13946685e-01 7.18089104e-01 5.84989369e-01 -4.32872385e-01
-1.31378019e+00 -1.28744805e+00 4.79622275e-01 2.90460438e-02
1.49505273e-01 -2.80781806e-01 -1.40457690e+00 6.22672975e-01
1.61266536e-01 4.19816911e-01 5.06665289e-01 -1.21924743e-01
-4.91197437e-01 -3.33583593e-01 -1.37085557e+00 4.85710174e-01
9.86248910e-01 -7.79424310e-01 -5.27208924e-01 4.76928055e-02
6.31460607e-01 -1.91639170e-01 -1.16171908e+00 6.31444335e-01
4.48976219e-01 -1.00898778e+00 1.23929584e+00 -3.36028934e-01
5.68746150e-01 -3.43061119e-01 -2.47362405e-01 -1.06662405e+00
-5.42084634e-01 -3.89132589e-01 -2.92805374e-01 1.41505265e+00
-4.88775130e-03 -4.88850623e-01 5.72678089e-01 4.09846872e-01
-4.28185612e-02 -7.96163857e-01 -6.72291458e-01 -3.16542566e-01
-5.65570071e-02 2.51090407e-01 4.77136523e-01 1.09747505e+00
-3.17071557e-01 5.54446995e-01 -7.13380575e-01 3.56175154e-01
9.72513139e-01 5.90002872e-02 5.84316611e-01 -9.56009209e-01
-3.21628690e-01 -1.58136889e-01 -1.08573876e-01 -1.02842760e+00
3.31173748e-01 -7.54178226e-01 -6.43516928e-02 -1.49466336e+00
3.68739218e-01 -3.34947556e-02 -2.57933080e-01 4.13621336e-01
-1.59449622e-01 4.85756129e-01 -1.20465308e-01 3.83943886e-01
-2.37913564e-01 9.86044645e-01 1.46853304e+00 6.74561411e-02
-1.65536240e-01 -2.83005774e-01 -8.08441043e-01 8.07075202e-01
6.01333678e-01 -6.43661544e-02 -4.34095472e-01 -4.18353081e-01
-2.48388678e-01 3.66020471e-01 6.24005377e-01 -8.32627535e-01
2.69923121e-01 -8.98726583e-02 8.27933609e-01 -1.05951622e-01
6.96437001e-01 -6.24584377e-01 1.88387096e-01 1.62904367e-01
-4.11698580e-01 -4.74905312e-01 -8.83451011e-03 6.67391062e-01
-4.93795067e-01 1.83543056e-01 1.31863892e+00 -2.05115393e-01
-4.47430998e-01 5.91246307e-01 1.08808018e-01 -1.66596934e-01
6.42071366e-01 -2.33697310e-01 -5.45287430e-02 -4.70623642e-01
-1.04736412e+00 -2.13666186e-02 4.38672304e-01 3.02163959e-01
1.08909929e+00 -1.45173573e+00 -8.77754688e-01 5.11936247e-01
-1.95515200e-01 1.18877497e-02 4.94094402e-01 7.94723213e-01
1.94127765e-02 2.32697707e-02 -4.38097030e-01 -3.53437543e-01
-1.00071526e+00 6.63101494e-01 4.48095500e-01 -8.26934054e-02
-6.96414769e-01 7.97537267e-01 5.91051161e-01 -2.67232448e-01
1.56456485e-01 2.84416676e-01 -5.30885935e-01 -2.37172574e-01
7.32147872e-01 2.14741275e-01 -4.58883464e-01 -9.11184967e-01
-9.27036479e-02 9.60125148e-01 -6.03670403e-02 -3.58921513e-02
1.49838686e+00 -3.21572363e-01 -4.08309579e-01 1.79258183e-01
1.31249225e+00 1.01535451e-02 -1.40832531e+00 -4.98106956e-01
-4.60693419e-01 -6.78340018e-01 3.52932036e-01 -3.20322931e-01
-1.35709548e+00 1.04458237e+00 7.74906218e-01 -2.62721390e-01
1.40389633e+00 2.27484293e-02 6.75316095e-01 1.45674963e-02
2.26653561e-01 -7.84500897e-01 5.54884136e-01 2.12817520e-01
1.25326991e+00 -1.41607046e+00 -3.90028395e-02 -5.45693576e-01
-3.10505599e-01 1.02709293e+00 8.54662001e-01 -3.57304990e-01
8.94069076e-01 1.42096162e-01 -2.52335787e-01 -3.57827581e-02
-5.23272157e-01 -1.35031939e-01 2.49503449e-01 5.63120067e-01
4.43906605e-01 -2.95806140e-01 -1.80674836e-01 7.88688719e-01
7.81325623e-02 1.05746120e-01 2.80221760e-01 3.39020193e-01
-6.07658625e-01 -9.31433558e-01 -3.92126322e-01 1.82785958e-01
-4.60192382e-01 -1.11322984e-01 4.98540178e-02 3.92992258e-01
2.43724138e-01 8.34053516e-01 -7.26809874e-02 -2.94176847e-01
2.40594238e-01 -8.52432624e-02 5.42436957e-01 -7.15971768e-01
2.53904536e-02 4.81572032e-01 -3.91709864e-01 -6.97397828e-01
-5.64003348e-01 -2.28641406e-01 -1.08099103e+00 -2.91417897e-01
-2.42124572e-01 3.33264433e-02 3.64200234e-01 7.58046687e-01
3.20145994e-01 4.24668700e-01 7.30386853e-01 -1.14618421e+00
-4.52725708e-01 -9.71312344e-01 -7.30005503e-01 1.71403974e-01
5.59021950e-01 -6.40087008e-01 -4.34523582e-01 7.34039545e-02] | [12.838478088378906, -0.02319193072617054] |
4489e24e-abb4-4a31-8791-b0dd84b44474 | humans-in-4d-reconstructing-and-tracking | 2305.20091 | null | https://arxiv.org/abs/2305.20091v2 | https://arxiv.org/pdf/2305.20091v2.pdf | Humans in 4D: Reconstructing and Tracking Humans with Transformers | We present an approach to reconstruct humans and track them over time. At the core of our approach, we propose a fully "transformerized" version of a network for human mesh recovery. This network, HMR 2.0, advances the state of the art and shows the capability to analyze unusual poses that have in the past been difficult to reconstruct from single images. To analyze video, we use 3D reconstructions from HMR 2.0 as input to a tracking system that operates in 3D. This enables us to deal with multiple people and maintain identities through occlusion events. Our complete approach, 4DHumans, achieves state-of-the-art results for tracking people from monocular video. Furthermore, we demonstrate the effectiveness of HMR 2.0 on the downstream task of action recognition, achieving significant improvements over previous pose-based action recognition approaches. Our code and models are available on the project website: https://shubham-goel.github.io/4dhumans/. | ['Jitendra Malik', 'Angjoo Kanazawa', 'Jathushan Rajasegaran', 'Georgios Pavlakos', 'Shubham Goel'] | 2023-05-31 | null | null | null | null | ['pose-tracking', 'action-recognition-in-videos', 'human-mesh-recovery'] | ['computer-vision', 'computer-vision', 'computer-vision'] | [-1.50693715e-01 -8.71597379e-02 -1.27801448e-01 3.22865532e-03
-5.60889304e-01 -4.39315826e-01 4.79848772e-01 -6.29177272e-01
-2.94781625e-01 5.13438642e-01 4.87417966e-01 2.09990159e-01
2.72794545e-01 -3.37561429e-01 -7.46188521e-01 -2.52609462e-01
-9.74114910e-02 8.85894001e-01 3.86793941e-01 -2.04023004e-01
-1.84241995e-01 7.96250761e-01 -1.81604457e+00 4.43581492e-01
-1.53134778e-01 7.19423592e-01 -4.46804076e-01 8.41347277e-01
5.26480794e-01 9.36256468e-01 -4.46510494e-01 -4.94906634e-01
6.35204256e-01 -9.80362296e-02 -8.88523221e-01 1.11213021e-01
1.09182501e+00 -6.22114718e-01 -9.15927768e-01 6.75339520e-01
7.09687471e-01 2.19628215e-01 1.52041748e-01 -1.43030882e+00
-2.22957268e-01 2.84759730e-01 -4.96364862e-01 1.16850719e-01
1.12514472e+00 4.15514141e-01 4.80952501e-01 -9.11966860e-01
1.12681246e+00 1.58669209e+00 1.15248811e+00 1.03687811e+00
-1.14845824e+00 -7.26955295e-01 4.92907055e-02 1.19774073e-01
-1.60817182e+00 -1.04187644e+00 3.10010850e-01 -5.01064420e-01
1.03226852e+00 1.52142718e-01 1.18109035e+00 1.57245898e+00
-3.61925401e-02 1.19090044e+00 6.91157460e-01 -2.27601707e-01
-2.82748401e-01 -4.36473161e-01 -1.95164800e-01 9.02312040e-01
9.39012915e-02 2.87342131e-01 -1.01232791e+00 -6.09130822e-02
1.06184804e+00 3.63237523e-02 -3.51441383e-01 -6.77817583e-01
-1.40821803e+00 3.14182818e-01 1.35604098e-01 -7.66104534e-02
-2.69319087e-01 4.94428754e-01 3.37583393e-01 2.22868443e-01
4.57302064e-01 8.12331289e-02 -2.68348068e-01 -3.29844654e-01
-1.03083336e+00 6.38518870e-01 6.83967710e-01 1.06085420e+00
1.38301373e-01 -7.98859969e-02 -1.50262713e-01 5.04710376e-01
2.38553360e-01 5.60744464e-01 4.48396094e-02 -1.54512143e+00
4.28531058e-02 4.33464319e-01 2.79129833e-01 -8.56256068e-01
-4.52112019e-01 2.15011984e-02 -4.49460030e-01 4.75445956e-01
7.20559299e-01 1.35151274e-03 -9.41874743e-01 1.51292658e+00
7.17183471e-01 2.62923688e-01 -1.33040980e-01 1.01165199e+00
8.92653346e-01 5.27365804e-02 -8.37921277e-02 2.95137376e-01
1.29000187e+00 -1.04315984e+00 -4.68094081e-01 -1.31292373e-01
4.32894468e-01 -8.26023102e-01 5.52495301e-01 5.93389988e-01
-1.33434796e+00 -6.39055192e-01 -7.07194865e-01 -9.27113965e-02
-1.60464615e-01 2.48146087e-01 5.30585647e-01 5.65013528e-01
-1.28491473e+00 9.57012475e-01 -1.22795486e+00 -7.77974427e-01
5.95060885e-01 4.15966243e-01 -8.15322638e-01 -4.02359925e-02
-8.59490037e-01 8.99301946e-01 1.95084736e-01 4.55936193e-02
-9.95407879e-01 -5.21750391e-01 -7.79510081e-01 -5.97868443e-01
6.30746961e-01 -9.66210127e-01 1.57923532e+00 -8.17247629e-01
-1.25102329e+00 1.24541247e+00 -1.23598345e-01 -5.25915146e-01
1.16503835e+00 -7.91227102e-01 -4.34194684e-01 4.51319218e-01
1.76947325e-01 8.80287647e-01 8.22217345e-01 -8.32793295e-01
-5.99962354e-01 -3.89207453e-01 2.39510294e-02 2.09032029e-01
1.91934794e-01 3.82472426e-01 -9.25908267e-01 -7.60577679e-01
-8.15051049e-02 -1.19694757e+00 -7.31487423e-02 6.55431449e-01
-3.70628178e-01 -6.53403178e-02 6.99472904e-01 -9.70800638e-01
7.14794815e-01 -1.96373582e+00 3.90773952e-01 -2.33640000e-02
3.66551816e-01 3.30418706e-01 1.60660613e-02 3.54703814e-01
-1.73952371e-01 -4.31146502e-01 2.06574276e-01 -8.53411794e-01
4.41994928e-02 -3.86600681e-02 7.69679546e-02 9.17709112e-01
-9.25200507e-02 9.10423398e-01 -7.24160135e-01 -6.02534950e-01
4.44715410e-01 6.81881547e-01 -4.58586335e-01 2.31164262e-01
-1.66319504e-01 7.92748392e-01 -8.49464834e-02 1.17296278e+00
3.32264781e-01 -3.12957078e-01 2.54894525e-01 -2.75733680e-01
-7.64219984e-02 -8.87098014e-02 -1.44924998e+00 2.09631252e+00
3.08851391e-01 6.65916622e-01 1.29773915e-01 -4.16632026e-01
4.02386606e-01 5.45238674e-01 9.64665174e-01 -3.49194527e-01
2.94588268e-01 -9.89099890e-02 -4.61478055e-01 -3.12159419e-01
5.35549283e-01 3.85350913e-01 8.27015787e-02 4.30792123e-01
-1.37162684e-02 5.11855662e-01 2.40531191e-01 2.44048342e-01
1.38474429e+00 9.88355815e-01 3.80639404e-01 1.43714368e-01
4.06815588e-01 2.35803902e-01 5.48724473e-01 6.74141169e-01
-4.57518578e-01 8.27819824e-01 -2.62366217e-02 -8.22708070e-01
-1.11023021e+00 -1.12366486e+00 4.81584817e-02 1.01897740e+00
1.25204874e-02 -6.56370759e-01 -6.74403310e-01 -4.85620886e-01
2.81022102e-01 2.69461274e-01 -7.40604460e-01 9.48200524e-02
-9.19865489e-01 -1.51906922e-01 1.00525713e+00 8.40843976e-01
5.44867933e-01 -1.11858737e+00 -9.68129039e-01 -2.83256192e-02
-3.08052838e-01 -1.17933726e+00 -6.11656249e-01 -5.10162234e-01
-5.24337232e-01 -1.48163962e+00 -8.73347104e-01 -3.86569887e-01
4.09857452e-01 1.92161813e-01 1.32113278e+00 2.67005920e-01
-7.32352376e-01 9.48346019e-01 -3.05178672e-01 -1.32699743e-01
-3.58014852e-01 -1.33372292e-01 5.06779075e-01 -1.16374455e-01
3.81407380e-01 -5.19694507e-01 -7.44748294e-01 5.21496296e-01
-3.58548433e-01 2.53983997e-02 1.57642707e-01 2.84391046e-01
5.60348392e-01 -2.61991918e-01 -1.58786431e-01 -5.65965295e-01
-2.18227431e-02 -5.06401211e-02 -6.19318247e-01 2.24628866e-01
-1.50069401e-01 -2.06448644e-01 7.25744218e-02 -5.79088092e-01
-9.24965501e-01 5.90549171e-01 -1.35160342e-01 -1.08756757e+00
-3.81612629e-01 -4.12781864e-01 -5.39717413e-02 -2.50712574e-01
6.34365082e-01 5.80288544e-02 2.99634576e-01 -6.59387589e-01
3.13509703e-01 4.47353005e-01 9.44421053e-01 -4.76952374e-01
7.59708941e-01 8.48731756e-01 -2.95601208e-02 -5.83136261e-01
-8.69504213e-01 -6.10104203e-01 -1.10616624e+00 -7.79462337e-01
9.92166638e-01 -1.24626541e+00 -1.14451039e+00 8.81292880e-01
-1.04607058e+00 -5.26436865e-01 -3.78900468e-01 2.58107185e-01
-7.74597704e-01 3.61890435e-01 -7.11564660e-01 -7.66005695e-01
-3.34145039e-01 -7.39235938e-01 1.37983298e+00 -5.44550046e-02
-5.82820117e-01 -6.78387225e-01 3.66578311e-01 6.59468353e-01
-2.67428402e-02 5.96429467e-01 -2.43329048e-01 -4.08142447e-01
-7.53728688e-01 -2.49872327e-01 1.91972032e-01 1.30770914e-02
-1.16263069e-01 -5.54423630e-02 -7.95187652e-01 -5.68364203e-01
-4.45450932e-01 -4.96165514e-01 7.16189206e-01 3.69039059e-01
7.50828326e-01 -3.79455909e-02 -6.48675799e-01 7.29763627e-01
7.93397844e-01 -1.46476850e-01 7.75059044e-01 3.62923354e-01
9.98936832e-01 5.02107263e-01 6.10411286e-01 5.34278512e-01
3.20922673e-01 1.08188975e+00 9.42309871e-02 5.09877838e-02
-5.69866717e-01 -3.73778492e-01 5.02217352e-01 2.53609151e-01
-7.26811469e-01 7.51668513e-02 -9.90739167e-01 3.07953715e-01
-2.12870836e+00 -1.39314425e+00 -1.32183030e-01 2.11344123e+00
5.90103388e-01 -7.35895038e-02 7.14585423e-01 -1.78178236e-01
8.85331154e-01 1.64020404e-01 -5.27269959e-01 4.50568527e-01
4.86329906e-02 1.81768686e-01 5.55999756e-01 3.02981168e-01
-1.34709823e+00 1.18094134e+00 6.78366089e+00 4.17502016e-01
-6.61062419e-01 1.46162659e-01 -1.42710194e-01 -6.16778135e-01
5.41896880e-01 -1.12697750e-01 -1.05869687e+00 1.52428031e-01
6.92826509e-01 1.16830781e-01 5.78873575e-01 7.90164173e-01
7.34722316e-02 -4.43865210e-02 -1.38917327e+00 1.17576504e+00
4.60468560e-01 -1.19317472e+00 -2.04260796e-01 6.13410138e-02
3.61299455e-01 -4.71544936e-02 -1.30896807e-01 7.02652112e-02
5.72870016e-01 -9.95157361e-01 1.02021992e+00 8.52176487e-01
9.31287110e-01 -6.44250989e-01 3.72511804e-01 2.07457915e-01
-1.41945612e+00 1.84264883e-01 -1.29933104e-01 2.00539771e-02
5.89641929e-01 -1.26041770e-02 -6.38810098e-01 4.59492326e-01
1.04095125e+00 1.24842978e+00 -7.98654199e-01 9.53171134e-01
-1.35180548e-01 -7.05679953e-02 -3.30167264e-01 4.64742243e-01
-3.97092253e-01 2.43813798e-01 7.74509072e-01 9.25877869e-01
1.36739820e-01 1.87843412e-01 5.77365577e-01 3.55396360e-01
-1.03346035e-01 -3.30056578e-01 -6.89595222e-01 1.24656111e-01
3.42256695e-01 9.29794967e-01 -6.71944916e-01 -3.87340933e-01
-1.88216940e-01 1.19214964e+00 3.50432485e-01 -1.29290760e-01
-9.80744720e-01 2.19881982e-01 7.55141616e-01 4.83821779e-01
3.86751860e-01 -1.89739808e-01 3.82698506e-01 -1.41806030e+00
2.25589387e-02 -1.36282420e+00 6.34071469e-01 -1.01650155e+00
-1.09536862e+00 4.16576296e-01 1.71068862e-01 -1.37126398e+00
-3.58510494e-01 -6.24107957e-01 -2.23971321e-03 3.66972476e-01
-6.30453169e-01 -1.47175443e+00 -4.62992251e-01 1.07408059e+00
5.71810067e-01 -2.23761424e-01 8.71147633e-01 6.15022659e-01
-6.02905691e-01 5.36834478e-01 -4.73929554e-01 4.13846970e-01
9.42580402e-01 -9.80613470e-01 8.83554101e-01 9.67073500e-01
2.89330602e-01 5.52771270e-01 7.15926111e-01 -1.14449537e+00
-1.42499363e+00 -1.03015542e+00 4.63988632e-01 -1.27777958e+00
4.27146554e-01 -4.03583765e-01 -4.96559471e-01 1.34853566e+00
4.72004013e-03 5.64584546e-02 5.12767494e-01 6.57576770e-02
-5.24218917e-01 2.10274279e-01 -1.03618717e+00 7.73028493e-01
1.80964005e+00 -4.12279069e-01 -6.87081635e-01 5.47203004e-01
3.60166281e-01 -9.55312490e-01 -9.42396104e-01 2.50570655e-01
1.13838995e+00 -1.07206213e+00 1.41845620e+00 -6.51291847e-01
9.78840366e-02 -4.82305110e-01 -1.38027474e-01 -7.09089637e-01
-3.41041237e-01 -8.22123408e-01 -6.20493352e-01 7.82119215e-01
-5.68857267e-02 -4.06720877e-01 1.13897967e+00 5.51899731e-01
1.61955521e-01 -3.12506527e-01 -8.74406517e-01 -8.00662100e-01
-4.16138291e-01 -3.34506154e-01 3.55944544e-01 6.30392909e-01
-3.61712217e-01 1.75603162e-02 -9.23034728e-01 2.23433748e-01
1.05908215e+00 -4.75187078e-02 1.28299797e+00 -1.29869676e+00
-4.98570204e-01 1.82497241e-02 -6.17539704e-01 -8.81217182e-01
2.33691528e-01 -7.18337357e-01 -4.88288999e-02 -1.34282124e+00
1.69041350e-01 1.76766720e-02 2.09202036e-01 8.04822564e-01
3.40091020e-01 6.94748044e-01 4.35310394e-01 6.44533396e-01
-9.80031550e-01 2.93703467e-01 9.70200896e-01 1.19079739e-01
2.11524617e-04 6.09094054e-02 -2.81182468e-01 9.58303034e-01
8.77460241e-01 -5.29612243e-01 7.60073513e-02 -4.23260480e-01
-1.26057550e-01 1.25421584e-01 9.43327546e-01 -1.54102325e+00
4.03664023e-01 2.18513593e-01 9.20288682e-01 -9.08406675e-01
8.60358059e-01 -7.22862959e-01 9.75632370e-01 6.27543688e-01
-1.16281740e-01 3.67513925e-01 8.68773684e-02 4.42694098e-01
3.87553096e-01 2.07503542e-01 7.54354477e-01 -5.12022018e-01
-8.65220428e-01 3.89407516e-01 -1.89655617e-01 1.13595843e-01
1.05729008e+00 -3.68057251e-01 -2.42028251e-01 -3.32994670e-01
-1.15438044e+00 1.95675641e-01 9.21994925e-01 5.89638829e-01
6.24443531e-01 -1.47485268e+00 -6.56111419e-01 3.18730921e-02
-8.98185838e-03 -3.61954987e-01 2.21720830e-01 8.65589321e-01
-8.48430991e-01 1.40746325e-01 -5.35059869e-01 -7.20053792e-01
-1.81510520e+00 3.98694485e-01 5.21229565e-01 -1.43142324e-02
-1.15032935e+00 6.34755671e-01 -3.57959569e-01 -4.49340373e-01
4.78238195e-01 3.74440312e-01 2.31265184e-02 -3.82376127e-02
8.59740973e-01 7.84515262e-01 -2.58952975e-01 -9.55747843e-01
-6.64485812e-01 6.73225880e-01 -2.15008836e-02 -2.66950548e-01
1.33702898e+00 -5.37385233e-03 9.11696255e-02 2.46063665e-01
7.53951073e-01 -2.35039964e-02 -1.54380596e+00 -1.15543045e-01
-3.07347864e-01 -7.09253967e-01 -3.45141143e-01 -6.58660591e-01
-1.15630734e+00 4.09519374e-01 6.43468261e-01 -3.46704304e-01
7.64782190e-01 1.23027816e-01 8.30218434e-01 4.13661003e-01
6.69414759e-01 -1.11804652e+00 2.01826304e-01 3.85139257e-01
7.75049388e-01 -9.87977028e-01 3.70748490e-01 -3.86579961e-01
-4.30960029e-01 9.86104667e-01 6.81426585e-01 -1.70112073e-01
3.59232306e-01 3.71946335e-01 1.36136681e-01 -4.74620700e-01
-4.63883877e-01 -3.24092299e-01 2.15128332e-01 7.72823155e-01
1.87469661e-01 -5.80289923e-02 3.36104751e-01 -3.30447108e-02
-3.18318486e-01 3.25426102e-01 2.89070696e-01 1.25324214e+00
-1.29577667e-01 -1.12755704e+00 -7.13582158e-01 9.83584486e-03
-5.26302159e-01 2.28563368e-01 -5.81227839e-01 1.04953027e+00
1.26717195e-01 6.45376205e-01 -2.48334184e-02 -4.00716394e-01
7.52695024e-01 1.61524251e-01 9.70404387e-01 -4.71713990e-01
-5.76530397e-01 7.34778717e-02 4.66287136e-01 -1.29031646e+00
-8.28859150e-01 -1.12925339e+00 -9.17903364e-01 -8.46123278e-01
1.63953424e-01 -4.92095143e-01 6.54254183e-02 5.62037051e-01
4.91773099e-01 3.15751523e-01 -7.59661496e-02 -1.32076752e+00
-3.43034893e-01 -6.35592520e-01 -3.55817467e-01 5.15096664e-01
1.67744353e-01 -9.88865614e-01 4.06382512e-03 3.26742113e-01] | [7.076631546020508, -0.8994901776313782] |
aceb2d94-7d8e-467d-8aad-070808a73ce8 | spatio-temporal-covariance-descriptors-for | 1303.6021 | null | https://arxiv.org/abs/1303.6021v1 | https://arxiv.org/pdf/1303.6021v1.pdf | Spatio-Temporal Covariance Descriptors for Action and Gesture Recognition | We propose a new action and gesture recognition method based on spatio-temporal covariance descriptors and a weighted Riemannian locality preserving projection approach that takes into account the curved space formed by the descriptors. The weighted projection is then exploited during boosting to create a final multiclass classification algorithm that employs the most useful spatio-temporal regions. We also show how the descriptors can be computed quickly through the use of integral video representations. Experiments on the UCF sport, CK+ facial expression and Cambridge hand gesture datasets indicate superior performance of the proposed method compared to several recent state-of-the-art techniques. The proposed method is robust and does not require additional processing of the videos, such as foreground detection, interest-point detection or tracking. | ['Andres Sanin', 'Conrad Sanderson', 'Mehrtash T. Harandi', 'Brian C. Lovell'] | 2013-03-25 | null | null | null | null | ['interest-point-detection'] | ['computer-vision'] | [ 2.14148551e-01 -6.49427474e-01 -2.83102721e-01 -5.56898117e-01
-6.73816323e-01 -3.75644445e-01 9.21338320e-01 -1.69552132e-01
-8.90872717e-01 4.26947057e-01 7.10388198e-02 1.75953899e-02
-3.91028911e-01 -2.76666403e-01 -3.01720917e-01 -1.06495643e+00
-3.16117674e-01 1.01098493e-01 5.98859608e-01 -9.83924512e-03
5.18320203e-01 9.58453596e-01 -1.73108137e+00 3.27447712e-01
3.24600667e-01 1.26924169e+00 -8.91875178e-02 8.49804878e-01
5.85852712e-02 7.67652273e-01 -2.46204093e-01 -1.12715133e-01
3.35491300e-01 -3.65429640e-01 -5.54706514e-01 2.78826773e-01
5.77931881e-01 -3.19059104e-01 -2.18894839e-01 7.81655550e-01
4.85727727e-01 8.70160460e-01 7.95709252e-01 -1.08814013e+00
-2.70487130e-01 -3.30328226e-01 -7.70032942e-01 5.29912531e-01
3.77967715e-01 -3.25944662e-01 8.66227388e-01 -1.19133639e+00
9.05929446e-01 1.36702967e+00 4.38622981e-01 5.78933477e-01
-9.85220015e-01 -6.46540225e-01 2.07776010e-01 5.31464577e-01
-1.38738883e+00 -2.10069567e-01 7.18673408e-01 -3.84290963e-01
9.17369008e-01 3.92701417e-01 6.16326511e-01 9.04099166e-01
8.51502717e-02 9.27926958e-01 1.22184491e+00 -5.40778756e-01
2.56656259e-01 -3.93762887e-01 3.32601070e-01 1.04509282e+00
-3.33878219e-01 1.85954019e-01 -5.00020862e-01 -4.17154819e-01
7.44484127e-01 1.79871336e-01 -6.17465563e-02 -8.71265471e-01
-9.90914166e-01 8.38663518e-01 2.95455873e-01 5.02005577e-01
-5.55887938e-01 2.44322315e-01 4.77472246e-01 -1.21050432e-01
5.74495614e-01 -4.40837741e-01 -1.13818176e-01 -4.35098529e-01
-1.12935710e+00 2.14468926e-01 3.00666243e-01 5.53553224e-01
4.25408721e-01 -1.64192483e-01 4.21145409e-02 7.20973313e-01
4.84297246e-01 3.20220351e-01 4.88277555e-01 -9.54539955e-01
3.08927596e-01 5.59995353e-01 5.88661293e-03 -1.00550175e+00
-4.70279455e-01 2.76379138e-01 -4.63273197e-01 6.06163383e-01
6.69423044e-01 2.10017130e-01 -9.12006855e-01 1.24846900e+00
7.41381288e-01 3.40022087e-01 -1.74785778e-01 1.01893127e+00
4.24377263e-01 3.46231222e-01 2.31840938e-01 -2.51207817e-02
1.12140512e+00 -7.94427574e-01 -5.85095465e-01 5.34171581e-01
4.54438627e-01 -7.45834887e-01 5.01711965e-01 4.92071539e-01
-9.09780383e-01 -4.33707803e-01 -9.12005901e-01 -8.30022211e-04
-4.70474601e-01 4.07676399e-01 7.12193906e-01 8.16671431e-01
-8.22721124e-01 7.28879392e-01 -1.15312600e+00 -5.87442815e-01
4.34510082e-01 4.87399042e-01 -5.37695110e-01 1.67731434e-01
-5.40478289e-01 8.57535839e-01 2.13655770e-01 2.16721848e-01
-5.97709596e-01 -2.94867158e-01 -7.98989952e-01 -4.53685164e-01
3.69024603e-03 -3.99239771e-02 8.26609671e-01 -1.20764279e+00
-1.70320439e+00 7.54739881e-01 -3.19898933e-01 -1.37426838e-01
7.13804901e-01 -2.56581366e-01 -3.00673276e-01 7.80415833e-01
-3.34837198e-01 3.97756249e-01 1.07916212e+00 -6.62740290e-01
-8.24980319e-01 -6.80983484e-01 -9.99304000e-03 3.39225709e-01
-1.77271485e-01 5.54774404e-01 -4.13668692e-01 -7.21078873e-01
3.61112595e-01 -1.14049685e+00 -2.47839373e-02 3.24578285e-01
1.86002329e-01 -5.08216143e-01 1.26204562e+00 -7.71177113e-01
8.98247659e-01 -2.18975663e+00 5.04287302e-01 5.66732287e-01
-3.20917279e-01 3.16451490e-01 -1.80792645e-01 1.00522742e-01
-1.79803610e-01 -2.87732333e-01 1.72329638e-02 -1.79192230e-01
-1.68141201e-01 1.22011885e-01 5.19042909e-02 1.09811342e+00
1.92848146e-01 5.31209886e-01 -8.20843935e-01 -6.89750314e-01
6.18367732e-01 8.51629078e-01 -3.12876433e-01 -5.91010526e-02
2.90733486e-01 4.83217329e-01 -6.48605585e-01 8.33033264e-01
5.32907307e-01 2.80012131e-01 2.01111473e-02 3.25509682e-02
-6.18513450e-02 -4.55613956e-02 -1.42641723e+00 1.76333666e+00
-1.38383567e-01 6.23588383e-01 6.91722706e-02 -1.13616538e+00
8.56843710e-01 1.61477551e-01 7.06861377e-01 -5.80619574e-01
2.18265802e-01 2.30999991e-01 -2.81954169e-01 -5.07901549e-01
4.51549858e-01 -1.60499290e-01 4.13390756e-01 3.63159329e-01
2.15355847e-02 4.31934386e-01 1.90182105e-01 3.53464894e-02
9.64020431e-01 6.91654444e-01 2.17787176e-01 -3.23855847e-01
9.72672164e-01 -2.99319953e-01 3.15075934e-01 2.67193407e-01
-4.89078641e-01 4.36888576e-01 2.43138760e-01 -5.48180103e-01
-6.24750733e-01 -9.51708555e-01 -2.90736854e-01 1.41062438e+00
4.28326502e-02 -1.32458255e-01 -7.43512273e-01 -1.00305581e+00
6.31270977e-03 2.66805619e-01 -9.70688999e-01 2.06255794e-01
-8.19433570e-01 -5.48587024e-01 3.45130116e-01 8.72757971e-01
2.80730456e-01 -9.79238868e-01 -1.02696788e+00 1.86112504e-02
8.29614922e-02 -9.20127094e-01 -4.64894354e-01 -1.39779910e-01
-1.25820267e+00 -1.22832763e+00 -1.13312805e+00 -6.75417483e-01
7.06433773e-01 2.78487444e-01 3.61761153e-01 2.47636754e-02
-4.96231824e-01 9.30892646e-01 -5.89985371e-01 1.54204115e-01
2.04955488e-02 -4.28270161e-01 8.31326991e-02 5.90146780e-01
5.71677387e-01 -3.39328796e-01 -5.16450703e-01 4.28700149e-01
-7.75781155e-01 -3.76431972e-01 3.32409292e-01 9.06546533e-01
6.24281049e-01 -3.97887975e-01 4.19162735e-02 -2.08371535e-01
2.75561482e-01 -3.52566540e-02 -5.60826719e-01 4.17220980e-01
-6.99843094e-02 8.32303390e-02 -8.81346911e-02 -8.50842416e-01
-1.19891596e+00 4.73624080e-01 2.60295212e-01 -6.17057025e-01
-3.58716957e-02 1.42750651e-01 1.10244930e-01 -5.75987279e-01
3.38637680e-01 2.21997410e-01 9.72924456e-02 -5.57143033e-01
4.67063755e-01 6.98362947e-01 2.12246850e-01 -3.64202380e-01
4.28510666e-01 9.14010108e-01 3.53357881e-01 -1.20315218e+00
-2.08234459e-01 -9.81040597e-01 -1.24135828e+00 -5.49181938e-01
1.10632479e+00 -4.26982164e-01 -7.56781220e-01 4.13923621e-01
-1.05657578e+00 -9.89796594e-02 -3.36525589e-02 8.44054222e-01
-7.88649380e-01 7.52300739e-01 -5.01051962e-01 -1.28708184e+00
-2.32296750e-01 -9.82757330e-01 1.32104599e+00 1.23716287e-01
-1.66032881e-01 -8.67182136e-01 1.02028973e-01 2.76938885e-01
1.37681082e-01 4.15303200e-01 5.55825293e-01 -3.12021941e-01
-4.91672486e-01 -4.06426936e-01 -2.23327473e-01 3.78190041e-01
5.00989221e-02 1.31161079e-01 -7.65702605e-01 -1.63310841e-01
-3.13121080e-01 -1.88767865e-01 1.02174532e+00 3.93791616e-01
9.35753047e-01 2.39923701e-01 -2.52695084e-01 4.71376091e-01
1.18950176e+00 4.29640949e-01 5.54403186e-01 3.26386601e-01
5.68630159e-01 6.40954077e-01 9.94320869e-01 4.56761956e-01
-3.01849861e-02 9.56380367e-01 1.99561805e-01 2.71597076e-02
4.09558192e-02 1.48106337e-01 4.66142178e-01 3.80453080e-01
-8.19953084e-01 4.71378505e-01 -5.74622393e-01 3.77733439e-01
-2.14212489e+00 -1.13785672e+00 -8.83842185e-02 2.18509960e+00
3.55591565e-01 -3.56396914e-01 3.68736148e-01 3.40704620e-01
5.78287423e-01 1.87543020e-01 -1.70198202e-01 -5.10573685e-01
-2.28424352e-02 6.14214242e-01 4.79340643e-01 5.23182333e-01
-1.60015762e+00 8.36589336e-01 6.31397390e+00 7.57065237e-01
-9.59855258e-01 2.38017350e-01 1.84142008e-01 -2.81201422e-01
5.33172250e-01 -3.15594584e-01 -5.09026349e-01 5.41345216e-02
6.38877034e-01 2.68545449e-01 1.67731524e-01 1.05086422e+00
1.62514642e-01 -4.38995034e-01 -8.33394468e-01 1.06492984e+00
5.00320852e-01 -9.87738013e-01 -1.67697266e-01 1.35672092e-01
4.61479872e-01 -1.34980604e-01 -8.94112065e-02 -5.58633879e-02
-8.47146884e-02 -6.48383439e-01 7.13609576e-01 7.51632571e-01
4.15084630e-01 -8.09079409e-01 3.54696542e-01 -3.83944474e-02
-1.47084355e+00 -8.84615555e-02 -3.07418704e-01 8.89503509e-02
-6.92038685e-02 -1.85092509e-01 -3.22279543e-01 3.10373247e-01
6.50201499e-01 7.64417529e-01 -4.89053547e-01 9.08526599e-01
-1.05381958e-01 3.28357637e-01 -3.34159523e-01 -3.40744257e-01
5.11102915e-01 -4.68854606e-01 5.89464188e-01 1.51735759e+00
2.66751736e-01 3.70113373e-01 1.36125863e-01 2.31207773e-01
4.57747370e-01 4.63626295e-01 -4.74465251e-01 1.32331520e-01
-4.20458078e-01 1.35449088e+00 -1.05266857e+00 -3.96081686e-01
-5.32984316e-01 1.31094992e+00 2.32124299e-01 3.94158572e-01
-6.98010087e-01 -3.16601723e-01 7.69216716e-01 -1.55051053e-01
8.52220178e-01 -5.41950226e-01 2.90446728e-01 -1.19981909e+00
1.81889132e-01 -6.12834811e-01 5.87349534e-01 -4.75229830e-01
-8.56636405e-01 3.96502018e-01 7.37279430e-02 -1.26762211e+00
-3.54038477e-01 -9.59694266e-01 -3.25968295e-01 6.53172374e-01
-1.26354480e+00 -1.20352101e+00 -1.79707348e-01 1.04240739e+00
5.55032432e-01 -1.07985526e-01 9.13960397e-01 3.75896424e-01
-2.30717734e-01 3.25440407e-01 1.76309079e-01 2.08126754e-01
4.89618033e-01 -1.04914534e+00 -1.26677632e-01 7.52559662e-01
3.22912872e-01 4.21155483e-01 3.79224330e-01 -4.95357394e-01
-1.34094346e+00 -7.70072699e-01 8.50686431e-01 -5.45011759e-01
7.44899988e-01 -1.99469313e-01 -6.33750558e-01 4.89137053e-01
-1.38508514e-01 4.74417388e-01 6.72935009e-01 -5.82362600e-02
-4.97147292e-01 -1.58785641e-01 -1.21038687e+00 2.85153925e-01
1.04076409e+00 -5.11642814e-01 -6.39014006e-01 3.16636801e-01
-2.73106813e-01 -3.67144734e-01 -6.25977099e-01 1.85254321e-01
1.13617718e+00 -9.58618402e-01 9.98392045e-01 -8.98680925e-01
-6.94604516e-02 -1.75523609e-01 -3.74224067e-01 -8.02985907e-01
-2.51578480e-01 -5.35064042e-01 -2.56621931e-03 8.37827384e-01
1.20142661e-02 -3.29745322e-01 8.02845180e-01 5.33668518e-01
3.23986351e-01 -6.56005204e-01 -1.46725106e+00 -9.32998776e-01
-2.62323439e-01 -6.01703525e-01 5.88898771e-02 5.83456457e-01
2.18870193e-01 -3.75101596e-01 -3.62032682e-01 -1.47424296e-01
7.88900852e-01 1.85173422e-01 6.74766123e-01 -1.10145223e+00
-8.89619663e-02 -3.43343645e-01 -1.17414427e+00 -7.03323603e-01
3.18206996e-01 -7.79812813e-01 -1.62318021e-01 -1.05701315e+00
1.65193439e-01 -7.41685182e-02 -5.72515249e-01 3.54768872e-01
8.32516775e-02 6.09953642e-01 3.78617495e-01 -2.29286104e-02
-7.27071106e-01 5.10099530e-01 9.47325289e-01 3.44070978e-02
-1.52341172e-01 4.28679995e-02 1.85388029e-01 9.03262854e-01
4.61849183e-01 -3.19974810e-01 1.26132637e-03 1.61293060e-01
-5.08796036e-01 -9.15511027e-02 5.58172524e-01 -8.86244774e-01
3.31393927e-02 -2.13644221e-01 5.99674642e-01 -6.29941881e-01
7.48871922e-01 -9.08907115e-01 -2.24540025e-01 5.12039602e-01
-3.27855825e-01 9.19976607e-02 1.26691386e-01 7.27176547e-01
-5.06761894e-02 -1.43008083e-01 9.78194654e-01 1.44672841e-01
-9.75561082e-01 1.23076759e-01 -4.85520780e-01 -3.99871826e-01
1.39530408e+00 -3.57944012e-01 3.11167687e-01 -2.61217624e-01
-1.01181054e+00 -2.35231116e-01 1.70185700e-01 5.47719300e-01
8.12844694e-01 -1.45590866e+00 -3.86702389e-01 1.76233888e-01
-4.75315563e-02 -9.63334262e-01 1.41995445e-01 1.18094897e+00
-5.42799532e-01 3.32562000e-01 -4.46319371e-01 -7.36967981e-01
-2.08843327e+00 4.88875926e-01 3.57792705e-01 -1.11467980e-01
-8.25343788e-01 6.51188672e-01 -2.48244815e-02 -6.96836337e-02
4.83119994e-01 -4.28957731e-01 -2.84221649e-01 3.38992268e-01
9.31381464e-01 8.02811384e-01 -7.11863628e-03 -1.27487981e+00
-8.49989355e-01 1.06293845e+00 2.33479366e-01 -4.18186575e-01
1.36153245e+00 1.03658251e-01 8.04112777e-02 4.58767503e-01
1.38998139e+00 -4.33208913e-01 -1.33228290e+00 -1.27531275e-01
2.80297488e-01 -8.46339762e-01 1.72272563e-01 -4.16418105e-01
-1.11893392e+00 9.72144842e-01 1.15994990e+00 -1.94315925e-01
1.24692380e+00 -1.55126050e-01 3.05523396e-01 2.65226007e-01
4.27340627e-01 -1.37147999e+00 1.64689533e-02 2.64319360e-01
1.04920101e+00 -1.13796687e+00 2.53878027e-01 -3.27803552e-01
-6.43552244e-01 1.45683634e+00 1.67220309e-02 -5.06887913e-01
9.77153361e-01 -1.13390744e-01 1.49420515e-01 2.35121530e-02
-2.80677259e-01 -4.81293887e-01 7.04828084e-01 6.29666746e-01
3.83514911e-01 5.83218150e-02 -8.07515562e-01 -1.31704230e-02
4.53315377e-01 1.38049915e-01 -8.30322132e-03 1.26957142e+00
-2.33721927e-01 -1.22101617e+00 -4.06166673e-01 1.51153624e-01
-5.67783117e-01 3.16802800e-01 -4.30788308e-01 1.01154721e+00
-2.38829404e-02 6.48834884e-01 8.16239603e-03 -2.69760370e-01
3.01626176e-01 4.08760220e-01 8.68637204e-01 -1.07108191e-01
-6.37922287e-01 3.65010291e-01 2.50905603e-02 -1.22039735e+00
-1.10003102e+00 -1.26998687e+00 -1.45621490e+00 1.89507961e-01
-3.86070609e-01 -1.86485663e-01 9.63853955e-01 9.80658174e-01
1.48891658e-01 -6.34248555e-02 4.80009198e-01 -1.41116643e+00
-4.72777635e-01 -9.51867342e-01 -7.99862027e-01 5.79785585e-01
2.17315525e-01 -1.18509567e+00 -2.92063266e-01 7.93930218e-02] | [8.01049518585205, 0.3413868248462677] |
01d4aa12-c3b6-4de7-8568-682b34dbe83a | hsr-l1-2-regularized-sparse-representation | 1409.6448 | null | https://arxiv.org/abs/1409.6448v1 | https://arxiv.org/pdf/1409.6448v1.pdf | HSR: L1/2 Regularized Sparse Representation for Fast Face Recognition using Hierarchical Feature Selection | In this paper, we propose a novel method for fast face recognition called L1/2 Regularized Sparse Representation using Hierarchical Feature Selection (HSR). By employing hierarchical feature selection, we can compress the scale and dimension of global dictionary, which directly contributes to the decrease of computational cost in sparse representation that our approach is strongly rooted in. It consists of Gabor wavelets and Extreme Learning Machine Auto-Encoder (ELM-AE) hierarchically. For Gabor wavelets part, local features can be extracted at multiple scales and orientations to form Gabor-feature based image, which in turn improves the recognition rate. Besides, in the presence of occluded face image, the scale of Gabor-feature based global dictionary can be compressed accordingly because redundancies exist in Gabor-feature based occlusion dictionary. For ELM-AE part, the dimension of Gabor-feature based global dictionary can be compressed because high-dimensional face images can be rapidly represented by low-dimensional feature. By introducing L1/2 regularization, our approach can produce sparser and more robust representation compared to regularized Sparse Representation based Classification (SRC), which also contributes to the decrease of the computational cost in sparse representation. In comparison with related work such as SRC and Gabor-feature based SRC (GSRC), experimental results on a variety of face databases demonstrate the great advantage of our method for computational cost. Moreover, we also achieve approximate or even better recognition rate. | ['Mengmeng Ma', 'Bo Han', 'Tingting Sun', 'Bo He', 'Amaury Lendasse'] | 2014-09-23 | null | null | null | null | ['sparse-representation-based-classification'] | ['computer-vision'] | [-3.95858660e-02 -4.40549880e-01 -3.31261992e-01 -2.36187682e-01
-5.66979051e-01 9.92802680e-02 2.37853266e-02 -4.54481393e-01
6.60362616e-02 5.55933058e-01 3.09417129e-01 3.86607975e-01
-3.30835670e-01 -9.81468558e-01 -3.50910187e-01 -9.77932811e-01
3.83625701e-02 -1.51174366e-01 -9.31989104e-02 -1.08335748e-01
2.04557180e-01 9.41963911e-01 -1.97124565e+00 2.64196813e-01
5.04059672e-01 1.23584592e+00 1.09688029e-01 1.36419311e-01
-1.62428081e-01 5.89205146e-01 -3.21237057e-01 -2.52315635e-03
2.37615436e-01 -4.87262070e-01 -3.91355336e-01 5.55480778e-01
3.09067190e-01 -1.84456825e-01 -6.82331324e-01 1.03235137e+00
5.64427853e-01 2.64158040e-01 5.35217643e-01 -8.60720336e-01
-8.45310390e-01 7.34482612e-03 -8.23245049e-01 1.20333359e-01
5.36346555e-01 -3.43949258e-01 6.03831232e-01 -1.50197971e+00
5.87057173e-01 1.51654005e+00 6.23373568e-01 2.67175585e-01
-9.56871748e-01 -5.12131691e-01 4.33497643e-03 3.70296031e-01
-1.92590761e+00 -7.41050482e-01 1.04091156e+00 -7.99033046e-02
9.61401761e-01 4.14280951e-01 5.32474220e-01 5.77539921e-01
3.06235198e-02 4.41196680e-01 9.66199100e-01 -5.87598026e-01
1.25241920e-01 -1.15244433e-01 -9.59625244e-02 1.04210055e+00
3.29892218e-01 1.10123470e-01 -8.58495831e-01 -3.64728212e-01
1.23782623e+00 4.42711085e-01 -3.24335903e-01 -6.57336861e-02
-8.44134986e-01 1.01223695e+00 2.35983357e-01 6.29652441e-01
-6.03546917e-01 -2.07910120e-01 2.73663759e-01 4.66539443e-01
2.98126012e-01 -2.89306015e-01 1.35802543e-02 1.65721983e-01
-1.01653707e+00 -2.57054359e-01 5.76341391e-01 7.67037570e-01
9.45286274e-01 6.35831714e-01 -3.30159552e-02 1.19960678e+00
3.55968446e-01 7.50330269e-01 9.11961079e-01 -8.03572774e-01
1.33987710e-01 5.23836553e-01 -2.66185045e-01 -1.70253146e+00
-1.50831923e-01 -3.96994859e-01 -1.26190710e+00 -1.65844142e-01
-2.62846053e-01 3.45690846e-01 -7.71554351e-01 1.24319935e+00
4.50942516e-01 4.44685012e-01 2.40863875e-01 8.36701393e-01
1.01415479e+00 7.02029705e-01 -2.91943252e-01 -6.19079471e-01
1.50398004e+00 -7.53079712e-01 -8.51739526e-01 1.16202109e-01
2.59759128e-01 -9.86889422e-01 5.80320716e-01 8.52531642e-02
-7.85717309e-01 -7.78651357e-01 -8.15081716e-01 1.31655738e-01
8.47127140e-02 4.02745605e-01 7.84061670e-01 5.72694361e-01
-9.57336187e-01 2.42341295e-01 -7.36632705e-01 -4.70394976e-02
4.63351488e-01 4.87014651e-01 -8.22210431e-01 -5.91674805e-01
-7.52386451e-01 4.59736973e-01 3.36352549e-02 1.62865400e-01
-6.19466424e-01 -3.64319235e-01 -1.16517305e+00 2.68988043e-01
-1.92155375e-03 -2.24308148e-01 4.06483471e-01 -6.69412136e-01
-1.51604366e+00 3.51321340e-01 -8.33668232e-01 1.30075216e-01
-4.25386608e-01 3.04267406e-01 -6.66291535e-01 6.08035147e-01
8.88950229e-02 1.84725463e-01 1.35454762e+00 -8.65540981e-01
-2.81302333e-01 -5.88512659e-01 -4.24971670e-01 3.71976495e-02
-6.26635492e-01 3.96480076e-02 -2.00476244e-01 -9.56835747e-01
7.71375775e-01 -5.62771380e-01 -1.40547290e-01 -1.79003507e-01
1.96855217e-01 -2.36817613e-01 1.29710639e+00 -7.48763859e-01
1.27484643e+00 -2.63169050e+00 3.02928209e-01 3.90828639e-01
1.21878482e-01 1.50338039e-01 -3.81321251e-01 4.62674648e-01
-1.99591771e-01 -2.00077772e-01 -4.42956714e-03 -1.08708829e-01
-5.47124505e-01 3.47437501e-01 -4.08344179e-01 7.25838184e-01
3.51010203e-01 5.70322871e-01 -5.38950562e-01 -5.70255339e-01
2.29585424e-01 1.07761180e+00 -6.40149236e-01 8.02023709e-02
5.64775229e-01 3.15690875e-01 -6.97306395e-01 1.09741461e+00
8.97408426e-01 -5.21971174e-02 2.68934518e-02 -3.90937269e-01
-1.01903789e-02 -3.23402643e-01 -1.55482233e+00 1.37301719e+00
-7.22155511e-01 2.43865266e-01 3.44784148e-02 -1.35683393e+00
1.34683204e+00 5.67863405e-01 5.11576355e-01 -7.69891918e-01
-1.55931199e-02 3.48462045e-01 -5.34019709e-01 -4.35461074e-01
5.41497543e-02 -9.48388129e-02 3.91404778e-01 1.16372131e-01
2.64589578e-01 3.05847168e-01 2.23095138e-02 -1.86205506e-01
7.73695111e-01 -2.48704210e-01 4.08521384e-01 -2.46097371e-01
1.13440800e+00 -5.02800286e-01 7.51295924e-01 1.74985096e-01
1.68376625e-01 4.41025168e-01 7.78207853e-02 -7.15484560e-01
-6.68152928e-01 -4.08580482e-01 -4.43674207e-01 7.35747695e-01
1.50064185e-01 -4.35790002e-01 -3.38583201e-01 -4.30437803e-01
6.81705996e-02 3.16221341e-02 -4.47848022e-01 -1.54334784e-01
-7.84171700e-01 -5.80160737e-01 9.53395814e-02 3.67154121e-01
6.31390989e-01 -1.04980028e+00 -2.45304987e-01 2.17845127e-01
1.04181007e-01 -1.02552032e+00 -4.14756060e-01 -1.38811812e-01
-1.16717803e+00 -9.12301540e-01 -7.83732235e-01 -1.12833047e+00
1.07172704e+00 8.85611057e-01 5.61228633e-01 3.26405287e-01
-6.16246700e-01 3.92430663e-01 -6.51844740e-01 3.02089036e-01
1.89822540e-01 -4.17138696e-01 2.58336633e-01 5.02008379e-01
2.83549786e-01 -8.48749280e-01 -5.62038839e-01 3.66618305e-01
-9.09485519e-01 -4.69772220e-01 5.61358869e-01 1.13272345e+00
9.85203028e-01 4.05858994e-01 8.18058789e-01 -5.33849537e-01
2.52906919e-01 -4.08721119e-01 -5.77199936e-01 9.31340754e-02
-3.07668418e-01 -8.53507966e-02 8.33050728e-01 -3.61697346e-01
-1.01258349e+00 1.80405945e-01 -1.37651771e-01 -5.55619001e-01
9.37217623e-02 5.14777958e-01 -6.00811094e-02 -9.25913990e-01
3.56711954e-01 9.88289595e-01 2.33934715e-01 -7.04347432e-01
1.33786753e-01 8.13456833e-01 1.59377128e-01 -3.90307903e-01
9.16780710e-01 5.75934231e-01 4.08834666e-01 -1.35242522e+00
-4.74105954e-01 -4.55449671e-01 -3.41806293e-01 1.54569238e-01
2.61217833e-01 -1.27573979e+00 -6.69309914e-01 2.75080800e-01
-7.20861077e-01 5.91448069e-01 -3.43806595e-01 6.52403951e-01
-3.68746847e-01 6.79417133e-01 -4.85572428e-01 -9.61583257e-01
-2.78726637e-01 -1.15650964e+00 1.15029407e+00 2.94101357e-01
3.12459141e-01 -6.85345531e-01 -3.33806753e-01 2.46688530e-01
7.60037184e-01 1.53109044e-01 7.85902977e-01 -1.41060680e-01
-4.29235101e-01 -5.49173534e-01 -1.36926204e-01 2.87901700e-01
4.54387307e-01 -3.46729785e-01 -6.19364500e-01 -6.89825296e-01
3.86776417e-01 -3.54414254e-01 7.79835522e-01 3.00992906e-01
1.27813506e+00 -5.21731079e-01 -1.73167750e-01 8.28904986e-01
1.49516928e+00 1.83058858e-01 7.60786712e-01 -1.08811418e-02
6.22400761e-01 3.32303345e-01 5.09674311e-01 7.46551275e-01
1.85057133e-01 8.20220530e-01 -1.16068879e-02 -1.64849311e-01
-4.64358240e-01 -5.69268689e-02 4.07357901e-01 1.22175181e+00
-3.25243324e-01 5.80685675e-01 -8.80513787e-02 4.33434218e-01
-1.54903829e+00 -1.09008956e+00 3.01970392e-01 2.08420753e+00
6.08149409e-01 -8.13993633e-01 1.91628579e-02 4.72383291e-01
6.70924127e-01 2.73900270e-01 -2.58119524e-01 -2.08361834e-01
-2.43392885e-01 5.58731377e-01 -8.25786740e-02 5.72187185e-01
-7.40262628e-01 8.64811957e-01 5.85905123e+00 1.28151584e+00
-1.31453073e+00 1.38163060e-01 3.95008445e-01 9.26438570e-02
-1.68315262e-01 -8.64965990e-02 -1.00314224e+00 3.59169453e-01
5.82358420e-01 -5.89254536e-02 6.93046749e-01 1.07780147e+00
-8.31276849e-02 2.14693621e-01 -5.35315871e-01 1.57336771e+00
2.62518942e-01 -1.25511801e+00 3.63218844e-01 1.35202661e-01
7.42978156e-01 -6.04260147e-01 5.11142612e-02 2.03638703e-01
-4.07363027e-01 -1.05322111e+00 1.65287644e-01 3.54663402e-01
1.14599013e+00 -1.03136194e+00 6.98393703e-01 2.14644298e-02
-1.68966758e+00 -1.90399244e-01 -1.03948224e+00 2.04346642e-01
-2.03952983e-01 9.01651144e-01 -9.45196822e-02 6.67779446e-01
5.53070664e-01 8.09095800e-01 -3.07206601e-01 7.90452659e-01
7.85765126e-02 3.49136353e-01 -3.58917981e-01 3.37820888e-01
2.45334823e-02 -2.66737223e-01 4.95819777e-01 9.82763112e-01
6.12393737e-01 5.79737663e-01 4.14350390e-01 2.26381153e-01
-2.57224143e-02 4.89638209e-01 -6.65036380e-01 -1.46933436e-01
5.36657095e-01 1.22757709e+00 -6.00749791e-01 -1.56407073e-01
-6.05770051e-01 1.01941490e+00 2.45886773e-01 4.48137522e-01
-2.91206896e-01 -6.43833518e-01 3.54131699e-01 -7.01559782e-02
8.84248137e-01 -3.38406116e-01 1.20161682e-01 -1.37136877e+00
1.58168629e-01 -1.07619476e+00 2.44394422e-01 -6.16899692e-02
-9.77801561e-01 9.83070731e-01 -4.37280238e-01 -1.21952653e+00
-6.10710010e-02 -2.54967153e-01 -1.77293688e-01 9.17606056e-01
-1.66996586e+00 -1.24267066e+00 -3.04924548e-01 1.10686088e+00
6.29775584e-01 -7.12180197e-01 1.14263701e+00 4.54663903e-01
-5.67161024e-01 7.69048929e-01 1.96860150e-01 -5.57107441e-02
2.56275415e-01 -3.52270782e-01 -2.73416430e-01 6.39287889e-01
3.07265610e-01 6.92321718e-01 -1.78994294e-02 -5.95544696e-01
-1.88301134e+00 -1.09253907e+00 9.33769822e-01 3.27128261e-01
1.40105620e-01 9.60823447e-02 -8.77095819e-01 3.55186701e-01
-4.28632826e-01 5.93798161e-01 1.00868011e+00 -3.90493274e-02
-6.89274132e-01 -4.29786414e-01 -1.40171742e+00 1.22577704e-01
9.95151341e-01 -6.10566139e-01 -2.70340919e-01 6.28871322e-01
4.52761978e-01 -1.57963917e-01 -1.24071276e+00 5.02131760e-01
6.69862628e-01 -8.01640570e-01 1.15286815e+00 -4.19353753e-01
1.90691665e-01 -4.42619801e-01 -7.06939399e-01 -9.22729194e-01
-8.52774382e-01 -3.56691211e-01 -4.66676295e-01 1.04946065e+00
-1.97151408e-01 -8.37829411e-01 5.33786535e-01 -9.32312161e-02
1.84372857e-01 -1.03929925e+00 -1.37332880e+00 -8.83819282e-01
-5.80382109e-01 2.23759651e-01 6.71556056e-01 8.95591795e-01
-1.92836642e-01 -4.31992933e-02 -5.78212082e-01 3.12108606e-01
8.16116035e-01 6.43293023e-01 3.40240061e-01 -1.06052196e+00
-1.49215713e-01 -8.61319304e-02 -9.50573981e-01 -8.20214450e-01
2.11322054e-01 -8.75661016e-01 -4.32085335e-01 -1.11547351e+00
2.28677049e-01 -4.53981459e-01 -2.51604348e-01 7.01028585e-01
3.50840464e-02 7.40411162e-01 7.32077062e-02 5.90262949e-01
-3.08467686e-01 9.46601629e-01 1.24735010e+00 6.27691448e-02
-2.56082743e-01 -2.35848144e-01 -6.87933564e-01 4.36697334e-01
3.77928436e-01 -4.54027504e-01 -3.59823108e-02 -2.46653736e-01
-3.97885859e-01 3.63852859e-01 1.03517026e-01 -9.71575975e-01
1.23161368e-01 -1.21758014e-01 7.68751562e-01 -1.29866436e-01
4.42583799e-01 -9.19997215e-01 2.24284679e-01 5.53002834e-01
1.88555717e-01 -2.82780617e-01 1.05319612e-01 5.56504548e-01
-7.81094849e-01 -1.60418123e-01 9.31399047e-01 -3.52404565e-02
-5.87579072e-01 4.90102798e-01 -2.60988116e-01 -6.43077552e-01
9.07230377e-01 -4.21304226e-01 1.24113873e-01 -2.53916234e-01
-7.64237702e-01 -3.36620390e-01 3.26081067e-02 2.59479135e-01
1.13924372e+00 -1.82740688e+00 -8.65358651e-01 1.03820288e+00
-1.39988109e-01 -2.40493536e-01 5.65361261e-01 7.66319692e-01
-3.75776768e-01 5.53305805e-01 -2.93815732e-01 -5.25378346e-01
-1.25257814e+00 4.48006690e-01 -2.19263703e-01 9.34176669e-02
-7.99558401e-01 8.87193918e-01 -1.97480321e-02 1.75122827e-01
6.90468252e-02 3.94377202e-01 -3.77472132e-01 -9.25738644e-03
1.00486314e+00 3.61394167e-01 1.23729095e-01 -1.24919546e+00
-5.26693463e-01 1.16812670e+00 -1.65384173e-01 3.68173331e-01
1.68693411e+00 4.78188209e-02 -7.36407101e-01 -2.53561974e-01
1.51517367e+00 3.54729980e-01 -7.57912934e-01 -3.56163740e-01
-5.49942017e-01 -1.03150821e+00 4.18941259e-01 -7.52607659e-02
-1.52469969e+00 5.47181606e-01 7.55571783e-01 -1.24070263e-02
1.47967196e+00 -1.70441955e-01 8.19571435e-01 3.83614838e-01
7.66816139e-01 -5.07283986e-01 1.63370967e-01 2.50760823e-01
1.23748422e+00 -1.06022871e+00 1.28477037e-01 -8.59742224e-01
-3.27927828e-01 1.33563125e+00 2.80192673e-01 -5.51227570e-01
7.67226934e-01 1.68181896e-01 -3.06454241e-01 -2.30159387e-01
-2.39985466e-01 -3.20449807e-02 3.71106118e-01 6.14428461e-01
2.77916372e-01 -1.16043761e-01 -3.89274746e-01 5.79427660e-01
-3.68229523e-02 3.93343456e-02 -2.19229367e-02 9.21334624e-01
-7.19161212e-01 -1.20971525e+00 -5.73552251e-01 4.25213099e-01
-2.49950841e-01 -1.03238605e-01 1.70632958e-01 2.44731933e-01
-4.22012545e-02 9.41926122e-01 -2.70470351e-01 -3.66104275e-01
1.40984923e-01 -3.82742099e-02 7.07572460e-01 -6.05351806e-01
-5.49532846e-02 3.89589965e-01 -3.83427143e-01 -7.46306598e-01
-3.44525158e-01 -3.88978332e-01 -1.03180802e+00 -2.10521922e-01
-4.95518029e-01 4.44331706e-01 5.19398153e-01 7.04057574e-01
7.26954758e-01 1.74076319e-01 1.21021676e+00 -9.06915426e-01
-5.27594924e-01 -6.32325947e-01 -9.56964552e-01 1.81202888e-01
3.53196144e-01 -9.50212181e-01 -5.11244357e-01 8.55928361e-02] | [12.532699584960938, 0.42020609974861145] |
9fdccb6d-f563-47cc-a235-7c21febccd34 | selffed-self-supervised-federated-learning | 2307.01514 | null | https://arxiv.org/abs/2307.01514v1 | https://arxiv.org/pdf/2307.01514v1.pdf | SelfFed: Self-supervised Federated Learning for Data Heterogeneity and Label Scarcity in IoMT | Self-supervised learning in federated learning paradigm has been gaining a lot of interest both in industry and research due to the collaborative learning capability on unlabeled yet isolated data. However, self-supervised based federated learning strategies suffer from performance degradation due to label scarcity and diverse data distributions, i.e., data heterogeneity. In this paper, we propose the SelfFed framework for Internet of Medical Things (IoMT). Our proposed SelfFed framework works in two phases. The first phase is the pre-training paradigm that performs augmentive modeling using Swin Transformer based encoder in a decentralized manner. The first phase of SelfFed framework helps to overcome the data heterogeneity issue. The second phase is the fine-tuning paradigm that introduces contrastive network and a novel aggregation strategy that is trained on limited labeled data for a target task in a decentralized manner. This fine-tuning stage overcomes the label scarcity problem. We perform our experimental analysis on publicly available medical imaging datasets and show that our proposed SelfFed framework performs better when compared to existing baselines concerning non-independent and identically distributed (IID) data and label scarcity. Our method achieves a maximum improvement of 8.8% and 4.1% on Retina and COVID-FL datasets on non-IID dataset. Further, our proposed method outperforms existing baselines even when trained on a few (10%) labeled instances. | ['Marius George Linguraru', 'Syed Muhammad Anwar', 'Kapal Dev', 'Sunder Ali Khowaja'] | 2023-07-04 | null | null | null | null | ['self-supervised-learning', 'federated-learning'] | ['computer-vision', 'methodology'] | [ 2.01662898e-01 2.21445844e-01 -3.38007510e-01 -5.17631710e-01
-8.72595012e-01 -2.23544031e-01 3.56142521e-01 2.32525945e-01
-4.79132205e-01 8.36666644e-01 3.20499748e-01 -1.23511352e-01
-1.38451800e-01 -5.49512982e-01 -4.90637004e-01 -8.93271625e-01
6.79273456e-02 4.56063807e-01 1.77303106e-01 2.22843379e-01
-4.26784664e-01 1.37481987e-01 -1.37507391e+00 4.50263023e-01
9.38356817e-01 1.27118909e+00 2.30882764e-01 2.48356968e-01
-2.54391760e-01 1.28426528e+00 -3.50098968e-01 -2.52089113e-01
5.25561929e-01 -3.65279615e-01 -8.23282123e-01 1.50005460e-01
4.02698696e-01 -2.77779222e-01 -1.00039206e-01 9.58195210e-01
9.52193379e-01 -2.74673611e-01 3.64930630e-01 -1.24351454e+00
-2.62216032e-01 7.32197940e-01 -5.38747549e-01 2.65673012e-01
-5.61817229e-01 8.52504447e-02 6.74769640e-01 -4.91567641e-01
6.34915233e-01 7.89433956e-01 6.96276307e-01 6.57893181e-01
-9.04496193e-01 -8.48429799e-01 -1.69366345e-01 1.21791646e-01
-1.03249705e+00 -4.31114912e-01 7.49423385e-01 -3.89269263e-01
5.51423192e-01 -1.65653735e-01 2.10905999e-01 8.95715833e-01
2.02630490e-01 7.69883335e-01 1.47686267e+00 -3.99782807e-01
5.09972990e-01 2.14769900e-01 5.09099029e-02 7.83874094e-01
1.01313382e-01 -3.41157131e-02 -4.50497359e-01 -3.12192261e-01
4.18577909e-01 3.08922321e-01 1.31842077e-01 -4.38199043e-01
-1.23883665e+00 6.95657730e-01 5.33613443e-01 4.73610491e-01
-6.51805162e-01 -5.67985810e-02 8.41961086e-01 4.57929015e-01
6.52795494e-01 7.27735236e-02 -6.99420929e-01 8.89216736e-02
-9.53094840e-01 -4.11056370e-01 5.03832400e-01 7.95594692e-01
6.26121938e-01 -4.49879691e-02 -3.28438222e-01 1.00027990e+00
1.32855907e-01 3.80352497e-01 8.95116866e-01 -8.36051345e-01
3.56969208e-01 7.13663757e-01 -1.66363999e-01 -5.28694153e-01
-6.18190050e-01 -1.06596375e+00 -1.11456871e+00 3.86487767e-02
2.59457052e-01 -5.74872077e-01 -8.92988861e-01 1.86863136e+00
6.49740994e-01 2.21990600e-01 2.30955556e-01 7.18523681e-01
8.80937278e-01 1.91249594e-01 2.83478051e-01 -4.23697829e-01
1.43555045e+00 -1.35464084e+00 -8.62590551e-01 2.06637233e-01
8.49928319e-01 -4.95664537e-01 7.56053448e-01 2.83419788e-01
-8.54067266e-01 -2.99219519e-01 -8.51230323e-01 3.19900721e-01
-9.63626504e-02 1.89534143e-01 5.05244732e-01 6.44593894e-01
-1.01568210e+00 3.99934500e-01 -7.29956508e-01 -3.55704039e-01
1.07333803e+00 5.77208638e-01 -5.41875064e-01 -2.02340767e-01
-9.01642501e-01 4.73776549e-01 1.89226314e-01 -4.45947886e-01
-1.00982451e+00 -9.36903536e-01 -4.70682442e-01 -2.20914874e-02
4.06590164e-01 -8.73190939e-01 1.18223035e+00 -1.13784778e+00
-1.18378758e+00 8.06587219e-01 8.81948173e-02 -7.63040721e-01
5.98324418e-01 1.37810513e-01 -4.69345748e-01 2.84777343e-01
2.47049645e-01 5.20157337e-01 8.21024895e-01 -1.04417288e+00
-8.50020409e-01 -6.17357135e-01 -1.86683491e-01 1.98599800e-01
-6.65254354e-01 -1.11893520e-01 -1.00729503e-02 -6.05362236e-01
-1.54197872e-01 -8.81065905e-01 -2.70452976e-01 5.15132248e-02
-3.00238758e-01 -3.08660507e-01 9.69078481e-01 -2.82633603e-01
9.44603622e-01 -2.14823747e+00 -3.27796072e-01 -1.60075292e-01
5.22771358e-01 4.08898234e-01 -7.86877871e-02 2.94240326e-01
-9.92406383e-02 -9.31518301e-02 -1.38200685e-01 -5.34127712e-01
-2.73941100e-01 1.05148204e-01 2.39427030e-01 4.64559466e-01
-8.51367973e-03 6.53201878e-01 -1.03098595e+00 -9.51171339e-01
1.25822891e-02 4.21076745e-01 -4.40521270e-01 1.88205302e-01
-1.30686283e-01 8.24863434e-01 -4.38054472e-01 8.13687801e-01
7.08984554e-01 -6.10267282e-01 1.96947306e-01 -4.60621387e-01
-1.55131668e-02 -1.49956895e-02 -8.64372790e-01 2.07010746e+00
-6.62688553e-01 2.74636131e-02 2.79215425e-02 -1.18263733e+00
6.53651059e-01 6.70150936e-01 1.14496052e+00 -7.87750483e-01
3.25797617e-01 5.01750350e-01 9.50971097e-02 -4.97302204e-01
-2.20332786e-01 -3.71627092e-01 2.35014334e-01 7.56259739e-01
4.86122936e-01 7.12753892e-01 2.88081076e-02 3.30530375e-01
1.62176776e+00 -7.19905943e-02 2.66029179e-01 -3.74244511e-01
5.71372747e-01 3.25607844e-02 8.97474587e-01 6.50468409e-01
-5.35500348e-01 3.09146792e-01 1.13497689e-01 -5.92484891e-01
-8.68400931e-01 -7.44535506e-01 -2.50293911e-01 8.75362575e-01
-5.61998934e-02 -2.94780225e-01 -5.36052227e-01 -1.26551569e+00
-6.27176762e-02 1.63902059e-01 -6.33770704e-01 -9.54468325e-02
-1.96809039e-01 -8.69456589e-01 4.07802343e-01 3.24988186e-01
9.11022902e-01 -1.06332600e+00 -8.28231514e-01 1.85834199e-01
-1.27540812e-01 -1.09292114e+00 -3.85572761e-01 5.42241931e-01
-1.02274549e+00 -1.06636953e+00 -7.86525190e-01 -7.70459771e-01
6.84327364e-01 2.91826844e-01 9.98110116e-01 -4.90240268e-02
-3.67922425e-01 2.65009314e-01 -5.46230495e-01 -5.29277205e-01
-3.98112804e-01 1.35705352e-01 -6.09021895e-02 4.17561293e-01
3.43403995e-01 -6.13390446e-01 -1.08726120e+00 4.21155035e-01
-9.38062906e-01 -3.95164266e-02 7.95211554e-01 1.21400857e+00
7.18629062e-01 1.72589734e-01 1.07049978e+00 -1.35876548e+00
2.24221885e-01 -9.62143540e-01 -2.57005364e-01 3.31715554e-01
-9.96593833e-01 1.53211057e-01 8.44994962e-01 -1.81421503e-01
-1.16219079e+00 3.71791780e-01 8.86462554e-02 -4.39447224e-01
-1.47707537e-01 2.79837102e-01 -1.17554612e-01 -1.17136203e-02
7.41451263e-01 5.33845946e-02 2.78556406e-01 -6.22005999e-01
1.36206433e-01 1.21807504e+00 2.34225124e-01 -3.86436224e-01
5.53741753e-01 8.14824820e-01 2.54538544e-02 -3.43611360e-01
-9.38573718e-01 -7.10382879e-01 -3.14334869e-01 -8.14078152e-02
6.04449868e-01 -1.20503497e+00 -5.04106820e-01 5.52573383e-01
-5.77392459e-01 -2.38265231e-01 -6.76081657e-01 5.72499394e-01
-4.94724393e-01 1.81262791e-01 -5.37205994e-01 -6.11878574e-01
-9.68838274e-01 -1.04629922e+00 9.42522287e-01 2.26952359e-01
2.12598845e-01 -7.87670553e-01 1.30702451e-01 6.26551986e-01
7.25303829e-01 3.08438480e-01 8.57158720e-01 -1.02043891e+00
-2.13957235e-01 -6.38815993e-03 -2.88128376e-01 5.18747091e-01
7.19490051e-01 -8.11723471e-01 -1.05574441e+00 -5.23996353e-01
2.62321144e-01 -7.56127715e-01 7.00403988e-01 1.70658603e-01
1.01301539e+00 -3.09473783e-01 -3.45348686e-01 5.05437016e-01
1.68046260e+00 1.33513853e-01 2.44276151e-01 3.54741246e-01
4.94890124e-01 4.56671774e-01 5.20782471e-01 6.40064418e-01
3.42497021e-01 4.51052904e-01 5.91401339e-01 -3.74284536e-01
-5.51360607e-01 -2.80603692e-02 -1.79190174e-01 8.36113632e-01
4.15084958e-01 -9.25157517e-02 -8.60175252e-01 7.99876571e-01
-1.97621775e+00 -6.76734507e-01 1.80483684e-01 2.07527614e+00
1.01853323e+00 -2.09357679e-01 2.40246490e-01 7.97459036e-02
7.22911954e-01 -2.22780555e-01 -1.05767500e+00 -5.46376742e-02
-1.54821277e-01 2.53286988e-01 6.30102873e-01 -1.48485214e-01
-1.07816291e+00 5.49857080e-01 4.91963768e+00 7.82539487e-01
-1.26998866e+00 8.02059054e-01 7.99667239e-01 -1.48370564e-01
-2.46181972e-02 -2.20751032e-01 -4.20798302e-01 7.27123082e-01
8.93837214e-01 -7.32958987e-02 -4.30447422e-02 7.71785378e-01
9.90832895e-02 7.29953647e-02 -6.89578891e-01 1.14612758e+00
-8.84146318e-02 -1.41186464e+00 -2.62588680e-01 1.23240784e-01
9.36925530e-01 6.05961740e-01 -7.15648681e-02 2.52576191e-02
4.25471544e-01 -6.48002923e-01 2.64070034e-01 2.71369845e-01
9.33045089e-01 -6.45565271e-01 9.99590993e-01 4.93704200e-01
-7.93305159e-01 -3.92902613e-01 -1.76767603e-01 3.98532242e-01
-1.20480157e-01 1.03070402e+00 -8.86676669e-01 6.97015047e-01
7.55251527e-01 7.22320914e-01 -4.11382258e-01 1.22325706e+00
2.13818207e-01 7.85341561e-01 -2.67630488e-01 3.78529400e-01
8.53447616e-02 2.24148467e-01 2.40294322e-01 9.04798746e-01
-4.54374850e-02 -3.71256191e-03 1.89538747e-01 3.60543579e-01
-3.89714718e-01 4.66144949e-01 -3.57722849e-01 2.53728598e-01
4.87000942e-01 1.53332758e+00 -6.22618437e-01 -3.49923104e-01
-4.94802386e-01 7.39404261e-01 3.19463581e-01 -6.84034601e-02
-6.05026841e-01 -1.57947913e-01 1.35793298e-01 8.42956901e-02
3.11292440e-01 3.60612005e-01 -2.35908199e-02 -1.15950847e+00
-8.31434354e-02 -9.86973464e-01 7.05695331e-01 -2.53902853e-01
-1.60679543e+00 9.62281346e-01 -3.54106605e-01 -1.45640743e+00
-1.60899997e-01 1.73328053e-02 -2.97348559e-01 4.60003227e-01
-1.76665258e+00 -1.47020853e+00 -4.84762311e-01 9.66670871e-01
3.79019976e-01 -5.54604709e-01 9.41104352e-01 7.20783532e-01
-4.75996554e-01 7.79645026e-01 3.33007365e-01 5.46968393e-02
9.92522836e-01 -1.15906000e+00 -2.88998514e-01 5.42386353e-01
-5.02785109e-02 3.38627577e-01 2.24130884e-01 -6.45193338e-01
-1.20277941e+00 -1.33035779e+00 6.79804862e-01 8.07743445e-02
3.72246623e-01 -1.20764226e-01 -5.13143897e-01 3.86305630e-01
3.16162288e-01 8.25045705e-01 1.10571527e+00 -1.57086328e-01
-3.27221841e-01 -6.67992711e-01 -1.73745787e+00 -4.99560349e-02
8.83305907e-01 -1.62775204e-01 -1.27557844e-01 6.78823709e-01
6.10171020e-01 -4.84062396e-02 -1.16207874e+00 4.96282667e-01
3.65268230e-01 -8.78934562e-01 4.13752168e-01 -4.78828251e-01
2.91697443e-01 -2.72741437e-01 -2.05545723e-01 -1.23772466e+00
-1.91390097e-01 -7.90614247e-01 -6.93423301e-02 1.25630987e+00
2.22456127e-01 -8.06379974e-01 1.04717731e+00 2.93150336e-01
-1.27845317e-01 -9.31839824e-01 -1.01793933e+00 -7.99226701e-01
1.88083891e-02 1.87226664e-02 5.37715554e-01 1.05169439e+00
-4.61798385e-02 3.89109910e-01 -5.14020026e-01 -3.04865748e-01
1.08314562e+00 1.74001917e-01 5.28818965e-01 -1.18870020e+00
-2.84423441e-01 2.82183588e-01 -3.59893501e-01 -4.64372873e-01
-1.18565224e-02 -1.00430524e+00 -2.54316151e-01 -1.39998662e+00
4.66279507e-01 -9.83276725e-01 -8.00998509e-01 9.09127235e-01
6.88504800e-02 3.46259445e-01 1.58138037e-01 4.71438676e-01
-8.82110238e-01 4.49786484e-01 1.13813019e+00 -2.90698290e-01
4.83864062e-02 -3.68219875e-02 -7.21677065e-01 3.56026202e-01
9.90961969e-01 -8.28931570e-01 -6.83379114e-01 -5.18471956e-01
-2.32414827e-01 9.67447162e-02 1.67632356e-01 -1.19307482e+00
3.58056247e-01 2.91796118e-01 1.60922036e-01 -3.22266936e-01
-1.67818174e-01 -1.17709136e+00 1.28933847e-01 7.40322411e-01
-2.94962674e-01 -1.83289722e-01 -1.06320627e-01 5.91301501e-01
-2.40976200e-01 2.81053549e-03 9.50032771e-01 -1.90604731e-01
-3.86556715e-01 5.33248305e-01 1.08596787e-01 1.16086654e-01
1.15295529e+00 9.45620909e-02 -4.89675552e-01 -1.62499338e-01
-6.91648662e-01 3.71142596e-01 2.30707034e-01 2.83865482e-01
2.54864782e-01 -1.18696332e+00 -8.15629423e-01 7.52142370e-02
2.36465156e-01 -1.75848324e-02 5.57700455e-01 1.16415012e+00
-2.18792930e-01 3.53823066e-01 -3.58120710e-01 -7.24282324e-01
-1.13067174e+00 5.57592332e-01 3.88169706e-01 -6.08839333e-01
-7.23603189e-01 6.16656542e-01 3.52177694e-02 -3.80336642e-01
3.35333496e-01 9.35924873e-02 -8.14381093e-02 -2.61824690e-02
4.01570350e-01 3.14221889e-01 4.80242670e-01 -6.27190351e-01
-5.65326512e-01 2.63591170e-01 -3.05123448e-01 2.80757129e-01
1.50181854e+00 -1.05378263e-01 -1.23658981e-02 2.69600362e-01
1.43417287e+00 -1.70615196e-01 -1.09634578e+00 -6.67727828e-01
-1.44581556e-01 -1.28753096e-01 2.57790267e-01 -1.24551404e+00
-1.55974352e+00 4.69017923e-01 1.16901004e+00 -1.03869639e-01
1.39834976e+00 -1.70450900e-02 1.15783548e+00 3.38171050e-02
5.81574261e-01 -8.86730373e-01 -9.36240982e-03 -7.77329877e-02
2.30862007e-01 -1.49650443e+00 -1.00661479e-01 -2.04176441e-01
-6.84708655e-01 7.03870416e-01 4.75807667e-01 4.75141816e-02
8.35068107e-01 4.42161858e-01 3.44635487e-01 -3.12522322e-01
-9.03979361e-01 -1.75730601e-01 -1.41648874e-01 4.41527128e-01
3.48642051e-01 -6.04617223e-02 -4.76208001e-01 6.51123285e-01
2.66062409e-01 5.97274721e-01 1.32102355e-01 1.13370860e+00
-1.18768997e-01 -1.34786260e+00 -5.67309335e-02 7.74485409e-01
-9.27082002e-01 -8.31021182e-03 1.06145196e-01 3.32022756e-01
5.46846151e-01 1.09143591e+00 -1.89505041e-01 -3.96021873e-01
1.21330068e-01 -4.40425240e-03 2.28600889e-01 -6.42576158e-01
-1.01900387e+00 1.29716024e-01 -1.04728736e-01 -5.76351285e-01
-7.91079521e-01 -5.64490199e-01 -1.33062959e+00 3.42770994e-01
-4.07179773e-01 3.55218381e-01 8.93923461e-01 8.20889175e-01
7.62882590e-01 5.72753549e-01 1.07422745e+00 5.78405671e-02
-8.07411194e-01 -8.76096308e-01 -6.05178237e-01 5.93942404e-01
3.58073711e-01 -5.59726954e-01 -1.89192191e-01 -2.88126096e-02] | [6.042026996612549, 6.451242446899414] |
b6246a67-91f4-495c-b4dd-cbc5778a8852 | a-self-organising-eigenspace-map-for-time | 1905.05540 | null | https://arxiv.org/abs/1905.05540v1 | https://arxiv.org/pdf/1905.05540v1.pdf | A self-organising eigenspace map for time series clustering | This paper presents a novel time series clustering method, the self-organising eigenspace map (SOEM), based on a generalisation of the well-known self-organising feature map (SOFM). The SOEM operates on the eigenspaces of the embedded covariance structures of time series which are related directly to modes in those time series. Approximate joint diagonalisation acts as a pseudo-metric across these spaces allowing us to generalise the SOFM to a neural network with matrix input. The technique is empirically validated against three sets of experiments; univariate and multivariate time series clustering, and application to (clustered) multi-variate time series forecasting. Results indicate that the technique performs a valid topologically ordered clustering of the time series. The clustering is superior in comparison to standard benchmarks when the data is non-aligned, gives the best clustering stage for when used in forecasting, and can be used with partial/non-overlapping time series, multivariate clustering and produces a topological representation of the time series objects. | ['Jacek Brodzki', 'Donya Rahmani', 'Damien Fay'] | 2019-05-14 | null | null | null | null | ['time-series-clustering'] | ['time-series'] | [-7.08539262e-02 -3.66096795e-01 3.87712657e-01 -3.08410794e-01
6.82749078e-02 -7.87361503e-01 9.28818762e-01 2.12356776e-01
-7.90474340e-02 9.72254649e-02 4.47487235e-01 -4.85330611e-01
-1.05580676e+00 -5.70050299e-01 -3.83796878e-02 -9.00711417e-01
-1.15249348e+00 6.78939939e-01 1.61062673e-01 -4.19426560e-01
3.64930302e-01 6.92333639e-01 -1.98833978e+00 4.33998168e-01
4.55493361e-01 1.07088947e+00 -4.08536494e-02 7.95793116e-01
-2.25878045e-01 3.51618350e-01 -5.08128643e-01 5.46010077e-01
5.22416890e-01 -2.92418778e-01 -6.75388575e-01 1.16470642e-01
-1.37728781e-01 4.40698683e-01 1.84091240e-01 7.32725859e-01
4.53460664e-02 4.82142836e-01 1.04354775e+00 -1.51591563e+00
-2.46012971e-01 3.57144177e-01 -3.27614129e-01 4.96895909e-01
3.24202299e-01 -5.85193217e-01 7.44125247e-01 -6.02888227e-01
7.60387838e-01 1.27531230e+00 8.93000305e-01 -1.87586576e-01
-1.65301013e+00 -7.09150732e-02 -4.52630103e-01 3.67100179e-01
-1.19774568e+00 -7.66459703e-02 9.96441901e-01 -9.53023374e-01
1.16901791e+00 8.29632938e-01 5.56130290e-01 3.73350710e-01
5.48624575e-01 1.34803727e-01 1.48184001e+00 -5.50835609e-01
5.77176452e-01 -2.44002178e-01 2.80348331e-01 -1.97673868e-02
-2.08454058e-01 1.98217303e-01 4.65663709e-02 -5.53344429e-01
4.64428365e-01 3.35223347e-01 4.58474085e-02 -7.97795415e-01
-1.47710705e+00 7.13636458e-01 2.48265207e-01 1.02439857e+00
-5.79338133e-01 -3.93467516e-01 6.61489189e-01 9.54197586e-01
5.94043672e-01 4.17856336e-01 -4.42756295e-01 -3.34423244e-01
-1.08319783e+00 -6.69274153e-03 9.13648129e-01 2.29348660e-01
5.77693403e-01 7.79952928e-02 3.36568207e-01 3.72361183e-01
1.71258613e-01 3.07359725e-01 9.21728253e-01 -7.09978700e-01
2.40988448e-01 9.64467049e-01 -2.97437161e-01 -1.64589477e+00
-9.59595740e-01 1.89440977e-02 -1.17077971e+00 2.81058967e-01
2.19945669e-01 1.15109876e-01 -6.88006818e-01 1.17327619e+00
2.22512662e-01 1.72177013e-02 4.30357978e-02 6.52357697e-01
2.39415616e-01 1.03137231e+00 -4.49722916e-01 -7.25222230e-01
9.82731700e-01 -3.02316695e-01 -6.94556892e-01 6.33560240e-01
5.78844368e-01 -7.18860388e-01 5.92260480e-01 3.18320036e-01
-6.16315961e-01 -5.74404299e-01 -9.46648777e-01 6.85872197e-01
-8.84432077e-01 -2.54972517e-01 2.65279651e-01 4.88920122e-01
-1.44189131e+00 1.02443647e+00 -1.05204165e+00 -6.13618195e-01
-5.44480801e-01 4.72439557e-01 -6.84250593e-01 5.84492266e-01
-9.73976791e-01 6.82644427e-01 8.17051113e-01 6.96819741e-03
-5.68138249e-02 -2.85147518e-01 -5.72804928e-01 -5.68085536e-02
-3.90171856e-01 1.71499630e-03 4.30114806e-01 -9.91060138e-01
-1.31260967e+00 6.15121186e-01 -2.11216658e-02 -4.03202355e-01
1.98765010e-01 5.05440295e-01 -9.71030056e-01 1.40997097e-01
1.15980273e-02 2.85825551e-01 1.05632424e+00 -8.60688984e-01
-3.22813392e-01 -3.65843922e-01 -6.85334265e-01 -1.96471393e-01
-5.65009356e-01 1.82110161e-01 4.26889807e-01 -8.25602531e-01
7.05832601e-01 -1.13054800e+00 -2.86935300e-01 -8.97655308e-01
-2.36481681e-01 -4.14117455e-01 1.44490707e+00 -6.47760332e-01
1.64353096e+00 -2.42676139e+00 5.06543875e-01 9.99550164e-01
1.87950283e-01 -3.05290073e-01 1.25940159e-01 1.28793132e+00
-9.33450937e-01 -6.27244636e-02 -4.90104675e-01 1.86534062e-01
7.63938054e-02 2.13451073e-01 -4.24504697e-01 7.05770552e-01
-2.15616867e-01 4.59328651e-01 -5.77975333e-01 -3.32040608e-01
4.17159677e-01 1.89688087e-01 -7.95394927e-03 -4.28415164e-02
3.29314679e-01 3.55188698e-01 4.36268449e-02 1.13900729e-01
5.13127148e-01 -1.32141352e-01 6.47745654e-02 -1.00557104e-01
-5.69367111e-01 -9.03208032e-02 -1.46081221e+00 1.06283855e+00
2.21808165e-01 9.59688783e-01 -2.61257917e-01 -1.25450361e+00
1.39714527e+00 4.75894213e-01 1.10937130e+00 -3.13670456e-01
8.82415697e-02 1.29942954e-01 2.62855351e-01 -3.95534933e-01
3.55942875e-01 3.74911539e-02 -1.96801387e-02 7.57081211e-01
5.36703430e-02 1.14625558e-01 4.27247733e-01 -6.70861676e-02
1.09047925e+00 -2.07543567e-01 3.82224679e-01 -1.13194251e+00
7.19374418e-01 -9.91722289e-03 5.89956678e-02 5.14151342e-02
-7.90468138e-03 4.29626971e-01 5.22056997e-01 -9.40831304e-01
-1.27007127e+00 -1.02188861e+00 -4.14223075e-01 9.81466055e-01
-4.31838095e-01 -6.43479288e-01 -5.54862738e-01 -5.60514107e-02
-4.64973710e-02 3.28606516e-01 -8.85666728e-01 1.53588131e-01
-3.91742736e-01 -6.53484821e-01 1.09173963e-02 6.85496554e-02
-1.46110818e-01 -1.10655975e+00 -8.27359021e-01 4.02143359e-01
-1.06849419e-02 -3.93111289e-01 -1.45347655e-01 5.30992150e-01
-1.41672671e+00 -7.63120174e-01 -5.42136312e-01 -7.50514627e-01
5.30889511e-01 1.38885677e-01 9.51985061e-01 -3.32358450e-01
-2.17071250e-01 5.90480447e-01 -5.32729745e-01 -3.69384795e-01
-4.82345432e-01 -8.27143043e-02 4.70726758e-01 3.37339282e-01
5.80917597e-01 -1.13101673e+00 -3.74263823e-01 7.54755259e-01
-1.11202753e+00 -3.04417610e-01 -6.16773847e-04 5.36121845e-01
3.22287023e-01 8.41368973e-01 2.82061279e-01 -1.19877063e-01
9.38829422e-01 -5.67573428e-01 -6.68926895e-01 1.23639725e-01
-6.73831403e-01 9.04333815e-02 7.83426523e-01 -4.01313066e-01
-2.10882574e-01 1.68941930e-01 4.86676157e-01 -5.41799963e-01
-4.15539682e-01 5.76900840e-01 5.40248990e-01 -1.37966320e-01
8.80747139e-01 4.16342407e-01 2.23399520e-01 -6.49498701e-01
2.18834683e-01 7.28618503e-01 3.95924509e-01 -2.73037434e-01
9.08444881e-01 7.63152480e-01 2.90725827e-01 -1.01686394e+00
4.19671953e-01 -9.64339197e-01 -1.24345887e+00 -4.97368217e-01
8.29234421e-01 -2.26101026e-01 -5.67508817e-01 1.37660041e-01
-7.64172733e-01 9.74551141e-02 -3.55009258e-01 6.69400871e-01
-5.59440017e-01 4.29008514e-01 -2.89336830e-01 -1.00455570e+00
-5.44438697e-02 -7.13840663e-01 8.78555596e-01 -2.84375817e-01
-5.20524025e-01 -1.49864769e+00 6.97879076e-01 -4.46554840e-01
4.75595862e-01 9.32706594e-01 1.01968861e+00 -9.49893713e-01
2.59334624e-01 -4.00193959e-01 3.69006604e-01 1.79060370e-01
1.53954327e-01 4.25658822e-01 -6.94637537e-01 -4.80378360e-01
2.63881534e-01 3.79598141e-01 5.47273993e-01 4.64232206e-01
7.01497734e-01 -2.10776135e-01 -3.40044558e-01 3.47146183e-01
1.22778726e+00 7.05183506e-01 4.08648014e-01 7.71126926e-01
4.09814209e-01 1.25932789e+00 1.83701336e-01 5.75916231e-01
2.38474578e-01 4.84506398e-01 1.26929149e-01 -8.07889923e-02
8.54808331e-01 3.19843769e-01 3.93841356e-01 1.64147604e+00
-4.25123036e-01 3.95052195e-01 -1.48526251e+00 5.27683675e-01
-2.03207135e+00 -1.22631598e+00 -6.12591445e-01 2.05641508e+00
3.52878690e-01 4.09112945e-02 6.26512170e-01 7.78424740e-01
8.79591405e-01 1.79892272e-01 -3.77193233e-03 -7.20247567e-01
-2.48015389e-01 -3.36228125e-02 2.91305274e-01 1.49913162e-01
-1.49315608e+00 4.39763591e-02 7.11309862e+00 5.50800562e-01
-1.16944468e+00 -8.60544294e-02 3.45200777e-01 3.39677095e-01
-3.16209942e-02 -1.37481660e-01 6.23672828e-02 5.30295670e-01
1.48760200e+00 -4.57444727e-01 5.23996770e-01 7.01909781e-01
3.04166138e-01 2.78368324e-01 -1.06025004e+00 1.13590658e+00
-9.15683135e-02 -1.07818139e+00 -3.52434844e-01 2.77432442e-01
8.45775843e-01 -9.16530786e-04 4.68590930e-02 2.43238136e-02
1.00076776e-02 -8.92288864e-01 5.20252287e-01 5.85732996e-01
3.61050934e-01 -7.66976535e-01 7.25368619e-01 3.27906132e-01
-1.65537012e+00 -2.86971450e-01 -2.19831705e-01 -3.99243742e-01
2.11515591e-01 5.06649375e-01 -6.30873501e-01 7.49046385e-01
1.07639265e+00 8.84366512e-01 -6.37294292e-01 1.03406453e+00
8.11863959e-01 4.36808884e-01 -7.33403623e-01 7.86584243e-02
5.80698431e-01 -6.77707136e-01 7.76020825e-01 1.25010788e+00
5.02396107e-01 -7.69347474e-02 1.76852360e-01 3.69555652e-01
9.15528476e-01 4.25418437e-01 -8.13500285e-01 -3.13266397e-01
2.00108096e-01 1.30782187e+00 -1.45038414e+00 -2.61082202e-01
2.83474643e-02 4.85141039e-01 -3.29254150e-01 3.36687505e-01
-2.18271479e-01 -5.83554149e-01 3.75518650e-01 4.20626290e-02
3.73268187e-01 -3.83237451e-01 -1.56554312e-01 -9.17912185e-01
4.81726676e-02 -6.57839358e-01 6.99197173e-01 -7.26764798e-01
-1.50813103e+00 1.11934018e+00 4.47356135e-01 -1.94589412e+00
-9.16286051e-01 -7.28925824e-01 -7.02594936e-01 6.61279798e-01
-5.34894288e-01 -4.45107400e-01 -6.57797605e-03 8.92170131e-01
-1.13036726e-02 -5.73374093e-01 1.13938236e+00 5.57089485e-02
-9.09771249e-02 -2.20745116e-01 9.11757112e-01 -2.30478961e-02
2.13838816e-01 -1.68219435e+00 4.67675954e-01 6.37626588e-01
2.09149808e-01 6.24528825e-01 8.91879916e-01 -5.01861036e-01
-1.17052925e+00 -7.88573623e-01 8.55803192e-01 -5.59345663e-01
1.14494622e+00 -4.72975105e-01 -9.60789442e-01 3.23629409e-01
3.12054336e-01 -3.29050869e-01 9.39869404e-01 5.77404238e-02
-2.32313544e-01 -1.70265719e-01 -8.21611643e-01 3.84370387e-01
5.35577655e-01 -6.74573600e-01 -1.14895809e+00 5.37794530e-01
3.66163611e-01 1.02776870e-01 -1.41427422e+00 2.12590501e-01
4.50689495e-01 -1.29921865e+00 9.75225925e-01 -5.18986583e-01
9.28636566e-02 -5.69794059e-01 -1.57966763e-01 -1.42214644e+00
-8.20194602e-01 -1.02690482e+00 8.48129392e-02 9.02918100e-01
1.61947578e-01 -9.01183069e-01 2.78562039e-01 1.54880524e-01
1.01179697e-01 -3.54121238e-01 -1.29797769e+00 -1.25210571e+00
-2.75600612e-01 -3.83547306e-01 7.88844287e-01 1.56272101e+00
4.56063837e-01 -6.58644736e-03 -2.58125621e-03 -1.59096316e-01
5.58141112e-01 3.15845460e-01 3.63495320e-01 -1.93930769e+00
-2.88546626e-02 -8.45184624e-01 -1.19297874e+00 -1.14033692e-01
-4.52652425e-02 -9.55945373e-01 -2.45165482e-01 -1.09410584e+00
-5.12497008e-01 -1.29390210e-01 -4.57809597e-01 2.18521222e-01
6.24146700e-01 8.66432190e-02 1.80415630e-01 8.48996520e-01
-3.71581823e-01 1.61007226e-01 6.14714980e-01 1.58628255e-01
-6.12440705e-01 6.42405124e-03 2.17574462e-01 4.53451574e-01
7.32283652e-01 -2.95967162e-01 -3.79095912e-01 3.56290191e-01
1.18121527e-01 1.94855899e-01 1.39184624e-01 -1.17487919e+00
3.21496993e-01 3.63545083e-02 2.35339269e-01 -1.01866746e+00
-1.55689716e-01 -1.43963242e+00 9.20777082e-01 4.04796809e-01
-1.68053985e-01 9.88365710e-01 2.29492024e-01 4.19486254e-01
-5.83024204e-01 1.24703094e-01 3.91947925e-01 1.64482102e-01
-8.92478704e-01 -1.04978137e-01 -5.42296231e-01 -4.13108468e-01
1.15934002e+00 -7.74248600e-01 4.75355536e-02 -4.45864737e-01
-1.05553257e+00 -4.76470701e-02 3.85193765e-01 5.58591425e-01
2.47408658e-01 -1.78404903e+00 -5.49844086e-01 5.07742703e-01
8.14116597e-02 -6.57745302e-01 1.61815658e-01 1.18116260e+00
-7.68238306e-01 7.31028259e-01 -6.90479338e-01 -1.23827994e+00
-1.25843275e+00 9.86055851e-01 1.41862169e-01 -1.58998042e-01
-8.05433691e-01 6.91121593e-02 -2.15488032e-01 -6.52769685e-01
1.53625086e-01 -5.29600620e-01 -7.15411365e-01 4.94209796e-01
5.46719313e-01 7.22087324e-01 2.09605530e-01 -1.16769326e+00
-5.35641491e-01 8.46571326e-01 5.48725069e-01 -2.92037606e-01
1.72688639e+00 -1.48593917e-01 -1.02623308e+00 1.25320637e+00
1.60422218e+00 -4.56044972e-01 -7.76715755e-01 5.52913807e-02
7.64060378e-01 -2.03806698e-01 -1.72894076e-01 -2.42772043e-01
-5.09329259e-01 7.26401150e-01 8.96514833e-01 1.60088205e+00
1.43338752e+00 -2.45129496e-01 6.93932027e-02 5.36975384e-01
1.96175516e-01 -1.14093924e+00 -3.98077548e-01 5.93107820e-01
9.47938919e-01 -7.10922837e-01 -1.25549719e-01 6.08903579e-02
-3.64579201e-01 1.74820006e+00 -2.62105197e-01 -3.55889469e-01
1.21291828e+00 2.20620751e-01 1.87087491e-01 -4.06635165e-01
-6.34950280e-01 7.81411678e-02 8.73577952e-01 6.00299418e-01
4.09150809e-01 3.63610566e-01 -1.66211322e-01 -9.26260129e-02
-6.86918259e-01 -5.19968331e-01 1.57758728e-01 8.56374621e-01
-4.13461626e-01 -8.45809937e-01 -1.01065230e+00 5.34127772e-01
7.21662724e-03 2.78227776e-01 -4.97280151e-01 8.45793366e-01
-9.19077918e-02 8.76495063e-01 4.96575296e-01 -7.55920470e-01
3.23641032e-01 5.19629419e-01 -1.88981920e-01 2.02429872e-02
-7.48555601e-01 4.29295063e-01 -4.22830313e-01 -6.76959872e-01
-7.44141757e-01 -9.80954289e-01 -8.97733748e-01 -4.64779973e-01
-2.78751249e-03 5.07334232e-01 8.52988780e-01 7.70627975e-01
4.00005758e-01 3.46817523e-01 1.10919201e+00 -1.25826621e+00
-1.38375297e-01 -1.00042081e+00 -9.09699798e-01 6.99317694e-01
5.11718810e-01 -6.50380909e-01 -6.44260287e-01 3.10146779e-01] | [7.262866973876953, 3.34808087348938] |
7808fad3-b175-49d9-9664-751ee2ceb6ae | it-is-ai-s-turn-to-ask-human-a-question | 2109.03423 | null | https://arxiv.org/abs/2109.03423v4 | https://arxiv.org/pdf/2109.03423v4.pdf | It is AI's Turn to Ask Humans a Question: Question-Answer Pair Generation for Children's Story Books | Existing question answering (QA) techniques are created mainly to answer questions asked by humans. But in educational applications, teachers often need to decide what questions they should ask, in order to help students to improve their narrative understanding capabilities. We design an automated question-answer generation (QAG) system for this education scenario: given a story book at the kindergarten to eighth-grade level as input, our system can automatically generate QA pairs that are capable of testing a variety of dimensions of a student's comprehension skills. Our proposed QAG model architecture is demonstrated using a new expert-annotated FairytaleQA dataset, which has 278 child-friendly storybooks with 10,580 QA pairs. Automatic and human evaluations show that our model outperforms state-of-the-art QAG baseline systems. On top of our QAG system, we also start to build an interactive story-telling application for the future real-world deployment in this educational scenario. | ['Zheng Zhang', 'Ying Xu', 'Mo Yu', 'Toby Jia-Jun Li', 'Tongshuang Wu', 'Dakuo Wang', 'Bingsheng Yao'] | 2021-09-08 | null | null | null | null | ['question-answer-generation'] | ['natural-language-processing'] | [-1.92936473e-02 5.94551980e-01 4.15966958e-01 -5.62893748e-01
-1.22754943e+00 -1.02655172e+00 4.57225621e-01 3.79068315e-01
1.30438954e-01 7.14750767e-01 5.22237599e-01 -7.70848811e-01
-9.66145918e-02 -1.37572646e+00 -5.50203979e-01 2.07421795e-01
4.50721145e-01 1.03404927e+00 8.96577179e-01 -8.93952489e-01
2.95741320e-01 -6.39278516e-02 -1.89268875e+00 8.59616637e-01
1.43924260e+00 4.60167170e-01 9.91409719e-02 1.27151096e+00
-4.80317265e-01 1.83716428e+00 -1.03506887e+00 -8.89442801e-01
-2.98708618e-01 -8.77504647e-01 -1.71367824e+00 -3.02681625e-01
8.55777979e-01 -7.22873390e-01 -1.91502366e-02 6.44489408e-01
5.19884527e-01 3.60988259e-01 4.94084090e-01 -1.12214303e+00
-7.36905634e-01 8.21386278e-01 2.94582903e-01 1.67630594e-02
1.19884741e+00 2.09953889e-01 1.01832020e+00 -5.65537214e-01
6.18280351e-01 9.71936166e-01 2.68227041e-01 9.02111411e-01
-6.74084783e-01 -3.65936667e-01 -2.44331345e-01 8.19631100e-01
-7.67304420e-01 -1.11812316e-01 7.37340629e-01 -5.12074172e-01
6.76663220e-01 5.16307592e-01 1.07005620e+00 7.73390889e-01
-4.73974571e-02 1.15769422e+00 9.41078067e-01 -4.96586561e-01
1.86099127e-01 3.31191383e-02 4.14417565e-01 8.70331347e-01
-2.28205994e-01 -5.91813803e-01 -7.22703934e-01 1.35583967e-01
5.39473176e-01 -7.68626392e-01 -3.08598131e-01 -2.01737389e-01
-8.46507847e-01 1.05390096e+00 -1.03565723e-01 1.41354308e-01
1.53188363e-01 -7.67091960e-02 2.41221309e-01 7.85341859e-01
-8.69577453e-02 1.16257203e+00 -3.06709528e-01 -7.71460056e-01
-7.26766586e-01 1.02037930e+00 1.31363678e+00 1.05054176e+00
-8.49643201e-02 -3.38245034e-01 -6.51131213e-01 8.15968335e-01
2.21445411e-01 9.96405408e-02 4.75705415e-01 -1.43648934e+00
6.04646027e-01 1.04262388e+00 1.69607744e-01 -6.20273292e-01
-1.79475442e-01 -1.99522048e-01 -8.32061768e-02 8.83332938e-02
8.62573326e-01 -1.51208565e-01 -4.21680629e-01 1.34660053e+00
5.82877815e-01 2.37019733e-01 2.80305207e-01 7.38522410e-01
2.00352597e+00 7.03050792e-01 2.71332208e-02 2.36820325e-01
1.86311305e+00 -1.26301813e+00 -7.39072740e-01 -3.17724943e-02
9.75579202e-01 -6.76410854e-01 1.64735174e+00 8.18789661e-01
-1.73154330e+00 -4.99695927e-01 -9.50093806e-01 -5.08697808e-01
-1.10529386e-01 -6.11704914e-03 1.80693999e-01 9.84332263e-01
-8.99872780e-01 1.12312846e-01 -3.75090092e-01 -1.01084076e-01
3.02428246e-01 -1.84226319e-01 -4.49880511e-02 -4.82954472e-01
-1.20614433e+00 8.96007359e-01 5.67198992e-02 -5.53473234e-01
-1.19803607e+00 -1.12981415e+00 -9.95103598e-01 4.54483926e-01
4.35120225e-01 -8.29282165e-01 2.11736679e+00 -2.11725637e-01
-1.98356199e+00 9.27763879e-01 2.62150526e-01 -1.86624974e-01
5.25264382e-01 -5.26685953e-01 -2.23381538e-02 4.67044771e-01
1.96482733e-01 7.32335508e-01 -7.89168552e-02 -6.85700238e-01
-4.45214480e-01 -4.33079042e-02 9.70179319e-01 5.87713361e-01
-1.36355147e-01 1.06253829e-02 -3.91818322e-02 -4.33051318e-01
2.92292506e-01 -2.91727096e-01 -2.70992592e-02 -3.75202358e-01
8.86513367e-02 -7.87078202e-01 5.39207637e-01 -8.15916240e-01
1.35709202e+00 -1.37416840e+00 -2.58573622e-01 -3.27295363e-01
1.44597173e-01 3.76402587e-01 -2.44615272e-01 7.38547444e-01
5.87509610e-02 -2.06961527e-01 5.82612492e-02 2.66068662e-03
1.25397027e-01 -4.71972786e-02 -4.21695024e-01 -3.34553838e-01
1.97826877e-01 9.52570856e-01 -1.47669423e+00 -7.74800718e-01
2.25605950e-01 -2.55964965e-01 -7.41422832e-01 1.04048586e+00
-6.76834822e-01 4.55571949e-01 -5.13834298e-01 4.45414603e-01
1.95762500e-01 -4.02328037e-02 -3.02983522e-01 7.76664615e-01
2.51834035e-01 1.10373902e+00 -9.69933093e-01 1.96847868e+00
-6.10425055e-01 7.57574141e-01 -2.39947662e-01 -6.87611997e-01
9.57995713e-01 5.99700451e-01 -7.78479651e-02 -6.89590991e-01
6.47564381e-02 -1.27820820e-02 8.21802467e-02 -9.80274081e-01
7.04794765e-01 1.71173349e-01 -2.50246912e-01 8.43301296e-01
3.85255665e-01 -1.24309886e+00 6.44900262e-01 6.13947392e-01
1.50167012e+00 3.24637175e-01 7.38145262e-02 -1.48515284e-01
8.21545899e-01 2.94499129e-01 -6.03384785e-02 8.46169651e-01
-1.20155111e-01 7.46678412e-01 7.00371385e-01 -4.05272722e-01
-4.06941026e-01 -1.13156688e+00 3.00406635e-01 1.42486668e+00
-2.15636417e-01 -5.82266927e-01 -1.13908100e+00 -8.69820714e-01
-8.34607542e-01 1.37956655e+00 -2.25272149e-01 5.12117520e-02
-6.99138463e-01 2.64714748e-01 7.73780227e-01 2.72133410e-01
7.84974575e-01 -1.19003379e+00 -1.08684587e+00 5.22791386e-01
-6.41615689e-01 -9.26070988e-01 -6.41948804e-02 -6.12721741e-01
-4.74467039e-01 -1.29414344e+00 -5.01239121e-01 -9.21380758e-01
3.55347157e-01 6.07316047e-02 1.83597970e+00 4.23094571e-01
1.11381158e-01 7.83223331e-01 -7.64332712e-01 -5.33651114e-01
-7.44070411e-01 1.95941627e-01 -6.81824684e-01 -9.13057923e-01
3.58416110e-01 -6.46653414e-01 -7.34012842e-01 2.04355836e-01
-7.38065600e-01 6.93306029e-01 -9.85523909e-02 6.44031823e-01
-3.52871679e-02 -1.99828744e-01 7.80995667e-01 -1.23413992e+00
1.09953785e+00 -3.69687468e-01 -6.31692171e-01 5.85653663e-01
-1.65917113e-01 -1.68473393e-01 5.87492585e-01 -3.19363117e-01
-1.36405373e+00 -3.93035412e-01 -8.73956144e-01 3.79945576e-01
-5.17610371e-01 4.49294180e-01 -3.10786784e-01 7.71062747e-02
9.30491388e-01 9.17989165e-02 -6.09925091e-01 -1.57422125e-01
6.84810162e-01 4.09066707e-01 9.85623479e-01 -1.25427246e+00
5.28176785e-01 -4.87700641e-01 -2.48697951e-01 -3.58859986e-01
-1.06956518e+00 -2.31904969e-01 -2.89118379e-01 -8.03653836e-01
7.35457182e-01 -7.63553560e-01 -1.21430433e+00 2.83430904e-01
-1.29368627e+00 -7.58365810e-01 -5.69730043e-01 1.15190044e-01
-7.70721912e-01 -6.12025931e-02 -7.48342812e-01 -6.25833452e-01
-2.27771759e-01 -1.11362553e+00 6.46518946e-01 8.04666042e-01
-5.96377552e-01 -9.23516750e-01 4.40125793e-01 1.39277613e+00
7.37002641e-02 3.92631218e-02 1.24911988e+00 -1.05874681e+00
-4.69637901e-01 8.95455405e-02 1.61549479e-01 1.16866380e-02
-5.13465941e-01 -8.99784267e-02 -7.95862734e-01 2.97980517e-01
-2.09489036e-02 -1.00776219e+00 1.43692687e-01 -4.55391034e-02
1.05672002e+00 -4.80637759e-01 4.08115566e-01 -1.06573917e-01
7.83620358e-01 2.50034720e-01 8.10351193e-01 2.45949686e-01
3.75476867e-01 1.01163113e+00 8.43044281e-01 5.90636320e-02
1.28059363e+00 3.70276392e-01 2.90365279e-01 4.52796608e-01
-4.39464033e-01 -6.52888596e-01 3.52320939e-01 9.42864656e-01
1.00217655e-01 -4.35137480e-01 -1.24853671e+00 1.17579913e+00
-1.61588407e+00 -1.00226855e+00 -7.89237916e-01 1.70158255e+00
1.24612212e+00 4.89011919e-03 6.95389956e-02 2.01948777e-01
-4.26683165e-02 -9.34966281e-02 -5.20750470e-02 -8.02451909e-01
4.57265675e-01 8.60471904e-01 -5.27578056e-01 5.74287593e-01
-4.39138025e-01 1.02599597e+00 6.15113449e+00 8.24306786e-01
-3.07800561e-01 1.03973731e-01 7.18878508e-01 2.10702538e-01
-6.28103733e-01 -1.92942306e-01 -5.31014800e-01 -4.12004516e-02
1.06762290e+00 -3.25768232e-01 1.26501873e-01 7.08442748e-01
-1.10072054e-01 -3.75672966e-01 -1.25231242e+00 5.07178724e-01
1.85624495e-01 -1.39914358e+00 -1.99456569e-02 -6.79524422e-01
9.83444870e-01 -7.34456003e-01 -2.27677330e-01 7.63677657e-01
9.59460437e-01 -1.26277792e+00 6.36411548e-01 4.12777603e-01
4.17037725e-01 -8.26758742e-01 8.21398675e-01 6.16891026e-01
-9.29292500e-01 3.87403890e-02 4.92298603e-02 -6.20251596e-01
4.77147065e-02 -7.39763603e-02 -1.48080230e+00 3.28645170e-01
6.64122581e-01 -1.20555378e-01 -1.00134909e+00 1.12253058e+00
-1.11936140e+00 1.19304371e+00 1.42267764e-01 -5.01044035e-01
1.50492311e-01 -9.24200565e-02 2.61650622e-01 7.14152813e-01
5.00674605e-01 1.17290556e+00 -1.39531776e-01 7.77823269e-01
-4.95748185e-02 3.22758526e-01 -3.56648684e-01 6.67070821e-02
6.97354376e-01 1.20598698e+00 -4.67232913e-01 -3.78649056e-01
-4.12273943e-01 7.05191433e-01 2.43828058e-01 8.28581229e-02
-4.69818890e-01 -5.91691613e-01 4.09830034e-01 3.06308836e-01
-2.07665503e-01 -1.04142763e-01 -5.06222248e-01 -8.69919240e-01
-9.90288332e-02 -1.35847795e+00 6.36810601e-01 -1.44310582e+00
-9.42507207e-01 2.30213925e-01 1.06615290e-01 -9.71941829e-01
-8.99276733e-01 -5.27843595e-01 -1.21584988e+00 6.87787950e-01
-9.16014135e-01 -1.16055548e+00 -5.33392787e-01 5.65332055e-01
1.07227194e+00 -7.48921558e-02 8.98806989e-01 -5.31322770e-02
1.36804879e-01 5.41293323e-01 -6.81816638e-01 2.33582616e-01
6.65918112e-01 -1.73046184e+00 6.48614883e-01 7.75300145e-01
2.41806537e-01 2.32928291e-01 1.03515768e+00 -4.83783185e-01
-1.09116054e+00 -6.63687348e-01 1.12505841e+00 -9.20052528e-01
6.22396827e-01 -1.28341734e-01 -1.14327729e+00 4.46129620e-01
6.83714628e-01 -6.55961573e-01 1.02283776e+00 -4.48639318e-02
-9.58447382e-02 2.94480056e-01 -1.21178412e+00 6.44262016e-01
9.43781912e-01 -3.49449039e-01 -1.42890239e+00 4.58083242e-01
9.71111834e-01 -1.03929865e+00 -1.03152978e+00 1.58636793e-01
3.23327333e-01 -1.00871420e+00 9.01714921e-01 -9.32584167e-01
1.21460199e+00 -2.67358571e-01 3.35384667e-01 -1.27820277e+00
2.75152236e-01 -4.69738454e-01 -1.17331751e-01 1.49995208e+00
2.82196194e-01 -8.27275589e-02 1.03093457e+00 9.98479426e-01
-5.47944069e-01 -8.90865088e-01 -8.37885439e-01 -4.79067564e-02
4.71028954e-01 -6.41445279e-01 1.12179172e+00 1.00718105e+00
4.32049125e-01 6.07968807e-01 2.48195529e-01 1.29556522e-01
3.51873636e-02 7.97749162e-02 1.02806234e+00 -1.17183542e+00
-3.64090353e-01 -4.90841687e-01 -1.51820779e-01 -1.18895769e+00
-6.33288473e-02 -4.95344102e-01 1.38836488e-01 -2.00190616e+00
-1.36807740e-01 -6.94325939e-02 5.82801521e-01 3.24466050e-01
-4.44999427e-01 3.45900208e-02 -1.73308910e-03 -7.44985878e-01
-9.41233933e-01 4.07355666e-01 1.79449117e+00 -5.55254035e-02
-8.32291618e-02 1.81053996e-01 -7.45562136e-01 7.94473827e-01
5.10432601e-01 -6.95572570e-02 -9.49762285e-01 -6.31144106e-01
6.72692776e-01 7.32869804e-01 1.81753770e-01 -1.09870458e+00
5.97098470e-01 -5.10699093e-01 5.77297509e-02 -7.29059637e-01
4.81277965e-02 -2.89111316e-01 -3.42150867e-01 9.40194577e-02
-5.79908669e-01 1.73062995e-01 7.34361038e-02 -4.03326929e-01
-4.47742164e-01 -9.98769581e-01 2.56315976e-01 -1.02713361e-01
-4.23413992e-01 -3.00206065e-01 -7.95960248e-01 5.58247685e-01
1.01345539e+00 -1.50202522e-02 -9.86417115e-01 -1.12949693e+00
-4.49656010e-01 9.76234794e-01 -5.08549809e-02 2.79037058e-01
8.87813807e-01 -1.31138194e+00 -1.11412835e+00 -1.57139510e-01
2.96855241e-01 4.96238172e-01 4.45051849e-01 -2.27775518e-02
-1.22234058e+00 4.61580843e-01 -2.06251740e-01 -2.22676992e-01
-1.56759894e+00 -9.82801169e-02 2.84843862e-01 -7.39962518e-01
-1.01636037e-01 1.43459678e+00 -1.39388248e-01 -8.01895618e-01
5.97844087e-02 -4.66663003e-01 -9.71567154e-01 -1.40993679e-02
9.69090998e-01 5.23639023e-01 2.78519452e-01 -9.32960138e-02
4.05278981e-01 1.22896535e-02 2.44288713e-01 -4.38301802e-01
1.10691094e+00 1.58208087e-01 2.00804994e-01 1.96312219e-01
3.77186000e-01 1.97654650e-01 -6.12640440e-01 8.79667997e-02
-4.85829376e-02 -3.74176890e-01 -4.97447550e-01 -1.27449143e+00
-3.86219680e-01 1.06493402e+00 1.59775361e-01 5.09662569e-01
1.24432814e+00 3.13533425e-01 8.10164094e-01 6.22203231e-01
4.62251723e-01 -7.60662079e-01 4.71622109e-01 8.51326525e-01
1.02831590e+00 -9.90917027e-01 -2.31473997e-01 -4.70649868e-01
-5.99123657e-01 1.17296946e+00 1.30230463e+00 1.70608476e-01
-2.27209225e-01 -2.57750124e-01 2.43563041e-01 -4.15216297e-01
-1.20669448e+00 -1.29346490e-01 4.31081831e-01 5.85924864e-01
9.33849454e-01 8.54116306e-02 -4.60568935e-01 1.24617922e+00
-1.30267203e+00 -1.26078382e-01 1.11898077e+00 9.62789774e-01
-6.08263254e-01 -1.27788615e+00 -6.33619845e-01 3.28839332e-01
-3.60766321e-01 -6.55725822e-02 -7.19414353e-01 5.99789560e-01
1.13759302e-02 1.58558726e+00 -9.14020240e-02 -1.21511467e-01
6.30033255e-01 3.75746727e-01 9.30324137e-01 -1.11811960e+00
-1.13105857e+00 -9.83818352e-01 6.58384502e-01 -4.60943878e-01
1.07072052e-02 -6.10329747e-01 -1.22334278e+00 -2.88935900e-01
-1.50979906e-01 6.11355305e-01 2.19598427e-01 1.08938134e+00
-2.18311757e-01 6.99938476e-01 2.56238449e-02 1.65260747e-01
-2.41544798e-01 -1.11569548e+00 8.17402005e-02 1.86083227e-01
-3.07830740e-02 -2.16072515e-01 2.73836941e-01 1.81894138e-01] | [11.56077766418457, 8.037120819091797] |
bcdff0db-4a5d-425e-9f6a-653b3cb44530 | assessment-of-the-local-tchebichef-moments | 1910.09758 | null | https://arxiv.org/abs/1910.09758v1 | https://arxiv.org/pdf/1910.09758v1.pdf | Assessment of the Local Tchebichef Moments Method for Texture Classification by Fine Tuning Extraction Parameters | In this paper we use machine learning to study the application of Local Tchebichef Moments (LTM) to the problem of texture classification. The original LTM method was proposed by Mukundan (2014). The LTM method can be used for texture analysis in many different ways, either using the moment values directly, or more simply creating a relationship between the moment values of different orders, producing a histogram similar to those of Local Binary Pattern (LBP) based methods. The original method was not fully tested with large datasets, and there are several parameters that should be characterised for performance. Among these parameters are the kernel size, the moment orders and the weights for each moment. We implemented the LTM method in a flexible way in order to allow for the modification of the parameters that can affect its performance. Using four subsets from the Outex dataset (a popular benchmark for texture analysis), we used Random Forests to create models and to classify texture images, recording the standard metrics for each classifier. We repeated the process using several variations of the LBP method for comparison. This allowed us to find the best combination of orders and weights for the LTM method for texture classification. | ['Teo Susnjak', 'Napoleon Reyes', 'Andre Barczak'] | 2019-10-22 | null | null | null | null | ['texture-classification'] | ['computer-vision'] | [ 1.60608724e-01 -4.51945662e-01 -1.91477254e-01 -3.41090590e-01
-2.70944178e-01 -4.27519709e-01 9.12839115e-01 5.07275701e-01
-5.61182141e-01 5.69174647e-01 -1.33648723e-01 -2.41288856e-01
-5.70372880e-01 -1.12830520e+00 -3.34261537e-01 -1.07754230e+00
-2.36799777e-01 5.02611518e-01 9.22084153e-01 -2.86951005e-01
7.73011506e-01 8.05110753e-01 -1.95486820e+00 6.64696872e-01
4.62870002e-01 1.28409505e+00 2.70541459e-01 5.61899245e-01
-2.49279886e-01 4.20274824e-01 -6.44451678e-01 -2.56328881e-01
1.24380246e-01 -3.62608433e-01 -9.68583643e-01 4.33842018e-02
2.21384734e-01 2.70910531e-01 4.51236188e-01 7.97121167e-01
3.32846105e-01 1.98487490e-01 1.04389560e+00 -8.74394357e-01
-1.55485660e-01 3.25842053e-01 -5.59611976e-01 2.03606963e-01
3.59315604e-01 -2.67412633e-01 5.37896812e-01 -5.75640917e-01
7.44353473e-01 1.36727071e+00 7.52682209e-01 -3.97662073e-02
-1.56209278e+00 -2.75087059e-01 -4.94056731e-01 7.99735844e-01
-1.30429924e+00 -2.96775639e-01 4.03939843e-01 -5.72354972e-01
5.32695651e-01 5.42409062e-01 7.12799072e-01 8.40489984e-01
4.87872124e-01 1.05753817e-01 2.10220313e+00 -1.11950219e+00
3.42534989e-01 2.47046947e-01 1.35955721e-01 4.08864677e-01
1.46004602e-01 -5.34305684e-02 -1.45395517e-01 -4.22957361e-01
7.04715252e-01 -3.76899898e-01 1.00267373e-01 -5.83695710e-01
-9.96829748e-01 9.75945115e-01 7.10018277e-02 8.33539367e-01
-1.66264996e-01 -2.59554148e-01 5.20551503e-01 4.39526826e-01
5.11284113e-01 4.07672167e-01 -4.00575697e-01 -1.76021770e-01
-9.22917128e-01 2.81639397e-01 8.66080463e-01 -8.99128467e-02
8.68922472e-01 -6.32879078e-01 -4.80491579e-01 1.25567043e+00
2.64484175e-02 1.98597163e-01 6.34413421e-01 -7.31780946e-01
-3.96685451e-02 3.56657922e-01 -8.97753537e-02 -1.39150178e+00
-3.64427447e-01 1.50697082e-01 -6.09254479e-01 5.52750766e-01
9.81384039e-01 3.91807467e-01 -7.99676120e-01 1.15552151e+00
2.96960115e-01 -2.78558701e-01 -2.86374867e-01 6.98126912e-01
3.41922730e-01 4.67546731e-01 -2.86430661e-02 3.87359485e-02
1.39618993e+00 -5.20882905e-01 -3.95364672e-01 3.31145585e-01
5.98255098e-01 -1.24068177e+00 9.82498765e-01 6.84852004e-01
-4.33322102e-01 -4.70640570e-01 -8.62902403e-01 3.78479630e-01
-7.66363204e-01 2.37597793e-01 4.52115983e-01 6.68138802e-01
-1.03898335e+00 1.04920769e+00 -6.72386706e-01 -7.84353197e-01
-1.75320029e-01 2.31179237e-01 -5.76671720e-01 1.18508868e-01
-9.57088590e-01 1.24577630e+00 4.58414048e-01 1.81800261e-01
-5.63550694e-03 1.33930683e-01 -6.78823233e-01 -1.11208811e-01
1.93681642e-01 -6.17222823e-02 7.53182113e-01 -8.82924795e-01
-1.82510746e+00 1.09228921e+00 -6.15021922e-02 -2.99738616e-01
5.02264917e-01 3.31622511e-01 -1.92505077e-01 1.56766862e-01
-4.11987817e-03 4.41565633e-01 8.17909598e-01 -1.09257472e+00
-3.99221152e-01 -1.99131951e-01 -2.21107036e-01 -4.03308630e-01
4.68564518e-02 2.79191155e-02 -1.50602967e-01 -5.57507634e-01
1.74432695e-01 -9.77387428e-01 1.27747906e-02 -3.76377404e-01
-2.45847218e-02 -2.08484352e-01 6.21982872e-01 -5.92897654e-01
9.64590847e-01 -2.07461381e+00 2.00017244e-02 6.70381665e-01
-2.59934664e-01 2.21445486e-01 -1.06610492e-01 6.79849267e-01
-1.00696549e-01 5.00928005e-03 -1.38928786e-01 3.12551409e-02
-1.39088199e-01 3.78978878e-01 5.06453728e-03 4.40331966e-01
7.50710666e-02 3.55168223e-01 -4.90240633e-01 -6.30776823e-01
4.18316960e-01 4.95898694e-01 -2.07284138e-01 -2.25898519e-01
1.38848335e-01 3.88057739e-01 -1.01533577e-01 3.84985864e-01
7.95471311e-01 2.21527770e-01 2.43447751e-01 -1.94057420e-01
-3.07844013e-01 6.23182096e-02 -1.24815691e+00 7.32759953e-01
-6.08480036e-01 5.32325268e-01 -2.91828185e-01 -1.12101161e+00
1.50037110e+00 1.32451281e-01 3.93655866e-01 -7.82704294e-01
4.13228832e-02 4.91315335e-01 2.03570545e-01 -4.83615249e-01
2.84430355e-01 -2.76745021e-01 1.60554111e-01 1.26289040e-01
1.95257030e-02 -1.45401418e-01 5.53828657e-01 -6.29095495e-01
6.63008094e-01 3.41377072e-02 4.25027579e-01 -7.11365402e-01
8.54454637e-01 -8.83376822e-02 5.46042807e-02 7.30093837e-01
4.81910966e-02 5.76982796e-01 9.89967763e-01 -5.63738525e-01
-9.71243083e-01 -7.29819477e-01 -8.17063630e-01 1.02981603e+00
-1.79642364e-01 -2.32541919e-01 -6.74549580e-01 -4.59276080e-01
-2.05643475e-02 2.53465265e-01 -7.94131339e-01 6.93785027e-02
-2.84879744e-01 -8.47952485e-01 2.07501188e-01 -9.27630812e-02
4.10618037e-01 -1.35047030e+00 -7.14300156e-01 -9.06276926e-02
-8.34541768e-02 -6.31829023e-01 1.79713026e-01 5.53148210e-01
-9.60112751e-01 -1.08894825e+00 -7.27911234e-01 -5.54866493e-01
4.46541607e-01 -1.21757701e-01 9.56194639e-01 1.57614723e-01
-2.45035559e-01 2.07001537e-01 -7.19745696e-01 -2.25982890e-01
-6.28561854e-01 2.16020659e-01 -3.58100802e-01 3.38660985e-01
1.83217481e-01 -3.54311645e-01 -2.15501755e-01 6.10913634e-01
-1.12852418e+00 -2.53406465e-01 7.32288957e-01 9.27449644e-01
6.23759270e-01 3.77986044e-01 5.26199490e-02 -8.79420638e-01
6.10304475e-01 -2.17158750e-01 -5.75207651e-01 1.76759884e-01
-4.72070724e-01 2.36931831e-01 5.60557604e-01 -5.22501707e-01
-6.25479996e-01 -1.39204875e-01 -2.88159460e-01 9.51894969e-02
-4.47430044e-01 5.13007522e-01 5.45301102e-02 -6.44778669e-01
4.54497069e-01 1.82042047e-01 2.88429677e-01 -7.34985888e-01
-1.39847860e-01 7.06250608e-01 3.49588483e-03 -7.40449727e-01
2.03995228e-01 3.50044399e-01 4.52279657e-01 -9.68140304e-01
-3.58745188e-01 -3.53165597e-01 -8.83035183e-01 -1.58174783e-01
6.31719649e-01 4.44130525e-02 -6.05679989e-01 7.38520145e-01
-8.05516601e-01 -4.61449474e-01 -1.43629700e-01 3.56682032e-01
-5.57752430e-01 4.95416164e-01 -3.75961304e-01 -7.45237231e-01
2.38629311e-01 -1.31034577e+00 1.03511620e+00 -3.09529882e-02
-1.55317992e-01 -1.08901203e+00 2.56995410e-01 -3.20753232e-02
7.98036277e-01 4.84879047e-01 1.32683265e+00 -4.01902944e-01
2.39938930e-01 -2.03996032e-01 -6.63881823e-02 3.42047095e-01
2.11968228e-01 2.48896256e-01 -6.97390616e-01 -1.34918004e-01
2.60200053e-01 -2.46805307e-02 1.13727677e+00 3.70178312e-01
1.37395501e+00 -2.27364898e-01 -1.71631172e-01 3.34090292e-01
1.47784555e+00 2.40888745e-01 1.01714087e+00 1.01269281e+00
5.07842153e-02 8.41156185e-01 5.02273262e-01 2.30412379e-01
-9.23999846e-02 1.21253490e+00 8.49837884e-02 -9.22632962e-02
6.13136217e-02 1.71090454e-01 8.46136585e-02 3.24282140e-01
-4.45613265e-01 9.76079926e-02 -8.83927584e-01 1.87352046e-01
-1.71751702e+00 -9.80258942e-01 -3.42956066e-01 2.56648636e+00
6.68037057e-01 2.04681754e-01 4.15467143e-01 6.77803993e-01
6.82833850e-01 8.62884820e-02 4.59553778e-01 -1.02542722e+00
-1.50744349e-01 6.13064706e-01 5.10317922e-01 5.09830654e-01
-1.25896358e+00 4.96642560e-01 6.38256931e+00 1.24812710e+00
-1.57368350e+00 -7.83731788e-02 6.81685388e-01 5.55214107e-01
2.29249030e-01 2.88401276e-01 -6.02132678e-01 7.01673090e-01
7.86071956e-01 4.50490296e-01 4.67040837e-01 5.80256343e-01
1.69837222e-01 -8.98039520e-01 -5.39793134e-01 8.29016089e-01
-1.14854105e-01 -8.21870506e-01 -7.05186501e-02 2.19398499e-01
4.30113405e-01 -3.71072829e-01 -1.68273404e-01 -1.30781248e-01
-4.89354461e-01 -9.75969374e-01 6.29119337e-01 7.78822362e-01
4.78395909e-01 -6.58279181e-01 1.15619636e+00 8.29410180e-02
-9.03110385e-01 -7.11251870e-02 -5.02642572e-01 -2.14288205e-01
-2.95170903e-01 9.60241973e-01 -7.52790630e-01 4.88032311e-01
8.73342812e-01 2.16480166e-01 -1.09397578e+00 1.17678034e+00
9.10503417e-02 6.30172729e-01 -5.50480306e-01 -1.43352762e-01
8.67544189e-02 -4.54247385e-01 3.96713048e-01 1.39298368e+00
4.34501678e-01 -7.05665231e-01 -1.30383447e-01 5.25547385e-01
8.14248979e-01 4.94411647e-01 -4.20168817e-01 2.32679233e-01
1.41238004e-01 1.48791432e+00 -1.40819430e+00 -1.99920144e-02
-1.34849604e-02 7.01061845e-01 1.33774117e-01 6.76140562e-02
-4.86830771e-01 -4.65120763e-01 3.71865593e-02 5.26396453e-01
5.86327374e-01 -2.98525691e-01 -2.35905871e-02 -7.12236881e-01
-3.33472937e-02 -8.62354159e-01 3.32032979e-01 -5.31138062e-01
-1.48192823e+00 6.22655213e-01 5.30659497e-01 -9.23387110e-01
-1.53720692e-01 -9.77770269e-01 -5.04596412e-01 9.49257076e-01
-1.07144344e+00 -9.70089555e-01 -8.05959478e-02 3.29345614e-01
-1.12201996e-01 5.71942516e-02 9.02075589e-01 2.09758431e-01
-1.78377494e-01 2.28932500e-01 2.48455256e-01 -1.63502112e-01
8.68101299e-01 -1.32779860e+00 -1.30856961e-01 3.53985369e-01
1.29993588e-01 4.14906174e-01 8.29447091e-01 -3.08349401e-01
-6.87127173e-01 -4.05252904e-01 1.09112597e+00 -2.53779322e-01
5.44154406e-01 -4.12064940e-01 -9.05704200e-01 2.34163571e-02
-4.72413339e-02 -1.31517619e-01 5.23413718e-01 1.29273489e-01
-1.40635088e-01 -2.73744971e-01 -1.17951930e+00 2.79220670e-01
2.88003713e-01 -2.69472957e-01 -3.61407727e-01 5.07250763e-02
-2.92507678e-01 -7.51401559e-02 -9.50508893e-01 4.93092418e-01
8.90124261e-01 -1.66810703e+00 6.79866672e-01 -2.85995509e-02
3.51114035e-01 -3.69737715e-01 -1.00511484e-01 -1.28303432e+00
-4.55105513e-01 -8.48134086e-02 3.19873184e-01 1.27426589e+00
2.15114802e-01 -1.04318750e+00 2.76971757e-01 -1.04028456e-01
4.66122895e-01 -1.08046126e+00 -1.10828733e+00 -8.93541396e-01
3.24677713e-02 -1.88794315e-01 5.75842440e-01 5.32137990e-01
-1.89691484e-01 -7.42990151e-02 -2.10091978e-01 -3.98961365e-01
1.58983961e-01 1.60276845e-01 7.60904491e-01 -1.39866793e+00
-4.50345218e-01 -8.90170157e-01 -8.57936859e-01 -3.04190397e-01
-1.34394076e-02 -7.33418703e-01 -2.18169801e-02 -1.06911778e+00
1.99875042e-01 -6.87150538e-01 -3.49750340e-01 5.08511364e-01
7.97476992e-02 5.26879430e-01 2.90361106e-01 2.68628180e-01
-5.73658869e-02 -4.95557077e-02 1.12210655e+00 5.07422164e-02
-2.56795675e-01 1.45317450e-01 -2.00119942e-01 5.17843604e-01
7.58967042e-01 -5.17499506e-01 8.90019462e-02 1.87209368e-01
1.19020075e-01 -2.98846245e-01 5.80385149e-01 -1.11969817e+00
-2.83535153e-01 -3.13131958e-01 6.16901755e-01 -1.93942249e-01
1.06611677e-01 -7.48620927e-01 2.76232481e-01 4.65009302e-01
-7.84472525e-02 9.22942758e-02 1.92998454e-01 4.47182842e-02
-3.51314485e-01 -6.65639460e-01 1.02342618e+00 2.61620395e-02
-7.51302302e-01 -3.51012439e-01 -5.56620240e-01 -6.46640658e-01
7.92931676e-01 -3.62355798e-01 -1.12577900e-01 -2.41790906e-01
-7.68483043e-01 -5.08742392e-01 7.16577113e-01 2.13110253e-01
8.40838924e-02 -1.16000152e+00 -4.39409703e-01 2.98690379e-01
1.13458112e-01 -6.19742870e-01 6.89710379e-02 1.40104330e+00
-9.00274515e-01 3.22568297e-01 -7.10600913e-01 -7.83135951e-01
-1.29324377e+00 5.38001180e-01 4.33539301e-01 -5.57982385e-01
-2.75015026e-01 1.03128731e-01 -2.58169323e-01 -3.54983121e-01
-2.82038152e-01 -3.80073071e-01 -6.31616175e-01 2.55429000e-01
2.79881746e-01 5.70819557e-01 4.83119011e-01 -7.25548863e-01
-4.15488273e-01 9.68552828e-01 3.94679904e-01 -2.63903737e-01
1.33480513e+00 -1.36348596e-02 -6.71394110e-01 6.96473718e-01
1.19076836e+00 3.19221288e-01 -6.56120956e-01 3.27796012e-01
3.91124964e-01 -6.10047817e-01 -5.60482703e-02 -7.29415596e-01
-7.27053106e-01 7.31453896e-01 8.52408588e-01 6.90483451e-01
1.26038706e+00 -1.23441353e-01 2.97303230e-01 2.36530639e-02
6.48839593e-01 -9.30337429e-01 -2.88093895e-01 7.43075550e-01
8.20074379e-01 -8.99457455e-01 4.03257720e-02 -4.08249021e-01
-4.21422094e-01 1.69965518e+00 -4.13064510e-02 -1.40331700e-01
6.83220148e-01 1.30960181e-01 1.62945643e-01 -4.02739048e-02
-3.35181594e-01 -3.80817533e-01 4.46116298e-01 5.18574476e-01
6.41976833e-01 2.14208141e-01 -1.15659118e+00 6.74297884e-02
-2.80501217e-01 9.16285142e-02 3.10027361e-01 8.07180882e-01
-3.23693514e-01 -1.80786860e+00 -8.82297873e-01 6.51638210e-01
-5.33841252e-01 3.30597013e-01 -5.08824170e-01 8.30194473e-01
6.51290059e-01 8.89490306e-01 1.29192531e-01 -5.47185659e-01
2.07487404e-01 2.63902485e-01 7.88763821e-01 -2.28521928e-01
-3.69835407e-01 4.98716272e-02 -2.23916750e-02 -3.56568217e-01
-5.92485130e-01 -7.04517424e-01 -6.10913336e-01 -3.07746619e-01
-3.57515484e-01 3.41936469e-01 8.66084039e-01 1.02392101e+00
-2.65121460e-02 -1.36354081e-02 6.15290701e-01 -1.15657127e+00
-2.58971572e-01 -1.23233044e+00 -8.90937746e-01 5.96416652e-01
-6.65222434e-03 -1.14188325e+00 -4.01175916e-01 -1.39794007e-01] | [10.295022010803223, -0.3919641673564911] |
58766fe4-b2c4-4839-b01b-36deca166954 | predicting-intubation-support-requirement-of | 2011.01787 | null | https://arxiv.org/abs/2011.01787v1 | https://arxiv.org/pdf/2011.01787v1.pdf | Predicting intubation support requirement of patients using Chest X-ray with Deep Representation Learning | Recent developments in medical imaging with Deep Learning presents evidence of automated diagnosis and prognosis. It can also be a complement to currently available diagnosis methods. Deep Learning can be leveraged for diagnosis, severity prediction, intubation support prediction and many similar tasks. We present prediction of intubation support requirement for patients from the Chest X-ray using Deep representation learning. We release our source code publicly at https://github.com/aniketmaurya/covid-research. | ['Aniket Maurya'] | 2020-10-28 | null | null | null | null | ['intubation-support-prediction'] | ['computer-vision'] | [-3.31207782e-01 -2.91425958e-02 -6.55881584e-01 -7.12487817e-01
-9.69594657e-01 -4.37020123e-01 -1.72692448e-01 4.68607366e-01
-1.50387660e-01 5.26162624e-01 7.11777031e-01 -1.06270969e+00
-5.13609588e-01 -5.34654677e-01 -3.17864060e-01 -5.46502471e-01
-1.34865254e-01 9.10124779e-01 -2.85261452e-01 1.79098487e-01
1.66718438e-02 5.79788506e-01 -8.60356450e-01 8.97050500e-01
2.39636809e-01 1.03350222e+00 5.72850443e-02 1.06862926e+00
2.87898421e-01 1.66644287e+00 -2.42269412e-01 -6.96085840e-02
2.17192471e-01 -1.39342368e-01 -1.16699326e+00 -3.64207506e-01
1.08002916e-01 -9.50915456e-01 -7.14764178e-01 2.86638349e-01
1.15975344e+00 -2.19480135e-02 8.73933613e-01 -6.83679998e-01
-3.60918254e-01 4.51079309e-01 -2.78458059e-01 9.98779833e-01
3.86326104e-01 2.71392912e-01 4.17675793e-01 -5.79204619e-01
2.58982360e-01 6.30276918e-01 9.53795910e-01 7.69669175e-01
-3.66545320e-01 -6.05445564e-01 -4.64049131e-01 4.16669726e-01
-8.15493762e-01 -6.55167550e-02 -4.58617229e-03 -6.87477171e-01
1.04605854e+00 2.93737561e-01 4.25766766e-01 1.43819010e+00
4.55655754e-01 6.83498144e-01 7.70527065e-01 1.80180334e-02
-3.15288424e-01 -2.39782184e-01 3.85826886e-01 1.06153977e+00
3.02344352e-01 2.77971745e-01 -5.66104613e-03 -4.09187049e-01
7.51275957e-01 6.52141809e-01 -3.34896177e-01 1.65839076e-01
-1.06429648e+00 9.91197884e-01 5.94134748e-01 2.09477663e-01
-8.50897908e-01 3.57657880e-01 8.33815515e-01 2.55689263e-01
2.05308720e-01 2.38556042e-01 -8.30884755e-01 -3.13489407e-01
-7.58009434e-01 1.36583626e-01 7.57762134e-01 5.31912982e-01
-1.72413304e-01 1.81124881e-01 -2.84778357e-01 8.06034267e-01
1.74543887e-01 2.45122164e-01 1.12856138e+00 -9.33702528e-01
2.81048119e-01 2.65407801e-01 1.19225951e-02 -3.03220332e-01
-1.09636784e+00 -5.55011928e-01 -7.73533046e-01 -1.47055704e-02
-1.20009229e-01 -7.95279682e-01 -8.86369050e-01 8.35887671e-01
2.13163227e-01 2.30573848e-01 9.86859649e-02 8.65030348e-01
1.41066134e+00 2.87930071e-01 7.97479004e-02 5.54514863e-02
1.37057376e+00 -9.70990837e-01 -5.25726914e-01 8.13982710e-02
1.09068692e+00 -7.30385125e-01 3.41710806e-01 6.24553740e-01
-1.24697435e+00 -1.95825651e-01 -5.96602798e-01 1.48696676e-02
-8.86512101e-02 1.15071610e-01 7.60613739e-01 3.69384885e-01
-1.15250695e+00 8.43324184e-01 -9.60149944e-01 -2.44752139e-01
5.66660881e-01 5.24559021e-01 -2.90574431e-01 -3.62367511e-01
-9.71675158e-01 1.29672611e+00 1.62868798e-01 -2.44706988e-01
-1.02264047e+00 -5.28375804e-01 -7.61487246e-01 -2.06858050e-02
1.47305757e-01 -1.52057564e+00 1.89622808e+00 -5.15226796e-02
-9.75490034e-01 9.76212502e-01 4.48233970e-02 -7.91603327e-01
2.71100491e-01 -5.23197174e-01 -3.25935245e-01 3.51254284e-01
-2.49870852e-01 2.75007546e-01 6.08252406e-01 -5.81776142e-01
-6.81567073e-01 -3.52788389e-01 -1.97237477e-01 1.02385273e-02
1.14836404e-02 3.54277700e-01 1.36738539e-01 -6.23474181e-01
-7.71965086e-02 -8.09289455e-01 -5.53944588e-01 -1.93619832e-01
-2.84320742e-01 -2.21717134e-01 5.18601656e-01 -9.17686462e-01
1.19774139e+00 -1.68844914e+00 -6.03894107e-02 -9.54899862e-02
5.92761695e-01 5.41254878e-01 1.82964966e-01 4.40330267e-01
-6.17941737e-01 2.31025040e-01 6.36914074e-02 -4.69384529e-02
-5.78864694e-01 2.69262552e-01 2.48931237e-02 5.55495143e-01
1.09746329e-01 1.18012917e+00 -8.63469601e-01 -2.88310468e-01
6.89014018e-01 6.18292511e-01 -4.79321688e-01 7.15980530e-01
1.67402387e-01 8.74458849e-01 -5.34088314e-01 8.63596678e-01
9.17547941e-02 -7.25327790e-01 -2.00850084e-01 5.06583489e-02
3.95822898e-02 6.12267613e-01 -4.26231623e-01 1.38627183e+00
-5.93154848e-01 2.86496997e-01 -9.92498025e-02 -1.02682745e+00
4.75859344e-01 8.71675313e-01 1.06111312e+00 -4.95210066e-02
7.49389708e-01 9.31942537e-02 3.57895523e-01 -1.26705062e+00
-2.78066427e-01 -3.94568801e-01 2.69336194e-01 8.33795190e-01
-6.22548386e-02 -1.76079005e-01 -4.33676749e-01 -6.14703074e-02
1.54302227e+00 -2.85006970e-01 7.23957598e-01 1.54795304e-01
1.42239332e-01 -1.43925682e-01 2.81117082e-01 6.45850420e-01
-3.69228989e-01 7.16655314e-01 1.05636753e-01 -1.02676642e+00
-4.86193180e-01 -8.81455421e-01 -4.30406153e-01 7.79927611e-01
-7.53677666e-01 -1.55414090e-01 -3.59907001e-01 -8.21337819e-01
1.52194068e-01 5.69170058e-01 -7.95130968e-01 -2.55590707e-01
-5.39708674e-01 -5.20365953e-01 6.76792562e-01 1.18031812e+00
-3.98120463e-01 -1.23855913e+00 -1.01949096e+00 1.28103524e-01
-1.86582044e-01 -6.44873083e-01 -1.43230045e-02 4.93541569e-01
-1.07487893e+00 -1.46598971e+00 -1.03029227e+00 -6.94003999e-01
1.62841588e-01 -2.42887549e-02 1.21348357e+00 6.10278130e-01
-9.22842979e-01 7.07273066e-01 -4.76477474e-01 -7.86441267e-01
-5.12299836e-01 -1.38452919e-02 3.79398823e-01 -8.30919683e-01
2.81445891e-01 -4.42681462e-01 -1.27691591e+00 6.22877926e-02
-6.71911120e-01 -2.63623089e-01 3.37677240e-01 9.69385743e-01
6.28712535e-01 -4.38848913e-01 7.54697144e-01 -1.00782323e+00
8.02684069e-01 -1.33779705e+00 1.74631439e-02 -1.07670538e-01
-6.94436312e-01 -2.70136505e-01 9.99980941e-02 1.23846784e-01
-4.57254529e-01 -2.07312718e-01 -8.21840227e-01 -9.95819390e-01
-4.82740104e-01 6.34764016e-01 1.12111545e+00 2.24061593e-01
6.66848481e-01 -1.53863013e-01 1.54010698e-01 -4.81499583e-01
-6.93630725e-02 9.25872803e-01 4.08596277e-01 -2.99803585e-01
8.78037363e-02 2.22457394e-01 5.02239540e-02 -7.08674341e-02
-1.10348046e+00 -8.02062690e-01 -7.15346694e-01 1.67161345e-01
1.01499951e+00 -8.36697817e-01 -6.22335494e-01 -1.20898746e-01
-9.71380115e-01 -3.42032850e-01 -4.38686371e-01 8.26073468e-01
-5.39177358e-01 1.11971930e-01 -8.75479043e-01 -5.25470257e-01
-1.00086546e+00 -1.28494608e+00 8.09815288e-01 -1.66995600e-01
-6.23445868e-01 -1.33614147e+00 2.42153943e-01 6.53574824e-01
4.51666385e-01 4.94220257e-01 8.56701434e-01 -1.19804347e+00
-4.45565991e-02 -7.20725417e-01 -7.45454654e-02 4.57532793e-01
4.56939697e-01 3.02968286e-02 -8.27176332e-01 -3.30894470e-01
2.06096247e-01 -4.81048703e-01 8.52153361e-01 8.77275884e-01
2.05750537e+00 -3.43717575e-01 -1.64169818e-01 9.81622219e-01
1.01069045e+00 4.03483570e-01 2.31039673e-01 1.68130919e-01
6.66703701e-01 3.56301904e-01 3.13266098e-01 9.15556431e-01
4.27819669e-01 2.52805930e-02 6.59770668e-01 -1.55686155e-01
-1.04724847e-01 4.17469949e-01 -3.26894969e-01 9.32769954e-01
-2.88325787e-01 -2.61060864e-01 -1.62958944e+00 4.70017970e-01
-1.66657984e+00 -9.56934392e-01 -3.20631415e-02 1.76292408e+00
4.96539056e-01 -3.17592502e-01 3.01431984e-01 8.86073411e-02
4.11200374e-01 -2.54886597e-01 -6.27377391e-01 -7.49879479e-01
6.06401265e-01 5.30298591e-01 3.81507725e-01 1.80869117e-01
-9.77801085e-01 4.62486178e-01 7.36528158e+00 1.62328646e-01
-1.30341446e+00 5.74277580e-01 8.33913863e-01 -3.90114546e-01
-6.73229918e-02 -7.03668118e-01 -4.14852291e-01 4.29840714e-01
1.26776671e+00 -2.24141628e-02 9.76827219e-02 8.50731909e-01
2.80109942e-01 2.51258612e-01 -1.22460747e+00 9.35528278e-01
-2.10956167e-02 -1.46082759e+00 -1.44885063e-01 -3.23949847e-03
3.71094376e-01 1.01573205e+00 2.45335639e-01 1.80108443e-01
2.37394288e-01 -1.38711357e+00 -2.48753190e-01 4.10118073e-01
1.38583684e+00 -4.53086942e-01 1.17480111e+00 1.44654378e-01
-6.50976837e-01 -5.24650514e-01 -2.21820757e-01 1.01454638e-01
1.37817383e-01 1.18670642e-01 -1.59682667e+00 4.69049096e-01
6.80122674e-01 8.09070051e-01 -2.11425260e-01 1.26823783e+00
-1.09057829e-01 1.00928962e+00 -7.48501252e-03 3.88686210e-01
2.40710825e-01 3.03953916e-01 3.98076713e-01 1.33470273e+00
3.06841254e-01 6.11562669e-01 2.20498726e-01 1.91608936e-01
-1.25766054e-01 2.36546844e-01 -1.19135857e+00 1.77156717e-01
2.09715277e-01 1.19025528e+00 -3.45322043e-01 -4.74376589e-01
-5.69284379e-01 6.18420899e-01 3.55204083e-02 6.17691828e-03
-8.62532258e-01 2.58279294e-01 4.63411123e-01 2.80289114e-01
1.24081083e-01 3.90266567e-01 -5.35557151e-01 -1.04644370e+00
-6.20713890e-01 -9.48591888e-01 9.60379779e-01 -9.50876832e-01
-1.26395226e+00 7.26625323e-01 -1.48177907e-01 -1.48042142e+00
-8.19434822e-01 -9.03106749e-01 -9.51204240e-01 6.34802282e-01
-1.54790390e+00 -8.57227743e-01 -2.44524896e-01 7.56978869e-01
8.08234215e-01 -5.15497327e-01 1.34291196e+00 2.78488338e-01
-6.80282295e-01 4.72750425e-01 -6.00416213e-02 1.97034016e-01
6.43386424e-01 -1.24692249e+00 2.46332482e-01 1.77153066e-01
-3.14166099e-01 6.46107554e-01 2.40066573e-01 -4.63525504e-01
-9.83364940e-01 -1.25466800e+00 4.04137433e-01 -9.12128150e-01
5.71573198e-01 8.33211482e-01 -8.17428827e-01 8.82931709e-01
2.77806729e-01 3.07957292e-01 1.34133196e+00 8.63603503e-02
-1.70135900e-01 2.49752551e-01 -1.28649962e+00 8.39133654e-03
5.58128774e-01 -4.50925469e-01 -6.24297261e-01 9.37555373e-01
6.90254688e-01 -7.26887286e-01 -1.43966842e+00 9.48964119e-01
6.82101667e-01 -7.41007507e-01 1.08075237e+00 -1.27800548e+00
8.25371742e-01 3.69525760e-01 -2.64458526e-02 -1.16797769e+00
-5.77212453e-01 -3.25371921e-01 -5.76861799e-01 1.19398177e-01
3.54602814e-01 -7.19998419e-01 4.81830359e-01 5.83886802e-01
-4.13023114e-01 -1.76112115e+00 -7.46305764e-01 -1.82697028e-01
4.12027568e-01 -5.46346366e-01 5.13363600e-01 9.93521929e-01
5.47743076e-03 6.40661120e-02 -2.32992053e-01 2.60411978e-01
8.09739679e-02 5.98928379e-03 1.41933322e-01 -1.14566433e+00
-6.29399598e-01 -3.90002817e-01 -2.42631137e-01 -1.01665616e-01
-6.53729364e-02 -1.21244371e+00 -3.74192923e-01 -2.18230200e+00
1.74233764e-01 -3.40421766e-01 -1.03717506e+00 8.03226471e-01
-2.72088265e-03 1.14426464e-01 -2.24222958e-01 3.03839207e-01
-2.57833511e-01 -1.07520759e-01 1.15651989e+00 7.42792562e-02
2.40150332e-01 5.86251795e-01 -6.87105119e-01 1.06897736e+00
1.71105278e+00 -5.20364583e-01 -2.93744028e-01 -6.89043522e-01
-9.46751144e-03 8.38567436e-01 1.31290615e-01 -1.12718093e+00
-1.49531156e-01 -8.67456198e-02 6.72148883e-01 -5.50629437e-01
2.21867770e-01 -7.18155980e-01 -3.09613019e-01 9.84799266e-01
-4.24939305e-01 6.81514800e-01 3.70613068e-01 2.76578993e-01
-1.22283865e-02 -4.24070150e-01 6.74539864e-01 -4.69375134e-01
-4.13525999e-01 8.63789916e-01 -7.31143534e-01 4.40952927e-02
9.64526415e-01 2.24554524e-01 -4.69612718e-01 -5.38341701e-01
-8.99641216e-01 2.34169692e-01 -9.12926644e-02 4.37549353e-01
1.02048099e+00 -7.63811409e-01 -9.66141343e-01 1.10214120e-02
1.15940452e-01 1.03245385e-01 2.73329586e-01 1.18163061e+00
-9.17259514e-01 5.87543070e-01 -1.62247807e-01 -3.89756441e-01
-1.48001969e+00 8.18842769e-01 5.90782583e-01 -2.86361545e-01
-8.26935589e-01 1.13067150e+00 1.17474116e-01 -2.51951247e-01
4.72082615e-01 -4.92122769e-01 -4.39184278e-01 -1.61872089e-01
6.67134523e-01 4.50153321e-01 2.77892351e-01 -1.65384889e-01
-4.24295723e-01 3.11570644e-01 -4.56177056e-01 4.96528596e-01
1.68451083e+00 1.20375350e-01 -9.19168815e-02 3.88708353e-01
1.08094227e+00 -4.62212771e-01 -3.12235653e-01 1.76304117e-01
-1.64294928e-01 -1.12375453e-01 2.54357249e-01 -1.16865659e+00
-1.22076333e+00 1.27168584e+00 9.76488709e-01 -1.30953461e-01
9.33798015e-01 1.57689467e-01 1.06375909e+00 6.40250742e-01
3.75929996e-02 -4.82900888e-01 1.08468212e-01 4.02258068e-01
9.94745195e-01 -1.69826090e+00 1.65013000e-01 3.89926881e-02
-8.81286323e-01 1.14911497e+00 5.67414343e-01 -2.96302944e-01
1.15137160e+00 5.21914721e-01 6.39167964e-01 -6.06552303e-01
-1.09017789e+00 -1.02969334e-01 2.24728644e-01 4.36035544e-01
1.10702753e+00 4.42651451e-01 5.60319163e-02 7.61410475e-01
8.84057358e-02 3.23979855e-01 5.23898661e-01 9.32211995e-01
-3.55752587e-01 -1.08578789e+00 -2.71333098e-01 1.19954598e+00
-1.21591270e+00 -4.82755154e-01 -1.43864647e-01 2.56851971e-01
1.31735355e-01 7.14835167e-01 -2.56178647e-01 -6.93801701e-01
1.56158164e-01 3.28419030e-01 3.43192399e-01 -1.06801105e+00
-8.62113535e-01 -1.81869939e-01 6.43134192e-02 -6.39788866e-01
-6.50096983e-02 -5.14150977e-01 -1.41909957e+00 -8.72401521e-02
-4.71954718e-02 -1.52948514e-01 7.07670629e-01 4.56166804e-01
4.89272177e-01 1.06382084e+00 6.10450685e-01 -5.56881070e-01
-7.38698304e-01 -1.08852935e+00 -1.38998717e-01 -9.46218595e-02
9.16079938e-01 -6.61524177e-01 -3.22087228e-01 -1.70696408e-01] | [15.251362800598145, -1.8276660442352295] |
88786129-1dcd-4525-afef-f6dd178b9193 | atf-towards-robust-face-alignment-via | null | null | https://dl.acm.org/doi/10.1145/3394171.3414037 | https://dl.acm.org/doi/10.1145/3394171.3414037 | ATF: Towards Robust Face Alignment via Leveraging Similarity and Diversity across Different Datasets | Face alignment is an important task in the field of multi-media. Together with the impressive progress of algorithms, various benchmark datasets have been released in recent years. Intuitively, it is meaningful to integrate multiple labeled datasets with different annotations to achieve higher performance on a target landmark detector. Although numerous efforts have been made in joint usage, there yet remain three shortages in recent works, e.g., additional computation, limitation of the markups scheme, and limited support for the regression method. To address the above problems, we proposed a novel Alternating Training Framework (ATF), which leverages similarity and diversity across multi-media sources for a more robust detector. Our framework mainly contains two sub-modules: Alternating Training with Decreasing Proportions (ATDP) and Mixed Branch Loss (mathcal LMB). In particular, ATDP trains multiple datasets simultaneously to take advantage of the diversity between them, while mathcal LMB utilizes similar landmark pairs to constrain different branches of corresponding datasets. Extensive experiments on various benchmarks show the effectiveness of our framework, and ATF is feasible for both heatmap-based network and direct coordinate regression. Specifically, the mean error even reaches 3.17 on the experiment on 300W leveraging WFLW, which significantly outperforms state-of-the-art methods. Both in an ordinary convolutional network (OCN) and HRNET, ATF achieves up to 9.96% relative improvement. Our source codes are made publicly available at https://github.com/starhiking/ATF. | ['Jian Cheng', 'Cong Leng', 'Fangzhou Xiong', 'Qinghao Hu', 'Xing Lan'] | 2020-10-12 | null | null | null | acm-mm-2020-10-1 | ['robust-face-alignment', 'face-alignment'] | ['computer-vision', 'computer-vision'] | [ 5.23781031e-02 -1.40997514e-01 -3.43718648e-01 -5.44041991e-01
-9.79440451e-01 -1.02898076e-01 4.18529958e-01 -2.20838308e-01
-2.89750725e-01 4.57871705e-01 1.80058181e-01 -1.43116070e-02
1.51277590e-03 -5.22811770e-01 -6.96992576e-01 -6.32898986e-01
2.88517237e-01 2.70458937e-01 2.58371949e-01 -1.50837988e-01
3.42844203e-02 5.30125678e-01 -1.19460690e+00 9.92829427e-02
5.88254571e-01 1.17143238e+00 8.83599222e-02 1.27453566e-01
-1.69413075e-01 4.30143327e-01 -2.42202312e-01 -7.76152372e-01
5.22474349e-01 -3.23663324e-01 -4.29297149e-01 -4.40260805e-02
9.16215539e-01 -3.30311149e-01 -4.02537078e-01 1.06598747e+00
9.05420601e-01 -1.38350055e-01 3.09596926e-01 -1.55922151e+00
-5.70127428e-01 5.43794870e-01 -1.30773747e+00 6.00892603e-02
3.68424468e-02 1.34448335e-01 1.07737577e+00 -1.39556098e+00
4.49277312e-01 1.21766388e+00 8.02037001e-01 5.08472383e-01
-1.03092968e+00 -1.35684180e+00 2.27542177e-01 2.28683501e-01
-1.61809874e+00 -7.06201077e-01 6.95363462e-01 -2.03607485e-01
6.46318972e-01 8.06356519e-02 3.81755352e-01 9.60581958e-01
-1.83026329e-01 8.48628879e-01 9.57469940e-01 -3.92736554e-01
-3.16232413e-01 3.70152406e-02 -4.82020453e-02 1.02758980e+00
2.79626876e-01 -2.03418344e-01 -7.62517333e-01 5.63235469e-02
7.53643751e-01 1.56961098e-01 -2.22646028e-01 -3.03734004e-01
-9.81613576e-01 6.80972219e-01 5.88995874e-01 1.55527622e-01
3.19229439e-02 6.69080392e-02 3.15578848e-01 3.76158543e-02
5.85242271e-01 -3.67502868e-02 -4.77388650e-01 2.45007262e-01
-1.00378263e+00 1.74456507e-01 3.89062524e-01 1.27378500e+00
7.68213451e-01 -1.58877924e-01 -2.60784805e-01 1.13302350e+00
4.09631461e-01 6.17178142e-01 2.00319231e-01 -5.89351058e-01
9.16614413e-01 6.10720873e-01 -4.24900889e-01 -1.17196727e+00
-5.47439694e-01 -4.85044897e-01 -1.13499629e+00 -6.72181919e-02
4.93259013e-01 -2.00833201e-01 -9.10031438e-01 1.87453222e+00
5.01012981e-01 7.10621536e-01 -4.61354315e-01 9.07456577e-01
9.99503493e-01 4.98266190e-01 -9.15292129e-02 -2.26880193e-01
1.27737129e+00 -1.25041652e+00 -5.61211526e-01 -2.06871569e-01
6.32688284e-01 -1.05867064e+00 8.29710722e-01 5.88424951e-02
-9.73053992e-01 -6.61686361e-01 -1.03536344e+00 -1.54463544e-01
-1.98722199e-01 3.78345072e-01 6.18113041e-01 4.15711373e-01
-1.10292792e+00 3.08575690e-01 -6.49839342e-01 -2.82695830e-01
7.96314597e-01 5.80983043e-01 -5.08489072e-01 -2.62582541e-01
-9.18261290e-01 4.99192327e-01 5.28392605e-02 3.99724871e-01
-3.81935179e-01 -7.85640180e-01 -8.11774969e-01 -1.80156186e-01
4.74329233e-01 -4.43931848e-01 1.14544427e+00 -7.34312892e-01
-1.29588234e+00 8.41780186e-01 -2.88097769e-01 -5.71293645e-02
7.74083197e-01 -3.77734721e-01 -5.11997104e-01 -6.28761947e-02
2.48441041e-01 8.52349520e-01 6.76924229e-01 -9.78177249e-01
-8.10268044e-01 -4.70817298e-01 -2.35044360e-01 3.39344777e-02
-6.76094532e-01 3.52574855e-01 -1.28497827e+00 -6.74831808e-01
2.38909647e-01 -1.04014754e+00 -8.86827260e-02 2.51162052e-01
-6.54488027e-01 -3.20583284e-01 6.60287023e-01 -6.36489451e-01
1.40677488e+00 -2.15614414e+00 3.51351462e-02 1.81947395e-01
3.37982267e-01 4.74331170e-01 -4.03116852e-01 1.39165938e-01
-2.32961118e-01 4.03032936e-02 -1.40458956e-01 -7.38357306e-01
-9.27902535e-02 -1.03438139e-01 -2.49292612e-01 6.70241058e-01
3.87752026e-01 8.78342211e-01 -5.71751952e-01 -7.35273600e-01
-2.08779555e-02 5.32727718e-01 -5.24644792e-01 8.32965225e-02
1.97484419e-01 3.76045167e-01 -4.21433926e-01 1.01646650e+00
1.03325558e+00 -2.99210906e-01 4.07497510e-02 -4.87134367e-01
-8.13668370e-02 -4.77546528e-02 -1.33526278e+00 1.71849060e+00
-2.92593539e-01 5.75158060e-01 -6.81216642e-02 -6.78748310e-01
9.01467502e-01 2.08641663e-01 5.93480825e-01 -8.09784055e-01
5.76105826e-02 3.22783560e-01 -5.17535023e-02 -2.65135318e-01
3.12447160e-01 3.00109714e-01 2.30101556e-01 1.66872233e-01
1.36257291e-01 5.06039083e-01 3.71852368e-02 3.65262814e-02
9.11358953e-01 2.15316087e-01 1.87562093e-01 -4.50798236e-02
5.41467786e-01 -3.60588580e-01 9.87565517e-01 4.20545548e-01
-3.14525366e-01 8.92199278e-01 3.42754036e-01 -4.44712311e-01
-9.61395562e-01 -6.07061803e-01 -3.18066508e-01 1.12025106e+00
3.31569195e-01 -5.63807487e-01 -6.60824299e-01 -7.22647309e-01
4.18090075e-02 8.83999318e-02 -5.26278377e-01 1.46597117e-01
-1.02068460e+00 -1.10024655e+00 6.99443758e-01 6.29592359e-01
8.27140808e-01 -8.10337484e-01 4.45903130e-02 -5.83854280e-02
-2.02585123e-02 -1.27934921e+00 -7.84537852e-01 -2.44491771e-01
-6.10793531e-01 -1.06319606e+00 -7.20301569e-01 -8.51205289e-01
7.52619624e-01 6.64644837e-01 8.93422723e-01 3.12280416e-01
-2.92892218e-01 -2.93616729e-04 -1.28978640e-01 -3.29433918e-01
1.86750457e-01 3.59650731e-01 -2.20721848e-02 2.83981591e-01
4.88399655e-01 -4.68271881e-01 -7.70517707e-01 6.24984205e-01
-5.37040114e-01 1.76244199e-01 7.66870081e-01 7.97941744e-01
5.77823997e-01 -3.77298236e-01 5.35400093e-01 -8.80049348e-01
1.93899587e-01 -5.42257488e-01 -5.95009446e-01 3.52157533e-01
-6.57985568e-01 -2.88391620e-01 3.52060348e-01 -3.80928308e-01
-8.75989616e-01 1.44767731e-01 -2.29738593e-01 -5.51592588e-01
7.01608956e-02 5.62624410e-02 -4.34034735e-01 -3.12030882e-01
2.76165098e-01 -8.87988657e-02 2.82244645e-02 -4.66518968e-01
1.56846836e-01 6.13618076e-01 4.08467203e-01 -5.60752034e-01
1.01532733e+00 2.83503205e-01 4.44165766e-02 -5.32178342e-01
-9.10607874e-01 -5.03622890e-01 -6.48953021e-01 -2.54505157e-01
5.97089171e-01 -1.09664345e+00 -6.24473214e-01 6.00474119e-01
-1.16246593e+00 -1.10080436e-01 4.00613874e-01 2.60541230e-01
-1.71772335e-02 2.43651256e-01 -6.02956533e-01 -4.71162260e-01
-4.50691730e-01 -1.33251214e+00 1.15184486e+00 5.00231504e-01
1.50204897e-01 -7.20086217e-01 -2.68770695e-01 5.04560292e-01
3.50466520e-01 1.84641823e-01 5.22907972e-01 -7.16282070e-01
-7.45258093e-01 -2.30860025e-01 -5.92097402e-01 6.53502643e-02
1.05716839e-01 1.71089202e-01 -1.02276468e+00 -4.23276335e-01
-3.50355804e-01 -3.61138701e-01 9.58791614e-01 1.97286814e-01
1.36879253e+00 -2.04234987e-01 -5.53265214e-01 9.59510386e-01
1.35575020e+00 7.08605396e-03 5.57480693e-01 3.77437860e-01
1.00752473e+00 5.20918489e-01 6.67295516e-01 2.74522603e-01
4.31731433e-01 1.20179343e+00 2.86784202e-01 -4.30701494e-01
-3.08653593e-01 -2.64850378e-01 2.53477424e-01 7.77163923e-01
-2.03498363e-01 -4.96049523e-02 -8.78392518e-01 2.40356892e-01
-1.96435666e+00 -7.76888847e-01 -1.09912574e-01 2.17633510e+00
8.73801053e-01 1.76276401e-01 1.43384993e-01 -1.31285727e-01
9.55807209e-01 3.13511878e-01 -6.68915689e-01 2.08364293e-01
-1.88346565e-01 1.44143701e-01 6.01526558e-01 3.37060064e-01
-1.25389564e+00 1.00185406e+00 5.15870190e+00 1.16819263e+00
-1.11939883e+00 3.58015329e-01 9.34648156e-01 -2.12112173e-01
1.37594596e-01 -2.46786267e-01 -1.50378299e+00 4.92804021e-01
3.36041003e-01 1.09529547e-01 1.54011250e-01 8.20046723e-01
-2.13992987e-02 2.27393985e-01 -9.35689092e-01 1.16644657e+00
3.05275857e-01 -1.33286762e+00 -6.74826503e-02 8.32959116e-02
7.56327927e-01 1.97971731e-01 2.49159306e-01 2.32756093e-01
6.94452301e-02 -9.10655975e-01 5.93529701e-01 1.60081938e-01
9.42547143e-01 -7.16970921e-01 7.48028278e-01 7.07528740e-03
-1.49076092e+00 4.24157903e-02 -4.75594521e-01 3.97000611e-01
5.22422977e-02 6.16251767e-01 -6.21851087e-01 6.42378092e-01
6.69523239e-01 7.56255269e-01 -7.71081507e-01 1.14180410e+00
-1.67426303e-01 5.02481937e-01 -4.30968165e-01 2.70769984e-01
1.57698393e-01 -1.39535695e-01 2.56866813e-01 1.30882490e+00
3.57298672e-01 -1.59032509e-01 3.66114140e-01 3.66329163e-01
-4.34296459e-01 3.62815052e-01 -2.77717888e-01 3.86181504e-01
7.10219085e-01 1.69996560e+00 -6.21115208e-01 -4.53003421e-02
-7.61967361e-01 7.64451504e-01 5.95038772e-01 2.61811674e-01
-1.24938130e+00 -1.96020529e-01 7.03267753e-01 9.74592343e-02
1.83327869e-01 -1.26932487e-01 -2.58416593e-01 -1.02087307e+00
2.99020380e-01 -9.00445938e-01 4.52490896e-01 -3.32318515e-01
-1.26834381e+00 8.10807228e-01 -1.01585843e-01 -1.33805001e+00
2.00554729e-01 -5.83877981e-01 -4.89860266e-01 7.66297698e-01
-1.75201440e+00 -1.55535543e+00 -5.42344153e-01 5.87991357e-01
6.00958884e-01 -4.12522048e-01 4.34883773e-01 7.57944763e-01
-1.21520793e+00 1.21388125e+00 -1.64235875e-01 4.76026773e-01
1.17037082e+00 -8.56889427e-01 3.86211812e-01 9.91584182e-01
3.29583138e-01 6.89096689e-01 1.79425210e-01 -4.70567614e-01
-1.26224768e+00 -1.33185780e+00 7.16821849e-01 -2.18417868e-01
5.93822837e-01 -3.99792761e-01 -9.07280862e-01 6.92812502e-01
1.09125145e-01 4.12356257e-01 7.45133281e-01 1.84861869e-01
-6.98358357e-01 -5.63435376e-01 -8.11209857e-01 7.04170585e-01
1.32712054e+00 -2.74468780e-01 -1.11581376e-02 5.25036097e-01
6.30974710e-01 -7.05442727e-01 -6.30886734e-01 6.44016445e-01
5.90722322e-01 -1.04958284e+00 9.43696380e-01 -4.17524755e-01
5.30369163e-01 -4.67076570e-01 -1.93817779e-01 -8.69475722e-01
-3.97700608e-01 -6.69061422e-01 -6.54761344e-02 1.64862478e+00
5.40971458e-01 -6.64356589e-01 7.33201206e-01 5.81053495e-01
-1.30118534e-01 -1.24035835e+00 -9.80740964e-01 -5.75244129e-01
-1.54311791e-01 -3.03375602e-01 7.12458074e-01 9.49788690e-01
-4.62901354e-01 3.09207737e-01 -6.56905472e-01 3.43539923e-01
5.22938251e-01 1.71976864e-01 1.02306080e+00 -9.87695217e-01
-4.24373196e-03 -6.33062959e-01 -4.48895901e-01 -1.25926113e+00
1.57601342e-01 -9.17615950e-01 -1.34921223e-01 -1.21409452e+00
4.45528448e-01 -7.12286532e-01 -3.86546999e-01 8.71835232e-01
-3.72219950e-01 7.44582772e-01 3.59448284e-01 4.01473701e-01
-6.72745645e-01 4.15981114e-01 1.17451155e+00 -3.23213115e-02
-4.85785380e-02 -5.04472032e-02 -7.70203888e-01 7.70948470e-01
8.30237508e-01 -4.38008815e-01 -1.39824584e-01 -6.06275260e-01
1.10459939e-01 -3.12090665e-01 3.00637186e-01 -9.62281644e-01
5.64490914e-01 4.17928547e-02 6.49648786e-01 -5.71125805e-01
3.10873836e-01 -7.03795969e-01 9.55525264e-02 8.91515762e-02
-1.29392773e-01 2.15481266e-01 2.80673623e-01 4.18767333e-01
-2.60617435e-01 1.27906561e-01 9.03476059e-01 3.64754766e-01
-6.80060685e-01 8.48434985e-01 7.16443598e-01 -1.12652697e-03
1.03031194e+00 -9.35610384e-02 -4.70244706e-01 -7.79973269e-02
-2.04623923e-01 3.92397106e-01 2.20035091e-01 5.92301905e-01
5.09292841e-01 -1.58006072e+00 -8.49453211e-01 2.53381997e-01
1.22316524e-01 8.11159685e-02 2.20546141e-01 1.11065078e+00
-3.71182919e-01 3.10813427e-01 -1.32283956e-01 -6.47174060e-01
-1.37861514e+00 1.96023345e-01 2.76525140e-01 -2.05459297e-01
-5.05005002e-01 1.07288110e+00 2.03903243e-01 -3.81803811e-01
3.85656357e-01 1.14360144e-02 -1.54415637e-01 2.37420440e-01
7.64208078e-01 3.24553519e-01 1.06049448e-01 -8.43528390e-01
-3.89591992e-01 9.58623588e-01 -4.84265566e-01 2.46611506e-01
1.38843262e+00 5.31697273e-03 -1.98958620e-01 -1.84059478e-02
1.27589178e+00 2.99980789e-01 -1.38682854e+00 -5.26027024e-01
2.48329453e-02 -6.42269015e-01 -1.12447321e-01 -4.69552338e-01
-1.62609673e+00 7.53925800e-01 6.01299763e-01 -2.26965696e-01
1.31197929e+00 -6.30774125e-02 9.09563065e-01 2.52663232e-02
2.17896342e-01 -8.73027623e-01 1.48715079e-01 2.47672573e-01
8.90554309e-01 -1.49552703e+00 2.02732116e-01 -7.38367617e-01
-4.40289348e-01 1.08528197e+00 1.05974317e+00 2.07340971e-01
6.16014302e-01 3.41882259e-01 1.16671488e-01 -8.83750096e-02
-3.86819154e-01 -1.95288718e-01 3.65653932e-01 1.56078875e-01
5.05562127e-01 -2.02010289e-01 -2.03368753e-01 4.86459762e-01
-2.76035871e-02 -1.03450269e-01 -1.87281352e-02 6.42700911e-01
-1.91186354e-01 -1.20334363e+00 -2.71133363e-01 3.34385484e-01
-5.42955101e-01 -1.78983405e-01 -1.45108670e-01 9.39381003e-01
3.43855530e-01 7.64142394e-01 -7.57409409e-02 -5.01751304e-01
3.43855977e-01 -1.64924949e-01 2.60678560e-01 -4.20590013e-01
-4.14078951e-01 2.69356757e-01 -1.41935781e-01 -7.34039366e-01
-4.89877105e-01 -6.69965684e-01 -1.05800593e+00 -3.82692099e-01
-5.30316949e-01 -1.91539511e-01 4.47606653e-01 8.17147911e-01
5.25120199e-01 3.42393309e-01 7.47722864e-01 -7.48208106e-01
-3.39655399e-01 -1.14006341e+00 -2.70550340e-01 1.84673131e-01
1.48731083e-01 -8.43997896e-01 -1.63132831e-01 -2.50834525e-01] | [13.434258460998535, 0.5031903982162476] |
f7767a1b-1710-4bf7-91cb-54f5e0202a2e | trans-dimensional-generative-modeling-via | 2305.16261 | null | https://arxiv.org/abs/2305.16261v1 | https://arxiv.org/pdf/2305.16261v1.pdf | Trans-Dimensional Generative Modeling via Jump Diffusion Models | We propose a new class of generative models that naturally handle data of varying dimensionality by jointly modeling the state and dimension of each datapoint. The generative process is formulated as a jump diffusion process that makes jumps between different dimensional spaces. We first define a dimension destroying forward noising process, before deriving the dimension creating time-reversed generative process along with a novel evidence lower bound training objective for learning to approximate it. Simulating our learned approximation to the time-reversed generative process then provides an effective way of sampling data of varying dimensionality by jointly generating state values and dimensions. We demonstrate our approach on molecular and video datasets of varying dimensionality, reporting better compatibility with test-time diffusion guidance imputation tasks and improved interpolation capabilities versus fixed dimensional models that generate state values and dimensions separately. | ['Arnaud Doucet', 'Tom Rainforth', 'Valentin De Bortoli', 'Christian Weilbach', 'William Harvey', 'Andrew Campbell'] | 2023-05-25 | null | null | null | null | ['imputation', 'imputation', 'imputation'] | ['computer-vision', 'miscellaneous', 'time-series'] | [ 2.31183782e-01 1.75741911e-01 -3.05142462e-01 -2.25329906e-01
-7.62859941e-01 -7.01205969e-01 1.16990280e+00 -3.70204329e-01
-2.85696894e-01 1.03594530e+00 3.93026114e-01 -4.61290419e-01
-4.86828983e-01 -7.60054767e-01 -1.01952434e+00 -1.11888468e+00
-2.70079315e-01 1.27196145e+00 -1.19125858e-01 2.44134232e-01
2.88302541e-01 6.08388007e-01 -1.14917231e+00 7.03082234e-02
7.16147423e-01 5.09648979e-01 6.28866851e-02 1.18058574e+00
-1.72306806e-01 3.46720517e-01 -5.63551843e-01 -3.25846761e-01
4.60231692e-01 -7.53765523e-01 -5.84171057e-01 1.24743648e-01
2.37307072e-01 -5.78550756e-01 -4.19849545e-01 5.64134181e-01
5.53166151e-01 1.31423116e-01 1.27473772e+00 -1.13897264e+00
-9.78950799e-01 5.17739952e-01 -4.08704937e-01 2.27189749e-01
1.81255162e-01 4.62454557e-01 5.20361245e-01 -4.26221073e-01
1.12305593e+00 1.17214453e+00 6.50202692e-01 9.56658185e-01
-1.93597412e+00 -3.67001414e-01 7.25246742e-02 -3.84038895e-01
-9.12360251e-01 -1.94718063e-01 7.15905547e-01 -7.05982387e-01
9.97588694e-01 6.79567680e-02 8.96529555e-01 1.93103480e+00
4.23316449e-01 6.91454232e-01 9.97618139e-01 2.22424254e-01
7.29733229e-01 -4.99326885e-01 -1.45840287e-01 4.56610739e-01
1.39339074e-01 4.30274874e-01 -3.98838401e-01 -6.18560314e-01
1.13511443e+00 1.86145529e-01 7.65593052e-02 -7.35818148e-01
-1.38956404e+00 7.26814747e-01 -1.60387069e-01 -2.31860787e-01
-5.99454701e-01 5.01409471e-01 2.23795786e-01 2.20721066e-01
3.62870395e-01 3.00424159e-01 -4.42186534e-01 -5.62525094e-01
-9.95945334e-01 6.37199581e-01 9.21099961e-01 9.89144444e-01
4.19574380e-01 8.82388130e-02 -4.84693944e-01 4.31307405e-01
2.59839505e-01 6.07718408e-01 4.51306731e-01 -1.16212237e+00
4.24957782e-01 1.72111467e-01 3.60782117e-01 -3.06136161e-01
-4.01290327e-01 -4.78636503e-01 -1.15884852e+00 1.05088085e-01
5.22021294e-01 -3.95628929e-01 -1.26701772e+00 2.10542297e+00
5.45543790e-01 6.21325135e-01 1.03812605e-01 4.53808665e-01
9.01881903e-02 7.92815447e-01 1.59144655e-01 -3.03979933e-01
8.95649314e-01 -4.39995021e-01 -6.46118343e-01 1.83080971e-01
6.79951906e-01 -1.33395731e-01 8.63785207e-01 5.50851882e-01
-1.41194773e+00 -3.02610278e-01 -7.06584811e-01 -1.76996931e-01
-2.57261485e-01 -1.84145689e-01 8.38061035e-01 5.25170982e-01
-1.07697034e+00 8.85744095e-01 -1.27500498e+00 -7.24341208e-03
6.78027987e-01 1.69988394e-01 -1.56903595e-01 2.44713962e-01
-8.42258036e-01 5.02941430e-01 1.15175452e-03 -1.60558924e-01
-1.41750646e+00 -1.10869348e+00 -3.70884478e-01 -3.01694065e-01
-2.41453052e-01 -1.39942086e+00 7.24242926e-01 -6.11333586e-02
-1.42149436e+00 6.90564454e-01 -3.02826226e-01 -8.64177346e-01
9.39999640e-01 -1.85428653e-02 -1.56377524e-01 -2.72902787e-01
-4.26864699e-02 8.03153515e-01 1.25521290e+00 -1.19863892e+00
-1.05116285e-01 -4.60383624e-01 -4.28665608e-01 1.05175413e-01
1.22010216e-01 -9.00370896e-01 -3.76064956e-01 -7.28525460e-01
1.45463526e-01 -1.16122055e+00 -3.16922933e-01 4.16260622e-02
-7.46824205e-01 6.70594797e-02 5.15070617e-01 -7.51485646e-01
1.25083363e+00 -1.52376795e+00 9.32453096e-01 8.23978111e-02
2.94582576e-01 -2.30021626e-02 -3.18177104e-01 6.10617280e-01
-1.91466976e-02 2.11103737e-01 -6.05405390e-01 -7.48887300e-01
2.76723318e-02 6.34328425e-01 -5.78885078e-01 3.58198524e-01
2.92360127e-01 9.46158886e-01 -1.00833690e+00 4.56001535e-02
-9.62736979e-02 7.74643242e-01 -8.31336856e-01 1.82158440e-01
-5.07533967e-01 9.12729144e-01 -4.23218757e-01 4.34910476e-01
7.38226891e-01 -1.08747259e-01 1.09128840e-01 4.43518832e-02
1.45283520e-01 9.71730500e-02 -1.17954516e+00 1.95097268e+00
-3.20412934e-01 2.98575759e-01 -4.89693582e-01 -6.03781283e-01
8.29543173e-01 -6.72495365e-02 4.73749965e-01 -3.89990211e-01
-2.13294163e-01 -1.76573306e-01 5.22042885e-02 -3.86349976e-01
4.09499824e-01 -2.96719193e-01 4.64697406e-02 6.87812507e-01
1.34435102e-01 -2.55628943e-01 1.20959915e-01 2.92970002e-01
1.38931239e+00 6.65399194e-01 -3.05721194e-01 5.24417162e-02
3.96563783e-02 -2.22160444e-01 2.54364103e-01 8.50385964e-01
1.25905648e-01 5.10198534e-01 9.40401375e-01 -3.90111566e-01
-1.72260356e+00 -1.55271232e+00 -3.86565357e-01 5.53478360e-01
-3.47660661e-01 -2.10812584e-01 -6.55106366e-01 -6.19841456e-01
3.00980717e-01 8.66100729e-01 -1.08361828e+00 -3.31000835e-01
-4.20654237e-01 -1.09922314e+00 7.23309875e-01 4.28902179e-01
7.82003254e-02 -5.35938621e-01 -1.51034594e-01 3.04284066e-01
1.09759890e-01 -4.02236581e-01 -3.29593748e-01 4.74877238e-01
-1.32642806e+00 -7.44957387e-01 -9.05904412e-01 -1.76285744e-01
6.85982883e-01 -4.18832004e-01 1.05446005e+00 -4.96039897e-01
-3.40131730e-01 3.46218497e-01 2.40379661e-01 -2.20322892e-01
-7.61449933e-01 6.82700127e-02 8.28253403e-02 -2.60847390e-01
7.85147920e-02 -7.29391098e-01 -7.34806120e-01 9.06368122e-02
-1.23370969e+00 1.05031408e-01 3.25158119e-01 1.04490471e+00
1.03929591e+00 -2.86213607e-01 5.20765305e-01 -8.17041457e-01
1.01731682e+00 -8.94249856e-01 -4.96782660e-01 -8.99703875e-02
-7.32383609e-01 6.89260662e-01 4.14919823e-01 -8.54918659e-01
-9.78833973e-01 -2.33256947e-02 -3.26419622e-01 -7.02200413e-01
-2.12681159e-01 -7.65217617e-02 9.92002636e-02 5.05831242e-01
6.18970811e-01 6.30512953e-01 2.76519626e-01 -5.65536797e-01
6.73150301e-01 8.19741189e-02 3.86584759e-01 -9.46602464e-01
7.42323577e-01 8.90019774e-01 6.91264451e-01 -5.77020288e-01
-4.21477973e-01 5.20978831e-02 -6.92620575e-01 1.18006431e-01
9.42256272e-01 -7.32942641e-01 -7.11198807e-01 6.22451961e-01
-8.95436347e-01 -7.46390283e-01 -9.17630076e-01 4.17427570e-01
-1.20087838e+00 2.24991515e-01 -6.21040881e-01 -7.79824913e-01
-1.13017410e-02 -9.69164968e-01 1.41108227e+00 -2.15153635e-01
-1.10860847e-01 -1.44011045e+00 9.17760491e-01 -1.45840615e-01
3.32058638e-01 4.91243035e-01 1.00102758e+00 -4.28241879e-01
-7.38108039e-01 -1.44807518e-01 3.32457006e-01 2.07245350e-01
3.59160826e-02 1.05152808e-01 -6.46975875e-01 -1.67254820e-01
-6.77107424e-02 7.54356235e-02 1.02911341e+00 8.53717685e-01
1.19593155e+00 -3.91122013e-01 -4.68469977e-01 1.07530534e+00
1.09149015e+00 2.32402608e-01 8.50500643e-01 2.04640374e-01
5.80883741e-01 3.44770998e-01 2.07233697e-01 6.47760451e-01
3.87034088e-01 4.15208042e-01 4.06585425e-01 7.56349713e-02
-1.57459483e-01 -7.10487187e-01 1.87868893e-01 2.78523743e-01
-2.72621542e-01 -5.59694290e-01 -7.67785132e-01 4.50624645e-01
-1.67632210e+00 -1.41069007e+00 -2.53909826e-01 2.26358342e+00
8.42945039e-01 3.45858559e-02 5.31877339e-01 -1.89379409e-01
2.89729357e-01 -7.64119178e-02 -1.19149005e+00 -2.92529404e-01
-2.14678898e-01 1.68516308e-01 4.27537411e-01 5.41233182e-01
-6.74757540e-01 4.60671812e-01 7.66524220e+00 4.68344450e-01
-7.28522182e-01 8.00856277e-02 7.33175576e-01 -4.77626741e-01
-7.86258578e-01 -3.11679598e-02 -1.02097237e+00 8.21967065e-01
1.27863753e+00 7.54179899e-03 5.40301919e-01 4.32639033e-01
2.78655469e-01 2.35165700e-01 -1.28508186e+00 7.88165033e-01
-3.84093791e-01 -1.82759655e+00 4.83452231e-01 7.67283499e-01
1.02570939e+00 7.87322074e-02 4.88634884e-01 2.09665149e-01
7.37830043e-01 -8.81000876e-01 4.44281548e-01 1.25545180e+00
5.77106357e-01 -7.12821066e-01 1.59809768e-01 6.16140842e-01
-4.70918447e-01 1.72858220e-02 -4.27834213e-01 1.53675094e-01
4.73892510e-01 6.37709379e-01 -8.91282320e-01 1.87314019e-01
1.02282017e-01 6.36260867e-01 -1.12362862e-01 7.74667978e-01
-7.00911926e-03 6.32178962e-01 -3.89772058e-01 3.49802941e-01
1.00375757e-01 -6.94163442e-01 7.43352950e-01 9.56195414e-01
7.26873100e-01 -2.81761527e-01 -3.34762812e-01 1.14731252e+00
-1.07691772e-02 -6.30257666e-01 -7.64161766e-01 -2.38630176e-01
4.99647081e-01 6.62399590e-01 -4.60374951e-01 -3.39105546e-01
1.34656787e-01 1.25530124e+00 3.29260558e-01 6.37226522e-01
-1.07814419e+00 2.07769796e-01 1.08390892e+00 2.86447495e-01
3.69823456e-01 -5.47099352e-01 -1.26536682e-01 -1.15304816e+00
-3.11588973e-01 -4.77514446e-01 3.49782318e-01 -6.69244468e-01
-1.55643666e+00 3.67799729e-01 8.13265964e-02 -1.19994497e+00
-7.09993243e-01 -2.51365960e-01 -4.43986684e-01 1.28853738e+00
-1.19313669e+00 -9.89455998e-01 2.72182733e-01 4.04751360e-01
3.68399501e-01 -2.23387367e-04 7.15262651e-01 -1.19088061e-01
-4.90499228e-01 3.91741663e-01 5.76497078e-01 -4.97016817e-01
4.45617378e-01 -1.40888441e+00 9.28132892e-01 3.93238753e-01
-1.44714400e-01 9.96207595e-01 8.18527520e-01 -1.09761727e+00
-1.76052463e+00 -1.07202291e+00 2.44958311e-01 -1.08649373e+00
7.83971250e-01 -4.04940158e-01 -1.03181541e+00 7.93966949e-01
-2.19810739e-01 -3.24811906e-01 7.57949769e-01 -3.14456895e-02
-2.44827792e-01 4.05770332e-01 -1.43614733e+00 5.73631406e-01
1.52505744e+00 -4.80230063e-01 -4.45120424e-01 4.01884168e-01
4.82068449e-01 -6.34067357e-01 -1.14546168e+00 -7.85540640e-02
4.69670177e-01 -7.49604404e-01 1.39751661e+00 -1.15778434e+00
4.37377542e-01 -1.71008483e-01 -5.66403903e-02 -1.39873731e+00
-4.18706417e-01 -8.38050663e-01 -9.93386209e-01 1.10642326e+00
2.93574959e-01 -6.48280442e-01 9.99678075e-01 5.08482575e-01
3.62623408e-02 -8.55885386e-01 -9.88067806e-01 -9.76037741e-01
4.32745785e-01 -4.94861603e-01 8.31933737e-01 6.88038588e-01
-5.83453715e-01 1.77293226e-01 -6.45794451e-01 1.21926792e-01
1.09113443e+00 -1.73776075e-01 8.72122049e-01 -1.14816701e+00
-7.01516390e-01 -2.08940431e-01 -3.44974369e-01 -1.46882093e+00
-5.58763975e-03 -9.16336656e-01 -3.60287964e-01 -1.61157620e+00
1.44147828e-01 -4.87321734e-01 1.66937098e-01 1.35951687e-03
5.31418733e-02 -9.79130492e-02 -2.29390845e-01 4.47777301e-01
-9.67627838e-02 6.48903668e-01 1.65939355e+00 1.50306702e-01
-3.69879633e-01 3.72193307e-02 -4.09156799e-01 3.44004393e-01
4.76702899e-01 -4.90612775e-01 -8.79602194e-01 -3.85013938e-01
4.49548692e-01 3.10878962e-01 5.68709016e-01 -7.93988168e-01
2.62508140e-04 -2.59918600e-01 7.99865246e-01 -8.65161657e-01
5.98100781e-01 -3.95528644e-01 8.46925616e-01 5.92115998e-01
-5.93969464e-01 -1.59248151e-02 3.02972663e-02 1.12175441e+00
4.55489218e-01 9.80042666e-02 5.63781917e-01 -3.10726855e-02
-1.71054468e-01 6.11091495e-01 -3.79992455e-01 6.60234541e-02
1.05996633e+00 -3.09865147e-01 -2.94688284e-01 -3.59755218e-01
-1.40738869e+00 -1.70581683e-03 7.15926111e-01 2.80981958e-01
4.87071753e-01 -1.30744457e+00 -4.47335392e-01 4.73592639e-01
-2.65308440e-01 -3.95806991e-02 5.21114707e-01 6.66866362e-01
-9.54288021e-02 1.30430952e-01 -1.57650843e-01 -8.05231750e-01
-6.01652741e-01 6.91051424e-01 3.90000284e-01 -3.32759976e-01
-5.43893397e-01 7.46604741e-01 -3.06208171e-02 -5.99939823e-01
-4.92062867e-02 -5.39822340e-01 3.44338387e-01 1.82005152e-01
4.02830064e-01 6.48390353e-01 -2.30767041e-01 -2.54794389e-01
1.07953899e-01 3.83178324e-01 6.90337569e-02 -3.17028522e-01
1.42242229e+00 -3.55921954e-01 3.30928802e-01 6.55692339e-01
1.02851474e+00 -3.63575220e-01 -1.97885644e+00 2.63113290e-01
-3.60035151e-01 -4.02886391e-01 -1.66337639e-01 -8.69459093e-01
-4.89299923e-01 7.99328685e-01 7.22701609e-01 3.31828117e-01
6.33681655e-01 1.86972946e-01 6.73547745e-01 1.24376051e-01
3.73038560e-01 -8.25851262e-01 5.06584346e-02 3.29690516e-01
7.46027768e-01 -7.66535223e-01 -2.25197479e-01 1.88058376e-01
-5.23644686e-01 7.64939129e-01 6.08228818e-02 -2.66402394e-01
5.31829655e-01 5.25367141e-01 -5.70753098e-01 -3.06709886e-01
-1.10683346e+00 -8.03905539e-03 1.65880561e-01 1.26032901e+00
2.65714079e-01 -3.32652628e-02 -1.70541942e-01 3.18319798e-01
-3.42463329e-02 2.30466768e-01 4.45994973e-01 8.32882643e-01
-1.17539071e-01 -1.34800148e+00 -2.76266903e-01 7.14883208e-01
-2.32264306e-02 -1.20498880e-03 -7.68836662e-02 7.05996156e-01
4.51421216e-02 1.91363707e-01 4.96878177e-01 -3.84653881e-02
2.19047770e-01 3.58078927e-01 7.19547629e-01 -2.48365253e-01
-1.64572611e-01 1.00780427e-01 -8.06490257e-02 -6.25876427e-01
-1.11157618e-01 -9.89834070e-01 -1.26901937e+00 -4.97484565e-01
2.08781451e-01 -2.81559732e-02 9.00744736e-01 6.04239285e-01
8.28969896e-01 6.72887444e-01 3.67154121e-01 -9.14551258e-01
-8.34505379e-01 -6.26972616e-01 -5.38235009e-01 5.47008932e-01
5.40247142e-01 -5.87723613e-01 -4.87663656e-01 2.13569537e-01] | [6.807392597198486, 3.9281530380249023] |
0bab169c-af19-440f-98eb-5d5b95c69533 | dmcnn-dual-domain-multi-scale-convolutional | 1806.03275 | null | http://arxiv.org/abs/1806.03275v2 | http://arxiv.org/pdf/1806.03275v2.pdf | DMCNN: Dual-Domain Multi-Scale Convolutional Neural Network for Compression Artifacts Removal | JPEG is one of the most commonly used standards among lossy image compression
methods. However, JPEG compression inevitably introduces various kinds of
artifacts, especially at high compression rates, which could greatly affect the
Quality of Experience (QoE). Recently, convolutional neural network (CNN) based
methods have shown excellent performance for removing the JPEG artifacts. Lots
of efforts have been made to deepen the CNNs and extract deeper features, while
relatively few works pay attention to the receptive field of the network. In
this paper, we illustrate that the quality of output images can be
significantly improved by enlarging the receptive fields in many cases. One
step further, we propose a Dual-domain Multi-scale CNN (DMCNN) to take full
advantage of redundancies on both the pixel and DCT domains. Experiments show
that DMCNN sets a new state-of-the-art for the task of JPEG artifact removal. | ['Yueyu Hu', 'Jiaying Liu', 'Xiaoshuai Zhang', 'Wenhan Yang'] | 2018-06-08 | null | null | null | null | ['jpeg-artifact-correction', 'jpeg-artifact-removal'] | ['computer-vision', 'computer-vision'] | [ 3.03878814e-01 -5.18421352e-01 -2.64287312e-02 -2.47365654e-01
-3.86210620e-01 8.00099373e-02 1.00380860e-01 -7.22531751e-02
-3.91632736e-01 5.46215892e-01 3.92563879e-01 4.01215442e-02
1.01953760e-01 -1.04601943e+00 -6.12002909e-01 -5.50359190e-01
2.08955318e-01 -6.06104553e-01 2.33803257e-01 -4.82787907e-01
4.26501572e-01 4.42516714e-01 -1.33451331e+00 4.91353273e-01
8.05850327e-01 1.09139013e+00 2.07031086e-01 4.32870239e-01
1.72220483e-01 9.55544710e-01 -6.83067441e-01 -5.00545681e-01
2.64457941e-01 -4.78377134e-01 -4.77552325e-01 -1.73911341e-02
3.05542380e-01 -8.23897243e-01 -1.01201296e+00 1.61939228e+00
7.65763104e-01 -2.07142681e-02 1.62168354e-01 -9.06328559e-01
-9.48932767e-01 3.74583334e-01 -6.84452832e-01 4.86992925e-01
-4.24336866e-02 2.34978721e-01 7.49611080e-01 -7.38934100e-01
3.89970660e-01 1.12062442e+00 7.28497922e-01 2.86098778e-01
-8.07585418e-01 -9.26064134e-01 -2.30141774e-01 6.82977736e-01
-1.32525671e+00 -4.09444124e-01 9.62395310e-01 1.15147218e-01
9.43372726e-01 4.51667905e-02 7.85988033e-01 1.03409386e+00
4.26575631e-01 7.78587282e-01 7.58090317e-01 -1.91857219e-01
-9.25810449e-03 -3.13317627e-01 -3.86006087e-01 3.58618557e-01
3.12734395e-01 2.09222722e-04 -5.67352057e-01 2.80681998e-01
1.20550001e+00 4.04351443e-01 -6.63278461e-01 -6.36655912e-02
-9.50615942e-01 6.88445628e-01 7.56286263e-01 4.90905225e-01
-3.22454393e-01 3.81158471e-01 6.94811940e-01 4.03843611e-01
3.47167045e-01 1.75069764e-01 -2.10660025e-01 -2.44422659e-01
-1.08736610e+00 1.02373980e-01 1.96436271e-01 8.50113988e-01
5.57264745e-01 1.49317011e-01 -1.11376822e-01 1.11208487e+00
5.63150533e-02 2.18439445e-01 6.96740925e-01 -1.11272526e+00
6.87971532e-01 4.83150631e-01 -1.20872594e-01 -1.41719162e+00
-2.54895650e-02 -6.47339106e-01 -1.49363363e+00 3.34593564e-01
-1.28402442e-01 2.89928287e-01 -6.80707574e-01 1.45648289e+00
-3.24673653e-01 9.26399529e-02 -1.10320687e-01 1.10829568e+00
6.92403793e-01 8.48663509e-01 -6.71476498e-03 -9.10610333e-02
1.03633511e+00 -1.05117977e+00 -1.01136303e+00 -1.82447046e-01
2.19645843e-01 -7.83772469e-01 8.65588427e-01 6.53366327e-01
-1.33510602e+00 -1.01588678e+00 -1.48539913e+00 -4.70702380e-01
-2.33247638e-01 -2.38286555e-02 4.68871057e-01 4.60737228e-01
-9.31379318e-01 1.09629178e+00 -7.13801980e-01 1.24412142e-02
7.34530151e-01 2.99680591e-01 -3.21323961e-01 -4.42139924e-01
-1.46998942e+00 6.48062050e-01 2.60959268e-01 2.31876358e-01
-9.77638125e-01 -3.60586792e-01 -6.62316799e-01 6.71224296e-01
1.57132417e-01 -4.56061393e-01 1.04363811e+00 -1.17743289e+00
-1.19041288e+00 5.30390859e-01 2.75605738e-01 -4.12352890e-01
5.84245443e-01 -3.67724925e-01 -7.26162255e-01 3.57090026e-01
-8.57369453e-02 6.14885449e-01 1.03570032e+00 -8.11831594e-01
-4.65092301e-01 -2.54732966e-01 8.45773742e-02 4.89304550e-02
-9.08357739e-01 -2.13628393e-02 -6.91135168e-01 -1.15870309e+00
-1.77949797e-02 -3.81694764e-01 -1.28412873e-01 3.73433977e-01
-1.49816498e-01 3.17386072e-03 8.08844924e-01 -9.49396849e-01
1.49479306e+00 -2.51089549e+00 2.22663619e-02 -2.18194813e-01
3.63723993e-01 5.92775047e-01 -3.20087433e-01 2.89765716e-01
-1.38037115e-01 3.47086549e-01 -3.50868106e-01 -6.14005141e-02
-2.64808148e-01 4.81328405e-02 -1.00024782e-01 4.49777991e-01
3.02179635e-01 6.94416165e-01 -6.72827125e-01 -4.67850268e-01
4.07626212e-01 7.29307592e-01 -7.51232028e-01 1.57282725e-01
1.52895093e-01 1.18487507e-01 -3.37284625e-01 4.52285647e-01
1.14301622e+00 -3.41823995e-01 1.58267599e-02 -4.56633866e-01
-6.18311996e-03 1.37560561e-01 -9.33877409e-01 1.97857499e+00
-5.14686465e-01 8.54709506e-01 4.00308566e-03 -1.03227806e+00
7.19709635e-01 2.77777612e-01 5.52545011e-01 -1.19079268e+00
2.96065122e-01 2.29622945e-01 7.34211430e-02 -6.61059201e-01
5.58889627e-01 -7.73804858e-02 2.71698922e-01 -1.91264208e-02
8.80903751e-02 1.07234456e-01 -1.67485699e-02 1.08299190e-02
1.09912360e+00 -1.43860996e-01 3.31013888e-01 4.74827774e-02
5.29712796e-01 -5.67357540e-01 7.23285437e-01 5.48449934e-01
-5.41619539e-01 1.08938694e+00 2.31129721e-01 -6.53174162e-01
-1.36903369e+00 -8.03578138e-01 -4.17915173e-02 6.14593387e-01
5.28529465e-01 -4.36083883e-01 -6.99843109e-01 -3.23816359e-01
-1.75515786e-01 2.67411232e-01 -4.04738009e-01 -5.79379261e-01
-7.79461622e-01 -5.82669616e-01 7.00769305e-01 5.60829222e-01
1.32537806e+00 -1.33011460e+00 -5.51021397e-01 5.12794197e-01
-5.30258298e-01 -1.19014788e+00 -5.59413195e-01 2.00918559e-02
-1.16351116e+00 -9.39379275e-01 -9.63810921e-01 -8.58091772e-01
4.11338270e-01 3.90990913e-01 8.93508315e-01 4.56622690e-01
-3.85473609e-01 -2.93969631e-01 -3.61040950e-01 -1.19436570e-01
-2.66685486e-01 -3.12640145e-02 -3.34789306e-01 -9.87135023e-02
3.50897938e-01 -8.36875379e-01 -9.54735160e-01 5.90537749e-02
-1.40608978e+00 9.65998992e-02 8.01118851e-01 7.35353887e-01
3.92752439e-01 6.51133001e-01 4.17774051e-01 -6.65320635e-01
7.84544051e-01 -1.95721522e-01 -2.75525272e-01 -1.34062111e-01
-5.50134838e-01 1.68973990e-02 1.09938955e+00 -1.88033849e-01
-8.97093713e-01 -3.74355048e-01 -4.95158225e-01 -5.51618874e-01
-1.44544318e-01 4.17196125e-01 -3.24926317e-01 -1.82094976e-01
3.36850137e-01 3.85938644e-01 -2.50186682e-01 -7.28124380e-01
-1.17571704e-01 7.10425735e-01 5.56373239e-01 3.81667763e-02
6.28638208e-01 4.24924046e-01 2.25781486e-03 -6.29790366e-01
-4.65048999e-01 -2.07109168e-01 -2.88124442e-01 -9.44931805e-02
7.07915843e-01 -1.09769893e+00 -5.60238540e-01 7.31193364e-01
-1.14668167e+00 6.61891177e-02 -6.65499493e-02 3.42083544e-01
-3.24311882e-01 7.82109499e-01 -8.87627244e-01 -2.75761724e-01
-4.08032268e-01 -1.30432701e+00 6.71208501e-01 3.59037220e-01
2.22648114e-01 -5.61403632e-01 -2.67564058e-01 1.75059080e-01
8.67866337e-01 -8.81989822e-02 1.01823986e+00 -4.99245077e-02
-6.40210509e-01 -1.34139478e-01 -7.10138679e-01 8.71046245e-01
1.64110363e-01 -3.34968686e-01 -9.51464534e-01 -4.94208843e-01
3.06398541e-01 -2.57463187e-01 1.29084873e+00 5.73845863e-01
1.89378977e+00 -3.91571641e-01 7.16726407e-02 1.03705657e+00
1.81147575e+00 1.51483223e-01 1.31377578e+00 5.64434290e-01
5.14066577e-01 -5.87980449e-02 2.58660078e-01 6.48534119e-01
-6.88162446e-02 4.59126651e-01 7.67738819e-01 -4.02409524e-01
-5.19397676e-01 -2.38986090e-01 2.68827260e-01 1.00616717e+00
-3.01096469e-01 -3.67719978e-01 -3.78272146e-01 5.34950972e-01
-1.58232605e+00 -9.85327542e-01 1.80585355e-01 1.87178779e+00
8.86116803e-01 3.74084711e-01 -3.68585318e-01 4.85729367e-01
7.92426288e-01 5.03227949e-01 -5.27753234e-01 -3.38980585e-01
-2.27554411e-01 4.13479090e-01 6.02330804e-01 5.11867478e-02
-1.22343683e+00 5.15822351e-01 6.13061047e+00 1.07390547e+00
-1.16387701e+00 -1.99234821e-02 7.40682960e-01 -1.17258634e-02
-1.17120016e-02 -1.53368115e-01 -2.83896178e-01 7.20463157e-01
7.54434943e-01 9.56749022e-02 5.74167550e-01 8.31710994e-01
2.90939420e-01 9.17553753e-02 -7.35258698e-01 1.29410541e+00
2.09129192e-02 -1.27640355e+00 2.83822119e-01 2.38476321e-02
7.55451143e-01 -6.17679320e-02 1.95869863e-01 1.48810565e-01
-1.90667391e-01 -1.13024700e+00 5.24215758e-01 3.11230838e-01
1.10224891e+00 -9.57537055e-01 1.14002419e+00 1.62196174e-01
-1.09027696e+00 -2.31451437e-01 -8.84927869e-01 -9.22435001e-02
2.62620803e-02 7.61780441e-01 5.74578866e-02 4.31399584e-01
1.18574846e+00 1.17790484e+00 -5.97261786e-01 1.37077653e+00
-9.09277201e-02 5.10174394e-01 5.99214211e-02 2.29176879e-01
2.38587737e-01 -1.35746170e-02 2.94045001e-01 1.27285171e+00
5.19946098e-01 1.89561788e-02 -2.44064853e-01 7.92060554e-01
-7.22813189e-01 -5.09588420e-02 -4.52532172e-01 5.79163432e-02
8.17203149e-02 9.93722141e-01 -4.29002494e-01 -4.28932190e-01
-6.29433036e-01 1.32176197e+00 5.30717941e-03 3.25557649e-01
-7.20770001e-01 -8.89249027e-01 6.71141744e-01 4.67905663e-02
6.09352946e-01 -3.50113697e-02 -2.09737599e-01 -1.39630473e+00
1.47624716e-01 -1.08343697e+00 1.63398020e-03 -7.96138883e-01
-1.13642478e+00 3.13687265e-01 -5.60672462e-01 -1.65343976e+00
3.63389969e-01 -5.64076662e-01 -3.84158403e-01 8.44236434e-01
-2.07363153e+00 -5.82718492e-01 -6.90384805e-01 6.55351996e-01
8.75685573e-01 -5.90534182e-04 6.32037580e-01 8.43231142e-01
-3.64581823e-01 4.74934906e-01 3.80433798e-01 4.44343030e-01
8.39031458e-01 -7.50404537e-01 3.82292122e-01 1.00435436e+00
-2.21582294e-01 4.79051143e-01 4.93582875e-01 -3.97541642e-01
-1.12304318e+00 -1.20846570e+00 6.80412471e-01 4.30774420e-01
1.44330144e-01 -9.60085094e-02 -9.75631297e-01 3.45315397e-01
5.19401133e-01 3.74043435e-01 2.82213628e-01 -4.32478368e-01
-4.13281143e-01 -4.71769154e-01 -1.13884318e+00 4.91306335e-01
9.43772674e-01 -6.56782269e-01 -2.32730910e-01 3.01615391e-02
6.66842461e-01 -1.34505481e-01 -8.15853953e-01 5.29495299e-01
5.96619904e-01 -1.53570259e+00 1.16229403e+00 -2.59442866e-01
1.21537173e+00 -1.93371758e-01 -2.27832705e-01 -1.31085408e+00
-6.54159725e-01 -1.89476222e-01 -1.42571017e-01 9.24772024e-01
-1.89898729e-01 -1.87079757e-01 6.14523053e-01 -6.97672293e-02
-2.63211340e-01 -5.84798574e-01 -8.92022312e-01 -5.62474072e-01
-1.00333095e-01 -2.86268383e-01 6.94909930e-01 7.20309079e-01
-1.85890049e-01 7.96879604e-02 -7.94378638e-01 -2.71178000e-02
4.52708483e-01 -1.04602471e-01 3.30401927e-01 -9.45562959e-01
-2.49977231e-01 -5.24422705e-01 -6.16436720e-01 -1.35412979e+00
-5.11837304e-01 -6.18321419e-01 -7.61594698e-02 -1.52633321e+00
4.31943327e-01 3.60662527e-02 -8.03779244e-01 3.34045380e-01
-1.34671062e-01 4.94672179e-01 2.59807974e-01 4.07993972e-01
-6.03854418e-01 8.12952518e-01 1.68424845e+00 -4.96228546e-01
2.16549605e-01 -2.09669724e-01 -6.81286812e-01 6.84568226e-01
8.22696388e-01 -4.77277577e-01 -2.21836179e-01 -9.60170567e-01
3.19164515e-01 1.42921492e-01 2.93221235e-01 -1.44466567e+00
1.90876201e-01 3.49517241e-02 8.95629823e-01 -4.92179453e-01
2.35904589e-01 -1.02738333e+00 -1.85526051e-02 6.25539005e-01
-4.92767572e-01 9.38799679e-02 1.78175718e-01 6.43235445e-01
-8.39905143e-01 -2.55055249e-01 1.14238834e+00 -5.22249818e-01
-7.09029794e-01 3.83146703e-01 -3.56200725e-01 -2.36629158e-01
4.84780580e-01 -1.67934924e-01 -1.64906010e-01 -5.41124821e-01
-3.03772002e-01 -2.59880930e-01 4.98878449e-01 4.14852738e-01
1.07687926e+00 -1.40500450e+00 -6.40392065e-01 4.10783350e-01
-5.66112809e-02 -2.21263260e-01 5.36523879e-01 5.41611612e-01
-8.40159416e-01 3.32842052e-01 -7.80021727e-01 -1.61283866e-01
-9.93246377e-01 6.86261117e-01 4.12431926e-01 -4.04777080e-01
-7.77905047e-01 7.22771823e-01 1.37429819e-01 2.18220115e-01
3.64336848e-01 -3.39127004e-01 -3.27584386e-01 -3.87706935e-01
5.93341827e-01 4.12105262e-01 1.62368238e-01 -4.56398964e-01
-1.03132404e-01 5.91215909e-01 -2.12557703e-01 3.46010566e-01
1.54845822e+00 -2.65368283e-01 -1.31275624e-01 -1.21859543e-01
1.52193034e+00 -1.71988502e-01 -1.25532770e+00 -1.22390352e-01
-4.52989101e-01 -8.48301291e-01 3.52323234e-01 -5.93892753e-01
-1.74885821e+00 1.39744294e+00 8.74090075e-01 2.35539734e-01
1.76661623e+00 -7.21122444e-01 1.35470319e+00 1.69017017e-01
2.23928019e-01 -1.26402819e+00 3.43437880e-01 1.28273845e-01
8.88459504e-01 -1.34834492e+00 2.22237170e-01 -3.16889107e-01
-2.70089895e-01 1.52051079e+00 5.97719848e-01 -4.84972924e-01
6.47592783e-01 1.64148569e-01 -7.89159387e-02 7.67005309e-02
-2.51075625e-01 3.40340972e-01 -1.16839260e-01 5.17378509e-01
5.45682669e-01 -2.09970132e-01 -3.70867789e-01 5.82602680e-01
1.39709949e-01 3.29241693e-01 6.09960198e-01 9.84459400e-01
-5.06267428e-01 -9.76316333e-01 -1.83414042e-01 5.17807066e-01
-7.88593411e-01 -4.52896029e-01 1.42725497e-01 5.09591877e-01
3.60944062e-01 7.88449287e-01 -6.72755167e-02 -4.61100608e-01
2.93055475e-01 -4.67041105e-01 2.63837278e-01 -2.10289091e-01
-6.03749156e-01 1.07003741e-01 -3.90462309e-01 -7.91081309e-01
-4.06543136e-01 -1.42862663e-01 -9.11063612e-01 -5.07416785e-01
-2.47069314e-01 -2.78021276e-01 4.04738873e-01 6.80683017e-01
3.12963516e-01 9.21844065e-01 6.69411063e-01 -6.92582607e-01
-5.59376419e-01 -1.04551458e+00 -8.08664203e-01 8.10127079e-01
5.78987420e-01 -2.65003294e-01 -2.65935957e-01 1.53996110e-01] | [11.298881530761719, -1.7291356325149536] |
994d1d76-7bca-4cb6-80cb-35c6fc24d5fb | bspell-a-cnn-blended-bert-based-bengali-spell | 2208.09709 | null | https://arxiv.org/abs/2208.09709v1 | https://arxiv.org/pdf/2208.09709v1.pdf | BSpell: A CNN-blended BERT Based Bengali Spell Checker | Bengali typing is mostly performed using English keyboard and can be highly erroneous due to the presence of compound and similarly pronounced letters. Spelling correction of a misspelled word requires understanding of word typing pattern as well as the context of the word usage. We propose a specialized BERT model, BSpell targeted towards word for word correction in sentence level. BSpell contains an end-to-end trainable CNN sub-model named SemanticNet along with specialized auxiliary loss. This allows BSpell to specialize in highly inflected Bengali vocabulary in the presence of spelling errors. We further propose hybrid pretraining scheme for BSpell combining word level and character level masking. Utilizing this pretraining scheme, BSpell achieves 91.5% accuracy on real life Bengali spelling correction validation set. Detailed comparison on two Bengali and one Hindi spelling correction dataset shows the superiority of proposed BSpell over existing spell checkers. | ['Mohammed Eunus Ali', 'Mohammad Rafsan', 'Samiha Zakir', 'Md. Hasibur Rahman', 'Chowdhury Rafeed Rahman'] | 2022-08-20 | null | null | null | null | ['spelling-correction'] | ['natural-language-processing'] | [ 0.6060491 -0.39525712 0.24209206 -0.29798222 -0.8619814 -0.69403315
0.10670859 0.72474885 -0.94236016 0.83825207 0.09745408 -0.64333904
0.3312745 -0.6499015 -0.8995885 -0.3234925 0.6896999 0.31767565
0.38232657 -0.42833516 0.546376 0.32384443 -1.0826082 0.5977179
1.2603197 0.32326493 0.71284455 1.251577 0.16170049 0.38208833
-0.958509 -0.52385587 0.00728049 -0.13203685 -0.6203095 -0.4429477
0.7390052 0.00554568 -0.08083922 1.1278689 0.8025187 0.02428958
0.46113887 -0.4800162 -0.9978083 0.78044444 -0.35069898 0.5662169
0.21974789 -0.05726412 0.7661607 -0.96792835 0.18855432 0.93359715
1.2227626 0.6811597 -0.75642055 -0.58839893 0.03143542 0.19417475
-1.2300525 -0.14342098 0.3337153 -0.10907177 1.3869035 0.59962523
0.14369795 0.72937363 0.32152596 0.92961246 1.149624 -0.9578654
-0.19621 0.14129174 0.7211322 0.654771 0.42151064 -0.09436601
-0.6405467 0.20220178 0.26827198 -0.38830408 -0.23605558 0.4540581
-0.9769606 0.38356218 0.07662369 0.10086858 -0.11046656 0.14284159
0.68603146 0.24545512 0.226399 0.4565042 -0.80746824 -0.31034666
-0.9944185 0.48383895 0.38153124 0.9752101 0.30122864 0.43462977
-0.63783044 1.1821376 0.03473835 0.8132207 0.5934898 0.10457548
0.71521276 0.39478636 0.17318967 -0.6058806 -0.34150288 -0.8082391
-0.72949845 0.01921815 0.35036337 -0.15442827 -1.313196 1.3728948
-0.13492648 -0.14302677 -0.07770111 0.7096229 0.6709092 0.6844688
0.1526236 0.02581137 1.3350115 -1.008437 -0.79850113 -0.20587663
0.7309242 -1.0442044 1.6317931 0.5870104 -1.0477957 -0.56520337
-1.2767472 -0.32569677 -0.5475726 0.4525099 0.06268299 1.1786075
-0.5811796 0.5612772 -0.60400116 -0.10400166 0.18864574 0.32098648
-0.13167165 -0.03504763 -1.3898904 1.0675279 0.5315403 0.23752366
-0.7449302 -0.9709138 -0.8445614 -0.03657571 -0.1366691 -0.11678843
1.4034556 -0.69368666 -1.1502959 0.97533685 -0.2877402 -0.62530655
0.63356537 -0.6067158 -0.69063306 -0.57805973 0.08504374 -0.11614332
0.7164043 -0.9350822 -0.9683053 -0.23257488 -0.46152264 0.34032324
-0.14246458 0.34477982 -0.29847327 -1.3214216 -0.30660284 -0.60752076
0.21864647 -0.8548724 -0.46231103 -0.23983468 0.60164696 -1.3739808
1.6929611 -1.627546 -0.28722936 0.19377527 -0.24093038 1.2217358
-0.15289594 0.24693745 -0.03195803 -0.05017634 -0.34183332 -0.35997584
-0.04969257 0.01030187 -0.16956677 0.47379032 0.10670385 0.8498723
-0.9755373 0.0152529 0.07894435 0.3603209 -0.57295614 0.07288879
-0.14264469 0.12400525 0.37101007 0.59984624 0.92156744 0.5726671
-0.10537171 -0.02199961 -0.12843658 0.6163131 -1.0595304 1.4205681
-0.8633999 0.5672349 -0.2783006 -0.4396283 0.7428055 0.11308361
-0.68416524 -0.8381936 0.20317598 0.49197325 0.0600623 -0.22164516
0.9798144 -0.00899178 -0.24521998 0.33864957 -0.02586721 0.08304766
0.07842544 -0.01254093 0.8169883 0.22175856 0.11813015 -0.32310995
0.68915826 -0.11454319 0.7079021 1.3926198 -0.17157866 0.68086433
-0.22954251 -0.2563803 -1.1245637 -0.8311415 -0.09012124 1.5472224
-0.0394888 -0.48829496 -1.0497782 -0.6432829 -0.23981382 1.073531
-0.6005647 -0.19505271 -1.1117327 -0.932224 1.1248884 0.7826815
0.64804107 -1.1434577 -0.24331519 0.31436858 0.10972782 -0.78636044
-0.88182807 0.43335366 -0.31139877 -0.70360714 -0.70412755 -1.1345869
0.37062535 0.02483328 0.8395683 0.43236953 -0.26734394 -0.36731836
-0.63292754 -0.9595823 -0.6407867 0.4545366 0.22176774 -0.23145118
0.7300108 0.07458418 -0.26175705 -0.12078025 -0.5651863 0.10988794
0.2727181 0.8844373 0.36801162 -0.12444302 0.481389 -1.3497894
0.8434172 -0.05978024 -0.48259526 0.5141073 -0.67767173 -0.0935988
0.79848737 -0.10246576 -1.1097432 -0.25835115 -0.7068577 0.3574867
-0.2464403 0.30657226 -0.36030516 -0.02174439 0.75981724 0.6124973
-0.59254164 -0.9207726 0.07827908 1.0577972 1.0015371 -0.46159536
0.61867964 -0.2867026 -0.60062677 -0.8321978 -0.83442783 -0.36811936
-0.6064363 0.17676148 0.60278213 -0.6951212 -0.66001105 1.3250557
-1.2884154 -0.35789204 0.12287175 0.06262433 -0.19139871 0.50999737
-0.7618601 -0.75646055 -0.77139574 -1.0095046 0.6512427 0.15858471
-0.3025949 -0.8840058 -0.06063527 0.4262306 0.38457966 -0.27855176
1.1112149 -0.8445479 0.19572836 -0.5685991 -0.31223562 0.78569114
0.10529816 -0.24441798 -0.9168015 -0.46083546 -0.30457258 -0.04207581
1.0273165 0.07856881 1.0445663 -0.3301214 0.11868025 0.7134879
1.4736754 0.22502653 0.59137493 0.5500192 1.1730651 -0.10524351
0.6909813 0.22120684 0.4040014 0.6385391 -0.02314897 0.08267313
-0.5473472 -0.41991055 0.22182831 1.1025052 -0.11305443 -0.5032897
-1.1690061 0.7481221 -1.4945947 -0.60875887 -0.58718306 2.3526242
1.35118 0.28330702 0.00844384 0.66605544 0.89475274 -0.27990818
-0.19631776 -1.2524885 -0.44210896 0.72485745 1.1394762 1.1696694
-1.0672567 1.4987596 5.1052485 1.2408919 -1.0313572 0.32217416
0.19629195 0.41384655 -0.2415813 -0.44378555 -1.3139527 0.59914565
0.80604523 0.32129836 0.2720195 0.31343985 0.11477868 -0.07732508
-0.70966595 0.70284694 0.33543012 -1.3518955 0.3569574 -0.8310663
0.78318465 -0.01631554 -0.05727101 0.30615497 0.25239572 -1.2906444
0.99082655 0.18817335 0.9983157 -1.0077232 0.8451307 0.37378702
-0.6611923 0.20042545 -0.21998769 -0.3025311 -0.20036213 -0.016981
-0.907993 0.18741725 0.42555985 0.30172706 -0.7768305 1.2398943
-0.5670624 1.1617811 -0.07142317 -0.20683797 0.24720995 0.46675467
0.5114468 2.0361025 0.03730549 -0.26976636 -0.28439993 0.06267742
0.05441541 0.23297581 0.17256296 0.04696611 0.64680564 0.67506856
-0.4588116 -0.22743398 -0.13367918 1.5912343 0.48085198 0.31602022
-0.6995692 -1.0775473 0.49521136 -0.01923591 0.22555947 -0.18480079
-0.9307393 -1.0450687 -0.07268375 -1.1540995 0.1970982 -0.43699646
-0.8941497 0.5979961 -0.5486648 -0.6154817 0.58531487 -0.84871954
-0.5298881 1.4160105 -1.5106487 -1.5128201 -0.23573367 0.3453916
1.1069531 -0.41226974 0.9227267 0.49316943 -0.61070573 1.4717922
0.18252279 0.36836877 0.78899807 -1.579364 0.9779723 1.381913
-0.0497609 0.68018764 0.86756873 -0.97670895 -0.8551781 -1.4079837
1.5078737 -0.6138198 0.25923172 -0.63011205 -1.1085125 0.6159131
0.11720736 -0.51408285 0.56739795 -0.04325424 -0.37477311 -0.01679959
-1.0307153 0.6114759 0.65473485 -0.5167939 -0.8501849 0.46250445
0.6536987 -0.7130787 -0.09901915 0.33538762 0.438972 -0.41694278
0.7058187 -0.9489725 0.26726982 -0.34971765 -0.2191013 -1.5680873
-0.57559335 -0.63904554 0.20388822 1.3393342 0.61842245 -0.07442493
0.52265453 0.15308292 -0.5086171 -0.3433697 -0.50403327 -0.60801345
0.34390408 -0.704305 0.5594826 0.81162095 -0.27060288 0.225621
-0.76716244 0.42935324 0.55026674 -0.49588636 0.35070485 -0.6061664
-0.21828176 -0.2743166 -0.22398895 -0.62432617 0.01393691 -1.0364119
0.40026963 -1.2618365 -0.09491035 -0.6965629 -0.4478416 0.46639043
-0.84429634 0.80276227 0.2280275 -0.32409954 -0.22444569 0.03496081
0.8317729 -0.10273844 -0.14237037 0.25321025 -0.34455225 0.77606595
0.9736668 -0.37224805 0.03046508 -1.0623173 0.534005 -0.42990613
0.06847898 -0.98576933 0.10548773 -0.20379615 0.329472 -0.72922164
-0.2179498 -0.19697155 -0.5600941 0.4516094 -0.5265408 0.3270962
0.463805 0.34783313 0.17762044 -0.58936065 1.0935136 -0.07455071
-0.88712865 -0.1581545 -0.66965073 0.10133601 0.50048596 -0.5738187
-0.25418025 0.29703242 -0.5232787 0.01984498 0.25063208 0.54290897
0.5821355 -1.0012045 -1.0320344 0.6003343 0.2252984 -0.25615963
0.19063266 0.41980216 -1.2930341 0.4597837 -0.34127238 0.11961947
-1.6550062 0.26970932 0.3749993 0.04256431 -0.36769745 1.647519
-0.2738044 -0.65748495 0.5189798 -0.36819348 -0.06000784 -0.19873835
0.7925704 0.5695612 0.82874894 -0.7160378 -0.5556035 0.32769498
-0.64215094 0.09891047 0.9511118 -0.23189136 -0.08855544 0.45446974
0.69339466 0.6411043 -0.6522582 -0.26229408 0.09521076 -0.40210077
0.06676926 -1.4319412 -0.35180008 1.0298113 0.64235324 -0.48380002
0.7871549 -0.86565083 1.2832915 0.28169197 0.03122775 -1.4981152
-0.46085373 1.236966 0.75155795 -1.1118296 -0.38642874 -0.13880482
-0.59014976 0.86350286 0.77857965 -0.37463316 0.31144062 0.25397804
0.308083 0.35510376 -0.1915326 0.09179182 0.35565117 0.8341046
0.6866326 0.33811057 -0.8360972 0.9049016 -0.5989882 -0.2826133
0.6678351 0.9315539 -0.51750255 -1.0795116 -0.37538993 0.3589666
-0.7819062 -0.7708195 -0.23926875 0.0991613 0.5083658 0.78193843
-0.08391348 -0.3443064 0.40879524 0.09818089 0.6703455 -0.8739572
-1.4006388 -0.3673071 0.22903468 0.02373328 0.33789518 -0.46469614
-1.0821975 -0.37162414 -0.39543104 0.07244877 0.6634375 1.1254548
-0.18489084 0.61959773 0.00822539 -0.30026412 -0.51236373 -1.1920646
-0.55667186 0.49444476 0.6701852 -0.10866605 0.3189364 0.01817822] | [10.95927906036377, 10.778428077697754] |
3bc0d526-2e79-482b-af06-e7c432999c88 | successor-predecessor-intrinsic-exploration | 2305.15277 | null | https://arxiv.org/abs/2305.15277v1 | https://arxiv.org/pdf/2305.15277v1.pdf | Successor-Predecessor Intrinsic Exploration | Exploration is essential in reinforcement learning, particularly in environments where external rewards are sparse. Here we focus on exploration with intrinsic rewards, where the agent transiently augments the external rewards with self-generated intrinsic rewards. Although the study of intrinsic rewards has a long history, existing methods focus on composing the intrinsic reward based on measures of future prospects of states, ignoring the information contained in the retrospective structure of transition sequences. Here we argue that the agent can utilise retrospective information to generate explorative behaviour with structure-awareness, facilitating efficient exploration based on global instead of local information. We propose Successor-Predecessor Intrinsic Exploration (SPIE), an exploration algorithm based on a novel intrinsic reward combining prospective and retrospective information. We show that SPIE yields more efficient and ethologically plausible exploratory behaviour in environments with sparse rewards and bottleneck states than competing methods. We also implement SPIE in deep reinforcement learning agents, and show that the resulting agent achieves stronger empirical performance than existing methods on sparse-reward Atari games. | ['Sam Gershman', 'Maneesh Sahani', 'Neil Burgess', 'Changmin Yu'] | 2023-05-24 | null | null | null | null | ['efficient-exploration', 'atari-games'] | ['methodology', 'playing-games'] | [-1.39017344e-01 3.04057211e-01 -3.49034131e-01 -8.50778893e-02
-4.73514646e-01 -4.87984389e-01 9.80580449e-01 1.59220561e-01
-9.61581588e-01 1.11881268e+00 4.92168337e-01 -1.91196725e-01
-2.83778459e-01 -7.57535219e-01 -5.50829113e-01 -7.44141519e-01
-7.41366744e-01 6.05577409e-01 1.68624952e-01 -6.29303277e-01
4.70189303e-01 2.78444111e-01 -1.72494459e+00 -3.42800587e-01
8.49904656e-01 4.43255812e-01 4.52130616e-01 6.63992405e-01
2.91663617e-01 1.14007974e+00 -4.79407728e-01 2.63166308e-01
4.15180236e-01 -7.35136271e-01 -8.71155679e-01 -1.00438081e-01
-8.37046385e-01 -5.88777006e-01 -2.14286953e-01 7.05676734e-01
4.06468332e-01 5.39697826e-01 4.17683005e-01 -1.16337740e+00
-3.72201711e-01 1.15249491e+00 -4.87990528e-01 3.89313936e-01
3.17955106e-01 6.30208015e-01 9.07064080e-01 -3.03233683e-01
7.26025999e-01 1.14600492e+00 5.89217246e-02 6.95903301e-01
-1.37434769e+00 -3.15802842e-01 3.84402990e-01 7.69249350e-02
-5.44241726e-01 -2.44642004e-01 6.92191482e-01 -3.22596729e-02
1.12611794e+00 3.95219587e-02 1.06988192e+00 1.22106481e+00
2.11408138e-01 1.03763342e+00 1.53393865e+00 -2.68919885e-01
9.29082870e-01 -1.46955892e-01 -3.79200965e-01 3.36573333e-01
-2.34666951e-02 1.14719546e+00 -7.26912856e-01 -4.28515486e-02
9.79407609e-01 -5.29405735e-02 3.16192597e-01 -8.14894319e-01
-1.31553817e+00 9.44826663e-01 5.00240982e-01 3.37661095e-02
-7.73872018e-01 5.84980309e-01 2.47362182e-01 6.51914895e-01
-1.09283924e-01 1.00112367e+00 -1.98815420e-01 -8.09754789e-01
-6.81715012e-01 5.10179400e-01 6.05152369e-01 6.50764525e-01
9.87478793e-01 3.58353883e-01 -1.72313467e-01 4.11374182e-01
2.51792789e-01 3.57498050e-01 7.12168694e-01 -1.45921469e+00
9.44385082e-02 3.04797173e-01 5.07757664e-01 -2.49648467e-01
-4.96888548e-01 -5.40176868e-01 -2.20954716e-01 7.88753510e-01
2.47086421e-01 -4.39297080e-01 -8.02225471e-01 2.08683681e+00
3.67430001e-01 -2.76567847e-01 3.67483795e-01 8.93320918e-01
9.83892083e-02 4.32870537e-01 2.17815861e-01 -2.83916146e-01
6.30637109e-01 -1.03778934e+00 -4.66136158e-01 -4.08726156e-01
7.69681215e-01 9.30126663e-03 9.59836423e-01 3.80630344e-01
-1.45555794e+00 -2.47839481e-01 -7.82626987e-01 4.23989415e-01
-7.24320933e-02 -3.57584208e-01 9.47614908e-01 5.24886072e-01
-1.38850021e+00 9.64534760e-01 -1.02637327e+00 -1.92444399e-01
2.88396776e-01 4.13118094e-01 -8.82036760e-02 5.45018554e-01
-1.20487702e+00 1.08178031e+00 5.43195784e-01 -1.68112949e-01
-1.66782618e+00 -9.79772806e-02 -7.21710682e-01 6.67195022e-02
8.78843427e-01 -4.20975149e-01 1.58539891e+00 -9.02517319e-01
-1.96960759e+00 2.77629554e-01 2.96536744e-01 -8.20588827e-01
4.37881917e-01 -1.53221473e-01 2.11035952e-01 3.73845734e-02
1.79833055e-01 1.08717680e+00 7.28723586e-01 -1.11037815e+00
-5.21802187e-01 -1.62852213e-01 4.05537009e-01 8.25765193e-01
-8.83155316e-02 -1.78532034e-01 2.56897330e-01 -3.20592880e-01
-2.14184791e-01 -1.02931178e+00 -9.84180033e-01 -4.87925202e-01
-4.32969108e-02 -2.54996955e-01 1.44092351e-01 1.34054437e-01
8.19920361e-01 -1.77852976e+00 3.60859364e-01 3.01274896e-01
1.70655414e-01 -1.15744315e-01 -4.19530630e-01 7.98690736e-01
1.72180220e-01 -3.17143053e-01 -1.91783383e-01 -4.62343805e-02
1.68313488e-01 4.58618253e-01 -4.40195501e-01 1.81068152e-01
3.90149020e-02 1.15595376e+00 -1.51804245e+00 -1.91638961e-01
6.22150339e-02 -1.22112237e-01 -7.03813314e-01 3.59268636e-01
-2.49098167e-01 5.85139573e-01 -6.76926017e-01 4.64740008e-01
6.32053167e-02 -1.15724958e-01 4.20586139e-01 1.01877439e+00
-4.14769351e-01 6.95929825e-01 -1.10979450e+00 1.70925426e+00
-3.24290663e-01 1.20461583e-01 1.28193244e-01 -7.57853746e-01
9.60398436e-01 -2.84265969e-02 5.34245491e-01 -1.03951812e+00
1.23686709e-01 1.98478445e-01 1.88384056e-01 -1.78398192e-02
9.88670707e-01 -2.77700946e-02 -1.83817297e-01 1.03118777e+00
-7.67986625e-02 -2.08044901e-01 4.13719356e-01 5.57447553e-01
1.48686683e+00 8.42475235e-01 4.38526124e-01 -4.10734862e-01
-4.51645106e-02 1.57520711e-01 4.04474854e-01 1.22617710e+00
-3.94839257e-01 -9.32410136e-02 7.63926268e-01 -2.69077778e-01
-1.01747310e+00 -1.25514412e+00 2.40866885e-01 1.40513635e+00
4.39376384e-01 -5.05270422e-01 -4.35621142e-01 -5.51758587e-01
-1.25181928e-01 7.09900975e-01 -9.02536213e-01 -2.61011451e-01
-6.22015476e-01 -6.72945082e-01 2.32232824e-01 7.09398866e-01
2.87672073e-01 -1.85901642e+00 -1.64199293e+00 5.17424107e-01
1.94789335e-01 -2.17737854e-01 -2.39339456e-01 9.58350122e-01
-1.20633543e+00 -6.20600462e-01 -6.55922830e-01 -2.69268066e-01
4.83489811e-01 6.39072135e-02 1.10594416e+00 -8.58077779e-02
-7.88611323e-02 5.85572958e-01 -4.78347927e-01 -8.88266135e-03
-3.56944144e-01 -8.23310018e-02 2.47514427e-01 -7.98460364e-01
2.49916732e-01 -8.01076710e-01 -6.89492762e-01 1.06287904e-01
-7.02802777e-01 1.38185054e-01 8.58992100e-01 1.20306408e+00
2.89295822e-01 -3.67116541e-01 8.44393075e-01 -5.29616773e-01
9.63611066e-01 -7.01879323e-01 -7.42458165e-01 -2.93348730e-01
-8.12341690e-01 6.83714271e-01 4.73321378e-01 -4.41412389e-01
-1.26301479e+00 -4.62545641e-02 8.37067813e-02 5.34494072e-02
-1.29318967e-01 4.05254960e-01 4.16066647e-01 1.50144488e-01
8.23009074e-01 3.81508529e-01 1.57766461e-01 -1.49009481e-01
6.36203647e-01 2.77270749e-02 1.74090549e-01 -1.02513671e+00
4.16771919e-01 3.59658360e-01 1.38218880e-01 -3.80096734e-01
-2.04177782e-01 -1.31609514e-01 -1.26258329e-01 -3.30028117e-01
2.33915284e-01 -8.21034253e-01 -1.17491436e+00 1.47256881e-01
-5.58406711e-01 -9.35255289e-01 -1.18427277e+00 6.36894643e-01
-1.31583548e+00 2.35714167e-01 -6.59261942e-01 -1.26798844e+00
1.96680769e-01 -1.42196238e+00 7.05477834e-01 3.79645616e-01
-1.97678000e-01 -6.71427131e-01 5.71124494e-01 -3.20747823e-01
5.84525764e-01 -5.45424744e-02 3.08692843e-01 -4.94136661e-01
-1.01972759e+00 5.35093963e-01 2.00611666e-01 -3.61089826e-01
-9.85871032e-02 -4.58759338e-01 -6.05722785e-01 -3.21369320e-01
-1.59322292e-01 -8.79610479e-01 1.02619708e+00 5.09241998e-01
6.40668869e-01 -4.28164989e-01 -2.25419104e-01 2.26241678e-01
1.19370544e+00 2.64383465e-01 6.65128291e-01 9.14297104e-01
6.17695216e-04 1.01866186e+00 9.74599779e-01 1.07879758e+00
5.18041492e-01 3.82344961e-01 9.99621451e-01 2.35277459e-01
3.85756999e-01 -3.63764822e-01 6.69252157e-01 3.24995577e-01
-3.18426132e-01 1.36403248e-01 -5.21236420e-01 1.00838006e+00
-2.04904342e+00 -1.19598401e+00 3.26824158e-01 2.26054287e+00
1.09297168e+00 2.32118785e-01 6.22195423e-01 -1.76269144e-01
2.43825600e-01 2.31696263e-01 -1.00169599e+00 -7.16371536e-01
-3.32880728e-02 2.07166582e-01 5.32234073e-01 5.51417649e-01
-6.49260700e-01 1.11707032e+00 7.42550659e+00 5.06913006e-01
-5.46635509e-01 -1.34954065e-01 5.06583571e-01 -5.32172203e-01
-6.18833005e-01 3.94483238e-01 -5.49973130e-01 2.09958851e-01
1.01717484e+00 -1.63139269e-01 8.76048028e-01 1.17022502e+00
1.52159780e-01 -6.76168919e-01 -9.61585939e-01 2.31196031e-01
-6.36750817e-01 -1.19289207e+00 -4.13993567e-01 3.83722812e-01
7.91765392e-01 2.18039885e-01 3.14734071e-01 5.03901958e-01
1.44866085e+00 -1.04898250e+00 6.91466570e-01 3.45235318e-01
3.74696732e-01 -1.00278270e+00 2.56495565e-01 5.57029843e-01
-9.64560509e-01 -4.69349235e-01 -2.44955048e-01 -6.11467600e-01
1.11027077e-01 1.22954305e-02 -8.52683127e-01 2.12910950e-01
5.83690941e-01 6.59416556e-01 -2.56751597e-01 7.43009806e-01
-4.09489095e-01 3.29659224e-01 -3.25962782e-01 -5.18979430e-01
7.71664739e-01 -4.75062758e-01 6.70519829e-01 6.44251525e-01
1.89913154e-01 3.02128959e-02 1.55372068e-01 1.16574252e+00
3.44067365e-01 -1.05293818e-01 -7.46618390e-01 -1.94048166e-01
3.44512820e-01 1.14189136e+00 -8.41217816e-01 -3.07887405e-01
1.57300562e-01 5.95907629e-01 6.34848654e-01 3.03822458e-01
-6.27913833e-01 -1.70966893e-01 5.21726370e-01 -2.79401779e-01
3.70793343e-01 -2.76894718e-01 -7.60299712e-02 -9.47735369e-01
-3.08292270e-01 -8.67528737e-01 2.66336113e-01 -5.55398524e-01
-7.30964839e-01 5.96806467e-01 1.06467158e-01 -9.42224681e-01
-9.49219882e-01 -1.21321216e-01 -5.97649932e-01 7.06613541e-01
-1.71119297e+00 -5.92223525e-01 1.68705776e-01 4.30685252e-01
5.64902663e-01 -1.16479002e-01 6.56193435e-01 -6.56127155e-01
-1.37599900e-01 1.44182444e-01 2.20522627e-01 -4.77320224e-01
4.51850235e-01 -1.67259824e+00 5.15815616e-01 6.51067078e-01
-6.27304241e-02 7.59200513e-01 9.86025989e-01 -9.54809070e-01
-1.39123690e+00 -4.60332125e-01 2.95424581e-01 -2.85950512e-01
7.50601232e-01 -3.14231254e-02 -5.64843953e-01 6.06096029e-01
4.72497940e-01 -3.23861122e-01 4.34988588e-01 3.21037710e-01
-2.71524061e-02 3.50777388e-01 -8.79203498e-01 9.46300387e-01
1.17281103e+00 -1.44171908e-01 -8.29664230e-01 -1.47046551e-01
5.23488283e-01 -3.00809979e-01 -4.89508599e-01 7.70470202e-02
5.47021806e-01 -1.23173666e+00 8.88804972e-01 -6.70622408e-01
4.27461356e-01 -2.04719692e-01 1.50607437e-01 -1.61508191e+00
-4.32365060e-01 -1.08818495e+00 -1.84582606e-01 4.72226709e-01
2.52149433e-01 -6.30407512e-01 1.01700735e+00 3.15835983e-01
-1.79676920e-01 -7.64520526e-01 -8.62166762e-01 -1.06134450e+00
1.31951526e-01 3.31864133e-02 7.42487371e-01 4.71437693e-01
6.09779537e-01 -6.16247840e-02 -6.15112126e-01 -4.09755349e-01
6.75278068e-01 2.75008023e-01 7.46758878e-01 -9.88050461e-01
-6.10323131e-01 -4.52764094e-01 2.55024314e-01 -1.22089195e+00
1.37611926e-02 -6.06336534e-01 4.68943417e-01 -1.32886600e+00
2.76283830e-01 -6.99825823e-01 -4.61135298e-01 5.24364829e-01
-8.64300355e-02 -1.43829003e-01 1.18390523e-01 1.86857104e-01
-1.00638068e+00 1.17113054e+00 1.56268513e+00 3.33713859e-01
-6.64409101e-01 -2.57152915e-01 -7.30410695e-01 4.71259654e-01
9.87629771e-01 -5.54433525e-01 -5.44893265e-01 3.86642553e-02
5.62743545e-01 3.49263012e-01 2.15151042e-01 -8.78258944e-01
2.17315137e-01 -7.39228368e-01 1.36548847e-01 -5.35470605e-01
5.04310310e-01 -2.84239352e-01 -2.74077803e-02 8.26108992e-01
-9.56277430e-01 2.92322785e-01 -1.81261133e-02 8.84681761e-01
1.72601774e-01 -3.09507579e-01 3.75409603e-01 -5.73680043e-01
-8.76608372e-01 1.14151627e-01 -9.83620703e-01 6.21271431e-02
1.07282376e+00 -3.73707145e-01 -3.21045905e-01 -5.49154282e-01
-6.81931674e-01 5.45855105e-01 6.16387784e-01 2.06523299e-01
7.28800535e-01 -1.17984664e+00 -4.46581662e-01 1.28199663e-02
9.60702002e-02 -3.83212298e-01 1.85599610e-01 7.74031520e-01
-7.46451467e-02 2.69717336e-01 -9.85914826e-01 -2.51704603e-01
-6.85758948e-01 6.91274047e-01 1.03988990e-01 -6.13638580e-01
-6.65111899e-01 6.61250651e-01 2.78880984e-01 -4.06751513e-01
8.23235065e-02 -9.18233842e-02 -4.48583484e-01 -1.28951371e-01
2.21326172e-01 4.66059625e-01 -6.21925056e-01 -2.38536894e-01
-1.38637885e-01 -4.13101949e-02 -1.81426466e-01 -7.59857774e-01
1.38757706e+00 -2.13693976e-01 1.87143415e-01 3.69013965e-01
2.58321077e-01 -2.13230476e-01 -1.93296635e+00 -9.98206288e-02
7.22053275e-02 -6.29309356e-01 4.14121673e-02 -7.32231796e-01
-5.63191295e-01 6.83367550e-01 2.33522668e-01 2.22471386e-01
9.32192564e-01 -1.15237951e-01 5.47452748e-01 7.38335490e-01
8.65699947e-01 -1.54979789e+00 6.99777365e-01 6.19225144e-01
6.26468658e-01 -1.14915359e+00 -1.31907195e-01 5.27250886e-01
-1.12514341e+00 8.96274149e-01 6.41686916e-01 -3.99236172e-01
5.03016300e-02 1.58766195e-01 -3.40516448e-01 -1.32670671e-01
-1.24350500e+00 -7.74262786e-01 -4.97587681e-01 8.26659620e-01
1.88410245e-02 1.92100734e-01 -2.92658508e-01 2.22213209e-01
-2.38914341e-01 -1.07757658e-01 9.99609113e-01 1.31149447e+00
-8.58702779e-01 -1.46653807e+00 -1.19344786e-01 4.65519041e-01
-1.72661245e-01 -9.98997539e-02 -8.24795440e-02 6.07903540e-01
-3.96064997e-01 8.17240477e-01 -2.04159077e-02 -7.02273622e-02
-3.59604776e-01 -4.25104886e-01 6.52653694e-01 -5.83798051e-01
-7.97890961e-01 2.17575952e-01 1.18546627e-01 -9.31770682e-01
-2.43450344e-01 -1.00634623e+00 -1.34527922e+00 -2.04726517e-01
-7.83289075e-02 4.88982260e-01 5.81860125e-01 7.26341069e-01
3.31999630e-01 5.94658375e-01 7.81392455e-01 -9.36194360e-01
-1.03466535e+00 -5.94098210e-01 -7.86532462e-01 3.06616500e-02
5.28497159e-01 -9.61699903e-01 -2.96923280e-01 -5.50193429e-01] | [4.0086493492126465, 1.7459741830825806] |
7662d949-eb5a-4f5f-8100-a2e9d935d05a | keyword-assisted-embedded-topic-model | 2112.03101 | null | https://arxiv.org/abs/2112.03101v1 | https://arxiv.org/pdf/2112.03101v1.pdf | Keyword Assisted Embedded Topic Model | By illuminating latent structures in a corpus of text, topic models are an essential tool for categorizing, summarizing, and exploring large collections of documents. Probabilistic topic models, such as latent Dirichlet allocation (LDA), describe how words in documents are generated via a set of latent distributions called topics. Recently, the Embedded Topic Model (ETM) has extended LDA to utilize the semantic information in word embeddings to derive semantically richer topics. As LDA and its extensions are unsupervised models, they aren't defined to make efficient use of a user's prior knowledge of the domain. To this end, we propose the Keyword Assisted Embedded Topic Model (KeyETM), which equips ETM with the ability to incorporate user knowledge in the form of informative topic-level priors over the vocabulary. Using both quantitative metrics and human responses on a topic intrusion task, we demonstrate that KeyETM produces better topics than other guided, generative models in the literature. | ['Fred Morstatter', 'J. Hunter Priniski', 'Bahareh Harandizadeh'] | 2021-11-22 | null | null | null | null | ['topic-models'] | ['natural-language-processing'] | [-2.01997042e-01 3.07612747e-01 -4.91973162e-01 -3.49563956e-01
-5.96848428e-01 -6.19057953e-01 1.07835555e+00 4.28931952e-01
-1.10777896e-02 2.96928138e-01 8.70976090e-01 -1.76641062e-01
6.83891103e-02 -8.95912051e-01 -7.99151137e-02 -4.26699400e-01
4.55548987e-03 5.74749410e-01 1.67532608e-01 2.39556879e-02
4.82595325e-01 -1.84214771e-01 -1.30111480e+00 9.43103954e-02
8.76077771e-01 3.89986992e-01 3.75413686e-01 1.07460074e-01
-8.18487763e-01 1.67253703e-01 -5.50329268e-01 -9.02018994e-02
-4.48583066e-01 -1.72789007e-01 -6.00342631e-01 3.46599877e-01
-2.27099165e-01 -3.47403497e-01 -2.74250478e-01 7.88412988e-01
7.22967088e-02 3.49650174e-01 1.16250932e+00 -1.43151414e+00
-7.95700967e-01 8.08266759e-01 -3.79060447e-01 -6.24649646e-03
2.84800053e-01 -4.11000490e-01 1.36968899e+00 -1.20228708e+00
7.12477624e-01 1.51840985e+00 2.17854574e-01 3.87385190e-01
-1.34977543e+00 -6.14108205e-01 4.71626610e-01 -1.32512406e-01
-1.15606701e+00 7.58885080e-03 1.12821865e+00 -6.82201147e-01
7.63566077e-01 -1.08629137e-01 5.72611690e-01 1.27209675e+00
1.72578350e-01 1.07901061e+00 8.07253957e-01 -6.19596362e-01
6.29179657e-01 5.51826477e-01 6.68357730e-01 1.40228897e-01
3.34116250e-01 -4.47781473e-01 -7.14326322e-01 -8.15939784e-01
4.68211353e-01 3.10736001e-01 -2.40191847e-01 -9.14187014e-01
-9.63308990e-01 1.53361094e+00 3.84968035e-02 3.76174301e-01
-6.33318961e-01 -1.61382947e-02 2.90371776e-01 -2.74826258e-01
1.13261449e+00 5.42438924e-01 9.76190716e-02 -1.10002428e-01
-1.07464516e+00 2.59718120e-01 8.46811175e-01 1.02987218e+00
8.83703113e-01 -1.69850647e-01 -1.32544205e-01 7.86222935e-01
9.05447066e-01 1.88953593e-01 9.01541531e-01 -5.76517344e-01
1.46646053e-01 8.21943521e-01 2.61752933e-01 -9.78905320e-01
-2.60342099e-03 -7.70695060e-02 -2.08697841e-01 -4.43599790e-01
-2.48048380e-01 -4.14436534e-02 -1.01408541e+00 1.71294093e+00
2.63083935e-01 1.17901703e-02 8.35729837e-02 4.28242177e-01
6.29479766e-01 1.12615228e+00 5.67851782e-01 -1.80660799e-01
1.46753561e+00 -6.34473264e-01 -1.06764245e+00 -2.55581528e-01
5.78016579e-01 -6.18662357e-01 1.21937060e+00 2.38357350e-01
-6.80276811e-01 -1.49766132e-01 -8.78520012e-01 -1.42566398e-01
-7.22649992e-01 -3.06788564e-01 7.89566934e-01 8.56254995e-01
-9.58992779e-01 -4.65213833e-03 -9.47269022e-01 -7.36773252e-01
3.07289034e-01 -1.84801623e-01 -5.50837554e-02 -1.92547262e-01
-1.29628968e+00 5.69678128e-01 4.59945828e-01 -6.48704708e-01
-1.10437155e+00 -7.11454391e-01 -1.03716528e+00 3.85550588e-01
2.71948516e-01 -5.12016654e-01 1.17835665e+00 -1.05398603e-01
-1.25696099e+00 4.86706942e-01 -6.19516015e-01 -3.36818308e-01
-3.73356372e-01 -5.07266998e-01 -1.97213590e-01 2.36871123e-01
3.36926550e-01 9.32265043e-01 8.25636148e-01 -1.33951294e+00
-4.02569264e-01 -2.41931304e-01 -1.15407586e-01 2.47314140e-01
-9.46278393e-01 1.15147464e-01 -4.30932701e-01 -8.16524982e-01
2.45353624e-01 -7.72336364e-01 -1.55475289e-01 -2.31492296e-01
-4.76350188e-01 -8.65681708e-01 1.28887761e+00 -4.80150282e-01
1.31060958e+00 -2.24255252e+00 -3.92649807e-02 2.62345254e-01
5.60984552e-01 -2.27085888e-01 5.82383573e-02 8.18584859e-01
1.73429891e-01 3.38031083e-01 -1.88243445e-02 -7.01363921e-01
4.39866692e-01 1.86877862e-01 -1.00201952e+00 1.68768629e-01
-2.25278020e-01 6.80568397e-01 -8.96349549e-01 -4.70002204e-01
1.38790205e-01 6.61279857e-01 -7.34840512e-01 7.23111778e-02
-5.09339273e-01 -1.30352437e-01 -7.33008265e-01 1.53337881e-01
2.40935981e-01 -4.49849963e-01 2.18116954e-01 1.50522619e-01
-9.77188498e-02 6.89057648e-01 -7.39965975e-01 1.77561891e+00
-2.95055717e-01 8.65852296e-01 -4.88238633e-01 -6.40257537e-01
1.04830658e+00 7.71170974e-01 5.76666951e-01 4.95419279e-03
-7.98776075e-02 -2.22698554e-01 -4.77403343e-01 3.89107037e-03
9.58683074e-01 -1.55765936e-01 -3.58573884e-01 1.13343930e+00
2.07291856e-01 -2.04082504e-01 1.37673587e-01 9.00129735e-01
7.62245595e-01 -1.73901066e-01 2.79222488e-01 -5.27070224e-01
-2.81356543e-01 8.67861211e-02 1.86112463e-01 7.32187510e-01
1.54744044e-01 3.18317384e-01 4.33054805e-01 -6.54550716e-02
-1.04851437e+00 -1.19536400e+00 -2.55681336e-01 1.25226557e+00
-1.20705953e-02 -9.76756454e-01 -5.96890390e-01 -4.48962420e-01
-8.34902450e-02 1.22494960e+00 -5.86563408e-01 -2.46100947e-01
-2.12843306e-02 -7.36132205e-01 2.10628416e-02 5.42033017e-01
1.07334852e-01 -7.54709959e-01 -4.52195555e-01 3.48435789e-01
-3.80417824e-01 -8.43447924e-01 -4.88292396e-01 -4.23357524e-02
-9.39298809e-01 -5.69851041e-01 -7.98745751e-01 -4.50644702e-01
6.33765996e-01 4.73469228e-01 1.02224982e+00 -6.79881692e-01
-1.98832422e-01 8.61148953e-01 -6.35800719e-01 -6.07257307e-01
-1.66604221e-01 1.23657927e-01 1.82376370e-01 -1.74666226e-01
1.01842940e+00 -6.57789767e-01 -4.28833991e-01 2.21654743e-01
-1.19646716e+00 7.00368136e-02 2.48112798e-01 5.62474012e-01
1.92384735e-01 2.97276229e-01 6.23134315e-01 -1.00373960e+00
1.09676647e+00 -9.18813288e-01 -3.43143761e-01 8.03093016e-02
-8.94542873e-01 7.44096711e-02 -2.07128853e-01 -6.40449166e-01
-1.15264571e+00 -6.92356884e-01 4.15189683e-01 -2.10561559e-01
-1.67208821e-01 8.09003294e-01 -2.48966679e-01 6.60210252e-01
4.52429652e-01 4.11395878e-01 -2.45720506e-01 -7.91831851e-01
7.57662296e-01 7.95317829e-01 2.51811817e-02 -6.36757433e-01
5.12984216e-01 6.36731565e-01 -5.87063551e-01 -9.47598219e-01
-8.44864428e-01 -1.06524765e+00 -4.36662912e-01 9.35429409e-02
9.93460238e-01 -9.96024787e-01 1.25522045e-02 -2.15744600e-02
-1.13256013e+00 -2.11967137e-02 -3.89619827e-01 7.44032919e-01
-4.34480339e-01 4.80540782e-01 -3.17413211e-01 -8.03107083e-01
-2.60670781e-01 -8.31895292e-01 1.17419982e+00 7.22598284e-02
-8.09147954e-01 -1.51464355e+00 2.60144711e-01 -2.52408069e-02
5.63404858e-01 -1.70854345e-01 1.31721699e+00 -1.10409868e+00
-2.97657073e-01 -2.62628138e-01 -6.68771639e-02 -1.06238329e-03
5.21538854e-01 -2.09235400e-01 -8.92191827e-01 -2.91943252e-01
2.09722593e-01 -3.73614319e-02 8.84139836e-01 4.44902658e-01
8.99134517e-01 -3.66838813e-01 -8.31998229e-01 7.13321790e-02
1.13518357e+00 2.11131200e-01 5.05620122e-01 2.81159163e-01
3.76528889e-01 7.06476331e-01 2.94337928e-01 6.48252249e-01
6.09215856e-01 4.23446417e-01 1.23309344e-03 1.47013709e-01
3.89437556e-01 -6.03629291e-01 3.33048463e-01 8.45864594e-01
5.20848870e-01 -5.91321230e-01 -1.09620166e+00 8.94241631e-01
-1.51096594e+00 -6.92272604e-01 2.64029205e-01 1.86510372e+00
8.68050694e-01 1.54771626e-01 -9.04006734e-02 -9.20392573e-02
7.38997519e-01 3.07918280e-01 -3.45012844e-01 -1.71417356e-01
2.13003844e-01 -9.53155607e-02 -5.81310019e-02 4.44297761e-01
-1.12416768e+00 9.97100353e-01 6.94403887e+00 7.11487830e-01
-6.44510269e-01 1.89283222e-01 1.95284158e-01 2.19268128e-01
-8.88723493e-01 4.60589319e-01 -1.16651762e+00 5.32670379e-01
9.43931282e-01 -6.58773899e-01 -3.00300926e-01 1.20476639e+00
2.02638894e-01 -9.78010073e-02 -1.07062864e+00 4.45273280e-01
2.70351619e-01 -9.92666423e-01 4.63843197e-01 4.93827969e-01
8.11419249e-01 -3.17508459e-01 3.82432222e-01 4.06708121e-01
8.13809514e-01 -6.53570056e-01 2.43202955e-01 3.55855912e-01
3.23943079e-01 -4.64179873e-01 4.01748329e-01 3.94848645e-01
-7.96860039e-01 1.34668574e-01 -5.65212250e-01 2.07250729e-01
4.53714877e-01 8.02616954e-01 -1.26802790e+00 1.19529009e-01
3.98137689e-01 7.39642620e-01 -2.06443503e-01 9.16777134e-01
-2.88483679e-01 1.10362792e+00 -3.71283650e-01 4.67788801e-02
4.37860757e-01 -4.84121293e-02 7.45376050e-01 1.06150138e+00
3.96446556e-01 -3.20363566e-02 4.51961130e-01 1.04406369e+00
1.24281667e-01 2.25743622e-01 -6.56094313e-01 -5.97231448e-01
7.92109847e-01 1.05524182e+00 -8.84969056e-01 -6.30637586e-01
-3.39786381e-01 5.52045584e-01 -2.54153699e-01 5.88077068e-01
-2.80455410e-01 -1.91043377e-01 5.86970091e-01 2.05104664e-01
2.54174978e-01 -5.94815433e-01 -1.05505109e-01 -1.17304969e+00
-4.38653886e-01 -3.59860122e-01 3.43259335e-01 -9.31834877e-01
-1.36141491e+00 3.28183711e-01 7.52017498e-01 -8.83163393e-01
-5.58964610e-01 -2.31628016e-01 -7.07921326e-01 9.26953435e-01
-1.28468132e+00 -1.11327744e+00 -5.87446094e-02 3.21454525e-01
9.01992381e-01 -2.16643080e-01 9.73166645e-01 -2.21726477e-01
-2.04046592e-01 1.46122901e-02 3.40310454e-01 -9.38896090e-02
8.09366465e-01 -1.50855958e+00 5.24034679e-01 5.58689594e-01
1.95090815e-01 1.19385195e+00 9.90353346e-01 -8.87449503e-01
-8.17493916e-01 -9.24279153e-01 1.00643718e+00 -5.49043417e-01
7.87611246e-01 -6.79257274e-01 -1.09709311e+00 9.50100362e-01
3.75451446e-01 -6.98373199e-01 1.44956088e+00 6.14971101e-01
-5.60962975e-01 6.03549242e-01 -6.79136276e-01 4.95791733e-01
3.67073298e-01 -6.24399960e-01 -1.11608624e+00 3.87979567e-01
1.13826263e+00 3.00265104e-01 -7.30123699e-01 -1.10480748e-01
2.27930114e-01 -3.07864070e-01 8.09773564e-01 -6.91180706e-01
3.75695229e-01 7.68541098e-02 -3.72923940e-01 -1.45019472e+00
-2.63998926e-01 -5.21714091e-01 -3.25355977e-01 1.52983618e+00
1.54918149e-01 -5.86236179e-01 7.58724213e-01 8.92486155e-01
-1.85284615e-01 -4.66995120e-01 -5.51052332e-01 -5.30391872e-01
1.82932034e-01 -5.15150428e-01 6.70158088e-01 9.44245160e-01
4.35311556e-01 2.72088110e-01 -1.20788492e-01 6.07089140e-02
6.31503582e-01 1.38750762e-01 6.61501229e-01 -1.62475085e+00
1.39547691e-01 -2.68336058e-01 -2.79560387e-01 -1.25716758e+00
3.00436676e-01 -6.18226588e-01 4.64804694e-02 -1.75256622e+00
5.04290819e-01 -4.68125463e-01 -5.47828786e-02 4.75382954e-01
-9.90497395e-02 -1.90007970e-01 -1.64010122e-01 6.75761878e-01
-7.04366803e-01 9.99455452e-01 6.53365910e-01 9.94374044e-03
-3.25482965e-01 -1.76146194e-01 -9.55087602e-01 9.06340599e-01
6.55010879e-01 -5.85839629e-01 -8.36773872e-01 -2.23959282e-01
2.47768655e-01 -3.13303977e-01 7.25824982e-02 -5.23691893e-01
3.07888955e-01 -7.21750483e-02 -1.59253161e-02 -9.15149689e-01
6.48008227e-01 -5.09591758e-01 -2.57899940e-01 -3.35084423e-02
-7.03698814e-01 -1.89374462e-01 8.34845304e-02 8.67990732e-01
-3.69386494e-01 -2.17545688e-01 2.20119223e-01 -3.56150307e-02
-7.57042944e-01 2.17812374e-01 -9.06002045e-01 -2.69417968e-02
8.22424710e-01 -8.55629295e-02 -2.54105687e-01 -7.37728775e-01
-7.68011689e-01 2.45044604e-01 4.20662999e-01 5.67771614e-01
6.53228700e-01 -1.34271002e+00 -3.50363195e-01 9.22920257e-02
3.49925846e-01 -6.99764043e-02 2.03239366e-01 2.03296304e-01
2.52891690e-01 9.84449029e-01 2.56143093e-01 -5.42609990e-01
-8.73906255e-01 6.26320004e-01 -4.07445878e-01 -3.18557411e-01
-7.77916491e-01 5.81432462e-01 8.91301870e-01 -1.61622316e-01
2.82067060e-01 -8.67879763e-02 -4.17745560e-01 3.46749723e-01
5.47159433e-01 9.76599976e-02 -4.50044215e-01 -2.79743582e-01
1.34050325e-02 1.76279843e-01 -5.03185511e-01 -5.59222400e-01
1.32069170e+00 -4.15858358e-01 -5.53519093e-02 8.17740560e-01
1.08817852e+00 -1.11130826e-01 -9.65614438e-01 -6.24170780e-01
3.51499736e-01 -4.29799438e-01 3.08912337e-01 -3.50159258e-01
-2.61136979e-01 1.15125203e+00 2.53267169e-01 4.70464349e-01
6.75843537e-01 4.04266983e-01 7.58734822e-01 1.81471244e-01
2.42352962e-01 -8.76721621e-01 4.60455418e-01 3.96052569e-01
5.58837116e-01 -1.01413870e+00 1.47576360e-02 -4.96176541e-01
-6.36983275e-01 8.69289279e-01 2.79540807e-01 1.12608738e-01
1.12755346e+00 -1.61793251e-02 -1.95978314e-01 -5.29297173e-01
-1.03866148e+00 -5.03362119e-02 4.32054341e-01 5.72039723e-01
5.49066007e-01 -4.46326844e-02 -2.20055401e-01 8.40639234e-01
-3.20065349e-01 -3.71637642e-01 5.13146579e-01 9.00494337e-01
-1.00564766e+00 -1.09552491e+00 -3.41201931e-01 3.35888565e-01
-3.68880749e-01 -1.35832295e-01 -1.99095249e-01 6.06908560e-01
-3.89731228e-01 9.78954256e-01 3.78437817e-01 -1.26949921e-02
-4.57507402e-01 5.79484880e-01 -2.36490384e-01 -1.34469163e+00
2.88988650e-01 4.01931286e-01 -2.68116027e-01 -7.70630613e-02
-2.83160210e-01 -7.39230931e-01 -8.70636284e-01 7.45382234e-02
-5.90443254e-01 8.04484606e-01 1.03001547e+00 1.09808147e+00
4.31622684e-01 2.95094669e-01 3.65084082e-01 -6.16600156e-01
-4.07658577e-01 -1.22339344e+00 -8.12138557e-01 2.62409598e-01
-2.28210792e-01 -8.82559717e-01 -3.92405689e-01 2.29334146e-01] | [10.366776466369629, 6.971808433532715] |
9bc64698-a6ab-42b5-8aab-8fb52f081beb | road-detection-via-on-line-label-transfer | 1412.3159 | null | http://arxiv.org/abs/1412.3159v1 | http://arxiv.org/pdf/1412.3159v1.pdf | Road Detection via On--line Label Transfer | Vision-based road detection is an essential functionality for supporting
advanced driver assistance systems (ADAS) such as road following and vehicle
and pedestrian detection. The major challenges of road detection are dealing
with shadows and lighting variations and the presence of other objects in the
scene. Current road detection algorithms characterize road areas at pixel level
and group pixels accordingly. However, these algorithms fail in presence of
strong shadows and lighting variations. Therefore, we propose a road detection
algorithm based on video alignment. The key idea of the algorithm is to exploit
the similarities occurred when a vehicle follows the same trajectory more than
once. In this way, road areas are learned in a first ride and then, this road
knowledge is used to infer areas depicting drivable road surfaces in subsequent
rides. Two different experiments are conducted to validate the proposal on
different video sequences taken at different scenarios and different daytime.
The former aims to perform on-line road detection. The latter aims to perform
off-line road detection and is applied to automatically generate the
ground-truth necessary to validate road detection algorithms. Qualitative and
quantitative evaluations prove that the proposed algorithm is a valid road
detection approach. | ['Antonio M. López', 'José M. Álvarez', 'Joan Serrat', 'Ferran Diego'] | 2014-12-10 | null | null | null | null | ['video-alignment'] | ['computer-vision'] | [ 5.16590714e-01 -4.73562293e-02 7.58363307e-02 -4.16740417e-01
-1.28173679e-02 -3.98960531e-01 8.69922936e-01 1.03960903e-02
-4.31780189e-01 6.34662032e-01 -2.15124458e-01 -5.26464701e-01
4.91340123e-02 -1.05512500e+00 -5.54787636e-01 -7.04384804e-01
1.43422917e-01 1.83912233e-01 8.10169041e-01 -4.15429354e-01
4.54931945e-01 8.44402671e-01 -2.11878252e+00 7.66193494e-03
1.03806591e+00 5.47102392e-01 5.30449927e-01 7.51438558e-01
-1.76702589e-01 5.15572369e-01 -1.89105049e-01 -4.42350321e-02
3.57676446e-01 -2.71888852e-01 -1.95203692e-01 3.10142756e-01
6.68702364e-01 -4.67073053e-01 -3.23475003e-01 8.82617235e-01
3.09395730e-01 2.18301594e-01 5.71409166e-01 -1.28994346e+00
4.73602623e-01 -1.83992431e-01 -5.85500598e-01 3.81734222e-01
2.94115812e-01 4.22784775e-01 5.00560105e-01 -8.65981042e-01
4.51979071e-01 1.05769801e+00 4.61099744e-01 -3.01282350e-02
-8.08145881e-01 -4.75723654e-01 1.41995683e-01 7.26456821e-01
-1.37002349e+00 -6.13241076e-01 8.88549030e-01 -4.24347579e-01
2.87837207e-01 2.17407197e-01 7.02666163e-01 4.10238594e-01
1.32147491e-01 6.90022469e-01 1.50456858e+00 -4.24631834e-01
4.91832849e-03 5.00152349e-01 4.02531683e-01 7.47765124e-01
4.67777878e-01 3.20975393e-01 -4.08469409e-01 3.70464057e-01
2.23652020e-01 -2.08883837e-01 -3.16846579e-01 -3.01577955e-01
-9.06595647e-01 4.95832950e-01 2.83611476e-01 1.72368959e-01
-7.73080230e-01 -2.15858817e-01 1.80421889e-01 3.72380428e-02
1.55154541e-01 -3.51823151e-01 1.32572129e-01 -1.26275783e-02
-9.92615700e-01 2.70035505e-01 5.76471627e-01 7.19858527e-01
1.14699221e+00 3.41399610e-02 2.79652923e-02 4.17669863e-01
2.07870513e-01 7.80053496e-01 -1.66529536e-01 -6.46745086e-01
4.35709476e-01 5.45226932e-01 2.51092792e-01 -1.25484598e+00
-4.09831375e-01 -2.88900584e-01 -4.56512690e-01 7.07186401e-01
5.41262686e-01 -1.39533192e-01 -1.04052365e+00 1.00703585e+00
6.55911684e-01 4.71244633e-01 9.08193514e-02 8.94049346e-01
6.83458090e-01 5.71582735e-01 -1.41671240e-01 -2.73648441e-01
1.35020196e+00 -7.10084975e-01 -8.81389618e-01 -4.15471524e-01
4.18569982e-01 -9.37909782e-01 6.75840080e-01 3.31621945e-01
-6.93572998e-01 -7.40221560e-01 -1.07471716e+00 2.41048738e-01
-6.30061626e-01 4.40996826e-01 2.34460056e-01 8.65607440e-01
-8.05545926e-01 9.94589999e-02 -1.35610476e-01 -5.32247782e-01
1.79228351e-01 -5.23919649e-02 -1.12560533e-01 -3.20055217e-01
-1.22335196e+00 1.13901615e+00 3.74593794e-01 6.76097155e-01
-8.01845789e-01 -3.16881120e-01 -6.82917595e-01 -3.54195029e-01
5.29936731e-01 -3.39048892e-01 6.81754231e-01 -8.92647326e-01
-1.27460432e+00 9.57966328e-01 -7.04213440e-01 -7.65142977e-01
9.45510030e-01 -3.18905026e-01 -6.72355950e-01 1.36374727e-01
5.05088866e-02 2.25404233e-01 8.16100001e-01 -1.59999037e+00
-1.20656753e+00 -2.24076733e-01 -3.20047438e-02 5.01978099e-01
3.61789227e-01 -1.67069569e-01 -5.42092383e-01 1.02436796e-01
-4.77404557e-02 -9.61871088e-01 1.92244258e-02 -2.38465771e-01
-5.00601470e-01 6.01959005e-02 1.13070512e+00 -7.78237879e-01
1.05848980e+00 -1.86166656e+00 -5.56368649e-01 6.64514899e-01
-4.87259813e-02 5.50902009e-01 1.20386481e-01 2.38511711e-01
3.07028621e-01 -4.32332218e-01 -2.59013116e-01 8.24221522e-02
-3.53690445e-01 2.15912923e-01 -2.43423194e-01 5.37769377e-01
1.07809469e-01 4.18378860e-01 -7.15968728e-01 -6.63008511e-01
7.96609759e-01 4.46305424e-01 3.23128462e-01 1.26915008e-01
1.11125112e-01 4.76368755e-01 -4.07949805e-01 3.90173912e-01
1.17580080e+00 8.08663368e-01 -7.63645023e-02 -3.43844533e-01
-8.78955960e-01 1.27550095e-01 -1.51569605e+00 6.50756299e-01
-4.89066511e-01 1.18920612e+00 8.99989605e-02 -7.72491336e-01
1.13331687e+00 -7.72629082e-02 7.67929852e-02 -1.03372335e+00
3.12459208e-02 2.58820385e-01 1.42862825e-02 -7.14194417e-01
7.48404562e-01 4.13897365e-01 5.50900340e-01 1.88904013e-02
-8.80913138e-01 -2.71579567e-02 2.51992106e-01 -9.83792990e-02
6.31521702e-01 2.29059100e-01 1.28749594e-01 -7.18976930e-02
1.05310428e+00 2.54967868e-01 4.79818523e-01 5.85531890e-01
-1.74261138e-01 2.83582866e-01 1.24234371e-01 -4.40415531e-01
-8.70337307e-01 -9.91273820e-01 -2.25215748e-01 6.09463453e-01
5.61896265e-01 2.07971647e-01 -8.10339808e-01 -4.21360493e-01
-1.66079342e-01 9.47589695e-01 -4.85232979e-01 1.90196276e-01
-7.38248765e-01 -6.10851049e-01 1.79967180e-01 1.13519520e-01
8.91756594e-01 -8.29749584e-01 -1.16248286e+00 5.32683320e-02
-1.90653041e-01 -1.27374637e+00 -5.91348335e-02 -2.82348037e-01
-4.93162036e-01 -1.36986959e+00 -4.02223468e-01 -4.93280977e-01
7.53540695e-01 1.21455228e+00 7.88303375e-01 8.69214758e-02
-5.01437306e-01 2.76148885e-01 -1.04249872e-01 -5.51099181e-01
-5.76083779e-01 -2.89808482e-01 -1.30550057e-01 5.26765406e-01
3.47379863e-01 -1.52473181e-01 -8.11343908e-01 8.19789410e-01
-3.45844746e-01 2.65459329e-01 7.48618245e-01 3.46673250e-01
5.04498959e-01 4.89117950e-01 3.35088134e-01 -8.41354430e-01
2.10333586e-01 -3.96951854e-01 -1.03566241e+00 2.45464146e-01
-7.03281701e-01 -2.80099839e-01 2.69281417e-01 1.50461480e-01
-1.38669395e+00 1.65726945e-01 7.21157938e-02 -4.86718453e-02
-6.45219266e-01 -8.21839646e-02 -3.72976631e-01 -9.56703424e-02
5.15294850e-01 4.07733202e-01 -5.29615805e-02 -1.22367386e-02
4.12909567e-01 6.96450233e-01 5.69043577e-01 6.51119575e-02
1.16274214e+00 8.97591174e-01 3.36985707e-01 -1.44993007e+00
-3.96073341e-01 -8.41806531e-01 -8.24125528e-01 -1.03600633e+00
8.15742552e-01 -6.99922740e-01 -5.90188682e-01 5.82043409e-01
-9.29118156e-01 -2.82483906e-01 1.48509353e-01 3.61110210e-01
-3.95277292e-01 4.36776847e-01 2.99716294e-01 -1.38522875e+00
-1.67380184e-01 -8.71483028e-01 6.45595133e-01 5.84335268e-01
2.56161392e-01 -8.07515442e-01 -1.20997638e-01 5.78365684e-01
2.67124385e-01 2.55064875e-01 4.40694243e-01 -9.35092717e-02
-9.36214507e-01 -1.94116682e-01 -4.65129375e-01 1.74999610e-01
1.84551835e-01 3.14316660e-01 -1.16640902e+00 2.28508890e-01
-4.41338062e-01 5.10095119e-01 9.32378292e-01 4.38195735e-01
5.52195907e-01 1.67439103e-01 -5.58517814e-01 2.68512875e-01
1.27981639e+00 2.81517535e-01 9.84543800e-01 5.30638218e-01
8.27437818e-01 1.20648408e+00 1.39921224e+00 4.43884134e-02
3.58404368e-01 8.67943943e-01 6.49418831e-01 -4.55616772e-01
-4.59362060e-01 -7.18288496e-03 4.33571219e-01 1.07531160e-01
-1.70543581e-01 -3.31859477e-02 -1.01212096e+00 7.53325343e-01
-1.78648400e+00 -1.17027438e+00 -1.13615310e+00 2.67819929e+00
1.31090030e-01 4.47095275e-01 2.25754052e-01 2.61293977e-01
9.82341707e-01 -4.12276536e-02 -3.31270993e-01 -4.36897367e-01
-3.00427973e-01 -3.35147679e-01 1.02570534e+00 7.39420056e-01
-9.83541131e-01 9.69918609e-01 4.99250364e+00 6.11233711e-01
-1.02638412e+00 -2.16968164e-01 3.49493057e-01 4.84720200e-01
-2.06989899e-01 6.45310879e-02 -9.97531950e-01 3.60148698e-01
7.48753250e-01 1.13484459e-02 1.08020075e-01 4.60283697e-01
1.01352763e+00 -8.84460509e-01 -4.57665622e-01 6.10392511e-01
9.57686529e-02 -9.92545247e-01 -2.19969958e-01 1.22751586e-01
5.39229155e-01 -6.64531067e-03 -2.02221200e-01 -2.06351712e-01
-1.17311105e-01 -7.40293086e-01 4.56164539e-01 9.21794832e-01
4.19115692e-01 -9.59816337e-01 8.67808163e-01 3.38650912e-01
-1.46582854e+00 1.26242280e-01 -7.65114799e-02 2.06850946e-01
4.28111076e-01 5.13993263e-01 -1.58511734e+00 6.60086632e-01
2.24753350e-01 5.13028502e-01 -5.42741358e-01 1.32453811e+00
-4.54421163e-01 4.66913700e-01 -2.26246983e-01 8.41717273e-02
2.61340678e-01 -6.60068393e-01 8.59024644e-01 1.46204150e+00
1.27482742e-01 -1.83337688e-01 7.52460584e-02 6.53458118e-01
4.29135740e-01 1.21356502e-01 -9.55034018e-01 5.10607421e-01
4.96945232e-01 1.34656131e+00 -8.70216668e-01 -2.71052480e-01
-4.76477325e-01 7.93197691e-01 -2.70714760e-01 6.04313731e-01
-9.19204891e-01 -4.43288505e-01 6.01488769e-01 6.81343317e-01
1.35935441e-01 -4.08177584e-01 -2.20116630e-01 -3.10095698e-01
3.03688869e-02 -3.14738184e-01 -6.44013882e-02 -7.41152585e-01
-3.98572505e-01 2.83137709e-01 1.08759545e-01 -1.38811672e+00
-1.15107149e-02 -4.51445490e-01 -8.36955011e-01 1.05783975e+00
-2.20612717e+00 -1.21733141e+00 -1.08149731e+00 5.89396954e-01
7.22523987e-01 -4.05159332e-02 2.54702479e-01 3.74092609e-01
-5.78067183e-01 1.56582311e-01 1.73131928e-01 -8.30599070e-02
4.40281063e-01 -1.05509973e+00 3.05065393e-01 1.49725425e+00
-1.92539036e-01 1.10905640e-01 1.04866123e+00 -7.81564951e-01
-1.17880547e+00 -1.06247091e+00 7.63565123e-01 1.43735418e-02
3.58416557e-01 -7.39913573e-03 -9.47571754e-01 8.08500648e-02
-4.51376960e-02 -2.52590626e-01 9.41966996e-02 -3.22675705e-01
2.28769526e-01 -5.06355107e-01 -9.24850643e-01 7.29356050e-01
7.46200979e-01 -4.20150816e-01 -4.68966454e-01 1.94532812e-01
-8.13926384e-02 -1.59730181e-01 -1.74918666e-01 3.99659634e-01
6.54279888e-01 -1.24262595e+00 8.77596915e-01 8.44755322e-02
-1.89155489e-01 -8.57763469e-01 2.52852142e-01 -1.01322508e+00
2.75173604e-01 -5.20903468e-01 2.11905017e-01 1.14669836e+00
1.86101794e-01 -8.09998989e-01 6.58702254e-01 2.77114064e-01
-2.41855457e-01 -2.50161201e-01 -8.05655718e-01 -6.49724662e-01
-6.60503566e-01 -5.02244174e-01 1.30530670e-01 5.31874180e-01
-7.23482311e-01 6.05206043e-02 -4.09918547e-01 6.52690053e-01
9.33749914e-01 1.56103581e-01 1.43416512e+00 -1.21186244e+00
3.18892837e-01 -4.32591677e-01 -5.66197932e-01 -8.19837809e-01
-7.48466551e-02 -2.33934298e-01 3.83720100e-01 -1.77443063e+00
-1.28132939e-01 -5.59281230e-01 1.49925888e-01 -1.19274169e-01
-3.30489248e-01 8.62123072e-02 -1.20724542e-02 1.35029927e-01
-1.37930885e-01 2.23500058e-01 9.22215581e-01 1.18239664e-01
-2.58421123e-01 4.81334448e-01 1.23724371e-01 8.47498000e-01
9.58096802e-01 -4.60265763e-02 -5.43499827e-01 9.26744789e-02
-6.17127940e-02 -2.50322759e-01 6.63764238e-01 -1.18431854e+00
3.03266644e-01 -2.24361539e-01 -5.11483029e-02 -1.19175184e+00
2.83880115e-01 -9.26400006e-01 1.50849130e-02 5.18935740e-01
2.06760168e-01 -1.92697465e-01 2.75793076e-01 7.91257799e-01
-1.02021188e-01 -3.33100945e-01 8.57293785e-01 1.93965554e-01
-1.42589986e+00 1.98555533e-02 -6.55452609e-01 -2.52923220e-01
1.49358308e+00 -9.90041494e-01 -3.03285539e-01 -3.91697645e-01
-3.14314485e-01 4.19928998e-01 4.24753755e-01 3.72490972e-01
8.70710313e-01 -7.92643428e-01 -8.46639812e-01 2.49518752e-01
2.99102038e-01 -1.61115289e-01 3.60693991e-01 1.05072486e+00
-6.84952497e-01 4.45385635e-01 -2.73779720e-01 -6.97391510e-01
-1.92416275e+00 2.21272871e-01 5.04823744e-01 2.86968768e-01
-7.51739264e-01 1.78587154e-01 3.20693314e-01 2.62219191e-01
-1.82595961e-02 -3.80289555e-02 -6.78961158e-01 2.28905976e-01
5.83616078e-01 8.20346475e-01 4.16620560e-02 -1.09019709e+00
-2.69010991e-01 9.23553348e-01 7.79797062e-02 -4.86616753e-02
7.51265645e-01 -7.19803989e-01 2.82291353e-01 5.43861210e-01
6.05077744e-01 3.60722780e-01 -1.24536765e+00 -2.08199009e-01
-7.00313896e-02 -6.99396551e-01 2.66306221e-01 -5.44541121e-01
-6.54823422e-01 8.97054553e-01 9.99423981e-01 9.46934745e-02
1.03843904e+00 -4.85704750e-01 6.31378055e-01 3.13519895e-01
3.95096093e-01 -1.40793586e+00 -4.54054922e-01 2.95644134e-01
4.54075426e-01 -1.35186350e+00 1.09497100e-01 -7.78496087e-01
-6.99460387e-01 1.31116807e+00 4.33507025e-01 3.87542814e-01
2.86072165e-01 -1.09668896e-02 -7.21608615e-03 -2.53876328e-01
-1.07129686e-01 -1.16822004e+00 3.82901996e-01 7.12339699e-01
-2.53507979e-02 2.05206081e-01 -2.91722894e-01 -4.86562937e-01
8.32728595e-02 -8.05659667e-02 7.52307951e-01 7.03265429e-01
-1.01969802e+00 -6.93120897e-01 -6.44006133e-01 2.70934135e-01
2.27662191e-01 1.84666708e-01 -2.75490940e-01 9.77776229e-01
4.48264569e-01 1.25092697e+00 -3.25617902e-02 -2.39399895e-01
4.80413437e-01 -5.31752072e-02 2.15423912e-01 -2.05927551e-01
-6.57847822e-02 -2.60576248e-01 6.33735299e-01 -4.28916365e-01
-5.16736627e-01 -9.05637443e-01 -1.25804722e+00 -2.26064995e-01
-2.37142473e-01 -8.16937089e-02 9.95293081e-01 9.32733774e-01
1.72569469e-01 4.55830634e-01 9.75997984e-01 -6.76516294e-01
2.43258715e-01 -5.57802975e-01 -3.44352126e-01 4.14388448e-01
3.63163024e-01 -9.28012848e-01 -3.75728458e-01 6.44402020e-03] | [8.02596664428711, -1.360795497894287] |
4b4c0cb4-c0e4-48ab-ae9a-8efc849cfe44 | learning-syntax-from-naturally-occurring | 2104.13933 | null | https://arxiv.org/abs/2104.13933v1 | https://arxiv.org/pdf/2104.13933v1.pdf | Learning Syntax from Naturally-Occurring Bracketings | Naturally-occurring bracketings, such as answer fragments to natural language questions and hyperlinks on webpages, can reflect human syntactic intuition regarding phrasal boundaries. Their availability and approximate correspondence to syntax make them appealing as distant information sources to incorporate into unsupervised constituency parsing. But they are noisy and incomplete; to address this challenge, we develop a partial-brackets-aware structured ramp loss in learning. Experiments demonstrate that our distantly-supervised models trained on naturally-occurring bracketing data are more accurate in inducing syntactic structures than competing unsupervised systems. On the English WSJ corpus, our models achieve an unlabeled F1 score of 68.9 for constituency parsing. | ['Lillian Lee', 'Igor Malioutov', 'Ozan İrsoy', 'Tianze Shi'] | 2021-04-28 | null | https://aclanthology.org/2021.naacl-main.234 | https://aclanthology.org/2021.naacl-main.234.pdf | naacl-2021-4 | ['constituency-parsing'] | ['natural-language-processing'] | [-2.44176798e-02 7.69186318e-01 -8.07237327e-01 -1.06692350e+00
-1.38838470e+00 -1.00184977e+00 2.90501565e-01 4.90318298e-01
-5.24739265e-01 7.53056288e-01 8.51713240e-01 -7.86322534e-01
3.97433043e-01 -6.09495461e-01 -8.92454028e-01 1.54771313e-01
8.25588927e-02 4.15432364e-01 5.18339157e-01 -3.33464265e-01
-3.41536440e-02 -2.35195160e-01 -8.91718447e-01 7.48628199e-01
1.00411010e+00 5.21773934e-01 2.96051919e-01 3.15854847e-01
-7.58938015e-01 1.26669180e+00 -5.24139404e-01 -6.77149653e-01
-1.54390857e-01 -2.63624907e-01 -1.21138537e+00 -7.35320896e-02
9.19537425e-01 -1.99307859e-01 1.84892893e-01 8.42089832e-01
-8.17961711e-03 -2.68234402e-01 3.96841615e-01 -4.69310701e-01
-7.40752935e-01 1.37837994e+00 -2.55625457e-01 6.24719024e-01
6.86487734e-01 -2.68936485e-01 2.21732903e+00 -9.01743889e-01
8.61993611e-01 1.20658076e+00 7.00630426e-01 4.95490640e-01
-1.51306581e+00 -3.70507866e-01 4.85519171e-01 -2.67334670e-01
-8.30257952e-01 -5.75985134e-01 7.94983447e-01 -3.42202038e-01
1.00185406e+00 2.93586820e-01 -4.71578650e-02 1.19440341e+00
6.10859059e-02 1.05112433e+00 1.14360392e+00 -8.47409725e-01
1.98783390e-02 8.24054033e-02 9.07242477e-01 9.45785642e-01
1.80868834e-01 -1.20743699e-01 -6.05173528e-01 -1.66077346e-01
2.69906491e-01 -5.87543249e-01 2.17140373e-03 1.91458493e-01
-7.51311362e-01 1.09461737e+00 3.83615702e-01 3.45385045e-01
-7.12334812e-02 -1.13625281e-01 2.77292013e-01 3.34067404e-01
6.32626235e-01 6.56130135e-01 -8.64022553e-01 -3.08517423e-02
-5.61549187e-01 2.36440927e-01 1.24069476e+00 1.15264189e+00
9.26444590e-01 -2.89202839e-01 1.74466390e-02 1.09901536e+00
3.18473488e-01 2.34906122e-01 4.08306479e-01 -1.15586841e+00
1.00290573e+00 6.08579278e-01 -3.39797623e-02 -6.62532330e-01
-6.19755626e-01 -1.42491832e-01 4.75589111e-02 -4.98632580e-01
7.77641714e-01 -4.66050267e-01 -6.34115815e-01 1.88751602e+00
3.85658890e-01 -6.00764871e-01 7.14401230e-02 6.72269940e-01
1.12181807e+00 7.10587680e-01 7.61993945e-01 -1.90120831e-01
1.71184051e+00 -8.25824857e-01 -7.89138436e-01 -7.31922388e-01
9.92264450e-01 -9.12168622e-01 1.54017460e+00 -1.93258092e-01
-1.16064227e+00 -3.06634992e-01 -7.46315181e-01 -4.77664649e-01
-1.06866129e-01 -5.74343428e-02 7.80312240e-01 5.85864365e-01
-6.88236237e-01 3.51948053e-01 -7.13414371e-01 -7.79597610e-02
2.24546287e-02 -5.79137392e-02 -2.37195283e-01 4.81994413e-02
-1.30020356e+00 8.36020350e-01 4.37747926e-01 -1.66325182e-01
2.70155761e-02 -6.72593117e-01 -1.27052450e+00 5.53417690e-02
5.05227447e-01 -2.34632999e-01 1.78463352e+00 -5.99065065e-01
-1.15510619e+00 1.29894590e+00 -4.17579919e-01 -4.61214125e-01
-1.38121858e-01 -2.32909143e-01 -2.59242654e-01 9.97225195e-02
6.00620747e-01 6.06324375e-01 2.64273763e-01 -1.13697457e+00
-8.62833083e-01 -2.43223265e-01 2.27419585e-01 2.71938354e-01
-2.42098525e-01 3.82629156e-01 -2.74822079e-02 -5.85532725e-01
5.26813745e-01 -7.02849627e-01 -1.44356951e-01 -5.29190660e-01
-5.27235031e-01 -8.89781535e-01 3.61660928e-01 -8.68881226e-01
1.26919103e+00 -1.86264455e+00 -4.23536807e-01 -1.18970890e-02
1.76886946e-01 -5.42705953e-02 -5.92269376e-02 3.88176769e-01
1.68425381e-01 4.00445163e-01 -1.24729834e-01 -2.41726801e-01
3.54292631e-01 5.60666144e-01 -4.25560474e-01 -6.50908500e-02
4.81069595e-01 1.01144958e+00 -9.49788511e-01 -7.57322550e-01
-2.82940716e-01 -1.93388253e-01 -6.80423737e-01 1.23378329e-01
-6.81597829e-01 1.18944198e-01 -5.37935436e-01 6.97131574e-01
1.97734401e-01 -2.66266376e-01 4.16665316e-01 4.57443371e-02
-4.43955176e-02 1.45164323e+00 -6.20032668e-01 1.44323170e+00
-6.10146046e-01 3.54993731e-01 1.50071204e-01 -6.22358561e-01
7.46383309e-01 2.39678085e-01 -9.71131325e-02 -6.88841522e-01
-1.05711736e-01 3.97479177e-01 1.81975305e-01 -5.71798027e-01
3.30943704e-01 -4.10972714e-01 -6.62867427e-01 3.97648871e-01
2.99778581e-01 -2.66169041e-01 4.90778446e-01 3.63512725e-01
1.22235179e+00 -3.00776549e-02 3.26403677e-01 -6.25052869e-01
2.45672941e-01 2.38835603e-01 8.54709983e-01 7.23529577e-01
-1.18716262e-01 3.40422004e-01 5.62007368e-01 -4.20887232e-01
-9.37123239e-01 -1.50812185e+00 -4.79655415e-01 1.59224260e+00
-2.66767800e-01 -7.17653751e-01 -7.66084790e-01 -1.11324823e+00
-1.43782794e-01 9.90250409e-01 -2.46974066e-01 4.62647229e-01
-1.08857620e+00 -3.40226084e-01 3.58457685e-01 7.28336692e-01
4.88897376e-02 -1.17967689e+00 4.39764559e-03 5.65873802e-01
-6.04934394e-01 -1.61248124e+00 -6.71188474e-01 5.68826199e-01
-8.54877412e-01 -1.02816188e+00 -9.91521925e-02 -1.49649942e+00
4.47551996e-01 -2.70372808e-01 1.81669068e+00 5.18437475e-02
1.17615022e-01 2.02718601e-01 -3.76739621e-01 -2.81825304e-01
-6.43567801e-01 4.08182144e-01 -2.60927320e-01 -6.54438734e-01
1.10966432e+00 -4.83613372e-01 -2.33603746e-01 -6.30786642e-02
-6.62729561e-01 -3.40115577e-01 2.70035386e-01 8.79905462e-01
5.32939017e-01 -6.87970996e-01 6.21540070e-01 -1.71814144e+00
8.14268410e-01 -4.94754821e-01 -6.79393589e-01 2.39586323e-01
-3.71383041e-01 2.61029333e-01 7.63109446e-01 -2.84853615e-02
-1.39120591e+00 8.35058019e-02 -6.57724202e-01 7.06269264e-01
-4.28033948e-01 6.74709320e-01 -3.39701883e-02 4.02242273e-01
9.11140382e-01 -2.25451976e-01 -3.41721177e-01 -6.48103595e-01
5.14842093e-01 5.57999253e-01 5.11813283e-01 -1.10188711e+00
6.56649530e-01 -8.90920162e-02 -6.36725903e-01 -8.80342185e-01
-1.88389528e+00 -6.61192775e-01 -3.75059217e-01 4.15728927e-01
9.92363274e-01 -9.46016133e-01 -2.80058861e-01 -2.37627491e-01
-1.37966788e+00 -2.45252252e-01 -2.23959103e-01 3.38487417e-01
-2.23290995e-01 4.71555978e-01 -1.34267855e+00 -5.17063618e-01
-2.02459514e-01 -5.27868330e-01 1.02320719e+00 5.89533634e-02
-8.51616561e-01 -1.25265074e+00 1.97403908e-01 7.01226354e-01
1.14299677e-01 -5.24898618e-02 1.34304821e+00 -9.85269070e-01
-3.81594241e-01 -3.26401480e-02 2.04824135e-02 3.23540479e-01
3.39850068e-01 -2.02838480e-01 -8.36862504e-01 2.94293225e-01
2.35855609e-01 -7.58522332e-01 7.54869938e-01 2.50736117e-01
6.64317489e-01 -7.40294933e-01 -4.72294614e-02 3.25445049e-02
1.18701267e+00 -1.73457891e-01 -2.58719865e-02 2.96773046e-01
4.34731752e-01 1.12106133e+00 2.79890001e-01 3.43466140e-02
8.37139666e-01 2.06392273e-01 -6.00135513e-02 7.34543428e-02
-1.31937101e-01 -6.95476055e-01 3.17793041e-01 1.39075375e+00
6.95465446e-01 -2.32141074e-02 -1.08491147e+00 6.01591170e-01
-1.38777828e+00 -7.16963112e-01 -3.65198761e-01 1.62464499e+00
1.57422519e+00 6.30298555e-01 -9.45754424e-02 -3.42587173e-01
8.41291368e-01 3.60879898e-01 -1.02440082e-01 -4.68457162e-01
-2.94603884e-01 4.69870806e-01 2.34029368e-01 1.04022348e+00
-1.26082718e+00 1.44344246e+00 6.91864538e+00 4.64403838e-01
-4.26342368e-01 1.30096659e-01 6.73494220e-01 3.62313062e-01
-7.46026337e-01 1.83301508e-01 -1.44077051e+00 4.30074364e-01
1.09075749e+00 1.89678755e-03 -5.22944815e-02 9.79610622e-01
7.52027705e-02 -4.73686233e-02 -1.34493947e+00 3.50454330e-01
-7.21627772e-02 -1.36359048e+00 -1.90924749e-01 -3.59248459e-01
7.28747606e-01 3.31893742e-01 -1.34636655e-01 4.66909945e-01
1.00230443e+00 -7.65686750e-01 6.07232213e-01 -3.65096211e-01
4.08036798e-01 -4.27187741e-01 6.99644983e-01 5.78279138e-01
-1.07712400e+00 1.86890721e-01 -3.81723076e-01 -3.60190958e-01
5.29250085e-01 6.64372742e-01 -7.75360942e-01 -2.12657511e-01
5.11113882e-01 1.42493948e-01 -5.70995867e-01 4.73195434e-01
-9.96836662e-01 1.35706890e+00 -6.76358163e-01 -5.15717208e-01
6.89158201e-01 -4.89657335e-02 2.67727822e-01 1.47948480e+00
-4.01562303e-01 4.15492535e-01 4.19351131e-01 8.37314069e-01
-4.47678149e-01 4.98604327e-01 -6.10030234e-01 -3.22466157e-02
6.30747557e-01 1.01587951e+00 -5.31614900e-01 -2.97218740e-01
-7.83065259e-01 4.60077077e-01 1.01374853e+00 4.06600952e-01
-4.23110425e-01 -2.08588257e-01 3.76903683e-01 1.05212621e-01
1.27860233e-01 -3.26430708e-01 -3.94349605e-01 -1.19633627e+00
5.74494898e-01 -8.70992005e-01 6.87205791e-01 -3.53783011e-01
-1.82471263e+00 6.29322886e-01 -2.16406912e-01 -6.29868925e-01
-4.08577889e-01 -9.36775148e-01 -6.65941715e-01 6.71583831e-01
-1.73382092e+00 -9.67560351e-01 3.96025658e-01 1.68927789e-01
7.25901425e-01 6.10798039e-02 1.06821942e+00 -5.68069518e-02
-2.93724775e-01 5.86755037e-01 -2.09158942e-01 7.11493015e-01
6.50469065e-01 -1.74181008e+00 7.03268766e-01 8.72168720e-01
8.21922243e-01 9.51792300e-01 7.53603399e-01 -5.55200100e-01
-9.15858090e-01 -7.76255608e-01 1.74945176e+00 -8.51604223e-01
1.07794714e+00 -5.93922734e-01 -1.08523846e+00 1.15198517e+00
4.80059862e-01 -1.27918571e-01 1.14525354e+00 9.62369204e-01
-7.68495798e-01 1.88530713e-01 -8.56002569e-01 6.46019161e-01
9.85314846e-01 -7.36958563e-01 -1.58473802e+00 7.06416965e-01
1.16337669e+00 -4.15156275e-01 -8.94316733e-01 2.86588252e-01
-1.02489009e-01 -6.22777581e-01 5.32048345e-01 -8.59291494e-01
6.75102830e-01 2.66938388e-01 -2.16974869e-01 -1.03941429e+00
-2.96048045e-01 -6.95239484e-01 6.55357465e-02 1.33841574e+00
1.23756039e+00 -4.84090388e-01 1.21320820e+00 9.76204515e-01
-2.34480351e-01 -4.65841085e-01 -8.36574435e-01 -6.19453430e-01
5.48112094e-01 -3.14516127e-01 9.96222370e-04 8.87273133e-01
7.86692441e-01 1.22669899e+00 3.02179068e-01 6.59623519e-02
7.07976639e-01 2.36799181e-01 1.94871157e-01 -1.21738863e+00
-3.45727891e-01 -2.16270387e-01 8.13124105e-02 -1.57429135e+00
6.56481802e-01 -1.15522754e+00 2.97793299e-01 -1.35074842e+00
-2.15007178e-02 -7.06715882e-01 1.12701096e-01 4.36402321e-01
-4.26255018e-01 -1.37623306e-02 -6.01368174e-02 -1.31093860e-02
-7.10599720e-01 2.95053273e-01 8.20477605e-01 4.03922275e-02
1.14971735e-02 -4.12477218e-02 -9.11102176e-01 1.26990187e+00
8.25421035e-01 -5.85667253e-01 -1.89916432e-01 -8.53788137e-01
3.06863099e-01 1.46342695e-01 -4.14836705e-01 -4.92841154e-01
2.39870995e-01 -8.75744149e-02 1.22198975e-02 -5.46282589e-01
1.18845947e-01 -4.85289842e-01 -9.48282301e-01 -1.77795030e-02
-7.56282210e-01 1.59098387e-01 8.07035267e-02 3.95957589e-01
-4.87161607e-01 -7.47110546e-01 3.46380055e-01 -5.51852286e-01
-5.50962329e-01 -4.19947803e-02 -6.00821733e-01 1.08303332e+00
3.54636759e-01 1.97519138e-01 -5.55296242e-01 -3.38285208e-01
-8.27217996e-01 3.23791683e-01 8.88643637e-02 2.79918641e-01
3.18857163e-01 -9.82608736e-01 -7.83452690e-01 -9.04453844e-02
2.01393917e-01 9.20760483e-02 -3.98111105e-01 8.22751075e-02
-5.06001472e-01 5.52166998e-01 4.45774078e-01 -4.87215847e-01
-8.99584413e-01 1.57653838e-01 -1.64172128e-01 -4.71915692e-01
-5.32072961e-01 1.21625817e+00 1.09770350e-01 -1.01492512e+00
3.29836756e-01 -6.86155319e-01 -1.11568920e-01 -1.81628481e-01
3.31103981e-01 -3.35851967e-01 5.78770898e-02 -3.54532361e-01
-2.08995417e-01 2.43574694e-01 -4.56994474e-01 -8.61315578e-02
1.26830721e+00 -1.97140962e-01 -2.30455160e-01 5.32415271e-01
1.18694293e+00 5.47839999e-01 -1.00826919e+00 -6.58318639e-01
1.15454519e+00 -5.39172580e-03 -3.95796835e-01 -7.28074133e-01
-3.01924616e-01 5.48684299e-01 -2.73316622e-01 5.05096734e-01
2.94421852e-01 7.00907052e-01 1.14760673e+00 8.75388861e-01
1.99648365e-01 -1.31397688e+00 -1.46271922e-02 1.03173792e+00
4.20921266e-01 -1.52150702e+00 -4.25093710e-01 -1.06929755e+00
-4.50543880e-01 9.28609192e-01 7.97881305e-01 -2.45646954e-01
6.28756166e-01 5.96315742e-01 4.70988005e-01 -2.17959002e-01
-9.26590323e-01 -1.54349968e-01 1.35818750e-01 4.01330680e-01
1.13528550e+00 6.26134081e-03 -7.41866410e-01 7.64401853e-01
-7.59405196e-01 -7.69117296e-01 4.59285676e-01 9.33591783e-01
-8.21764112e-01 -1.29590929e+00 1.23179397e-02 5.28679371e-01
-9.64549184e-01 -6.82247698e-01 -4.14844155e-01 6.02458596e-01
-4.32559311e-01 1.18467402e+00 7.79467076e-02 3.18349093e-01
3.53660256e-01 4.11977798e-01 3.63255471e-01 -1.37546682e+00
-5.51538169e-01 -1.04395010e-01 1.02064037e+00 -4.39296812e-01
-3.64776552e-01 -5.97721040e-01 -1.46953738e+00 1.28915101e-01
-3.75748426e-01 6.79476619e-01 3.46785128e-01 1.22143447e+00
-1.56768858e-01 2.63705216e-02 5.07919431e-01 -9.19774547e-02
-1.07410419e+00 -9.28270936e-01 -3.57490689e-01 6.46710992e-01
9.63049605e-02 -9.24835876e-02 -5.22813916e-01 1.76908359e-01] | [10.360138893127441, 9.63306999206543] |
ea283a4e-ce8b-4b14-9934-47c3d51f5b43 | pfgm-unlocking-the-potential-of-physics | 2302.04265 | null | https://arxiv.org/abs/2302.04265v2 | https://arxiv.org/pdf/2302.04265v2.pdf | PFGM++: Unlocking the Potential of Physics-Inspired Generative Models | We introduce a new family of physics-inspired generative models termed PFGM++ that unifies diffusion models and Poisson Flow Generative Models (PFGM). These models realize generative trajectories for $N$ dimensional data by embedding paths in $N{+}D$ dimensional space while still controlling the progression with a simple scalar norm of the $D$ additional variables. The new models reduce to PFGM when $D{=}1$ and to diffusion models when $D{\to}\infty$. The flexibility of choosing $D$ allows us to trade off robustness against rigidity as increasing $D$ results in more concentrated coupling between the data and the additional variable norms. We dispense with the biased large batch field targets used in PFGM and instead provide an unbiased perturbation-based objective similar to diffusion models. To explore different choices of $D$, we provide a direct alignment method for transferring well-tuned hyperparameters from diffusion models ($D{\to} \infty$) to any finite $D$ values. Our experiments show that models with finite $D$ can be superior to previous state-of-the-art diffusion models on CIFAR-10/FFHQ $64{\times}64$ datasets, with FID scores of $1.91/2.43$ when $D{=}2048/128$. In class-conditional setting, $D{=}2048$ yields current state-of-the-art FID of $1.74$ on CIFAR-10. In addition, we demonstrate that models with smaller $D$ exhibit improved robustness against modeling errors. Code is available at https://github.com/Newbeeer/pfgmpp | ['Tommi Jaakkola', 'Max Tegmark', 'Shangyuan Tong', 'Yonglong Tian', 'Ziming Liu', 'Yilun Xu'] | 2023-02-08 | null | null | null | null | ['2048'] | ['playing-games'] | [-4.11342621e-01 -5.44985831e-02 -9.67885703e-02 -2.66124427e-01
-9.38797295e-01 -4.26232249e-01 5.43181717e-01 -4.27582890e-01
-4.77114201e-01 8.70358706e-01 -7.10505713e-03 -4.45964515e-01
-5.40786684e-01 -1.01625192e+00 -6.08940661e-01 -1.06243408e+00
-5.22449374e-01 5.13015211e-01 6.39289469e-02 -6.00189827e-02
3.45578082e-02 4.20866311e-01 -9.62608516e-01 -3.06044042e-01
8.38008583e-01 8.77977550e-01 -1.09370403e-01 6.50949955e-01
6.87254034e-03 3.20946544e-01 -6.11562192e-01 -5.65283000e-01
5.22376239e-01 -4.66182411e-01 -3.80942822e-01 -1.42020240e-01
4.40841883e-01 -2.59046316e-01 -9.30787146e-01 1.05476773e+00
6.61655009e-01 3.50053042e-01 1.05496550e+00 -9.58673894e-01
-1.24775624e+00 6.09357238e-01 -9.13355708e-01 7.94218540e-01
-3.35856944e-01 5.92927158e-01 8.87998283e-01 -8.66867304e-01
4.94357616e-01 1.27704191e+00 5.71035922e-01 6.40768945e-01
-1.57628131e+00 -1.08596480e+00 4.49481398e-01 -3.26938629e-01
-1.58722591e+00 -1.57634974e-01 8.40085864e-01 -7.69537926e-01
9.22561288e-01 -4.70414273e-02 3.95469695e-01 1.04326403e+00
2.53811896e-01 3.10168773e-01 8.85713279e-01 -1.95221975e-01
3.49138498e-01 -2.82189995e-01 3.73751074e-01 6.85124874e-01
3.55492622e-01 2.94425219e-01 -3.98750722e-01 -3.95382822e-01
1.11164546e+00 -1.10762909e-01 -2.02288285e-01 -4.83111069e-02
-8.52556169e-01 1.17859316e+00 4.81836200e-01 1.44290939e-01
-1.98721752e-01 8.83705914e-01 -6.54374063e-02 -3.91729437e-02
4.93705779e-01 1.74926773e-01 -2.27051258e-01 -7.21511543e-02
-8.93965542e-01 6.59786284e-01 3.33806962e-01 1.08679008e+00
6.13995671e-01 5.80214679e-01 -4.21893477e-01 7.58919537e-01
6.17406607e-01 1.02492380e+00 -1.03563510e-01 -1.31847787e+00
5.25316298e-01 5.40956445e-02 1.70804918e-01 -7.31679022e-01
-2.80962020e-01 -7.59588659e-01 -1.08807576e+00 5.65717742e-02
6.81581438e-01 -3.93966466e-01 -1.10969639e+00 2.20721531e+00
1.28621981e-01 2.41376072e-01 -3.39453489e-01 6.54332757e-01
4.46676016e-01 7.33627677e-01 3.40355724e-01 -1.97440937e-01
1.16771674e+00 -5.21860123e-01 -3.02764565e-01 -2.96741337e-01
4.34236705e-01 -6.94018781e-01 1.15584481e+00 3.29433173e-01
-1.09744096e+00 -3.47294569e-01 -6.61451399e-01 2.72600830e-01
9.43650529e-02 -1.58021152e-01 5.89517713e-01 8.53109837e-01
-1.18875420e+00 5.34400046e-01 -1.03487825e+00 2.00394407e-01
7.90131271e-01 4.30649102e-01 4.20540661e-01 -1.87777251e-01
-1.17015648e+00 3.41256410e-01 -4.46692675e-01 -1.71441227e-01
-1.04943979e+00 -1.06423640e+00 -4.11382437e-01 -1.38611734e-01
-9.08893496e-02 -8.09432030e-01 8.70994747e-01 -2.57416695e-01
-1.08060455e+00 5.46038687e-01 -2.41937697e-01 -6.84000373e-01
3.69179875e-01 -9.68718156e-03 -4.54590619e-01 2.02030852e-01
7.04270825e-02 7.60967255e-01 9.12648499e-01 -1.08700740e+00
-1.63608819e-01 -3.49189281e-01 -7.71637782e-02 -2.27205679e-01
-1.96247086e-01 -1.28271058e-01 -3.32345754e-01 -9.68801320e-01
-7.15298858e-03 -1.24199903e+00 -3.22092563e-01 -2.01615784e-02
-4.07165140e-01 -6.31766096e-02 5.46592772e-01 -3.17055196e-01
1.57776308e+00 -1.94832134e+00 2.32060961e-02 3.82210046e-01
6.18955195e-01 2.42061198e-01 3.67258763e-04 2.27928445e-01
9.37066823e-02 4.46118623e-01 -5.73239505e-01 -3.90533656e-01
1.83566045e-02 8.71921182e-02 -2.49435186e-01 5.10982335e-01
-4.70863841e-02 7.63784885e-01 -6.58501446e-01 -1.32096604e-01
-8.95205662e-02 9.20949996e-01 -8.66636634e-01 -4.15164411e-01
-3.22582155e-01 5.36760449e-01 -7.19593883e-01 3.62310797e-01
8.76719415e-01 -6.15259826e-01 -1.47243291e-01 1.92857131e-01
8.17572623e-02 2.37026513e-02 -1.16665936e+00 1.50640476e+00
-7.78145045e-02 3.56013596e-01 1.57104179e-01 -8.73796403e-01
9.79717433e-01 3.10938172e-02 7.08961844e-01 -4.76937175e-01
1.32235467e-01 6.77799284e-02 2.97851384e-01 1.49671808e-01
2.09289834e-01 -5.00692606e-01 -1.08834960e-01 6.11601412e-01
6.55249059e-02 -2.20390707e-01 2.75252819e-01 3.77937734e-01
1.36664569e+00 -2.00670689e-01 -4.74752784e-01 -6.24662161e-01
-1.10182911e-03 -3.27929050e-01 6.19121194e-01 9.55230296e-01
-4.73907471e-01 5.50774992e-01 5.07815003e-01 -2.10782945e-01
-1.03453922e+00 -1.48574507e+00 -4.09393698e-01 6.55963004e-01
-2.27598567e-02 -3.10954332e-01 -8.20591629e-01 -4.68017906e-01
1.56455502e-01 8.66655111e-01 -6.30655766e-01 -2.51300801e-02
-6.61428154e-01 -1.61199594e+00 7.50708997e-01 4.86131072e-01
5.79728484e-01 -6.88891172e-01 -2.53081948e-01 2.62420058e-01
1.05093539e-01 -7.65181839e-01 -6.79388285e-01 7.04035014e-02
-8.88033807e-01 -6.73991799e-01 -1.03841007e+00 -3.25242847e-01
5.79630435e-01 3.47282924e-02 1.10082543e+00 -2.99486697e-01
-2.90907651e-01 4.62626278e-01 -1.78275302e-01 -3.11161637e-01
-1.29305080e-01 5.90917990e-02 1.47416264e-01 -2.70657182e-01
4.06912804e-01 -8.53883624e-01 -1.27411294e+00 3.73487025e-01
-9.23630297e-01 -4.80424970e-01 1.56837970e-01 7.18377948e-01
8.49635899e-01 -1.35732805e-02 5.99521816e-01 -8.40244770e-01
6.50715053e-01 -4.79339808e-01 -5.24772942e-01 -2.74045676e-01
-8.36512029e-01 2.40545720e-01 5.61322570e-01 -6.10825479e-01
-7.98265636e-01 -5.04338443e-01 -2.21697226e-01 -9.18538213e-01
1.50323227e-01 2.20532075e-01 2.39625439e-01 2.01370880e-01
1.02098691e+00 2.40049176e-02 -1.48175672e-01 -4.46938872e-01
5.37682295e-01 9.92055759e-02 2.07920358e-01 -8.82112205e-01
7.50053167e-01 6.97344124e-01 6.36870041e-02 -7.19183385e-01
-4.26707476e-01 5.91029897e-02 -1.60370603e-01 1.22017801e-01
7.50220180e-01 -9.63039279e-01 -5.44735610e-01 5.02822638e-01
-6.55240178e-01 -7.15091646e-01 -5.29881001e-01 6.97487712e-01
-5.46724617e-01 1.58840835e-01 -8.04422557e-01 -7.73814082e-01
-3.76816660e-01 -1.21171021e+00 6.64847672e-01 1.76699281e-01
-1.81638703e-01 -1.06361926e+00 2.58863896e-01 1.58173949e-01
8.38836432e-01 -9.40583833e-03 1.11172199e+00 -1.92632556e-01
-7.34748363e-01 8.48843157e-03 -1.73482060e-01 3.16668600e-01
-1.45464567e-02 9.17005315e-02 -7.76901722e-01 -3.41224939e-01
-4.98998202e-02 1.56845570e-01 1.07482898e+00 1.10171700e+00
1.20368171e+00 -2.53376782e-01 -4.68650758e-01 7.51137078e-01
1.19761670e+00 5.23151338e-01 6.84032679e-01 -1.99753150e-01
5.34304321e-01 4.86746840e-02 2.37051677e-02 6.34316981e-01
2.97510982e-01 5.01912415e-01 1.03099279e-01 9.56260413e-02
-4.22011375e-01 -1.89880803e-01 1.58325702e-01 6.68012023e-01
-8.87968391e-02 -4.92277205e-01 -1.12024426e+00 4.97383565e-01
-1.38797367e+00 -1.08584285e+00 -1.01047941e-01 2.14060688e+00
8.19462478e-01 3.45699757e-01 2.03840867e-01 -2.80843109e-01
6.99301064e-01 4.16866779e-01 -8.12749982e-01 3.44404648e-03
-4.57128547e-02 7.90307164e-01 6.65222824e-01 8.83937657e-01
-9.41895366e-01 8.98893058e-01 5.53455639e+00 1.27272522e+00
-1.08437657e+00 3.00431281e-01 1.08508742e+00 -7.08604038e-01
-6.78163588e-01 -4.47549559e-02 -1.24105811e+00 8.53520751e-01
1.01756990e+00 -2.75292695e-02 3.88404816e-01 6.32945895e-01
3.98784876e-01 8.53416137e-03 -7.41580904e-01 1.02962673e+00
-1.83688611e-01 -1.57661366e+00 -8.67260098e-02 3.97619843e-01
8.04410696e-01 4.24895912e-01 5.58663785e-01 1.94860414e-01
7.27369130e-01 -1.21545374e+00 5.40833175e-01 5.32049179e-01
9.94293272e-01 -6.69775844e-01 -4.63343821e-02 2.17865422e-01
-1.15276635e+00 9.09320191e-02 -4.24342901e-01 1.68409497e-01
5.45573711e-01 9.42669153e-01 -2.70539641e-01 6.50683194e-02
8.93524587e-01 5.33627689e-01 -9.05569568e-02 6.10224545e-01
4.11056690e-02 8.71379614e-01 -5.86597264e-01 6.50526062e-02
2.31826872e-01 -4.13481414e-01 7.40253866e-01 1.16902399e+00
6.63649976e-01 4.05312598e-01 -4.67241108e-02 1.25380933e+00
-1.57126859e-01 -3.45058531e-01 -5.15024900e-01 3.31460908e-02
6.87399268e-01 6.68479919e-01 -7.34997571e-01 -2.19914280e-02
2.42296271e-02 4.49652910e-01 1.22869574e-01 8.36693525e-01
-1.22321105e+00 -3.32937032e-01 9.65758622e-01 6.18161917e-01
5.69497585e-01 -6.82808101e-01 -2.32780218e-01 -1.20385706e+00
-2.29903281e-01 -4.59990770e-01 2.80709624e-01 -3.92933488e-01
-1.57389629e+00 7.74201214e-01 3.02757084e-01 -9.09424365e-01
-1.59503222e-02 -4.81164455e-01 -4.53189254e-01 1.19349921e+00
-1.14529133e+00 -7.85083354e-01 1.35051370e-01 5.86757243e-01
2.84562856e-01 -3.70045193e-02 6.16882682e-01 4.79452252e-01
-5.73518097e-01 7.07200825e-01 2.77344435e-01 1.72118574e-01
5.47242641e-01 -8.68848801e-01 5.35765409e-01 8.51388633e-01
8.22291970e-02 9.29950953e-01 7.89842963e-01 -6.69596851e-01
-1.05736327e+00 -1.07000911e+00 4.68331188e-01 -5.68115175e-01
6.06807649e-01 -3.56592506e-01 -8.33902299e-01 6.75236881e-01
-7.97956251e-03 2.48934016e-01 6.72242165e-01 -1.17882214e-01
-3.97169620e-01 -2.54102558e-01 -1.24572957e+00 5.22977352e-01
1.43009233e+00 -1.89722747e-01 7.51838684e-02 2.19324961e-01
6.43708825e-01 -2.85249680e-01 -1.11408997e+00 3.56565982e-01
2.51487404e-01 -7.73123920e-01 1.23804498e+00 -3.77160043e-01
1.70807377e-01 -1.23099275e-01 -4.46082205e-01 -1.15895903e+00
-4.81523395e-01 -8.37276876e-01 -1.84346661e-01 1.10135543e+00
7.35247791e-01 -7.86088169e-01 8.71265769e-01 8.90826821e-01
1.97396218e-03 -9.78635311e-01 -1.16723466e+00 -7.47182131e-01
1.04187417e+00 -6.77686214e-01 3.54075015e-01 6.89922452e-01
-5.47024846e-01 7.44898915e-02 -9.87311825e-02 3.97293642e-02
9.51824903e-01 -2.71098763e-01 3.63152087e-01 -1.01569974e+00
-6.49184525e-01 -6.21907353e-01 8.17128569e-02 -1.44967043e+00
2.10806988e-02 -7.61164308e-01 -4.02494371e-01 -1.29709303e+00
3.63598280e-02 -1.29951119e+00 -1.94140345e-01 2.15956360e-01
-4.93834680e-03 1.13774545e-01 1.48791343e-01 3.37290794e-01
-1.27175987e-01 7.89112449e-01 1.27963817e+00 -9.44627374e-02
-2.79970706e-01 -1.01426542e-01 -8.73253942e-01 6.05060518e-01
8.71579111e-01 -5.31177700e-01 -7.34699726e-01 -7.12306917e-01
1.93522394e-01 1.07753932e-01 4.00133491e-01 -9.14403260e-01
-3.90845239e-02 -3.52846771e-01 3.05090725e-01 -2.59093732e-01
5.38342535e-01 -1.49328083e-01 2.24652842e-01 2.66355664e-01
-1.86328158e-01 2.04471320e-01 3.24958086e-01 7.16501117e-01
1.56307146e-01 5.58941625e-02 1.17853916e+00 -1.66524813e-01
-9.41373780e-02 7.84314930e-01 -2.33576879e-01 6.51776075e-01
9.53162193e-01 1.99274972e-01 -6.83042705e-01 -5.88663638e-01
-8.71712923e-01 4.11112644e-02 3.00191343e-01 6.82036625e-03
3.55650991e-01 -1.16564476e+00 -6.42025113e-01 1.47247419e-01
-3.54508847e-01 2.33886674e-01 6.06592834e-01 7.82114983e-01
-4.77498472e-01 2.67599285e-01 3.46720457e-01 -6.81470573e-01
-5.16382992e-01 3.75467509e-01 5.43846488e-01 -4.32475895e-01
-4.85362679e-01 1.40491283e+00 3.69213551e-01 -1.73092514e-01
1.97006203e-02 -3.39545578e-01 5.18717468e-01 -2.11820260e-01
2.63527870e-01 5.08924544e-01 -6.71862736e-02 -4.48767811e-01
-3.52160066e-01 6.73330188e-01 -1.33853137e-01 -5.86539745e-01
1.24645996e+00 -6.29446581e-02 3.48090261e-01 -4.63972166e-02
1.01475012e+00 7.22068101e-02 -1.86315489e+00 -2.14870080e-01
-6.89789414e-01 -3.45682532e-01 9.12306458e-03 -7.57005990e-01
-1.33883560e+00 1.02358782e+00 6.83078170e-01 1.92386960e-03
7.50107884e-01 2.09448531e-01 7.11933494e-01 -2.18587920e-01
3.97716373e-01 -9.72215235e-01 2.12906733e-01 5.63646495e-01
5.71852624e-01 -6.61183059e-01 -6.31361455e-02 -2.12746456e-01
-5.60036480e-01 4.44072515e-01 4.91229177e-01 -5.09060919e-01
1.37181890e+00 4.08449799e-01 -7.85116404e-02 -4.64164108e-01
-5.13745129e-01 1.03548974e-01 1.11777231e-01 5.76613247e-01
4.27568376e-01 1.10415578e-01 -3.59063059e-01 6.35738611e-01
-2.66202301e-01 -4.41242866e-02 1.59049198e-01 7.99548686e-01
-4.07223582e-01 -1.24437690e+00 -2.76290923e-01 6.69325769e-01
-5.22979498e-01 -4.09386247e-01 2.94631988e-01 9.03914690e-01
1.50526524e-01 8.64797473e-01 3.05674970e-01 -1.57559633e-01
1.16612520e-02 1.90258846e-01 5.37435949e-01 -3.43911529e-01
-2.27463424e-01 4.49128658e-01 -3.54821473e-01 -3.28452557e-01
-3.10038894e-01 -7.06284642e-01 -1.50609958e+00 -7.14930713e-01
2.96220221e-02 3.87363769e-02 1.63769618e-01 6.42970860e-01
7.25886643e-01 3.44637573e-01 3.85678381e-01 -6.28634572e-01
-5.92500627e-01 -9.59768534e-01 -8.01556647e-01 2.59398788e-01
1.49774835e-01 -9.00123060e-01 -5.62736630e-01 1.63631756e-02] | [11.40469741821289, -0.4564675986766815] |
1a23955a-92cb-4da5-a713-e81dab150e52 | sampling-techniques-for-streaming-cross | null | null | https://aclanthology.info/papers/N15-1158/n15-1158 | https://www.aclweb.org/anthology/N15-1158 | Sampling Techniques for Streaming Cross Document Coreference Resolution | null | ['Victor Lavrenko', 'Miles Osborne', 'Luke Shrimpton'] | 2015-05-01 | null | null | null | hlt-2015-5 | ['cross-document-coreference-resolution'] | ['natural-language-processing'] | [-2.44508207e-01 3.89024585e-01 -2.65282035e-01 -2.15905145e-01
-8.60921741e-02 -7.76765764e-01 4.48510379e-01 -7.23253429e-01
-5.48377395e-01 1.31954515e+00 3.66348401e-02 -9.49533224e-01
-2.40340635e-01 -1.05564880e+00 -8.44053447e-01 -8.75781775e-01
-7.42435038e-01 6.86515033e-01 1.44298598e-01 -6.52004302e-01
8.47113907e-01 6.29996777e-01 -1.62287033e+00 5.77558100e-01
6.81926727e-01 5.49597681e-01 5.66466339e-02 1.04480565e+00
6.13576770e-02 1.59847176e+00 -6.05450153e-01 -4.56729174e-01
1.43710867e-01 -1.88022718e-01 -5.35770595e-01 -3.36825520e-01
7.91536197e-02 -5.88986814e-01 -3.10633808e-01 6.07356608e-01
1.14270830e+00 -5.53557873e-02 1.07025254e+00 -1.51067197e+00
-7.97060013e-01 5.44234335e-01 2.30399013e-01 1.20632313e-01
6.10446513e-01 -1.79472014e-01 3.51212233e-01 -1.36621380e+00
6.32680655e-01 8.64870071e-01 9.47046995e-01 5.99273622e-01
-1.20647264e+00 -5.60552299e-01 -7.17137158e-01 -2.57835120e-01
-1.59567785e+00 -6.30042374e-01 6.23617955e-02 -3.34735900e-01
1.69142139e+00 6.72587395e-01 1.35647905e+00 1.47341585e+00
1.36649036e+00 4.72517729e-01 1.13904178e+00 -3.16340059e-01
3.65351558e-01 6.43754780e-01 -2.29148138e-02 5.51889002e-01
1.19686353e+00 6.76312149e-01 -5.60917914e-01 -7.92779744e-01
1.11365557e+00 -3.12949389e-01 2.60771513e-01 -7.06989706e-01
-1.06566191e+00 7.88466871e-01 -1.33808568e-01 2.20882341e-01
-3.46203089e-01 -9.92989019e-02 2.24548295e-01 4.12961632e-01
-2.94082850e-01 4.77862865e-01 -8.89795244e-01 -1.87686309e-01
-1.01472771e+00 1.91863477e-01 1.30615628e+00 1.45811689e+00
4.72722203e-03 3.70100170e-01 1.66493103e-01 2.89523482e-01
5.29429734e-01 1.05927479e+00 4.36951250e-01 -1.33365822e+00
-9.55119133e-02 4.20878716e-02 6.12505674e-01 -1.13473701e+00
-7.12930143e-01 -2.41300568e-01 -9.54724610e-01 5.18771529e-01
-1.59270123e-01 -4.59180593e-01 -7.59686172e-01 4.79171664e-01
-1.81984559e-01 -3.19489211e-01 5.45619786e-01 2.69257158e-01
7.28577793e-01 1.45816103e-01 9.25893057e-03 -4.73371416e-01
8.38296890e-01 -1.26899445e+00 -1.11426139e+00 2.33304992e-01
6.82289064e-01 -1.11704338e+00 4.06794518e-01 4.29345578e-01
-1.87200487e+00 -9.69799384e-02 -1.08317137e+00 1.19022481e-01
-7.97166348e-01 -2.57194549e-01 6.66624963e-01 1.49922979e+00
-1.63972652e+00 6.47176981e-01 -6.05021298e-01 5.40335439e-02
-4.73218322e-01 9.07614470e-01 -3.47142309e-01 2.90655226e-01
-1.28751957e+00 9.52449858e-01 -1.04219764e-01 3.03568155e-01
-3.30324143e-01 -1.84501112e-01 -1.03385210e+00 -6.24127567e-01
-2.59417385e-01 -1.07312346e+00 1.36302292e+00 1.09815992e-01
-1.30379498e+00 1.00962746e+00 -4.14315462e-01 -1.14109404e-01
6.61359608e-01 -1.52096570e-01 -8.18174660e-01 1.58321753e-01
2.87593510e-02 3.47163439e-01 5.30214727e-01 -9.69232559e-01
-6.31655395e-01 -1.00153096e-01 -2.54188716e-01 2.48816490e-01
1.43904146e-03 2.05205917e-01 6.38460100e-01 -9.53604504e-02
1.91062525e-01 -8.61961663e-01 -3.21419090e-01 -7.61404037e-01
-1.77353188e-01 -3.84736449e-01 5.59762537e-01 -2.92163283e-01
1.33848476e+00 -1.84365213e+00 -4.54683900e-01 2.94456817e-03
2.53527164e-01 3.89016122e-02 2.33101144e-01 1.23831403e+00
-2.43849114e-01 1.06667292e+00 3.98064762e-01 9.52541176e-03
3.13382268e-01 5.07273376e-01 2.86674555e-02 2.51206756e-01
9.43330750e-02 1.15280068e+00 -1.09214199e+00 -5.20940363e-01
5.57267785e-01 4.60968643e-01 -3.84588838e-01 6.24593735e-01
9.30012882e-01 2.99172718e-02 -8.94021094e-02 1.33241999e+00
1.05377018e+00 -2.90812284e-01 -2.46628653e-02 3.44544262e-01
-6.44541264e-01 6.00222833e-02 -7.41804183e-01 9.49429095e-01
1.31349964e-02 1.99757457e-01 1.74998537e-01 -6.80728197e-01
3.94654661e-01 1.10545933e+00 4.45659161e-01 -9.24297392e-01
-1.69569388e-01 9.00310338e-01 1.03920102e-01 -9.62983370e-01
6.40755475e-01 -1.72503099e-01 7.69825354e-02 1.40852645e-01
-3.61913115e-01 -3.84120375e-01 5.52458428e-02 4.19574350e-01
1.00224769e+00 -1.33707494e-01 5.48685491e-01 -8.72328460e-01
6.88946426e-01 -2.01589778e-01 3.00597936e-01 1.23978651e+00
-2.90862411e-01 9.21478033e-01 3.19022417e-01 -5.37574828e-01
-5.72780192e-01 -1.20098352e+00 -4.80556756e-01 9.86326516e-01
-1.58388391e-01 -3.87671381e-01 -7.10155964e-01 -2.63155133e-01
-1.85071185e-01 4.93483543e-01 -5.93591571e-01 5.23159266e-01
-5.14705896e-01 -6.85349941e-01 7.65277922e-01 3.17972392e-01
6.49624616e-02 -1.49228525e+00 -5.70813596e-01 3.34754556e-01
-1.07132711e-01 -5.39811492e-01 -1.57590687e-01 6.12205923e-01
-1.19366693e+00 -1.01526415e+00 -8.75530541e-01 -1.03603637e+00
7.18889177e-01 2.44572356e-01 1.18463767e+00 2.84042627e-01
-2.87709564e-01 6.49904788e-01 -1.31030381e-02 -2.67939389e-01
-3.16364132e-02 -5.78750819e-02 5.01899779e-01 -1.07851994e+00
6.46393359e-01 -3.92214775e-01 -1.00622165e+00 5.35368681e-01
-3.85821253e-01 -2.04611585e-01 7.85597622e-01 1.04281425e+00
1.42742008e-01 -1.33182362e-01 2.41971418e-01 -1.16735983e+00
1.11498392e+00 -4.85333532e-01 1.30341025e-02 2.62960047e-01
-8.53362501e-01 -4.53584343e-01 2.27881029e-01 -1.33037493e-02
-6.33760452e-01 -2.51423597e-01 -4.14545417e-01 3.22794706e-01
-4.06866640e-01 3.34393084e-02 2.24873364e-01 -5.26000679e-01
5.21005154e-01 1.50258854e-01 -4.07556817e-02 1.21735156e-01
6.30975887e-02 3.75316888e-01 3.67786407e-01 -7.01224983e-01
5.19809723e-01 1.93325520e-01 -3.00995037e-02 -1.28791058e+00
2.66786069e-01 -1.64763972e-01 -6.60750508e-01 -2.69126564e-01
7.75121748e-01 -9.50198710e-01 -6.35705650e-01 5.21708906e-01
-9.54387486e-01 -1.36340424e-01 -5.25335252e-01 6.30965471e-01
-7.02474535e-01 2.94757709e-02 -4.17852998e-01 -1.17976546e+00
-4.78847146e-01 -1.38869667e+00 9.59160328e-01 3.85633826e-01
-2.05811903e-01 -1.19790471e+00 3.38975668e-01 8.99848714e-02
4.29175526e-01 7.39965662e-02 5.03531873e-01 -2.80004859e-01
-2.98304528e-01 -2.85470933e-01 1.59573574e-02 -1.65876746e-01
-3.09365243e-02 4.80770379e-01 -5.48391879e-01 -4.37962860e-01
2.44738594e-01 -4.86500829e-01 -7.51581416e-02 7.11018145e-01
4.09077644e-01 -6.02920949e-01 -7.12544680e-01 5.56552589e-01
1.27709007e+00 4.72497940e-01 5.88340998e-01 6.77552521e-01
9.28869769e-02 5.17871320e-01 4.48484391e-01 2.08113909e-01
1.07506551e-01 1.08483166e-01 1.14217520e-01 -2.17137560e-01
9.45900455e-02 -1.75644696e-01 4.44304377e-01 1.09980536e+00
-8.23837519e-01 -2.15223029e-01 -4.88283783e-01 6.68056428e-01
-1.69156313e+00 -1.31570101e+00 -6.70152605e-01 9.76623476e-01
3.78074080e-01 3.38884830e-01 1.58041969e-01 1.21232189e-01
5.47049284e-01 -4.76693302e-01 -1.27305150e-01 -9.16352808e-01
-3.82233024e-01 4.21136379e-01 7.14838982e-01 6.51479006e-01
-6.10602558e-01 5.08361042e-01 1.25505590e+01 5.49717486e-01
-5.10860123e-02 1.27314463e-01 3.14815581e-01 1.99914098e-01
-2.24020526e-01 1.62632689e-01 -1.04165065e+00 9.26906094e-02
1.64754760e+00 -4.36735392e-01 3.93212438e-01 6.44444764e-01
3.94769877e-01 -1.96429595e-01 -9.78620768e-01 5.21999180e-01
1.12709433e-01 -1.50397491e+00 -3.76861334e-01 9.64888573e-01
5.80093563e-01 -1.81895152e-01 6.81941509e-01 4.33209062e-01
7.39425182e-01 -1.14519978e+00 7.88427055e-01 3.18626672e-01
8.87266219e-01 -8.18826377e-01 1.07315242e+00 2.76328743e-01
-1.04355454e+00 1.15487054e-02 -5.90443075e-01 -9.22581375e-01
1.62244976e-01 2.02188566e-01 -5.61583102e-01 4.06869203e-01
6.19966745e-01 2.15250537e-01 -8.53338003e-01 1.45581341e+00
-4.29231405e-01 4.22687940e-02 -3.00976187e-01 -4.39922899e-01
4.83290881e-01 5.28384298e-02 3.26606274e-01 1.02295005e+00
2.89785862e-01 5.45581102e-01 -2.62208551e-01 2.92286038e-01
6.63503349e-01 5.84383309e-03 -1.38496268e+00 -8.68683681e-02
3.54620427e-01 9.69217539e-01 -3.99494410e-01 -3.56748998e-02
-6.11456513e-01 6.62115753e-01 -1.66942239e-01 5.75146377e-01
-2.92761147e-01 -8.89406264e-01 4.33969021e-01 -2.30117049e-02
-2.88469017e-01 -1.35338321e-01 -3.23372036e-01 -1.02459288e+00
-6.06416404e-01 -3.90122980e-01 7.32179061e-02 -8.63867998e-01
-1.79313409e+00 6.85844719e-01 -3.50327820e-01 -1.03178740e+00
-9.29034173e-01 -9.72984850e-01 -2.90242791e-01 6.67755604e-01
-6.59234166e-01 -1.30476952e+00 1.77773401e-01 4.65253919e-01
1.82401046e-01 -5.88519633e-01 1.09945750e+00 -1.42670095e-01
-3.18711132e-01 1.08882773e+00 3.51253241e-01 -3.05974871e-01
8.32965672e-01 -1.16271269e+00 4.23785180e-01 1.09189451e-02
-6.59259140e-01 1.28460062e+00 6.81860268e-01 -6.81785107e-01
-1.16639841e+00 -1.52893141e-01 1.38110602e+00 -9.15031374e-01
5.27874053e-01 -1.02810413e-01 1.46300867e-01 6.57190859e-01
9.07351732e-01 -5.09461224e-01 1.08935249e+00 -1.13616258e-01
4.93123859e-01 8.23448122e-01 -1.38805819e+00 5.99272728e-01
1.21222687e+00 -4.81854081e-01 -8.27020168e-01 3.28937650e-01
9.94227231e-01 -5.60786903e-01 -1.34066856e+00 6.99696779e-01
1.14132881e+00 -1.04883564e+00 1.54237401e+00 -1.16912270e+00
4.08816934e-01 1.70770600e-01 -2.18455836e-01 -6.52935803e-01
-5.24122715e-01 -8.07741940e-01 -3.97186369e-01 3.40213925e-01
8.64144087e-01 -1.16800272e+00 5.53630650e-01 1.39166510e+00
-3.58625412e-01 -4.51715857e-01 -8.63314152e-01 -1.06149745e+00
-4.66870656e-03 1.81816339e-01 3.85523647e-01 8.18638563e-01
7.09820271e-01 2.02331603e-01 -2.05201268e-01 -4.37674336e-02
6.59316301e-01 -4.87512946e-01 5.82342803e-01 -8.74011815e-01
2.49149442e-01 -3.85976404e-01 -2.94489831e-01 -2.90531307e-01
-3.26795012e-01 -4.87107635e-01 -8.59293997e-01 -1.53084815e+00
8.49664882e-02 1.97367743e-01 -2.33256146e-01 3.44385266e-01
6.77823246e-01 2.45789096e-01 -2.27382466e-01 1.17054820e-01
-1.86927155e-01 -9.06788632e-02 9.18564737e-01 1.10557920e-03
-2.17724726e-01 3.92412305e-01 -1.01716614e+00 9.60534811e-01
7.85404518e-02 -6.09580100e-01 -5.51487744e-01 1.64939865e-01
5.71885824e-01 3.57775450e-01 4.49405432e-01 -9.70509052e-01
9.87445056e-01 -4.32846308e-01 6.12257063e-01 -1.43385470e+00
-2.65485227e-01 -9.00021493e-01 3.61826897e-01 6.29400074e-01
4.08042461e-01 3.53796333e-01 2.26811334e-01 -9.13986787e-02
-2.01378226e-01 -6.44658387e-01 4.43001747e-01 -7.45609999e-01
-4.55086857e-01 -3.90635371e-01 -1.02313221e+00 -9.67486948e-03
7.88939476e-01 -4.61360693e-01 -6.47767961e-01 -1.95199717e-02
-1.44608676e+00 7.30553791e-02 8.33678424e-01 1.34732248e-02
6.98022604e-01 -1.44444084e+00 -1.11748077e-01 6.18802965e-01
-5.42432606e-01 -4.98194307e-01 2.20674574e-01 1.06156552e+00
-1.31528914e+00 1.26915324e+00 -5.21896303e-01 -2.05424115e-01
-8.96585643e-01 7.19095051e-01 3.15837234e-01 1.75098896e-01
-3.93049836e-01 7.00151443e-01 1.12239346e-02 -6.58594191e-01
1.24011397e-01 2.97559589e-01 -4.87763733e-01 -2.52128989e-01
9.22277749e-01 1.17918742e+00 1.32526740e-01 -3.91715407e-01
-5.58191121e-01 4.39446121e-01 2.08604142e-01 -6.12734079e-01
1.38662541e+00 -2.20989168e-01 -7.44191706e-01 5.00318348e-01
7.50003338e-01 -7.33495504e-02 -3.28966863e-02 1.27011871e+00
4.80256975e-03 -5.47239244e-01 -3.11336160e-01 -7.82959402e-01
-3.42039734e-01 5.93214929e-01 6.17733955e-01 4.59080935e-01
9.69997942e-01 -6.47576571e-01 7.70057857e-01 7.64313757e-01
7.83133745e-01 -1.22997761e+00 -6.36951804e-01 3.69483382e-01
9.39836800e-01 -9.30032432e-01 1.30400375e-01 -3.14221263e-01
-5.55534840e-01 1.08689797e+00 5.38173854e-01 -1.62142932e-01
1.17091513e+00 6.96656525e-01 -1.75611794e-01 -4.53250527e-01
-1.04549897e+00 9.83817875e-02 2.10727617e-01 1.32608259e+00
8.26061428e-01 1.41275764e-01 -1.40426433e+00 8.46834600e-01
-6.34404242e-01 4.76996809e-01 9.47268784e-01 1.55018985e+00
-5.12773991e-01 -1.41692734e+00 -4.41162825e-01 2.31142953e-01
-2.82854587e-01 -3.49930972e-01 -8.62966061e-01 1.09929097e+00
1.97083559e-02 1.37330377e+00 -4.48758155e-01 -4.63731140e-01
8.27405751e-01 3.19600284e-01 4.41484600e-01 -1.25142381e-01
-8.94688129e-01 6.48750722e-01 3.84649158e-01 -1.05672014e+00
-7.94483483e-01 -1.27049911e+00 -1.08188879e+00 -1.16740942e+00
-3.07217389e-01 4.65777248e-01 2.31674016e-01 3.19533348e-01
-2.09303293e-02 1.56313553e-03 6.67447627e-01 -1.00111675e+00
-1.97397545e-01 -6.07395887e-01 -1.20775104e+00 -5.56946039e-01
3.94425273e-01 -5.85869551e-01 -1.15223050e+00 -6.24734610e-02] | [-1.5392029285430908, 15.869206428527832] |
1d2a38b6-b969-4b14-8cca-64f8984f6a85 | the-2nd-place-solution-for-2023-waymo-open | 2306.15914 | null | https://arxiv.org/abs/2306.15914v1 | https://arxiv.org/pdf/2306.15914v1.pdf | The 2nd Place Solution for 2023 Waymo Open Sim Agents Challenge | In this technical report, we present the 2nd place solution of 2023 Waymo Open Sim Agents Challenge (WOSAC)[4]. We propose a simple yet effective autoregressive method for simulating multi-agent behaviors, which is built upon a well-known multimodal motion forecasting framework called Motion Transformer (MTR)[5] with postprocessing algorithms applied. Our submission named MTR+++ achieves 0.4697 on the Realism Meta metric in 2023 WOSAC. Besides, a modified model based on MTR named MTR_E is proposed after the challenge, which has a better score 0.4911 and is ranked the 3rd on the leaderboard of WOSAC as of June 25, 2023. | ['Minghao Tian', 'Di Xiu', 'Cheng Qian'] | 2023-06-28 | null | null | null | null | ['motion-forecasting'] | ['computer-vision'] | [-6.81833625e-01 -2.70217031e-01 -2.13606376e-02 2.63858587e-01
-6.30261123e-01 -4.85562384e-01 1.10983431e+00 -3.73322010e-01
-6.77031577e-01 8.38149607e-01 6.16538882e-01 -2.19005942e-01
-1.46975413e-01 -5.61848998e-01 -5.59530735e-01 -5.36374271e-01
-5.33096313e-01 6.71805203e-01 3.70852679e-01 -8.64714444e-01
-2.14647591e-01 1.53508022e-01 -1.27756906e+00 1.35659859e-01
6.63648188e-01 5.98103046e-01 2.42831588e-01 1.26675737e+00
2.77727753e-01 9.93555963e-01 -5.31880856e-01 -3.83396506e-01
2.50500798e-01 -2.89066762e-01 -8.37867558e-01 -2.59615034e-01
-2.46653602e-01 -2.89120734e-01 -6.03079796e-01 6.01403296e-01
3.82942796e-01 5.66660166e-01 4.76820081e-01 -1.95076597e+00
-4.59451377e-02 7.05827534e-01 -3.85703862e-01 5.22200316e-02
6.98516667e-01 6.45449460e-01 7.20899761e-01 -6.61477685e-01
8.90964508e-01 1.19470191e+00 5.68429112e-01 6.11740291e-01
-6.68779671e-01 -1.72216535e-01 2.76744306e-01 4.86798793e-01
-1.16923845e+00 -2.89346844e-01 4.80553240e-01 -3.71096939e-01
1.17093575e+00 6.97083771e-01 7.44508922e-01 1.38031328e+00
2.85639256e-01 9.47429240e-01 9.65296268e-01 1.14647612e-01
5.38987994e-01 -4.41701442e-01 1.67602628e-01 6.75402462e-01
-3.26755680e-02 3.33605468e-01 -7.10324526e-01 -5.67976177e-01
3.51363510e-01 -5.33415258e-01 1.23832524e-01 -1.13733329e-01
-1.77533376e+00 7.11414456e-01 1.54982939e-01 -2.15957537e-02
-8.55983555e-01 5.71658671e-01 3.82921606e-01 1.35680601e-01
2.95438558e-01 -7.33864456e-02 -1.29616752e-01 -9.14863884e-01
-5.23864567e-01 8.54789555e-01 7.63228357e-01 9.34306979e-01
1.32114008e-01 3.42704475e-01 -8.87676254e-02 4.69039410e-01
8.72467458e-01 7.40841210e-01 5.71018398e-01 -1.59183586e+00
2.51467645e-01 1.72652677e-01 5.12584269e-01 -7.18659580e-01
-8.38005483e-01 -2.44332418e-01 -7.75724173e-01 4.21157837e-01
1.00158677e-01 -7.21007168e-01 -3.61862600e-01 1.53176522e+00
5.84367335e-01 6.74721897e-01 5.23153782e-01 1.18261945e+00
1.17467284e+00 9.79381979e-01 1.19348980e-01 -7.53912330e-02
1.10641110e+00 -1.61828971e+00 -7.15578675e-01 5.90137169e-02
5.87065160e-01 -7.32511044e-01 6.67116642e-01 3.83721262e-01
-1.27911186e+00 -1.86385766e-01 -1.18715310e+00 3.89027148e-01
-3.36138606e-01 -3.83971661e-01 8.00819457e-01 6.09155834e-01
-1.48334527e+00 2.93227077e-01 -1.36102509e+00 -4.31166083e-01
-5.82152426e-01 2.59540051e-01 -2.14780360e-01 2.91649073e-01
-1.21942365e+00 8.38509142e-01 8.34723562e-03 -1.58201680e-01
-1.01271522e+00 -6.06275618e-01 -7.94721901e-01 -3.35389882e-01
3.95487577e-01 -8.68194759e-01 1.38357210e+00 -2.40410149e-01
-2.32457137e+00 3.79767627e-01 -2.30122164e-01 -7.27900863e-01
9.17651355e-01 -1.44111186e-01 -8.74457181e-01 -1.22366682e-01
1.09336875e-01 6.56078398e-01 4.46071774e-01 -1.29740584e+00
-7.70171762e-01 5.23324907e-02 3.08838993e-01 3.05204511e-01
3.44906718e-01 1.75680846e-01 -4.48024005e-01 -7.38749564e-01
-2.94223219e-01 -1.33494484e+00 -6.37653768e-01 -6.57849669e-01
-9.49924588e-02 -2.34061435e-01 6.36092842e-01 -7.32666016e-01
1.08159280e+00 -1.72592986e+00 5.82756400e-01 1.45654991e-01
1.50236085e-01 7.06482157e-02 -4.81106907e-01 8.18587542e-01
3.48457128e-01 -2.14010760e-01 -9.37029272e-02 -7.82584548e-01
5.16832650e-01 4.51117866e-02 -2.57544905e-01 5.64493179e-01
-3.32250744e-01 9.68810022e-01 -1.02522254e+00 1.79499574e-02
4.20974433e-01 2.52241701e-01 -5.57339549e-01 -8.64256695e-02
-3.50093782e-01 4.51420218e-01 -2.28735819e-01 3.49423945e-01
7.66976178e-01 -1.68683887e-01 2.22900242e-01 1.12164572e-01
-4.81897056e-01 -1.60681844e-01 -1.11648560e+00 1.79454887e+00
-2.66981661e-01 4.90152121e-01 2.47033149e-01 -9.43066627e-02
4.42925304e-01 3.15926641e-01 1.08409476e+00 -4.97395456e-01
4.87982109e-02 3.61552532e-03 -8.79410189e-03 -3.27535897e-01
1.14387906e+00 3.28579038e-01 -4.91725713e-01 5.28146386e-01
-1.40218988e-01 -6.71999604e-02 5.90966403e-01 5.87634325e-01
1.34352791e+00 5.00521004e-01 9.65413600e-02 -3.36103022e-01
6.50820792e-01 3.05718899e-01 5.13357639e-01 7.22526252e-01
-5.79899609e-01 2.96361834e-01 4.88541685e-02 -5.29469728e-01
-8.52243543e-01 -9.78872001e-01 5.77371657e-01 1.12183642e+00
4.81104106e-01 -9.20696080e-01 -6.80094421e-01 -6.58839881e-01
-1.48079634e-01 9.10448790e-01 -4.45017427e-01 2.45569423e-01
-6.55632377e-01 -9.21064556e-01 7.04036176e-01 2.21008927e-01
6.35200679e-01 -1.00533998e+00 -6.75241888e-01 2.60317892e-01
-5.27861238e-01 -1.30065668e+00 -4.99454290e-01 -6.40020549e-01
-1.75409272e-01 -8.20900261e-01 -8.20489109e-01 -1.60204992e-01
-9.14621130e-02 5.91913044e-01 9.82058883e-01 -4.97712791e-02
3.79797965e-01 6.68789506e-01 -6.18344903e-01 -9.61167663e-02
-4.85111028e-01 1.47193775e-01 4.54405844e-01 1.35024875e-01
-1.50810167e-01 -2.88693905e-01 -4.88430768e-01 3.98754388e-01
-6.96315110e-01 2.87763387e-01 1.83881279e-02 3.97604913e-01
4.48104888e-02 -3.37124765e-01 4.44828242e-01 -2.23944157e-01
6.46121979e-01 -7.30816305e-01 -9.54373002e-01 3.08221742e-03
-2.11315274e-01 -8.89562964e-02 5.08649349e-01 -1.99775293e-01
-1.03884482e+00 -2.26263896e-01 -2.98510849e-01 -3.31459463e-01
2.94731352e-02 5.70458412e-01 2.56460011e-01 -8.08129832e-02
8.02395344e-02 1.93843886e-01 -1.13335093e-02 -1.26839116e-01
4.14993316e-01 3.77218634e-01 7.36251295e-01 -4.64828879e-01
9.69803631e-01 8.09257984e-01 1.72987968e-01 -9.14487541e-01
-4.61571142e-02 -3.27588886e-01 -9.83828586e-03 -7.80509710e-01
1.19348621e+00 -1.11161816e+00 -1.45758402e+00 9.15412426e-01
-1.20389152e+00 -9.50543642e-01 1.42201617e-01 8.97082150e-01
-7.04381764e-01 3.49457681e-01 -7.09498644e-01 -9.05667365e-01
-1.95525318e-01 -1.28361404e+00 9.31042731e-01 1.12057209e-01
-4.06407773e-01 -1.16340172e+00 7.09364295e-01 5.55824220e-01
7.53388464e-01 2.71185428e-01 2.97781467e-01 -4.64244932e-01
-6.56776130e-01 -8.13095570e-02 2.91799039e-01 -2.77232260e-01
-3.06791216e-01 2.00594619e-01 -4.80046421e-01 -5.02093136e-01
-4.52975631e-01 5.54441698e-02 5.30294836e-01 3.65731090e-01
2.89323419e-01 6.42307773e-02 -5.60054854e-02 4.60958481e-01
9.79048073e-01 3.68288040e-01 6.97354078e-01 7.36302733e-01
5.08385003e-01 2.43002191e-01 7.46354997e-01 8.46988857e-01
1.26891744e+00 1.24174118e+00 5.59676290e-01 2.83673793e-01
1.04555115e-02 6.84489310e-03 1.13449311e+00 1.35489750e+00
-7.22274363e-01 -5.41442275e-01 -9.83091652e-01 1.50005043e-01
-2.36298490e+00 -9.69734788e-01 -6.97580576e-01 1.97626531e+00
-3.09476294e-02 -6.69903979e-02 5.49756229e-01 -5.76994240e-01
4.02097583e-01 5.86001873e-01 -1.69550940e-01 -2.20302865e-01
-4.52274084e-01 -3.09753060e-01 5.23912609e-01 1.08599913e+00
-9.46462154e-01 1.00527716e+00 6.90783453e+00 8.88201237e-01
-6.81005716e-01 2.70897418e-01 3.42899263e-02 -1.09852314e-01
-2.03325793e-01 6.32321462e-02 -6.19643271e-01 7.60178924e-01
1.26753986e+00 -6.36424065e-01 8.67169738e-01 4.27384466e-01
7.52923369e-01 -3.05369586e-01 -5.34868836e-01 8.51145983e-01
2.01522216e-01 -1.47853434e+00 -1.37046024e-01 3.46991241e-01
1.02177334e+00 4.93399411e-01 1.88536812e-02 7.64582574e-01
1.02244604e+00 -7.60915339e-01 8.27558994e-01 9.19091463e-01
-1.45135552e-01 -8.52247655e-01 9.87484753e-01 2.13077068e-01
-1.35757101e+00 1.82526752e-01 2.87842974e-02 -1.57209769e-01
8.59999001e-01 -1.23546675e-01 -4.10019666e-01 1.11370409e+00
4.14859265e-01 7.35993862e-01 -2.23325878e-01 1.07384431e+00
5.14110141e-02 5.26789308e-01 -3.61845255e-01 -2.27602363e-01
5.31371891e-01 -3.96583498e-01 1.31610394e+00 8.00451756e-01
6.20688438e-01 8.95553501e-04 3.36834848e-01 1.76027045e-01
1.40958026e-01 -2.30579421e-01 -3.27070653e-01 2.94108629e-01
2.63299644e-01 1.34518850e+00 -3.16010356e-01 -5.18401146e-01
-1.60395116e-01 1.01781118e+00 -3.06369483e-01 5.11568248e-01
-1.36154902e+00 3.59944105e-01 8.27594757e-01 -4.81324047e-01
-1.02072340e-02 -7.23782063e-01 1.31215677e-01 -1.21923947e+00
-3.28109354e-01 -7.96212375e-01 4.07558531e-01 -1.05052710e+00
-1.12730646e+00 7.97068954e-01 3.69466431e-02 -1.52952862e+00
-7.19328046e-01 -4.46296930e-01 -7.69323528e-01 3.77395540e-01
-1.03980315e+00 -1.24826038e+00 -2.52833605e-01 4.95423377e-01
7.24603057e-01 -8.40649486e-01 9.16093349e-01 2.93133527e-01
-5.47616780e-01 2.54350841e-01 2.98830867e-01 -5.68878174e-01
2.75925189e-01 -1.11229181e+00 8.33570600e-01 7.35809505e-01
-1.60469413e-01 2.23847687e-01 1.11439610e+00 -7.37358809e-01
-2.04958487e+00 -1.06008434e+00 5.75011611e-01 -3.80449027e-01
1.16665447e+00 -1.35799855e-01 -3.07872117e-01 8.32514942e-01
7.08962619e-01 -4.27587301e-01 3.13735723e-01 -2.91584581e-01
2.43029594e-01 2.76366979e-01 -8.01994741e-01 1.07783568e+00
8.67034018e-01 -1.22522578e-01 -1.07702382e-01 3.81384820e-01
9.13250864e-01 -5.90829492e-01 -9.15944934e-01 3.99703860e-01
3.61186832e-01 -8.06930661e-01 9.06383514e-01 -6.36008084e-01
1.84181314e-02 -6.39771640e-01 -5.85543633e-01 -1.61216271e+00
-1.14841603e-01 -1.26096082e+00 -3.71528924e-01 7.59162009e-01
4.48151439e-01 -6.04107499e-01 8.06362391e-01 4.54063207e-01
-2.01317713e-01 -2.09252208e-01 -1.21587086e+00 -1.05050206e+00
4.26511560e-03 -7.14350224e-01 7.32244611e-01 7.74395406e-01
1.01500891e-01 2.27814525e-01 -8.56808126e-01 2.22397074e-01
6.79011822e-01 -2.11614788e-01 1.35824084e+00 -6.53126538e-01
-5.52232981e-01 -3.93901855e-01 -4.28249508e-01 -9.78691518e-01
3.04247856e-01 -8.04219306e-01 -2.49972597e-01 -1.51889038e+00
1.82634238e-02 1.09583087e-01 -4.24177833e-02 3.14839021e-03
1.25172168e-01 1.51020989e-01 8.54181468e-01 5.84402531e-02
-1.27793765e+00 1.03457868e+00 8.68161738e-01 -1.77947372e-01
-1.32242382e-01 2.30808556e-01 1.11047213e-03 8.56816351e-01
7.07849979e-01 1.34164676e-01 -9.48051661e-02 -2.46331573e-01
1.12075709e-01 4.68116254e-01 5.98720133e-01 -1.10581517e+00
4.07027155e-01 -4.76244986e-01 -3.57742220e-01 -1.08556259e+00
9.29524243e-01 -5.99839389e-01 8.24547708e-01 6.59582734e-01
2.71616578e-02 7.82043040e-01 2.29139954e-01 2.78188884e-01
1.99227650e-02 -1.86369773e-02 6.59515634e-02 1.28249153e-01
-1.16880679e+00 1.35434031e-01 -1.07190478e+00 -1.79912612e-01
1.15507567e+00 4.47115660e-01 -8.22715282e-01 -9.07890201e-01
-9.52147424e-01 8.09900761e-01 3.13054651e-01 6.60023034e-01
6.34436071e-01 -1.76143944e+00 -1.05385518e+00 -4.74469930e-01
-9.45448577e-02 -8.98694813e-01 6.78833604e-01 1.02744818e+00
-7.71698773e-01 3.06759745e-01 -1.58847451e-01 -3.41157615e-01
-1.22743666e+00 2.52031833e-01 1.67116389e-01 -7.04044342e-01
-5.04834354e-01 1.78485677e-01 -3.17157328e-01 -7.41493881e-01
-1.14747874e-01 1.76070645e-01 -5.35294525e-02 -2.60581225e-01
7.00073063e-01 1.03297830e+00 -8.34780708e-02 -1.07705152e+00
-6.00081623e-01 2.61525214e-01 3.17021012e-01 -8.93218815e-01
1.48207474e+00 -3.70973229e-01 -1.41630828e-01 4.50308710e-01
7.53100157e-01 1.08715788e-01 -1.22943103e+00 1.47347808e-01
4.16831337e-02 -8.45767111e-02 2.52356082e-02 -8.09469998e-01
-9.25824821e-01 8.58389065e-02 5.80003500e-01 2.31833026e-01
5.42114079e-01 -2.46202365e-01 9.43333268e-01 3.61446440e-01
8.79165232e-01 -1.11166620e+00 -1.82234615e-01 1.10027409e+00
1.08555520e+00 -1.01558542e+00 -1.28915355e-01 -1.91707715e-01
-1.13195169e+00 6.89454615e-01 5.46572864e-01 -2.45296329e-01
4.11990076e-01 4.27139074e-01 1.96825176e-01 -8.80891532e-02
-1.33988082e+00 -2.94779122e-01 3.34368087e-03 7.19609678e-01
-4.88530844e-02 5.71811378e-01 -3.31078440e-01 6.68379486e-01
-4.20046568e-01 -1.23418950e-01 8.46739531e-01 7.88096070e-01
3.41929309e-02 -1.01807690e+00 -3.70303482e-01 -2.98994005e-01
3.21807675e-02 8.50445107e-02 -2.53906101e-01 8.14771473e-01
-4.04751956e-01 1.24615061e+00 -5.96485883e-02 -8.03647757e-01
2.49111876e-01 -4.08886641e-01 4.53788526e-02 4.06543970e-01
-7.06275523e-01 1.24814861e-01 6.34792507e-01 -9.81107235e-01
-5.44624269e-01 -7.29682922e-01 -1.40142703e+00 -1.07082784e+00
4.97642815e-01 2.48463124e-01 7.44606912e-01 8.83070111e-01
4.42344576e-01 7.11883485e-01 7.08497465e-01 -1.30217481e+00
-3.10024083e-01 -8.31164956e-01 -3.98363024e-01 1.89497933e-01
1.58135936e-01 -6.07717395e-01 -6.66113272e-02 -2.81256884e-01] | [5.782158851623535, 0.8607968091964722] |
9e6fd677-3588-44e2-8b8e-a5ca4a180de9 | a-metaheuristic-driven-approach-to-fine-tune | 2101.05795 | null | https://arxiv.org/abs/2101.05795v1 | https://arxiv.org/pdf/2101.05795v1.pdf | A Metaheuristic-Driven Approach to Fine-Tune Deep Boltzmann Machines | Deep learning techniques, such as Deep Boltzmann Machines (DBMs), have received considerable attention over the past years due to the outstanding results concerning a variable range of domains. One of the main shortcomings of these techniques involves the choice of their hyperparameters, since they have a significant impact on the final results. This work addresses the issue of fine-tuning hyperparameters of Deep Boltzmann Machines using metaheuristic optimization techniques with different backgrounds, such as swarm intelligence, memory- and evolutionary-based approaches. Experiments conducted in three public datasets for binary image reconstruction showed that metaheuristic techniques can obtain reasonable results. | ['João Paulo Papa', 'Leandro Aparecido Passos'] | 2021-01-14 | null | null | null | null | ['metaheuristic-optimization'] | ['methodology'] | [-2.90805161e-01 -4.75883812e-01 -7.72843808e-02 -2.62705892e-01
-1.09619632e-01 -1.32772684e-01 5.37290037e-01 -5.39696589e-02
-1.08777654e+00 1.00110590e+00 -1.15068540e-01 1.35272346e-03
-6.54476941e-01 -1.00324690e+00 -3.20977271e-01 -1.30094409e+00
-4.63916697e-02 7.46251166e-01 3.69667858e-01 -3.37571204e-01
8.16664517e-01 7.41125405e-01 -1.59381211e+00 -1.35379151e-01
1.02878129e+00 1.15134740e+00 1.58500642e-01 3.82138669e-01
-6.27471089e-01 4.26154792e-01 -8.21384430e-01 -5.41257977e-01
7.80834705e-02 1.03807107e-01 -8.01736593e-01 -6.79319739e-01
-1.86909154e-01 -1.16367852e-02 -7.25787058e-02 1.13033569e+00
7.22932220e-01 4.23446536e-01 6.80817902e-01 -1.20632339e+00
-4.26912487e-01 4.15246338e-01 -4.67005402e-01 7.25433588e-01
-2.29922384e-01 9.61139612e-03 5.71568251e-01 -3.15303892e-01
1.26380488e-01 1.22202253e+00 7.09670007e-01 7.55134225e-01
-8.38486135e-01 -7.32670426e-01 -2.25919947e-01 9.16361094e-01
-1.27137196e+00 -1.97430804e-01 7.43848622e-01 -1.71869248e-01
1.41531372e+00 3.11758250e-01 8.66606534e-01 1.11682367e+00
4.79205012e-01 6.21482491e-01 1.24020386e+00 -6.44204676e-01
8.57209384e-01 2.48776585e-01 2.77856946e-01 5.35309494e-01
5.34143806e-01 4.18048166e-02 -2.18623221e-01 -6.02058947e-01
6.16104007e-01 -2.58153737e-01 -4.77223061e-02 -5.44251442e-01
-5.35967410e-01 1.43435442e+00 5.40263891e-01 7.89511979e-01
-6.03035927e-01 1.31758884e-01 4.85088974e-01 -2.11898804e-01
3.48403424e-01 7.86265433e-01 -5.43617785e-01 -5.09219825e-01
-6.54073358e-01 2.21902221e-01 8.07082772e-01 2.09266320e-01
4.27536994e-01 1.09660022e-01 2.67322242e-01 1.20647192e+00
2.96039313e-01 2.64457941e-01 1.23102152e+00 -5.28217971e-01
4.07098472e-01 3.67609084e-01 6.08505085e-02 -1.35688889e+00
-5.52206933e-01 -4.39254403e-01 -9.93690610e-01 1.93139136e-01
9.32437032e-02 -2.09299758e-01 -1.01975071e+00 1.46400344e+00
4.21957821e-01 1.17906459e-01 3.37919265e-01 7.59048343e-01
9.48230743e-01 8.53178740e-01 4.09279317e-01 1.55129492e-01
1.00977802e+00 -1.06504571e+00 -4.93096977e-01 -4.43075299e-01
1.05790824e-01 -3.23369294e-01 7.29997516e-01 7.48498797e-01
-1.09649885e+00 -4.52758878e-01 -1.14197993e+00 5.68496048e-01
-8.24616373e-01 -4.18461055e-01 9.02039230e-01 1.24908698e+00
-1.29939115e+00 6.65836632e-01 -9.31916416e-01 -4.98948514e-01
5.57782173e-01 7.72550702e-01 1.30937710e-01 3.48910332e-01
-1.17719066e+00 1.21595931e+00 1.02629852e+00 5.78352064e-02
-3.66523355e-01 -3.16519290e-01 -2.80719638e-01 3.34506124e-01
1.00366369e-01 -7.69106686e-01 8.09791684e-01 -9.85889375e-01
-1.87590778e+00 7.11675048e-01 4.03395385e-01 -6.88259423e-01
3.40175331e-01 -2.55011041e-02 -1.65517733e-01 -7.48071223e-02
-6.69038236e-01 5.74272513e-01 8.71308029e-01 -9.88462329e-01
-6.46255910e-01 -5.44116735e-01 -3.63346264e-02 9.09796581e-02
-8.20586801e-01 -4.38083000e-02 -3.64560783e-01 -5.81455112e-01
6.38447423e-03 -9.42591369e-01 -6.05028152e-01 -7.29134619e-01
-5.50792292e-02 -3.47588360e-01 2.68640518e-01 -3.06427330e-01
1.12229204e+00 -1.82280016e+00 5.70600867e-01 2.37744018e-01
-1.39818832e-01 5.64831913e-01 -1.99302379e-02 1.66814812e-02
1.37627453e-01 2.15310112e-01 -6.30537644e-02 9.58806798e-02
-5.87124452e-02 7.62787640e-01 -1.38687745e-01 3.76855820e-01
-3.21284950e-01 6.09060526e-01 -4.76543039e-01 -6.78271592e-01
2.88692623e-01 3.86645675e-01 -6.44688070e-01 3.37107092e-01
-6.11135624e-02 5.40024601e-02 -6.90942109e-01 3.71261925e-01
9.16152418e-01 -4.03497927e-02 5.17022535e-02 -1.01390913e-01
-7.75083676e-02 -2.37928778e-01 -9.97118294e-01 1.41699493e+00
-3.00916135e-01 4.60172087e-01 -1.25510126e-01 -1.34123302e+00
8.70727539e-01 -9.00916755e-02 4.68673319e-01 -1.00034285e+00
4.52529222e-01 6.72863498e-02 1.37802541e-01 -6.09424591e-01
7.92890668e-01 -2.40979612e-01 2.27886170e-01 2.04158545e-01
3.35809350e-01 -3.44007425e-02 4.37710434e-02 -5.18741965e-01
8.70430350e-01 -2.46266276e-01 2.96685398e-01 -3.04126948e-01
6.94386244e-01 1.55082330e-01 6.09818280e-01 9.15864170e-01
-2.91210443e-01 4.21692789e-01 1.55106008e-01 -6.53811157e-01
-9.74550009e-01 -7.49320805e-01 -2.62982309e-01 1.29636180e+00
2.77613699e-01 2.27512166e-01 -1.01130533e+00 -3.22833747e-01
-1.46053419e-01 7.81385660e-01 -8.63609195e-01 -4.40558493e-01
-6.93864167e-01 -1.69644570e+00 6.22351527e-01 4.67549384e-01
8.17854583e-01 -1.30288184e+00 -1.30143964e+00 3.49449724e-01
-1.19843595e-02 -7.46758223e-01 6.88633084e-01 7.05846608e-01
-1.48546803e+00 -6.66604757e-01 -8.80929947e-01 -4.17727083e-01
3.40480626e-01 -7.71184787e-02 1.20109308e+00 1.86629802e-01
-5.48531651e-01 1.17899500e-01 -4.84045923e-01 -5.37793517e-01
-3.56395990e-01 5.78206003e-01 -1.71727285e-01 -3.87980372e-01
4.42974210e-01 -5.88714719e-01 -4.35276091e-01 2.12778464e-01
-9.72190857e-01 -1.96659997e-01 7.47660816e-01 1.02811217e+00
3.61393303e-01 5.65298915e-01 2.56940097e-01 -6.90543473e-01
9.31155026e-01 -6.10939682e-01 -7.51907408e-01 3.87468010e-01
-8.25791419e-01 4.38819766e-01 5.17161608e-01 -6.15208685e-01
-1.04186475e+00 -7.45036721e-01 -3.66843969e-01 -2.51724660e-01
-2.69700050e-01 3.51102442e-01 -1.71777815e-01 -4.89431888e-01
6.64908707e-01 4.71697628e-01 -3.12679470e-01 -5.74874759e-01
-2.27650836e-01 4.38452899e-01 1.75911680e-01 -7.41685331e-01
3.99363577e-01 3.19429338e-01 -2.82124549e-01 -6.82354093e-01
-4.33146745e-01 -2.83918202e-01 -3.19900900e-01 -4.14631292e-02
8.32074225e-01 -8.33988488e-02 -6.06612802e-01 8.78987193e-01
-1.01226425e+00 -2.67195135e-01 3.21366712e-02 4.71375525e-01
-4.41570759e-01 1.61557510e-01 -3.34847718e-01 -8.68806779e-01
-6.57781422e-01 -1.24017751e+00 2.25576058e-01 8.21878433e-01
1.86832268e-02 -1.17140079e+00 3.48761290e-01 5.31309605e-01
7.52174139e-01 5.74019272e-03 1.54135311e+00 -7.68598676e-01
-1.35855600e-01 1.71869949e-01 -4.37579341e-02 2.95764327e-01
-3.38837504e-01 3.25683728e-02 -9.51138794e-01 -2.09508866e-01
1.58661619e-01 -2.63260275e-01 8.14436853e-01 8.02403688e-01
1.38792777e+00 -7.52125084e-02 -2.58789092e-01 8.95777643e-01
1.65370417e+00 8.39995384e-01 9.38305557e-01 1.42424619e+00
2.15404212e-01 1.75182298e-01 5.24005055e-01 7.47056842e-01
3.22917067e-02 5.65730453e-01 9.42894518e-01 2.09708303e-01
4.28925633e-01 4.37948406e-01 -1.33784905e-01 6.43831789e-01
-4.62236702e-01 -4.27907646e-01 -1.15171278e+00 3.00703108e-01
-1.94096291e+00 -9.18238938e-01 4.50568199e-01 1.87507188e+00
6.26294851e-01 2.16972113e-01 7.25833103e-02 2.58665532e-01
6.12341583e-01 3.33738327e-01 -8.12145889e-01 -9.02219176e-01
-8.75075310e-02 4.86360550e-01 7.56060362e-01 -1.47990763e-01
-1.07456088e+00 1.12749028e+00 6.62310696e+00 1.20285666e+00
-1.32728148e+00 2.41278023e-01 6.50825322e-01 -2.19413623e-01
-1.62432287e-02 -5.66356599e-01 -7.69557834e-01 6.43057644e-01
1.05005670e+00 -6.53780578e-03 6.51227176e-01 9.25504327e-01
-2.83472121e-01 -4.88122255e-01 -4.80950236e-01 1.29668450e+00
1.26338094e-01 -1.40285027e+00 1.02112226e-01 1.52658492e-01
7.46137977e-01 3.01696599e-01 4.30973351e-01 4.93944317e-01
3.52130830e-01 -1.31414974e+00 6.79188550e-01 4.43459719e-01
-1.85175374e-01 -1.32071495e+00 1.13198018e+00 2.87825018e-01
-2.34863773e-01 -4.04364437e-01 -1.11352849e+00 1.42857134e-01
-3.78737718e-01 4.55842465e-01 -5.56924939e-01 3.35967243e-01
1.29823053e+00 -1.98723659e-01 -5.30411541e-01 1.58634996e+00
3.11495692e-01 6.39881372e-01 -4.08573717e-01 -8.11492085e-01
6.12751245e-01 -3.83256465e-01 3.60600978e-01 1.07430851e+00
1.85174406e-01 7.37266168e-02 -3.98593903e-01 7.56697595e-01
2.70261139e-01 3.14864159e-01 -5.07709477e-03 -5.14248461e-02
4.18572307e-01 1.14496243e+00 -1.02493894e+00 1.54389692e-02
-3.33385989e-02 5.77366889e-01 5.02617478e-01 1.33016974e-01
-1.18048048e+00 -2.71154583e-01 7.23458886e-01 -4.25759345e-01
4.12177175e-01 -1.36615291e-01 -4.97639269e-01 -7.52483308e-01
-3.85227382e-01 -8.66651535e-01 5.91193020e-01 -5.50031543e-01
-1.08883560e+00 7.92934656e-01 1.28978029e-01 -5.92827737e-01
-9.17512402e-02 -8.56386721e-01 -4.61399496e-01 5.60478926e-01
-1.49538910e+00 -5.62125802e-01 -4.49055612e-01 5.21319151e-01
4.45797712e-01 -6.42308295e-01 8.81709695e-01 1.85093135e-01
-8.58265877e-01 5.33775628e-01 6.39348686e-01 -1.37778774e-01
1.97206467e-01 -1.06935775e+00 1.08320169e-01 2.49508351e-01
1.54556051e-01 3.68502527e-01 9.37497556e-01 1.21524474e-02
-1.49482191e+00 -4.88940060e-01 1.94396975e-03 1.74807385e-01
2.83612818e-01 1.53616935e-01 -8.94456625e-01 -4.50008996e-02
2.66592294e-01 -2.44737908e-01 8.23374808e-01 3.41704786e-01
1.58855170e-01 -6.35692030e-02 -1.46718645e+00 5.09950936e-01
6.11685157e-01 1.91007987e-01 -7.25972235e-01 -2.87610460e-02
-1.33799285e-01 -4.40974504e-01 -6.38061166e-01 5.67453265e-01
5.97393751e-01 -1.35766268e+00 1.36919093e+00 -7.33257532e-01
1.08915575e-01 5.20296752e-01 -7.79267028e-02 -1.60023046e+00
-5.22447705e-01 2.15097815e-01 -4.22779098e-02 1.03341317e+00
1.29889280e-01 -9.27595079e-01 1.17089653e+00 7.71704376e-01
3.71078759e-01 -1.03987598e+00 -1.31582749e+00 -8.35002363e-01
4.23410296e-01 -1.06610015e-01 8.73301089e-01 6.84873462e-01
-5.58969617e-01 -2.28602663e-01 -1.08989701e-01 -1.24037161e-01
3.51521313e-01 1.44239128e-01 6.55686259e-01 -1.53334808e+00
-3.86456788e-01 -8.22497487e-01 -6.63556933e-01 -4.54875439e-01
2.76559085e-01 -3.89046162e-01 -2.16463637e-02 -1.41887414e+00
4.54087079e-01 -9.10392821e-01 -7.39772439e-01 2.64575928e-01
-3.67794093e-03 8.52795020e-02 -5.55315465e-02 2.59032268e-02
-5.72987616e-01 7.67412424e-01 7.61757135e-01 -1.84400588e-01
-1.03569508e-01 2.29997523e-02 -5.24830997e-01 9.89746749e-01
1.16286850e+00 -7.89035916e-01 -2.53370076e-01 -7.23232150e-01
2.33843014e-01 -2.76990235e-01 2.64885630e-02 -1.29064333e+00
2.49398261e-01 -4.98085082e-01 4.87787634e-01 -3.54787320e-01
5.94085038e-01 -8.57631326e-01 1.13893077e-01 6.96642935e-01
-1.16053455e-01 2.30731219e-01 4.92656767e-01 4.91816670e-01
-1.75663739e-01 -9.51767385e-01 1.13872147e+00 -2.63019979e-01
-1.07962215e+00 8.24771374e-02 -4.29692835e-01 2.22251192e-01
1.10455203e+00 -7.47413814e-01 -8.48257542e-02 1.13506531e-02
-3.77066165e-01 -3.36336106e-01 1.93636417e-01 4.54658449e-01
4.42955792e-01 -9.80596244e-01 -3.27257425e-01 -2.70967670e-02
-2.67740667e-01 -2.58159906e-01 2.45555401e-01 5.32641470e-01
-7.62768567e-01 6.30413592e-01 -8.09666634e-01 -5.56573153e-01
-1.37573600e+00 3.65790904e-01 5.70210218e-01 -5.79694629e-01
-3.99582833e-01 1.07701170e+00 -1.66042134e-01 -3.46289784e-01
5.93747556e-01 -1.97533548e-01 -5.66144943e-01 1.44973248e-01
2.78210819e-01 7.35008299e-01 2.92191982e-01 -4.06068057e-01
-5.87204814e-01 5.76765597e-01 -7.11221024e-02 -1.43618165e-02
1.83632684e+00 1.81269795e-01 -1.72684088e-01 4.17217165e-02
1.06192362e+00 -4.30536419e-01 -6.87959313e-01 6.25389367e-02
1.93302765e-01 -5.81596315e-01 6.83086097e-01 -9.31335270e-01
-1.42626905e+00 8.02696466e-01 1.04689813e+00 1.54580489e-01
1.23807955e+00 -4.01054293e-01 8.61199915e-01 6.61911905e-01
7.27808177e-01 -1.38164937e+00 -8.01773742e-02 6.38069510e-01
5.45375943e-01 -1.15078223e+00 -7.05822483e-02 3.26234996e-01
-3.09234947e-01 1.40374017e+00 6.90036833e-01 -1.92805305e-01
5.58961868e-01 3.81441265e-01 -1.40112534e-01 7.70307556e-02
-5.14733791e-01 1.79674879e-01 3.47550609e-03 6.86226010e-01
1.32167503e-01 -1.28804162e-01 -3.99443924e-01 6.20333314e-01
-2.29458258e-01 -5.79539128e-03 -3.22919451e-02 9.46405411e-01
-6.68048799e-01 -1.10239685e+00 -6.95206046e-01 2.58885026e-01
-6.75256491e-01 -5.53651378e-02 -1.26189217e-01 7.45414436e-01
7.32467100e-02 7.08413959e-01 -1.60134472e-02 -2.72440463e-01
6.45004660e-02 -8.27272758e-02 9.55045283e-01 -8.70996639e-02
-9.97771502e-01 -5.23079991e-01 -1.58045769e-01 -2.60957539e-01
-4.31930989e-01 -3.90780360e-01 -1.29414034e+00 -5.64202249e-01
-4.84496295e-01 7.30808794e-01 1.06274486e+00 8.96111190e-01
1.93597704e-01 3.67720097e-01 2.61377722e-01 -1.07788110e+00
-7.71763086e-01 -8.25125992e-01 -5.40968597e-01 6.44235462e-02
-2.34297499e-01 -1.04315424e+00 -2.30258480e-01 -7.59289682e-01] | [8.237793922424316, 3.276423692703247] |
cc0f053a-42d5-44ba-bb14-ad913d08c31f | collaborative-attention-memory-network-for | 2205.08075 | null | https://arxiv.org/abs/2205.08075v2 | https://arxiv.org/pdf/2205.08075v2.pdf | Collaborative Attention Memory Network for Video Object Segmentation | Semi-supervised video object segmentation is a fundamental yet Challenging task in computer vision. Embedding matching based CFBI series networks have achieved promising results by foreground-background integration approach. Despite its superior performance, these works exhibit distinct shortcomings, especially the false predictions caused by little appearance instances in first frame, even they could easily be recognized by previous frame. Moreover, they suffer from object's occlusion and error drifts. In order to overcome the shortcomings , we propose Collaborative Attention Memory Network with an enhanced segmentation head. We introduce a object context scheme that explicitly enhances the object information, which aims at only gathering the pixels that belong to the same category as a given pixel as its context. Additionally, a segmentation head with Feature Pyramid Attention(FPA) module is adopted to perform spatial pyramid attention structure on high-level output. Furthermore, we propose an ensemble network to combine STM network with all these new refined CFBI network. Finally, we evaluated our approach on the 2021 Youtube-VOS challenge where we obtain 6th place with an overall score of 83.5\%. | ['Jinpeng Tang', 'Yuandong Zhong', 'Yuwei Zheng', 'Fei Xie', 'Junli Zha', 'Zhixing Huang'] | 2022-05-17 | null | null | null | null | ['semi-supervised-video-object-segmentation'] | ['computer-vision'] | [ 5.03225267e-01 7.61020929e-02 -8.00749734e-02 -2.66754866e-01
-6.08401656e-01 -4.36207801e-02 3.86028260e-01 -3.10752034e-01
-5.53224146e-01 6.82919979e-01 1.09780664e-02 2.16927350e-01
3.40959609e-01 -5.89026630e-01 -9.43434834e-01 -6.91363573e-01
3.38957548e-01 1.66749865e-01 1.03982055e+00 -1.13611147e-02
1.86789557e-01 9.74252671e-02 -1.58853555e+00 6.35492325e-01
9.94262218e-01 1.29450095e+00 5.79409480e-01 4.14866656e-01
-2.38611907e-01 1.08593488e+00 -6.32394016e-01 -3.32567841e-01
1.80692896e-01 -2.88940459e-01 -8.74533951e-01 3.16734284e-01
7.82823503e-01 -4.68317449e-01 -3.65615726e-01 1.25921297e+00
2.91622698e-01 8.57770219e-02 4.81543332e-01 -1.25747454e+00
-7.21115112e-01 4.36702907e-01 -7.16932297e-01 5.26297748e-01
6.94646239e-02 2.90904492e-01 8.49286437e-01 -1.04567218e+00
6.01848364e-01 1.09496331e+00 5.45914114e-01 7.48477757e-01
-9.76740539e-01 -6.44460142e-01 6.78634644e-01 6.80992067e-01
-1.18183279e+00 -3.00557613e-01 9.98751700e-01 -3.32694650e-01
8.06287944e-01 1.60295263e-01 7.35542655e-01 1.04978752e+00
-9.26811695e-02 1.38637376e+00 8.56537163e-01 -6.59522638e-02
-1.68308601e-01 1.43589556e-01 3.17624778e-01 6.15452766e-01
1.17326722e-01 -9.96663570e-02 -2.46483222e-01 3.66189390e-01
6.69463158e-01 2.78508037e-01 -4.83928561e-01 -2.22954497e-01
-1.24961019e+00 4.65632766e-01 8.22148919e-01 4.89931166e-01
-3.59612435e-01 3.89108568e-01 1.76676616e-01 -1.64657354e-01
4.48016286e-01 -1.03078783e-01 -4.65229809e-01 3.63640040e-01
-1.21288586e+00 -8.05680156e-02 3.36933374e-01 1.04882026e+00
7.25628197e-01 1.28148153e-01 -6.09113872e-01 6.75735533e-01
4.24863726e-01 1.95584819e-01 3.75430673e-01 -9.38210428e-01
5.18691361e-01 7.43038654e-01 1.16240382e-01 -1.13410687e+00
-1.12598851e-01 -6.36837244e-01 -8.88841629e-01 2.19440803e-01
2.95303911e-01 -6.48868131e-03 -1.16330612e+00 1.50682616e+00
3.82189840e-01 8.47065330e-01 -1.11431368e-01 1.17126083e+00
1.23520648e+00 7.61929929e-01 3.67234200e-01 -2.12755844e-01
1.30201960e+00 -1.61630166e+00 -7.35771418e-01 -2.06654385e-01
1.73764706e-01 -6.33616686e-01 7.47706890e-01 4.58178043e-01
-1.19546187e+00 -1.13320792e+00 -9.52549815e-01 -2.10642159e-01
-1.68860182e-01 2.60234267e-01 4.34190243e-01 4.15597767e-01
-1.17796993e+00 5.67692161e-01 -7.01613665e-01 -7.33687952e-02
8.54939461e-01 3.94076467e-01 -1.08895414e-01 -5.97094335e-02
-1.01825118e+00 5.68405867e-01 5.86672425e-01 5.65486789e-01
-1.01344025e+00 -6.00004733e-01 -6.24436080e-01 1.38385549e-01
6.61196947e-01 -7.75572479e-01 9.38106596e-01 -1.47354376e+00
-1.20549786e+00 6.37227058e-01 -2.61541843e-01 -5.39739907e-01
6.78994536e-01 -3.59737992e-01 -2.91124582e-01 1.73707321e-01
6.84479102e-02 1.03847873e+00 9.61306512e-01 -1.33308673e+00
-1.09894919e+00 -2.97319829e-01 6.81015626e-02 2.14114934e-01
-2.34020203e-01 -1.33627711e-03 -9.03963268e-01 -6.09488368e-01
2.56098777e-01 -5.38014472e-01 -3.34553957e-01 -2.35050321e-02
-2.51718342e-01 -4.74124044e-01 1.21075130e+00 -7.50290215e-01
1.27855146e+00 -2.02582908e+00 2.20451370e-01 -2.65062332e-01
4.46674526e-01 6.16518378e-01 -2.35586558e-02 -2.81520694e-01
1.89024717e-01 1.61723286e-01 -3.45727563e-01 -5.36106408e-01
-2.48425394e-01 8.84125382e-02 -1.01839587e-01 2.11884141e-01
5.70487320e-01 9.94484425e-01 -8.46463978e-01 -8.77553940e-01
3.49531233e-01 5.00275373e-01 -6.75772727e-01 1.45962328e-01
-4.31075692e-01 5.25903642e-01 -3.46841246e-01 7.04691410e-01
7.52200365e-01 -4.51047450e-01 -2.34383449e-01 -6.24798536e-01
-9.16478410e-02 -2.33606353e-01 -1.19588053e+00 1.74923253e+00
-3.18945535e-02 5.54982066e-01 1.32394657e-01 -1.12973177e+00
6.76281989e-01 2.20673442e-01 4.56477582e-01 -5.26626587e-01
9.03176591e-02 1.51995420e-01 -2.28152946e-02 -6.13547564e-01
4.56113458e-01 2.66610324e-01 5.01843572e-01 -1.06072575e-01
2.26584226e-01 2.60372013e-01 1.44277036e-01 1.65202826e-01
8.18005085e-01 5.32781422e-01 -1.94541201e-01 -1.16054617e-01
9.87695634e-01 -5.72470874e-02 9.63636220e-01 5.52303016e-01
-6.47954404e-01 9.84732091e-01 1.94716543e-01 -5.31643629e-01
-6.98111296e-01 -7.02651441e-01 -5.91480210e-02 1.04753947e+00
7.06047118e-01 -1.11804008e-01 -6.73036277e-01 -9.23277974e-01
-3.07130128e-01 2.91727036e-01 -7.40651011e-01 1.24468289e-01
-7.30545342e-01 -7.00793624e-01 2.52690941e-01 7.78429210e-01
1.02093494e+00 -1.40360808e+00 -5.47261000e-01 4.12074059e-01
-5.20695806e-01 -1.21018660e+00 -6.03773177e-01 -9.76395756e-02
-7.59777248e-01 -1.18411756e+00 -1.05212772e+00 -1.13073242e+00
6.03897810e-01 3.56290132e-01 9.95623410e-01 2.96782404e-01
-3.14986765e-01 2.13658601e-01 -2.52200246e-01 -2.04535067e-01
1.16639072e-02 -7.97638856e-03 -3.51289093e-01 4.67014134e-01
3.69528949e-01 -1.71879426e-01 -1.10284793e+00 4.30898905e-01
-7.93934405e-01 2.68846840e-01 5.67428231e-01 8.79453659e-01
7.93357909e-01 -1.70847520e-01 6.05383754e-01 -7.33854473e-01
-1.12537153e-01 -3.52527291e-01 -4.70142841e-01 4.74874526e-01
-3.24317008e-01 -4.60724831e-01 3.12670171e-01 -3.61408770e-01
-1.25527740e+00 3.05270106e-01 -2.51181751e-01 -5.90217233e-01
-3.26818824e-01 -5.96219450e-02 -2.98991293e-01 -5.59821166e-02
1.92449063e-01 4.43496823e-01 -2.13107720e-01 -5.26672304e-01
1.92616194e-01 5.77320158e-01 6.13077402e-01 -3.64895254e-01
3.95359010e-01 4.64587152e-01 -3.92821699e-01 -6.59941375e-01
-9.92901385e-01 -5.11478066e-01 -7.14465141e-01 -4.36793625e-01
1.35387731e+00 -1.10534525e+00 -6.80317223e-01 4.96700436e-01
-1.38916826e+00 -2.64168471e-01 -4.20458876e-02 2.84816742e-01
-3.90597761e-01 6.30641341e-01 -6.46329343e-01 -7.72384644e-01
-5.01980424e-01 -1.34239888e+00 1.09375715e+00 5.61946690e-01
2.31459796e-01 -6.99054599e-01 -4.26870286e-01 6.09058321e-01
4.47401017e-01 2.72980511e-01 4.59171116e-01 -5.00022531e-01
-1.09917176e+00 2.52227932e-01 -7.04167068e-01 5.87355852e-01
-1.30955903e-02 1.32684469e-01 -1.04514611e+00 -2.51890868e-01
8.34392607e-02 -1.07104957e-01 1.32332885e+00 5.08436620e-01
1.26702726e+00 -4.38735075e-02 -5.14142811e-01 6.37395740e-01
1.42403781e+00 3.31217587e-01 5.99115193e-01 1.51511863e-01
1.06144285e+00 4.97667968e-01 6.56798780e-01 6.24631643e-02
2.93410122e-01 6.75822139e-01 6.46444321e-01 -2.91425079e-01
-6.00848258e-01 1.24154007e-02 2.90834248e-01 6.16937041e-01
-2.38869876e-01 -2.84574091e-01 -6.34815335e-01 7.26651251e-01
-1.98284495e+00 -1.03745699e+00 -5.33604562e-01 1.85320020e+00
5.50852001e-01 3.97655129e-01 6.92941695e-02 -4.15550992e-02
9.87667441e-01 1.67562708e-01 -5.80913067e-01 1.90781027e-01
-2.96370536e-01 6.74853697e-02 3.20549548e-01 2.91025788e-01
-1.36631024e+00 1.22031987e+00 5.12678528e+00 9.76895750e-01
-9.37963665e-01 4.17422503e-01 8.81876230e-01 8.79813731e-02
-1.47534862e-01 -1.46290988e-01 -1.00409603e+00 7.33105540e-01
2.99524784e-01 3.33421201e-01 1.39214117e-02 7.52387941e-01
9.44720488e-03 -6.58602789e-02 -9.24106538e-01 9.22330260e-01
2.14196756e-01 -1.35452533e+00 1.54419973e-01 -3.15588146e-01
9.03627515e-01 1.24555670e-01 3.23753990e-02 4.18499231e-01
-2.29077205e-01 -8.27683806e-01 7.08578527e-01 6.86958253e-01
6.16192400e-01 -5.39136350e-01 8.64576697e-01 3.34706157e-01
-1.44196033e+00 -1.62198752e-01 -4.66163903e-01 1.58121035e-01
2.36835137e-01 3.88313234e-01 -3.41820717e-01 6.39221311e-01
9.15237904e-01 9.18647349e-01 -6.69010818e-01 1.30685365e+00
-8.51296782e-02 6.36701643e-01 -2.48594165e-01 1.66437134e-01
4.15707082e-01 -1.27640128e-01 4.24531013e-01 1.23104441e+00
6.19973876e-02 2.37008333e-01 3.37914139e-01 9.49964881e-01
-4.97135371e-02 1.29402384e-01 -2.59433419e-01 4.41467077e-01
3.44758085e-03 1.25248730e+00 -9.29784954e-01 -7.42988288e-01
-6.28654003e-01 1.32110226e+00 1.97709069e-01 4.68883783e-01
-1.09376502e+00 -2.16721267e-01 4.21642631e-01 -3.60567383e-02
6.80572033e-01 9.32326242e-02 -1.30508021e-01 -1.26955557e+00
7.87979886e-02 -5.26425838e-01 2.61784285e-01 -8.39459538e-01
-1.20586038e+00 8.10395658e-01 -3.93772513e-01 -1.28044987e+00
3.78135026e-01 -5.14005244e-01 -6.71855450e-01 5.99409163e-01
-1.76329470e+00 -1.19481528e+00 -6.00291133e-01 6.82705104e-01
1.01893985e+00 4.36851457e-02 2.59726733e-01 8.11364830e-01
-8.57998192e-01 5.31068563e-01 -2.63645738e-01 2.36301765e-01
5.87337255e-01 -1.12986171e+00 1.77937016e-01 1.02395213e+00
3.01745206e-01 2.86431640e-01 2.64483124e-01 -8.89459193e-01
-8.62642467e-01 -1.43194973e+00 7.28876233e-01 -5.09287536e-01
3.44111234e-01 -7.12361112e-02 -1.10577857e+00 6.48908019e-01
3.08614880e-01 4.07334417e-01 1.16494164e-01 -3.74923646e-01
1.50757775e-01 -1.54314414e-01 -9.51745689e-01 5.07134974e-01
1.28917146e+00 -1.34558976e-01 -6.78656340e-01 2.42923364e-01
1.01929343e+00 -4.05876577e-01 -6.46884561e-01 7.65061796e-01
2.96104491e-01 -1.07741117e+00 9.83497202e-01 -4.68728095e-01
4.82623249e-01 -7.39598572e-01 -7.56971017e-02 -5.90604782e-01
-3.46810728e-01 -3.72924119e-01 -1.71019688e-01 1.42477286e+00
1.82619154e-01 -2.62665957e-01 1.03657329e+00 5.34404397e-01
-2.28868097e-01 -7.97967732e-01 -8.82205486e-01 -5.57184041e-01
-2.51744300e-01 -4.29559201e-01 3.78470540e-01 6.42487526e-01
-5.24864912e-01 3.31408232e-01 -5.84219813e-01 1.53110415e-01
6.66547775e-01 1.92468822e-01 6.75329268e-01 -1.10441422e+00
-2.72694141e-01 -4.22000855e-01 -4.61544693e-01 -1.49754238e+00
-2.89261211e-02 -7.82141507e-01 1.03073888e-01 -1.53910232e+00
3.59866500e-01 -3.69532168e-01 -7.95272827e-01 2.62340367e-01
-5.90983033e-01 6.47909462e-01 4.95648384e-01 1.61529019e-01
-1.16952121e+00 5.70013702e-01 1.39346075e+00 -3.87026370e-01
-1.58743486e-01 1.30227298e-01 -3.85272413e-01 9.43154156e-01
5.12765110e-01 -3.67085785e-01 -2.74451464e-01 -6.19834542e-01
-2.95560062e-01 1.03209438e-02 5.66999316e-01 -1.19045532e+00
4.54994321e-01 1.27056435e-01 6.25304401e-01 -9.83299315e-01
3.36385190e-01 -1.04927468e+00 1.82509765e-01 5.47412992e-01
-1.90266773e-01 -2.05816701e-01 7.63786882e-02 9.52682793e-01
-4.75873470e-01 -2.17829645e-01 6.53343618e-01 -1.78979740e-01
-1.19322348e+00 5.68452299e-01 -1.34847492e-01 -1.71407834e-02
1.20220840e+00 -4.69011992e-01 -1.65765971e-01 1.01951703e-01
-9.52504694e-01 4.77753162e-01 1.63710013e-01 3.90465468e-01
7.60110259e-01 -1.17659652e+00 -7.18456447e-01 1.64960250e-01
2.41514780e-02 2.97667772e-01 5.99209428e-01 1.01429296e+00
-4.45225507e-01 8.49861503e-02 -1.61752969e-01 -8.83563280e-01
-1.23897934e+00 6.54398620e-01 3.13085139e-01 -3.92458066e-02
-7.08345056e-01 1.19170952e+00 5.81947684e-01 1.27468035e-01
4.87566918e-01 -5.52414536e-01 -5.45465946e-01 1.23965867e-01
5.67496896e-01 2.28320301e-01 -1.88334003e-01 -8.23445499e-01
-3.80779892e-01 8.25990438e-01 -1.64528176e-01 2.99366862e-01
1.09053552e+00 -3.28982085e-01 4.20577005e-02 2.24898070e-01
1.15930963e+00 -3.70868891e-01 -1.65025711e+00 -4.57812220e-01
2.73511671e-02 -4.37097311e-01 -1.81609555e-03 -6.90476716e-01
-1.43556678e+00 1.02428901e+00 8.53826523e-01 -5.74118271e-02
1.16841984e+00 -1.70196742e-01 1.06738091e+00 1.14303641e-01
2.66987085e-01 -1.05359459e+00 8.91474858e-02 2.42536142e-01
7.11860299e-01 -1.72539139e+00 -1.15076400e-01 -5.32607377e-01
-7.21441150e-01 1.00691271e+00 1.10201144e+00 -3.19198012e-01
5.09031296e-01 -1.34206891e-01 -4.64421995e-02 3.40505093e-02
-6.37099981e-01 -6.01997733e-01 5.04222810e-01 5.17581999e-01
2.25551575e-01 -2.81621099e-01 -4.51066792e-01 8.06578636e-01
4.75677758e-01 2.09333539e-01 2.76921451e-01 4.88755614e-01
-5.59524953e-01 -7.25734472e-01 -1.38165012e-01 3.20156097e-01
-7.32989550e-01 -1.43932462e-01 -8.23853537e-02 6.32887721e-01
5.75314641e-01 8.37574542e-01 5.44778854e-02 -3.52818131e-01
1.74162731e-01 1.25343785e-01 2.80593038e-01 -5.43565333e-01
-8.25916231e-01 3.14246744e-01 -2.85175979e-01 -7.61536479e-01
-7.96727657e-01 -5.08426785e-01 -1.21320200e+00 2.50055879e-01
-4.30238783e-01 -1.06356926e-01 3.23172659e-01 9.01530564e-01
3.02214503e-01 8.36759090e-01 3.63052785e-01 -1.03963244e+00
-2.44955588e-02 -9.06414211e-01 -2.73062021e-01 5.14487147e-01
4.40710306e-01 -6.13217831e-01 -1.99936897e-01 3.49217534e-01] | [9.427870750427246, -0.1524517834186554] |
095cbfcb-e9f5-4cec-a919-617f6000a70f | cross-lingual-transfer-learning-for-check | 2211.05087 | null | https://arxiv.org/abs/2211.05087v1 | https://arxiv.org/pdf/2211.05087v1.pdf | Cross-lingual Transfer Learning for Check-worthy Claim Identification over Twitter | Misinformation spread over social media has become an undeniable infodemic. However, not all spreading claims are made equal. If propagated, some claims can be destructive, not only on the individual level, but to organizations and even countries. Detecting claims that should be prioritized for fact-checking is considered the first step to fight against spread of fake news. With training data limited to a handful of languages, developing supervised models to tackle the problem over lower-resource languages is currently infeasible. Therefore, our work aims to investigate whether we can use existing datasets to train models for predicting worthiness of verification of claims in tweets in other languages. We present a systematic comparative study of six approaches for cross-lingual check-worthiness estimation across pairs of five diverse languages with the help of Multilingual BERT (mBERT) model. We run our experiments using a state-of-the-art multilingual Twitter dataset. Our results show that for some language pairs, zero-shot cross-lingual transfer is possible and can perform as good as monolingual models that are trained on the target language. We also show that in some languages, this approach outperforms (or at least is comparable to) state-of-the-art models. | ['Tamer Elsayed', 'Maram Hasanain'] | 2022-11-09 | null | null | null | null | ['zero-shot-cross-lingual-transfer', 'cross-lingual-transfer'] | ['natural-language-processing', 'natural-language-processing'] | [-2.31807724e-01 8.84092003e-02 -5.73906481e-01 4.93448693e-04
-1.46591926e+00 -7.13647604e-01 1.05396092e+00 7.01125979e-01
-6.16993546e-01 1.00003731e+00 2.14402795e-01 -4.73060906e-01
3.64138782e-01 -7.96116054e-01 -8.80176067e-01 -9.35712755e-02
3.39656204e-01 9.16347325e-01 5.93890250e-01 -7.51702249e-01
4.16701317e-01 2.86864161e-01 -1.01966560e+00 6.64227664e-01
7.83688188e-01 5.83189130e-01 -3.84696096e-01 2.29261920e-01
-4.11013424e-01 1.34806931e+00 -7.07998931e-01 -1.23615301e+00
8.02396089e-02 -2.35483602e-01 -9.17975307e-01 -1.73627675e-01
6.31706417e-01 1.15863113e-02 1.03071779e-01 1.38511074e+00
3.03362608e-01 -7.84094512e-01 5.22344887e-01 -1.09281754e+00
-7.70771265e-01 1.02248859e+00 -6.67134404e-01 2.94468403e-01
3.14509600e-01 -2.32301697e-01 8.19213152e-01 -7.42415607e-01
1.13374591e+00 1.16857719e+00 9.59063768e-01 9.00605926e-04
-1.04456103e+00 -7.94518352e-01 -3.95899832e-01 3.89344580e-02
-1.25402522e+00 -3.15269917e-01 5.62122166e-01 -7.18082249e-01
6.43828630e-01 1.78180590e-01 1.99260727e-01 1.15326023e+00
2.53008425e-01 6.03867173e-01 1.78484428e+00 -4.25049305e-01
-4.10950005e-01 7.18095958e-01 1.09810300e-01 9.04627264e-01
5.06411672e-01 -3.98973703e-01 -6.12263203e-01 -5.29536366e-01
-8.02153870e-02 -4.44990963e-01 -2.95622293e-02 1.64679617e-01
-1.31402445e+00 1.24697673e+00 3.03315133e-01 8.12316418e-01
-3.47080290e-01 -2.73858994e-01 7.14836299e-01 6.52376175e-01
1.33333158e+00 4.26838547e-01 -3.92395169e-01 -5.81084378e-02
-1.23270488e+00 2.43381053e-01 1.11842966e+00 4.24988449e-01
6.62474036e-01 -3.90108854e-01 -1.27218977e-01 8.72117937e-01
-1.94746926e-01 6.77659452e-01 2.36331686e-01 -2.47179568e-01
1.05181026e+00 6.35963023e-01 2.73751169e-01 -1.62039602e+00
-3.07501942e-01 -5.36593974e-01 -6.61346436e-01 -1.60864040e-01
6.08911812e-01 -1.93388417e-01 -2.45389253e-01 1.20130086e+00
2.94580221e-01 -1.61595166e-01 -1.27497673e-01 7.04703569e-01
3.86787981e-01 6.48850858e-01 -1.75591305e-01 -3.11584771e-01
1.37043989e+00 -7.06452966e-01 -7.09269285e-01 -2.49688178e-01
8.94859672e-01 -1.26635003e+00 7.53008902e-01 2.37588376e-01
-8.55590820e-01 5.80903748e-03 -8.53298426e-01 -4.45962399e-02
-7.28817523e-01 -5.31003736e-02 3.20519298e-01 8.67125094e-01
-5.77962995e-01 5.21173358e-01 -4.06400651e-01 -3.36260766e-01
4.93335783e-01 -2.23928288e-01 -4.18439686e-01 -1.38666883e-01
-1.59498513e+00 1.31999695e+00 1.90378085e-01 -3.39375079e-01
-5.80895901e-01 -4.89305794e-01 -3.73434246e-01 -4.87844586e-01
3.87855917e-01 1.10653512e-01 8.69184673e-01 -9.22283590e-01
-8.25458288e-01 1.36600828e+00 1.81834474e-01 -6.71720266e-01
9.33357716e-01 -5.92464618e-02 -7.86290824e-01 -6.41570538e-02
7.63313234e-01 1.06491424e-01 7.13630974e-01 -1.10599422e+00
-5.66682816e-01 -2.46835172e-01 -2.71511339e-02 -3.46328348e-01
-4.71981198e-01 8.48265290e-01 -1.44277006e-01 -6.58659339e-01
-3.31509739e-01 -1.00142169e+00 2.45126337e-01 -2.96256840e-01
-6.47935033e-01 -7.73585811e-02 7.81626821e-01 -1.30788124e+00
1.27682412e+00 -1.76815200e+00 -3.13627571e-01 1.97685976e-02
4.08256911e-02 3.35272282e-01 2.47542217e-01 7.34173536e-01
3.46179783e-01 4.58102375e-01 -1.79380327e-01 -2.32246608e-01
-2.01757446e-01 -1.94627479e-01 -3.55092525e-01 9.59296405e-01
2.40854681e-01 6.92758679e-01 -8.70149195e-01 -6.58957601e-01
-4.35611516e-01 3.59046578e-01 -2.89087802e-01 -4.30909216e-01
-6.14382215e-02 3.02420855e-01 -2.65813917e-01 5.76068044e-01
6.43399358e-01 -2.74514616e-01 2.96329796e-01 -8.52417108e-03
-3.64921451e-01 7.51096845e-01 -6.18962049e-01 1.03037632e+00
-5.84230185e-01 9.47751880e-01 -2.73527857e-02 -6.54603243e-01
8.86358798e-01 1.69886813e-01 2.73941547e-01 -7.89399266e-01
1.84640944e-01 6.88928068e-01 -3.38535793e-02 -3.27006102e-01
5.22428036e-01 -3.88328135e-01 -4.31804508e-01 8.28072965e-01
-3.44007671e-01 -5.63127082e-03 2.53866047e-01 3.36183161e-01
8.29865038e-01 -4.06327069e-01 1.54159665e-01 -3.09813738e-01
7.14198530e-01 5.35933912e-01 6.64284602e-02 6.34650290e-01
-7.30021596e-02 1.55727014e-01 6.85056567e-01 -4.00941849e-01
-1.15421748e+00 -4.08812016e-01 -1.62286386e-01 9.56133187e-01
-1.61913007e-01 -1.11300744e-01 -7.51604676e-01 -9.41288888e-01
7.97355697e-02 8.67327988e-01 -5.02273738e-01 3.69319022e-01
-4.35283303e-01 -9.26396012e-01 1.04018450e+00 -3.53636146e-01
7.10424483e-01 -6.39652133e-01 -2.15086117e-01 2.50589758e-01
-6.57661319e-01 -1.36586595e+00 -1.86720565e-01 -3.42043847e-01
-3.12089771e-01 -1.14890265e+00 -6.92466438e-01 -4.66479212e-01
3.23265731e-01 1.66536197e-01 1.00836062e+00 1.83620408e-01
1.63466960e-01 -2.19452083e-01 -2.64310122e-01 -5.06018102e-01
-1.16203523e+00 3.24767947e-01 -6.58102185e-02 2.56554723e-01
2.73024112e-01 9.22840461e-02 1.08347982e-01 2.43960783e-01
-6.90829575e-01 -1.78972363e-01 3.95909429e-01 4.83810216e-01
1.00451417e-01 9.14679933e-03 7.30743408e-01 -1.49122953e+00
9.47681427e-01 -7.82032609e-01 -5.72455347e-01 3.70416373e-01
-4.45554882e-01 -1.60403661e-02 4.20620412e-01 -3.22472990e-01
-8.73318493e-01 -5.93091428e-01 -1.67758048e-01 6.75993040e-02
2.08506852e-01 8.90575349e-01 4.86358494e-01 -1.29042268e-01
7.82574058e-01 -1.20192982e-01 8.88327807e-02 -4.05949652e-01
2.78891027e-01 1.23026347e+00 -1.45574035e-02 -2.20013976e-01
8.04271221e-01 6.32810354e-01 -3.39339405e-01 -7.89769530e-01
-1.26491082e+00 -5.38998008e-01 -5.54157734e-01 -3.56231749e-01
5.78103483e-01 -9.14697587e-01 -2.96295166e-01 4.99205977e-01
-1.44957471e+00 -3.04725412e-02 3.14563572e-01 2.39369780e-01
-2.06433404e-02 3.91592443e-01 -9.95099962e-01 -6.80171847e-01
-1.69956908e-01 -9.96840775e-01 8.34924161e-01 -7.70506978e-01
-2.56854504e-01 -1.16002131e+00 3.24542969e-01 8.62838149e-01
5.39828062e-01 3.25735331e-01 6.17661059e-01 -1.13129735e+00
-1.73099697e-01 -7.16617584e-01 -4.61518526e-01 3.04610163e-01
1.10968411e-01 -1.51731938e-01 -7.94706464e-01 -1.11053303e-01
1.02838345e-01 -6.70050323e-01 8.09227586e-01 -3.10187668e-01
3.52284312e-01 -8.13898683e-01 -2.30452910e-01 -4.07533973e-01
1.32005918e+00 -6.43513560e-01 5.92501163e-01 7.12955773e-01
4.85558659e-01 8.54802012e-01 6.70325100e-01 1.72881514e-01
5.69385707e-01 7.95628965e-01 1.23745836e-01 5.45843989e-02
-1.34590179e-01 -3.71980637e-01 5.53191841e-01 1.13653708e+00
-3.72954421e-02 -1.44827932e-01 -1.18960822e+00 7.31221437e-01
-1.33931875e+00 -1.33472931e+00 -7.70638108e-01 2.11544061e+00
1.09863663e+00 4.13804054e-01 3.94711077e-01 4.41262149e-04
9.05971348e-01 1.97027653e-01 2.92901933e-01 -5.39191544e-01
-4.05556679e-01 -2.90529519e-01 9.96834159e-01 7.56639898e-01
-1.22120106e+00 1.03375292e+00 5.60090542e+00 1.11324978e+00
-1.35302722e+00 1.01020455e+00 6.83523417e-01 3.35052550e-01
-2.44264841e-01 -3.24170180e-02 -8.47854912e-01 6.09721243e-01
1.00782967e+00 1.25866488e-01 2.05348104e-01 5.23697078e-01
7.99368396e-02 2.40550682e-04 -3.18513989e-01 5.20214915e-01
4.94777083e-01 -1.66539538e+00 -2.48482689e-01 1.67637765e-01
9.53520656e-01 7.31341243e-01 -5.52816652e-02 2.57216394e-01
3.16148967e-01 -6.83992386e-01 1.14709878e+00 2.85392463e-01
6.56996429e-01 -7.11041987e-01 1.03971601e+00 9.33010638e-01
-3.20776761e-01 2.36050129e-01 -3.07701439e-01 1.40195981e-01
3.59341234e-01 1.00602615e+00 -9.08526659e-01 5.53646564e-01
4.05267239e-01 6.57834649e-01 -7.31488764e-01 6.23300314e-01
-3.02105635e-01 7.00186908e-01 -1.82179019e-01 -2.43473038e-01
5.88989913e-01 1.34555519e-01 4.33809966e-01 1.51169610e+00
4.20744985e-01 -4.16996479e-01 1.91283256e-01 6.50062263e-01
-4.17171866e-01 5.92010856e-01 -9.07623291e-01 -3.64935726e-01
1.50216877e-01 1.07536995e+00 -6.99957252e-01 -4.87043887e-01
-5.66279292e-01 9.20063615e-01 5.85546553e-01 -2.56449819e-01
-9.53228831e-01 2.43162345e-02 -4.05405574e-02 6.09981418e-01
2.42904909e-02 -1.15813740e-01 1.78747496e-03 -1.25226605e+00
-1.65679589e-01 -1.20965195e+00 3.70108932e-01 -5.06329417e-01
-1.61981881e+00 7.57108510e-01 -2.29382053e-01 -1.24638903e+00
-8.84108990e-02 -4.34843957e-01 -1.18938453e-01 7.45173216e-01
-1.63906789e+00 -1.44164026e+00 3.85071933e-01 4.24725354e-01
2.74389684e-01 -2.48891205e-01 4.72824037e-01 6.31415308e-01
-1.93283379e-01 2.87413299e-01 4.58833575e-02 3.69500548e-01
9.43600118e-01 -5.51143825e-01 2.23436743e-01 7.12447226e-01
4.92466450e-01 3.89991671e-01 8.41275573e-01 -1.09817064e+00
-9.03502822e-01 -1.07030225e+00 1.89621329e+00 -8.02213192e-01
1.60213387e+00 -3.97573471e-01 -9.23763752e-01 5.11977792e-01
5.08865893e-01 -4.00737494e-01 5.40337980e-01 2.79444188e-01
-7.69290745e-01 2.00971857e-01 -1.21606994e+00 2.36926451e-01
3.90727609e-01 -8.04241061e-01 -4.85090703e-01 1.04628670e+00
3.33083183e-01 -1.10372014e-01 -6.75792933e-01 -2.56662965e-01
3.71270210e-01 -1.06079137e+00 5.28839529e-01 -6.80484831e-01
6.85847461e-01 -7.83225745e-02 -3.67140695e-02 -1.28935218e+00
3.09992936e-02 -3.50712746e-01 5.77995360e-01 1.49942040e+00
1.02052009e+00 -8.28633368e-01 2.80676454e-01 -1.20850401e-02
2.86337107e-01 -1.52907208e-01 -1.06209910e+00 -1.04819191e+00
4.75713313e-01 -7.30881870e-01 3.05110097e-01 1.54563987e+00
1.06186546e-01 4.77717817e-01 -5.83083570e-01 1.98367178e-01
5.37459552e-01 1.25112191e-01 6.08954370e-01 -1.14060450e+00
-8.90358724e-03 -3.16584051e-01 -2.84054190e-01 -3.33699524e-01
4.19435591e-01 -1.15776551e+00 -2.65666664e-01 -1.29537475e+00
4.14851785e-01 -6.04959130e-01 2.18863204e-01 4.36296701e-01
2.25858003e-01 6.62570357e-01 1.80918470e-01 5.18825531e-01
-3.56233746e-01 -1.61311999e-01 1.06779563e+00 -4.59181726e-01
3.16542804e-01 -7.93779793e-04 -5.27199090e-01 7.94056892e-01
8.74815464e-01 -1.06209064e+00 3.14753771e-01 -3.36138874e-01
9.73466694e-01 -8.42991546e-02 4.90868717e-01 -7.51889169e-01
1.78098530e-01 -2.82189175e-02 -4.59363550e-01 -5.44044316e-01
3.86669748e-02 -6.37483358e-01 1.35357371e-02 6.75322533e-01
-3.39955032e-01 5.00441203e-03 -6.41586110e-02 4.66516852e-01
-4.13524449e-01 -3.12231302e-01 6.73126876e-01 -1.83783218e-01
-9.13294312e-03 -1.42618716e-01 -5.39994657e-01 3.71184647e-01
6.93015695e-01 4.15506959e-01 -9.43421006e-01 -5.36329031e-01
-1.35954604e-01 -3.31037492e-01 5.20352125e-01 3.19686383e-01
2.55358778e-02 -1.15871263e+00 -1.33474946e+00 -3.09447199e-01
3.14487368e-01 -1.17667079e+00 -1.62766472e-01 1.36090696e+00
-7.47327089e-01 7.06625164e-01 9.89584625e-03 -1.53344288e-01
-1.05334449e+00 7.46375024e-01 5.66140190e-03 -6.77845120e-01
-5.32368608e-02 6.17860138e-01 -6.27798319e-01 -7.51281381e-01
-5.33899128e-01 4.93322201e-02 -1.78766906e-01 5.52179456e-01
4.97680306e-01 4.20102984e-01 2.87180603e-01 -1.33231938e+00
-3.86954129e-01 1.80568039e-01 -1.01635233e-01 -2.86646485e-01
1.36317968e+00 -3.55174989e-02 -6.64871395e-01 7.03505158e-01
1.21535361e+00 9.28106308e-01 -2.52270788e-01 -1.13423258e-01
3.27570796e-01 -5.22217274e-01 1.13957651e-01 -9.16684747e-01
-9.16825473e-01 7.53060818e-01 3.89968231e-02 7.64245808e-01
3.42336595e-01 3.37170511e-01 6.90864503e-01 1.13701604e-01
7.77016044e-01 -1.12407267e+00 -9.38306525e-02 5.11971474e-01
9.74378169e-01 -1.43662310e+00 6.68742359e-02 -6.92081273e-01
-9.76174951e-01 7.61068940e-01 -2.05363646e-01 -1.96180269e-02
7.01199710e-01 8.40175897e-02 1.14823140e-01 -5.63031435e-01
-3.77629608e-01 -9.98456031e-02 3.23655933e-01 2.23982692e-01
7.56114125e-01 1.33475780e-01 -7.06176639e-01 1.43512353e-01
-1.15304820e-01 -1.91658616e-01 8.57004642e-01 6.44347429e-01
-4.76515323e-01 -1.15130520e+00 -5.63491762e-01 4.64982092e-01
-1.18757594e+00 -3.52946937e-01 -7.63310194e-01 1.01715040e+00
3.44844759e-01 1.03618956e+00 -3.00576150e-01 -1.89619392e-01
5.00720367e-02 -2.01231451e-03 2.19819337e-01 -4.86145020e-01
-8.51672709e-01 -1.11949287e-01 8.83353114e-01 -1.97835326e-01
-8.21014583e-01 -8.61014903e-01 -6.45441234e-01 -7.61124730e-01
-5.30978680e-01 1.76146179e-01 8.65577936e-01 1.10730433e+00
2.31598258e-01 -1.55522972e-01 4.30493802e-01 -2.16160476e-01
-4.88035679e-01 -1.07694995e+00 -4.28678721e-01 3.94344956e-01
1.20545745e-01 -4.97819573e-01 -4.04615283e-01 -5.41374795e-02] | [8.204413414001465, 10.260649681091309] |
b6c496ed-a218-4b43-98af-c60f32136e16 | direct-velocity-inversion-of-ground | null | null | https://doi.org/10.1029/2020JB021047 | https://www.researchgate.net/profile/Zi_Xian_Leong/publication/351760571_Direct_Velocity_Inversion_of_Ground_Penetrating_Radar_Data_Using_GPRNet/links/60b92d38299bf10dff91746c/Direct-Velocity-Inversion-of-Ground-Penetrating-Radar-Data-Using-GPRNet.pdf | Direct Velocity Inversion of Ground Penetrating Radar Data Using GPRNet | Ground penetrating radar (GPR) is used to image the shallow subsurface as evident in earth and planetary exploration. Electromagnetic (EM) velocity (permittivity) models are inverted from GPR data for accurate migration. While conventional velocity analysis methods are designed for multioffset GPR data, to our knowledge, the velocity analysis for zero‐offset GPR has been underexplored. Inspired by recent deep learning seismic impedance inversion, we propose a deep learning guided technique, GPRNet, that is based on convolutional neural networks to directly learn the intrinsic relationship between GPR data and EM velocity. GPRNet takes in GPR data and outputs the corresponding EM velocity. We simulate numerous GPR data from a range of pseudo‐random velocity models and feed the datasets into GPRNet for training. Each training data set comprises of a pair of one‐dimensional GPR data and EM velocity. During training phase, the neural network's weights are updated iteratively until convergence. This process is analogous to full‐waveform inversion in which the best model is found by iterative optimization until simulated data matches observed data. We test GPRNet on synthetic testing datasets and the predicted velocity models are accurate. A case study is presented where this method is applied on a GPR data collected at the former Wurtsmith Air Force Base in Michigan. The inversion results agree with velocity models established by previous GPR inversion studies of the similar area. We expect the GPRNet open‐source software to be useful in imaging the subsurface for earth and planetary exploration. | ['Tieyuan Zhu', 'Zi Xian Leong'] | 2021-05-20 | null | null | null | journal-of-geophysical-research-solid-earth | ['gpr', 'geophysics', 'seismic-imaging', 'gpr', 'seismic-inversion'] | ['computer-vision', 'miscellaneous', 'miscellaneous', 'miscellaneous', 'miscellaneous'] | [ 1.71926871e-01 -1.80473328e-01 2.13130057e-01 -1.58579588e-01
-8.06291521e-01 4.90716249e-02 2.88002223e-01 -3.42488140e-01
-5.97837090e-01 6.41026497e-01 9.73203182e-02 -1.07369101e+00
-3.82635415e-01 -1.28019834e+00 -6.84166014e-01 -7.09754765e-01
-5.88984787e-01 6.51895106e-01 -7.35490471e-02 -5.14913023e-01
5.35119534e-01 6.13857031e-01 -1.19491971e+00 6.84565827e-02
7.75815129e-01 1.07258856e+00 1.47614643e-01 8.37476671e-01
1.07139543e-01 5.17696142e-01 -5.68072274e-02 3.32188934e-01
6.74505457e-02 -1.48621812e-01 -9.92980421e-01 -7.95270562e-01
-3.41542400e-02 -4.46441948e-01 -4.10014153e-01 6.69611335e-01
7.07690775e-01 3.47327620e-01 9.29981232e-01 -6.22949004e-01
-3.03327858e-01 4.76632863e-01 -9.34237540e-01 6.83913171e-01
3.89887184e-01 -2.94231087e-01 6.67883694e-01 -1.01249337e+00
2.80670464e-01 7.57699609e-01 1.48580086e+00 1.63291857e-01
-9.34901953e-01 -7.54961133e-01 -4.66057390e-01 1.89718679e-01
-1.12268043e+00 -2.28463411e-01 7.32339859e-01 -5.25751173e-01
1.23637581e+00 1.01685040e-01 8.75973225e-01 1.24347281e+00
4.59293902e-01 4.42997068e-01 4.66906965e-01 -3.53850633e-01
2.21420243e-01 -3.60548258e-01 1.15395464e-01 4.38465983e-01
6.71755195e-01 6.80988312e-01 -3.63476247e-01 -2.51944274e-01
8.41346681e-01 -4.74959284e-01 -5.31070948e-01 2.05205120e-02
-7.42942452e-01 1.11396611e+00 4.09956366e-01 2.01809078e-01
-7.80118048e-01 4.18747753e-01 3.18765640e-01 2.56162584e-01
4.91804898e-01 5.10338485e-01 -5.65463960e-01 -2.91062921e-01
-1.31883061e+00 5.80991983e-01 1.01008570e+00 7.47855902e-02
8.53504002e-01 8.18174899e-01 5.99473298e-01 6.40836835e-01
8.37221861e-01 9.70178723e-01 7.20789909e-01 -6.36174023e-01
3.50816369e-01 -1.64493978e-01 1.93262219e-01 -1.07603741e+00
-7.23966300e-01 -3.80801201e-01 -7.06714511e-01 1.20335557e-01
1.78593129e-01 -6.36324286e-01 -1.00059509e+00 1.05302620e+00
7.98058603e-03 8.65010142e-01 6.40283167e-01 7.02431202e-01
1.22804141e+00 7.16290712e-01 7.29901195e-02 3.71879667e-01
1.04083025e+00 -1.79515287e-01 -2.70313591e-01 -6.82642519e-01
1.05655456e+00 -3.51153016e-01 3.64136100e-01 1.63987964e-01
-7.88320124e-01 -2.33595118e-01 -1.40684330e+00 6.94185078e-01
-2.40563065e-01 -2.39154607e-01 1.06903493e+00 8.41571748e-01
-8.27516794e-01 9.00193632e-01 -1.25027430e+00 -1.11466452e-01
3.29552531e-01 1.33960724e-01 -1.57504678e-01 7.62274712e-02
-1.85903263e+00 8.93890262e-01 2.37388715e-01 8.29361320e-01
-7.73796380e-01 -1.11089444e+00 -1.34638405e+00 -1.92177966e-01
-7.75561452e-01 -5.34595549e-01 1.34171033e+00 -2.19226032e-01
-1.37644362e+00 4.13320273e-01 2.75640994e-01 -5.86120963e-01
4.10559416e-01 -8.42564255e-02 -7.05516815e-01 1.21588029e-01
1.98034570e-01 1.75571620e-01 4.68235463e-01 -1.25215209e+00
-5.66211522e-01 -6.65448755e-02 -6.18303597e-01 -9.37159508e-02
2.05981746e-01 -4.11448777e-01 2.37560660e-01 -1.33477747e-01
7.80523002e-01 -6.26548171e-01 -3.58599722e-01 -7.51691103e-01
-1.07417673e-01 5.45971394e-01 8.03720295e-01 -1.02611971e+00
6.82923377e-01 -1.64284253e+00 -4.78122890e-01 8.02113175e-01
-1.58002630e-01 -9.15823355e-02 1.28081992e-01 2.85745859e-01
-4.01351303e-01 -9.28161517e-02 -6.16486013e-01 9.51568484e-02
-2.37206653e-01 -1.20734572e-01 -3.54804724e-01 9.05332804e-01
5.75475246e-02 8.74242544e-01 -7.75994778e-01 1.50069371e-01
1.36589780e-01 5.33000529e-01 -6.94876671e-01 -6.47517219e-02
3.24611604e-01 5.63992739e-01 -7.13741302e-01 7.48456597e-01
1.56824172e+00 3.80258299e-02 5.25322668e-02 -3.47348034e-01
-2.53027081e-01 8.98919106e-02 -1.22661459e+00 1.13701594e+00
-6.96113348e-01 7.67521322e-01 8.32516253e-02 -1.59584987e+00
1.42324400e+00 2.84054041e-01 5.20843267e-01 -1.30637515e+00
1.04162440e-01 7.45097458e-01 7.99865872e-02 -1.00458038e+00
8.34017038e-01 -7.56286621e-01 -8.19768459e-02 4.32892948e-01
-2.63264596e-01 -3.22985500e-01 -4.33362633e-01 -1.32886738e-01
9.00053382e-01 3.55566621e-01 -3.56304765e-01 -4.32901829e-01
1.40008166e-01 2.42562950e-01 1.94368735e-01 8.42090964e-01
3.21773082e-01 7.05308974e-01 6.66693598e-02 -4.22660470e-01
-1.08285952e+00 -1.14667988e+00 -7.56455004e-01 6.87980771e-01
2.89699584e-01 4.28503424e-01 -5.63167073e-02 2.22550064e-01
1.81703940e-01 6.51949227e-01 -8.15365851e-01 -3.61736029e-01
-6.12378418e-01 -1.47812569e+00 1.11861956e+00 8.69169712e-01
6.85354888e-01 -1.29705405e+00 -9.03062701e-01 5.56521475e-01
-2.99389422e-01 -9.89277542e-01 7.18102872e-01 3.82314771e-01
-1.07241404e+00 -9.12956178e-01 -6.70696974e-01 -6.90249503e-01
1.67161435e-01 -2.05633879e-01 9.11633730e-01 -4.56516147e-02
-3.01549315e-01 2.74085194e-01 -3.63082111e-01 -4.50646371e-01
-3.26583952e-01 2.52637137e-02 5.66156171e-02 -2.52312154e-01
3.92871767e-01 -6.54898584e-01 -7.37932801e-01 7.80515671e-02
-6.59295142e-01 -2.76009709e-01 4.00127113e-01 8.09339643e-01
1.01639412e-01 -1.99368745e-02 7.84667015e-01 -6.64415300e-01
7.79128313e-01 -1.01822042e+00 -5.05758703e-01 -3.28119189e-01
-1.03963673e-01 2.68463008e-02 7.12023303e-02 -2.48183325e-01
-1.31919229e+00 -6.18227363e-01 -9.12850022e-01 1.05637601e-02
1.11040503e-01 1.32324493e+00 5.35015106e-01 -5.50011575e-01
9.26949441e-01 1.99235424e-01 -1.30642429e-01 -3.27632785e-01
-3.32707882e-01 7.02689588e-01 6.93393290e-01 -6.82568312e-01
5.25151670e-01 6.80079281e-01 7.99816661e-03 -1.43815768e+00
-2.76977390e-01 -4.08845305e-01 -6.41638711e-02 -2.85913795e-01
6.43748045e-01 -8.66399646e-01 -5.78451455e-01 6.24161243e-01
-8.01333189e-01 -6.47172570e-01 -4.57317866e-02 1.25468528e+00
-4.18486863e-01 4.66222286e-01 -4.52520102e-01 -8.85818183e-01
-5.87345123e-01 -9.57709253e-01 8.70829761e-01 3.83678287e-01
-3.24001551e-01 -1.56343818e+00 4.29135889e-01 -2.27021560e-01
7.12531149e-01 3.97801518e-01 5.53308189e-01 -3.30559969e-01
7.24330097e-02 -4.82235759e-01 -3.43868017e-01 -9.12638456e-02
-3.14206332e-01 -1.53568774e-01 -1.11801481e+00 -3.45816314e-01
1.57672226e-01 -2.24531889e-01 1.24807823e+00 1.13741207e+00
8.19255292e-01 3.85868728e-01 -6.34169579e-01 1.36158729e+00
1.54838204e+00 3.74738574e-01 1.10704875e+00 1.07068956e+00
5.17143190e-01 1.04813650e-01 4.30564106e-01 7.44410515e-01
6.27903715e-02 -6.14711083e-02 4.78592843e-01 -3.47335041e-01
5.04042923e-01 -7.71875307e-02 -7.20190583e-03 3.51387322e-01
-5.72437108e-01 5.56450784e-02 -1.39112055e+00 6.47249818e-01
-1.37259996e+00 -1.12139320e+00 -3.30813110e-01 2.00162148e+00
7.67537355e-02 1.86814114e-01 -6.01237833e-01 2.66928703e-01
3.57516766e-01 7.71418512e-02 -3.19483191e-01 -5.57890058e-01
-4.04497832e-02 8.32686126e-01 1.01904178e+00 7.08514035e-01
-8.58201683e-01 5.16190112e-01 6.66121387e+00 3.33373770e-02
-1.58222210e+00 -2.07744181e-01 2.10992455e-01 5.95859170e-01
-8.62171173e-01 -8.54134038e-02 -5.76047361e-01 1.31531268e-01
1.20457137e+00 -5.63719943e-02 -2.47096196e-02 4.64622498e-01
2.53101438e-01 -4.00476068e-01 -5.49017429e-01 9.22897160e-01
-3.92754018e-01 -1.66924214e+00 -2.01073676e-01 -8.70969845e-04
4.11277562e-01 5.34711540e-01 -7.60133937e-02 4.53178495e-01
3.70709747e-01 -1.27442110e+00 5.32290995e-01 8.09355021e-01
9.70327795e-01 -1.02645755e+00 1.10874021e+00 6.48963600e-02
-1.34182882e+00 -1.75846256e-02 -4.20742363e-01 -2.36902669e-01
6.04099870e-01 5.02408087e-01 -8.77656400e-01 8.51509809e-01
8.99671614e-01 5.89923859e-01 2.16955781e-01 1.22922838e+00
2.09256321e-01 7.89767742e-01 -5.05676270e-01 2.01371014e-01
7.02123284e-01 -4.27228391e-01 3.69762033e-01 1.37338853e+00
1.03173304e+00 -7.20021501e-03 -4.13182229e-01 7.94573843e-01
2.86367804e-01 -2.67658174e-01 -8.29675317e-01 1.20681532e-01
4.25228983e-01 9.59366143e-01 -5.26067615e-01 1.04202151e-01
-2.20444873e-01 3.52995664e-01 -2.65986681e-01 6.74309850e-01
-9.58692908e-01 -7.19599426e-01 6.42354250e-01 2.79725462e-01
5.32339931e-01 -3.18478048e-01 -2.83018440e-01 -4.95297194e-01
-4.82345104e-01 -2.13321835e-01 2.63039917e-01 -8.19924414e-01
-9.84819055e-01 5.69983721e-01 1.22728676e-01 -1.11360013e+00
-3.78383398e-01 -8.84327173e-01 -1.06400573e+00 1.21414232e+00
-1.85413384e+00 -6.07027769e-01 -7.96912551e-01 1.05506368e-01
-1.68951079e-01 -8.82027522e-02 8.40124726e-01 4.81880337e-01
-2.04885304e-01 3.39786440e-01 4.59308028e-01 2.75846690e-01
1.23170786e-01 -7.31981516e-01 9.67280686e-01 3.98805290e-01
-5.64628959e-01 3.11390340e-01 1.22228968e+00 -9.97153342e-01
-1.48935211e+00 -9.21068907e-01 3.91244650e-01 1.40884385e-01
8.92219365e-01 2.72894770e-01 -1.41725469e+00 8.08818817e-01
-2.86124200e-01 2.32843459e-01 4.92048562e-01 -1.60551220e-01
2.78137058e-01 3.49954575e-01 -1.26580048e+00 2.11552575e-01
6.50807381e-01 -3.56102109e-01 -6.33618712e-01 -1.12843804e-01
9.72880572e-02 -7.85860777e-01 -8.54941308e-01 9.39475954e-01
8.28476131e-01 -7.85436332e-01 1.17324841e+00 -5.02983451e-01
4.77529407e-01 1.35583580e-01 -3.95730108e-01 -1.58065236e+00
-2.39910722e-01 -1.34785369e-01 4.14149500e-02 2.95453578e-01
6.45783186e-01 -1.09074974e+00 1.21690965e+00 -4.00645994e-02
-3.88572067e-01 -6.52559161e-01 -9.83380973e-01 -5.94711661e-01
3.30141097e-01 -9.65366721e-01 4.48155671e-01 9.54967558e-01
-7.57092834e-02 9.58190882e-04 -5.63456491e-02 6.77794397e-01
8.82819772e-01 2.93059181e-02 1.48463994e-01 -1.42819893e+00
-4.02727313e-02 -1.05431505e-01 -3.50998431e-01 -8.53510439e-01
7.77703077e-02 -8.38552296e-01 3.37080717e-01 -1.94528723e+00
-3.45173329e-01 -9.13240194e-01 1.66759416e-01 -5.77776320e-03
2.54951596e-01 4.15883362e-01 -7.37391174e-01 -3.33372355e-02
4.98150468e-01 5.75693965e-01 1.09816563e+00 -1.18973374e-01
-3.74669552e-01 1.14368439e-01 -4.11249369e-01 7.88909733e-01
8.83519888e-01 -3.13096344e-01 -1.15343250e-01 -5.06748140e-01
5.45982361e-01 4.50118303e-01 3.47088099e-01 -1.41551328e+00
2.23483279e-01 4.18699495e-02 7.77509809e-01 -8.16458762e-01
2.65722036e-01 -3.86138588e-01 3.88118684e-01 6.06363714e-01
5.06916285e-01 -3.51810604e-02 6.82382464e-01 2.37744644e-01
-2.17522547e-01 -4.82302964e-01 9.18211341e-01 -1.33047268e-01
-1.18034935e+00 5.16563654e-01 -5.97208500e-01 -3.94210510e-04
5.44108748e-01 -8.02299261e-01 -7.34487772e-02 -2.93464571e-01
-8.51592183e-01 -1.26108276e-02 -2.64198463e-02 -1.83729231e-01
1.12607253e+00 -1.22337651e+00 -9.47720587e-01 6.74838245e-01
-1.47177011e-01 1.89968094e-01 6.71602547e-01 9.77804780e-01
-1.25687873e+00 1.70414910e-01 -2.93663591e-01 -7.36567259e-01
-3.82642895e-01 -3.91083658e-01 1.18001318e+00 -9.02718827e-02
-1.11624646e+00 1.02504122e+00 -1.65082693e-01 -7.48359799e-01
-4.66339946e-01 -2.78513551e-01 -5.19370377e-01 1.29656315e-01
5.32748401e-01 3.78812313e-01 5.66039741e-01 -4.91840065e-01
-3.66516024e-01 7.90174246e-01 1.14984706e-01 -3.87456059e-01
1.74961257e+00 2.08480597e-01 3.99521232e-01 1.38122126e-01
1.35822403e+00 -4.56350714e-01 -9.84249711e-01 1.59932733e-01
-1.16374843e-01 -1.50893077e-01 6.06555462e-01 -1.88154787e-01
-1.21541953e+00 7.38157511e-01 6.31131530e-01 -2.66239457e-02
6.09858692e-01 -2.29976103e-01 7.86714613e-01 5.53334117e-01
1.48194179e-01 -9.22523558e-01 8.34691059e-03 1.02070999e+00
6.45306826e-01 -8.81603420e-01 2.41636261e-02 1.75394416e-01
-4.10079807e-01 1.16971326e+00 4.48115021e-01 -2.84549266e-01
9.87521350e-01 8.57934535e-01 1.90566748e-01 -5.82222819e-01
-1.11813350e-02 3.94252330e-01 -3.35271478e-01 7.94807792e-01
4.35404420e-01 -1.35829374e-01 1.62843272e-01 4.92619723e-01
-4.90232289e-01 2.43656933e-01 7.34939218e-01 1.24496484e+00
-6.67520404e-01 -3.42468858e-01 -7.15283215e-01 7.06328750e-01
-2.50051469e-01 -1.83635443e-01 7.84704506e-01 8.20136130e-01
-4.16015118e-01 3.15048069e-01 6.62835956e-01 -2.28378654e-01
3.38184327e-01 -2.78812706e-01 5.36067009e-01 -1.60766751e-01
1.41286597e-01 -4.57569987e-01 4.42736685e-01 -2.10750744e-01
-1.01404957e-01 -6.25328004e-01 -1.92451346e+00 -4.33964431e-01
-2.38727435e-01 5.64051330e-01 9.46574509e-01 9.27682459e-01
-3.87076646e-01 8.77881825e-01 4.93862331e-01 -1.55445850e+00
-3.89148831e-01 -1.08163655e+00 -1.30275118e+00 -1.21230014e-01
5.03661573e-01 -1.01811039e+00 -8.12078536e-01 -5.71062565e-01] | [6.818160057067871, 2.3868775367736816] |
009caa2a-f535-4319-91a3-dc1eb22faf88 | learning-to-count-objects-with-few-exemplar | 1905.07898 | null | https://arxiv.org/abs/1905.07898v1 | https://arxiv.org/pdf/1905.07898v1.pdf | Learning to Count Objects with Few Exemplar Annotations | In this paper, we study the problem of object counting with incomplete annotations. Based on the observation that in many object counting problems the target objects are normally repeated and highly similar to each other, we are particularly interested in the setting when only a few exemplar annotations are provided. Directly applying object detection with incomplete annotations will result in severe accuracy degradation due to its improper handling of unlabeled object instances. To address the problem, we propose a positiveness-focused object detector (PFOD) to progressively propagate the incomplete labels before applying the general object detection algorithm. The PFOD focuses on the positive samples and ignore the negative instances at most of the learning time. This strategy, though simple, dramatically boosts the object counting accuracy. On the CARPK dataset for parking lot car counting, we improved mAP@0.5 from 4.58% to 72.44% using only 5 training images each with 5 bounding boxes. On the Drink35 dataset for shelf product counting, the mAP@0.5 is improved from 14.16% to 53.73% using 10 training images each with 5 bounding boxes. | ['Yandong Guo', 'Rong Xiao', 'Jianfeng Wang', 'Lei Zhang'] | 2019-05-20 | null | null | null | null | ['object-counting'] | ['computer-vision'] | [ 2.54737884e-01 7.63967782e-02 -2.62763828e-01 -3.76806229e-01
-5.84708989e-01 -5.41873157e-01 4.95636165e-01 5.39123297e-01
-8.04273605e-01 6.78189695e-01 -5.20410538e-01 3.09698526e-02
2.58710861e-01 -6.59724355e-01 -7.81321824e-01 -5.29906571e-01
1.15189470e-01 6.57980263e-01 6.44422472e-01 4.13516939e-01
2.46716738e-01 3.46964657e-01 -1.61707616e+00 1.79020703e-01
7.47982264e-01 8.44129086e-01 4.09327805e-01 7.35579669e-01
-3.53584975e-01 9.13452506e-01 -8.25439334e-01 -4.87030804e-01
1.98658317e-01 8.78961757e-03 -5.82681477e-01 3.68867338e-01
7.85524428e-01 -5.38986146e-01 1.81811899e-01 1.27810955e+00
1.06464855e-01 7.88743943e-02 7.38881767e-01 -1.42775190e+00
-4.50087607e-01 4.89298463e-01 -1.15880740e+00 3.06279570e-01
-1.69646174e-01 -4.77288924e-02 9.26700175e-01 -1.17645478e+00
2.84153402e-01 1.16250312e+00 5.02256572e-01 5.83301008e-01
-1.25964880e+00 -9.06060994e-01 3.66273284e-01 5.33678383e-02
-1.56428230e+00 -1.57069772e-01 1.39261514e-01 -5.34739256e-01
6.31248176e-01 1.78292692e-01 4.52760637e-01 3.79208922e-01
-3.53999197e-01 1.04037023e+00 9.38654780e-01 -5.89014351e-01
1.97263330e-01 3.51532876e-01 8.50361645e-01 6.84593737e-01
9.82967198e-01 -1.12344868e-01 -8.72742236e-02 9.41293873e-03
5.24323523e-01 3.76211464e-01 2.26210043e-01 -1.74431816e-01
-9.81207788e-01 9.04874682e-01 4.99241590e-01 1.41566098e-01
-2.16854289e-01 3.51173609e-01 3.87111992e-01 -3.62189740e-01
4.65405732e-01 3.21121186e-01 -3.40380609e-01 3.12658876e-01
-8.30516517e-01 3.89487237e-01 6.71258509e-01 1.23234618e+00
8.01255643e-01 -2.98142314e-01 -4.61340249e-01 9.98688161e-01
4.14468208e-03 5.44410169e-01 -7.72736445e-02 -9.26690042e-01
5.10197878e-01 7.64538407e-01 5.90995491e-01 -7.29302764e-01
-5.64224839e-01 -5.27170658e-01 -5.30299604e-01 1.63120836e-01
8.02792549e-01 1.16153531e-01 -1.22491097e+00 1.46681333e+00
4.58841801e-01 -5.46711944e-02 -2.45527878e-01 8.09563935e-01
6.97912514e-01 4.47747678e-01 5.63202858e-01 -4.72763442e-02
1.66596913e+00 -1.08108735e+00 -6.13119602e-01 -5.64046979e-01
9.16704297e-01 -7.63819814e-01 1.23496675e+00 2.78079271e-01
-7.96928525e-01 -5.47015071e-01 -1.11950088e+00 5.49192317e-02
-3.53951424e-01 6.64629340e-01 5.95359981e-01 6.49772942e-01
-4.60036784e-01 3.37180227e-01 -8.36046457e-01 -2.38030866e-01
6.95139945e-01 3.93234462e-01 -3.66370119e-02 -6.06946170e-01
-5.13280809e-01 9.50146437e-01 7.75732756e-01 -3.87915485e-02
-7.04015613e-01 -5.64945877e-01 -6.45958841e-01 1.66495010e-01
6.83633149e-01 -2.39531964e-01 1.58935285e+00 -7.42729008e-01
-7.44566858e-01 9.53076363e-01 -2.84045070e-01 -4.65271652e-01
6.78797901e-01 -3.29848796e-01 -2.23756075e-01 -1.73548628e-02
5.71792483e-01 9.10243690e-01 6.55733049e-01 -1.50724244e+00
-1.21776223e+00 -4.96284872e-01 8.98275599e-02 -9.02498215e-02
-2.50789195e-01 1.73354726e-02 -5.65052748e-01 -3.56334955e-01
-4.77981335e-03 -1.02436256e+00 -2.41187707e-01 5.43823317e-02
-3.51492435e-01 -4.23836052e-01 8.24147642e-01 -3.65037590e-01
9.27557826e-01 -2.12522912e+00 -5.31293929e-01 -7.36441240e-02
3.43200684e-01 3.62770140e-01 -3.16098183e-02 -1.58924639e-01
2.66410142e-01 7.62322079e-03 -3.13009590e-01 -6.10722542e-01
-8.91155228e-02 3.97891402e-01 -8.27292725e-02 3.90058905e-01
5.49617708e-01 6.61585152e-01 -1.15599179e+00 -7.46194720e-01
2.79458314e-01 1.53882250e-01 -4.88145351e-01 -2.65840087e-02
-1.66568294e-01 1.79543391e-01 -3.60876441e-01 7.62228012e-01
1.08353484e+00 -5.86597264e-01 4.19123098e-02 -5.21402853e-03
7.06832930e-02 -1.72728226e-01 -1.36086309e+00 9.58138347e-01
-4.82033938e-01 4.24518883e-01 -9.99782458e-02 -6.96364045e-01
8.05464804e-01 -1.28092185e-01 1.36496902e-01 -6.48007512e-01
5.00850156e-02 3.07859331e-01 9.26049352e-02 -2.51377910e-01
6.86710417e-01 -1.78166032e-01 2.48683407e-03 2.77870059e-01
-6.64807037e-02 -1.36325639e-02 8.28027487e-01 2.07791135e-01
1.03899348e+00 -3.58350068e-01 4.94334817e-01 -1.17769666e-01
4.26841438e-01 3.84315789e-01 5.49976170e-01 1.08882916e+00
-3.57600838e-01 5.57275414e-01 3.33420277e-01 -3.72838616e-01
-1.10593033e+00 -8.89540255e-01 -4.12193477e-01 1.30868626e+00
2.76140660e-01 -2.41384014e-01 -6.63086355e-01 -8.46163273e-01
2.27460667e-01 9.82591867e-01 -6.33951724e-01 -1.79752111e-02
-5.83734393e-01 -9.15499270e-01 2.59426534e-01 8.96415830e-01
6.55343413e-01 -8.24003756e-01 -6.64981723e-01 1.82919621e-01
-1.43025920e-01 -1.41250694e+00 -4.19754982e-01 3.30942541e-01
-8.77272010e-01 -1.26307154e+00 -8.27107608e-01 -6.47816658e-01
9.82539773e-01 5.88835299e-01 1.22148347e+00 4.24971700e-01
-3.53542358e-01 2.19857786e-02 -3.46897423e-01 -7.76098430e-01
-4.51070309e-01 1.33447712e-02 -1.93797797e-01 -2.74628639e-01
7.54286706e-01 2.46402249e-01 -3.75584096e-01 5.28226912e-01
-7.22073376e-01 -1.70785233e-01 4.79608446e-01 9.23013031e-01
5.67366600e-01 8.30948651e-02 6.00570560e-01 -1.09741640e+00
-6.89195320e-02 -2.98883766e-01 -9.90968466e-01 1.28159523e-01
-4.38901126e-01 -1.80326685e-01 4.85021472e-01 -7.33310580e-01
-8.74960959e-01 4.48593467e-01 4.18602884e-01 -2.95949548e-01
-4.48107086e-02 -1.43066227e-01 7.23758936e-02 1.89125955e-01
6.96568251e-01 -7.26617202e-02 -4.19006348e-01 -5.00722766e-01
1.17216721e-01 5.11075556e-01 4.86125648e-01 -4.14556861e-01
6.06998503e-01 5.48313141e-01 -4.83966917e-02 -7.16829181e-01
-1.35358071e+00 -9.73011076e-01 -6.14449561e-01 -1.37889400e-01
7.81914830e-01 -8.95971298e-01 -8.43850195e-01 1.94232374e-01
-1.14129066e+00 -2.24295855e-01 -2.81952739e-01 4.83866513e-01
-2.18843639e-01 2.67210573e-01 -4.27917987e-01 -1.23875999e+00
-1.76271006e-01 -9.72937465e-01 1.26460350e+00 1.62272498e-01
-1.53681859e-01 -3.63321334e-01 -4.21621054e-01 3.18504155e-01
-1.21538065e-01 -5.05892932e-02 6.95120752e-01 -8.62213314e-01
-5.90720952e-01 -6.94979429e-01 -6.54452860e-01 5.02988935e-01
3.08021866e-02 -1.26467884e-01 -1.14693129e+00 -1.73178673e-01
-2.48458385e-01 -2.15738803e-01 1.10600317e+00 2.51434147e-01
1.18902481e+00 -3.67219821e-02 -5.61263919e-01 -1.81187108e-01
1.36556160e+00 2.03572914e-01 2.23181590e-01 2.44011372e-01
6.09893858e-01 5.64028382e-01 1.07094431e+00 4.96544749e-01
8.25750083e-02 5.85989237e-01 6.20410681e-01 -1.22344837e-01
-7.24726617e-02 -2.55131751e-01 -1.49180945e-02 1.05592906e-01
1.54525870e-02 -2.68131018e-01 -9.93039966e-01 7.17841864e-01
-1.75006831e+00 -6.46318793e-01 -5.12930393e-01 2.33598185e+00
6.24406636e-01 3.88097584e-01 4.25402313e-01 4.51394439e-01
1.17326343e+00 -3.26613754e-01 -4.48991030e-01 8.10523778e-02
1.96847081e-01 1.03484793e-02 9.51067746e-01 3.97280097e-01
-1.28893960e+00 8.03010404e-01 5.71503925e+00 8.02034378e-01
-5.13697863e-01 3.01811039e-01 6.30753875e-01 -9.43875089e-02
5.69633365e-01 -1.18797421e-01 -1.47795129e+00 6.02798939e-01
5.65234601e-01 6.46200031e-02 5.05439043e-02 1.22301710e+00
-5.34191057e-02 -6.70766771e-01 -1.16287398e+00 8.49969208e-01
-5.53543456e-02 -9.41694677e-01 -2.09810331e-01 -1.38487637e-01
7.77508676e-01 -2.49266982e-01 -9.69005451e-02 5.51566541e-01
4.59730804e-01 -8.67859304e-01 9.33642447e-01 3.10542621e-02
5.84142089e-01 -7.02635288e-01 9.92099226e-01 6.64244950e-01
-1.18746758e+00 -3.36409688e-01 -5.91074824e-01 -8.86226371e-02
1.07798189e-01 8.38245749e-01 -1.23284388e+00 -9.23428163e-02
7.15559065e-01 2.80792356e-01 -5.19622028e-01 1.29569387e+00
-2.78953254e-01 5.77760577e-01 -5.73421657e-01 -2.50605077e-01
2.75764376e-01 2.00946793e-01 1.54167861e-01 1.34963918e+00
1.08734109e-01 1.29585147e-01 3.95021230e-01 9.36933875e-01
-2.70528793e-01 -4.06130403e-03 -3.67371827e-01 2.94298798e-01
4.77810860e-01 1.35980165e+00 -1.14421284e+00 -7.48843849e-01
-4.50468749e-01 7.65663981e-01 4.42760587e-01 1.09044693e-01
-9.85385418e-01 -2.53357738e-01 9.55146998e-02 2.48903796e-01
5.94598770e-01 -4.46867235e-02 -3.08216751e-01 -7.37083018e-01
1.06766805e-01 -2.04646364e-01 4.61373240e-01 -4.98257995e-01
-1.34978580e+00 4.13550995e-02 1.80462450e-01 -1.05304086e+00
2.23786011e-01 -8.62024546e-01 -4.10386503e-01 6.54970288e-01
-1.49539554e+00 -1.00169516e+00 -5.91752112e-01 9.98910144e-02
6.88207090e-01 3.05185080e-01 4.64189947e-01 5.19493282e-01
-5.93901217e-01 5.81429243e-01 -2.67742015e-02 3.00906956e-01
4.63037461e-01 -1.27281725e+00 1.75165072e-01 6.55285060e-01
1.49056032e-01 4.26488966e-01 6.27492130e-01 -4.92281765e-01
-7.58942723e-01 -1.44377518e+00 9.43770051e-01 -5.81645906e-01
3.41611087e-01 -4.26979989e-01 -1.05878592e+00 6.35648608e-01
-5.27683556e-01 5.05414426e-01 2.19522104e-01 2.63523757e-02
-3.08204740e-01 3.15714292e-02 -1.42319715e+00 4.43851620e-01
9.45643306e-01 -2.65885945e-02 -4.15016264e-01 7.04234600e-01
4.56397980e-01 -2.60522157e-01 -4.64451730e-01 3.83476496e-01
2.59963900e-01 -5.73418200e-01 9.01718318e-01 -4.16244984e-01
2.60915220e-01 -6.34429097e-01 -3.30844611e-01 -7.08355546e-01
-2.89523453e-01 2.53790379e-01 -2.84565687e-01 1.16603839e+00
3.92918497e-01 -4.64795530e-01 9.01093006e-01 4.82866287e-01
1.29672661e-01 -2.42056325e-01 -7.99713135e-01 -1.20565701e+00
-2.02152990e-02 -3.30762029e-01 3.61404210e-01 6.91230834e-01
-2.43783697e-01 4.38061953e-01 -1.92268521e-01 2.96488047e-01
7.63755262e-01 8.62812772e-02 8.34718645e-01 -1.23508883e+00
-2.45557487e-01 -1.50457904e-01 -3.75576675e-01 -1.09236801e+00
-1.66033711e-02 -7.98645854e-01 3.21202636e-01 -1.25781953e+00
8.89130294e-01 -5.51154017e-01 -1.53422266e-01 5.04955173e-01
-5.69938123e-01 5.93680203e-01 5.18724918e-01 2.09968686e-01
-7.72772133e-01 8.08911175e-02 1.15425467e+00 -2.62559116e-01
7.38102011e-03 2.84364879e-01 -3.22343558e-01 9.40717101e-01
9.09700513e-01 -9.31037128e-01 -3.56951095e-02 -3.49198490e-01
-1.12528525e-01 -3.02294463e-01 5.77950716e-01 -9.31283355e-01
4.99983393e-02 5.99543490e-02 5.84547222e-01 -9.85642731e-01
4.34826881e-01 -9.89605725e-01 -2.48514041e-01 8.78879070e-01
-2.09520251e-01 -2.48668760e-01 3.36871654e-01 7.91616261e-01
1.76782701e-02 -7.09392011e-01 1.09528768e+00 -2.98468590e-01
-7.89629459e-01 -2.49055470e-03 -4.94075567e-02 8.42189193e-02
1.16343343e+00 -2.89146155e-01 -3.05922836e-01 1.31573156e-01
-8.15153480e-01 3.91847402e-01 2.29986116e-01 5.53820729e-02
1.41169101e-01 -1.24796891e+00 -6.61567152e-01 -2.97353677e-02
4.81543362e-01 2.83572763e-01 4.53016907e-02 6.31486177e-01
-2.46352106e-01 4.59129870e-01 1.72536090e-01 -1.00579190e+00
-1.41082323e+00 8.02057803e-01 2.24033058e-01 -3.56932431e-01
-4.82967645e-01 7.31482148e-01 4.64851707e-01 -3.20012629e-01
2.40593076e-01 -4.91657555e-01 -2.42920399e-01 5.89038059e-02
6.07228518e-01 8.15487623e-01 1.30398810e-01 -3.91590029e-01
-4.35931563e-01 4.24238890e-01 -4.85662699e-01 2.23386988e-01
1.02926743e+00 1.84363887e-01 3.71627271e-01 4.90716189e-01
9.95749295e-01 -1.80668727e-01 -1.26488519e+00 -3.19000572e-01
4.11673725e-01 -7.59427071e-01 -6.23738915e-02 -7.47382104e-01
-1.01521456e+00 7.76001394e-01 6.55116677e-01 2.28445739e-01
7.78295875e-01 1.76220268e-01 3.33971143e-01 4.63496327e-01
3.87103677e-01 -1.09660470e+00 2.06093490e-01 4.91002232e-01
5.17543852e-01 -1.66771245e+00 1.91522539e-01 -6.96767092e-01
-6.49120808e-01 8.90230179e-01 8.54728937e-01 -1.93277091e-01
2.09567949e-01 3.11968684e-01 -3.25214297e-01 -2.34602123e-01
-3.97030681e-01 -3.02833617e-01 1.19025894e-01 3.84068906e-01
3.20498407e-01 3.21058840e-01 -2.44687513e-01 5.25843620e-01
2.13076532e-01 1.01660430e-01 5.40548444e-01 1.02638495e+00
-7.71503389e-01 -5.82353115e-01 -6.02740049e-01 8.43023121e-01
-5.59624135e-01 -3.01824212e-02 -1.42855361e-01 1.15795958e+00
1.84930444e-01 1.01808131e+00 3.75357658e-01 1.93897635e-01
5.88027954e-01 -1.84977532e-03 5.03070593e-01 -9.49414253e-01
-3.32969159e-01 -1.34056121e-01 1.24671988e-01 -1.43204004e-01
-5.07451594e-01 -5.80993712e-01 -1.48783350e+00 -3.96966815e-01
-1.00515139e+00 -1.21967413e-01 5.41755319e-01 6.72568977e-01
-1.73754185e-01 4.62471455e-01 3.28871280e-01 -7.36308992e-01
-7.70650804e-01 -1.11485696e+00 -7.66147852e-01 2.61724919e-01
3.51210386e-01 -9.40487206e-01 -4.82448131e-01 -1.47823347e-02] | [9.072470664978027, 0.723052442073822] |
7dd5f53d-f901-4bc6-8d88-3ddc98bbac78 | meme-generating-rnn-model-explanations-via | 2012.06954 | null | https://arxiv.org/abs/2012.06954v1 | https://arxiv.org/pdf/2012.06954v1.pdf | MEME: Generating RNN Model Explanations via Model Extraction | Recurrent Neural Networks (RNNs) have achieved remarkable performance on a range of tasks. A key step to further empowering RNN-based approaches is improving their explainability and interpretability. In this work we present MEME: a model extraction approach capable of approximating RNNs with interpretable models represented by human-understandable concepts and their interactions. We demonstrate how MEME can be applied to two multivariate, continuous data case studies: Room Occupation Prediction, and In-Hospital Mortality Prediction. Using these case-studies, we show how our extracted models can be used to interpret RNNs both locally and globally, by approximating RNN decision-making via interpretable concept interactions. | ['Pietro Liò', 'Mateja Jamnik', 'Botty Dimanov', 'Dmitry Kazhdan'] | 2020-12-13 | null | null | null | null | ['occupation-prediction'] | ['natural-language-processing'] | [ 2.42962018e-01 1.01087260e+00 -3.44056077e-02 -7.20934451e-01
-2.17408583e-01 -2.91366894e-02 3.86349648e-01 8.09989274e-02
9.98057500e-02 9.10423696e-01 9.47937131e-01 -7.51994908e-01
-3.38167965e-01 -7.14746714e-01 -5.35219491e-01 1.18040890e-01
1.76957443e-01 9.64173794e-01 -9.19418275e-01 -1.41752958e-01
-2.17519969e-01 4.79285240e-01 -1.39957106e+00 8.80004883e-01
6.43648684e-01 5.68134725e-01 -1.78745881e-01 8.29188764e-01
-2.93008894e-01 1.99797153e+00 -8.06316376e-01 -2.68793106e-01
-3.49746287e-01 -5.36664665e-01 -1.09276688e+00 -4.43427593e-01
-1.39561504e-01 -1.79658487e-01 -4.78816688e-01 3.71632397e-01
1.65354479e-02 3.72005165e-01 1.07661366e+00 -1.03296900e+00
-9.89682138e-01 1.16190326e+00 4.29633647e-01 -1.02529414e-01
7.13696852e-02 -1.35133237e-01 1.05521560e+00 -7.84942865e-01
3.61539155e-01 1.61286163e+00 9.65493321e-01 1.00188076e+00
-1.49658823e+00 -3.74974698e-01 1.29569501e-01 2.69521754e-02
-1.09942985e+00 -1.90733582e-01 4.85135376e-01 -2.80570984e-01
1.59267259e+00 6.94330692e-01 7.14654624e-01 1.58107233e+00
3.69850755e-01 7.62606800e-01 8.51141751e-01 -3.41246754e-01
2.79507756e-01 -3.35199460e-02 4.67525512e-01 7.11241245e-01
1.83882281e-01 1.14259169e-01 -4.43053097e-01 -1.72453552e-01
9.49088097e-01 8.64349842e-01 2.19549108e-02 6.37634337e-01
-1.24409807e+00 7.06037641e-01 8.10339272e-01 2.97510356e-01
-7.75981128e-01 8.53952467e-01 3.15435559e-01 3.56703401e-02
8.22806358e-01 1.06470895e+00 -5.51101744e-01 -1.77434027e-01
-8.96500051e-01 -5.67705631e-02 1.01858985e+00 9.74277496e-01
3.73073637e-01 3.04911971e-01 -3.63427550e-01 7.23088503e-01
2.31934041e-01 4.47994262e-01 5.82038164e-01 -8.64669681e-01
2.07882822e-01 8.63214076e-01 -1.00059792e-01 -9.55708206e-01
-9.47729409e-01 -4.81522381e-01 -1.54580772e+00 -3.06876630e-01
-2.83110470e-01 -9.27645937e-02 -1.10944259e+00 1.57845259e+00
-6.15916789e-01 2.65059203e-01 4.02832747e-01 3.67647260e-01
9.93172526e-01 7.54338741e-01 4.54183370e-01 7.54563510e-02
1.39515436e+00 -1.18714249e+00 -1.05817389e+00 -4.16253686e-01
9.07699108e-01 -4.78105322e-02 7.04376280e-01 1.37015328e-01
-9.53148186e-01 -6.97895706e-01 -6.41635537e-01 -2.40452126e-01
-4.80008751e-01 1.54341221e-01 6.25636280e-01 -1.26647791e-02
-1.24679875e+00 7.36545086e-01 -1.12876678e+00 -2.60156780e-01
6.08210742e-01 5.63491464e-01 -8.36558491e-02 1.38288110e-01
-1.24484873e+00 1.26766634e+00 4.55605268e-01 4.99760866e-01
-7.24026859e-01 -7.07285285e-01 -8.96267593e-01 4.66217667e-01
-4.74176668e-02 -1.53587294e+00 1.37520182e+00 -9.48549807e-01
-9.36077058e-01 3.05940509e-01 -5.46685100e-01 -1.07418633e+00
1.31974384e-01 -3.24984372e-01 -6.61228240e-01 -2.79449612e-01
-2.51360208e-01 7.47888207e-01 4.30415362e-01 -1.26027286e+00
-2.22207323e-01 -7.94328526e-02 3.93261341e-03 1.05232306e-01
-3.86075526e-01 -2.36820802e-01 5.27035668e-02 -9.01042461e-01
2.76873633e-02 -8.25067103e-01 -5.45122504e-01 -3.50858361e-01
-1.00433934e+00 -1.57140717e-01 4.72484052e-01 -7.81511486e-01
1.54508531e+00 -1.62261760e+00 3.56422782e-01 3.56424719e-01
1.08774734e+00 8.89811292e-02 -8.47393796e-02 2.78701395e-01
-4.02313054e-01 6.89636946e-01 -4.22401428e-01 -9.14447665e-01
-3.74928005e-02 8.95748913e-01 -4.99922037e-01 -3.95454168e-01
5.39321005e-01 1.48744965e+00 -9.81247365e-01 -2.10701495e-01
4.10825729e-01 8.33034992e-01 -3.72560054e-01 4.37344283e-01
-2.25486442e-01 2.00758740e-01 -4.81929690e-01 5.08840263e-01
-1.85465470e-01 -5.44454336e-01 3.48340452e-01 3.99319874e-03
5.54823279e-01 4.85205203e-01 -4.43778306e-01 1.10796225e+00
-1.06911027e+00 1.30792284e+00 -7.75573492e-01 -8.49611163e-01
9.84740674e-01 5.50179064e-01 1.17509685e-01 -2.24061400e-01
-1.48750497e-02 4.24387306e-02 -1.98421195e-01 -5.30985475e-01
6.10921919e-01 -5.73757589e-01 2.81205624e-02 7.53653169e-01
-2.50052720e-01 1.35068879e-01 -4.70996946e-01 6.90339729e-02
1.39517415e+00 -1.06785879e-01 5.89295506e-01 -3.24173748e-01
2.34554648e-01 -9.74212140e-02 2.53182411e-01 9.84399259e-01
4.16958779e-01 8.29012454e-01 4.27030087e-01 -1.12044454e+00
-1.14472592e+00 -8.53887081e-01 2.01938540e-01 8.77305925e-01
-6.38514042e-01 -5.41860163e-01 -8.89511168e-01 -5.99635959e-01
-1.08749010e-01 1.41937935e+00 -1.28121960e+00 -4.43981916e-01
-6.09306931e-01 -5.62032282e-01 9.09678221e-01 1.32655334e+00
-5.88719808e-02 -1.49430978e+00 -6.14807427e-01 4.59341258e-01
-4.76783484e-01 -9.56087232e-01 -2.34849583e-02 5.23194373e-01
-1.31176805e+00 -6.96979702e-01 -2.59611547e-01 -4.00354862e-01
1.10846722e+00 -1.60903558e-01 1.87856591e+00 3.77237797e-01
-9.29853842e-02 3.74437511e-01 -1.11826889e-01 -8.56342733e-01
-1.03140223e+00 3.27583969e-01 3.64089847e-01 -3.70726705e-01
6.15022480e-01 -5.45163691e-01 -3.96541506e-01 1.00383937e-01
-1.14996636e+00 7.93432295e-01 6.35061860e-01 1.00193632e+00
4.93483841e-01 -3.88653576e-01 4.83141154e-01 -1.28789628e+00
8.81655812e-01 -6.75061941e-01 2.93482900e-01 5.14865398e-01
-7.25028813e-01 5.94732463e-01 9.06979859e-01 -1.59130096e-01
-8.65376055e-01 -2.57680327e-01 -2.01528221e-02 -4.86228734e-01
-2.87739009e-01 8.19169402e-01 3.20919514e-01 8.34729433e-01
8.97431135e-01 -1.22092143e-02 -9.82912257e-02 -3.74753982e-01
4.03588802e-01 7.00600505e-01 6.23453319e-01 -4.44044828e-01
2.53958613e-01 3.97202760e-01 1.98695183e-01 -4.48178142e-01
-1.08462906e+00 -3.86466831e-02 -6.31742835e-01 -3.51997279e-02
8.87712061e-01 -8.21820498e-01 -7.05636382e-01 -3.30259889e-01
-1.67608750e+00 -3.35745633e-01 -6.26866817e-01 3.30195695e-01
-7.19147980e-01 -4.57391381e-01 -6.75827146e-01 -1.11464870e+00
-6.62891567e-01 -7.63819456e-01 1.03520238e+00 -4.55507301e-02
-1.20405233e+00 -1.62355864e+00 -1.88458897e-02 3.11596066e-01
5.25240719e-01 4.14043009e-01 1.19233334e+00 -1.12568879e+00
-3.34197879e-01 -2.52632916e-01 -1.98626280e-01 3.63723099e-01
3.49865735e-01 -2.51354456e-01 -1.23677206e+00 3.25639874e-01
-1.12855464e-01 -2.23511321e-04 8.28498662e-01 6.06426060e-01
1.63104784e+00 -8.08855176e-01 -5.00264704e-01 5.27486563e-01
1.03634417e+00 -6.28248751e-02 6.97117209e-01 1.61190331e-02
1.16406274e+00 7.17173934e-01 4.57359590e-02 2.44472608e-01
3.61220270e-01 2.64441997e-01 3.94792140e-01 -6.68798089e-01
-3.17316055e-02 -5.37490189e-01 1.18447244e-01 1.01196957e+00
-7.79770911e-01 -5.08073509e-01 -1.41416776e+00 2.25743935e-01
-2.44593596e+00 -9.77831006e-01 -6.58506677e-02 1.35461724e+00
4.95907962e-01 4.67504710e-02 -3.13214868e-01 -1.64344430e-01
3.92740279e-01 -1.92618787e-01 -6.86257780e-01 -9.28124487e-01
-1.54143408e-01 2.15366676e-01 1.98024973e-01 4.43520069e-01
-6.95418298e-01 7.05610275e-01 7.32109833e+00 3.53389204e-01
-4.86700684e-01 -4.26074192e-02 1.41871881e+00 -1.56221017e-01
-6.64122403e-01 -3.59567434e-01 -2.82893986e-01 -8.10432360e-02
1.78249407e+00 9.40735042e-02 5.74386954e-01 8.99380624e-01
6.89254045e-01 6.53086782e-01 -1.75336921e+00 1.02587819e+00
5.22987405e-03 -1.82176554e+00 6.73933506e-01 -9.66883078e-02
6.84500396e-01 -2.76508778e-02 -4.18149233e-02 4.86511618e-01
5.12713373e-01 -2.19954586e+00 5.36299109e-01 1.40065873e+00
7.49842644e-01 -7.39396453e-01 1.10453582e+00 3.24872553e-01
-1.01747656e+00 -3.02812040e-01 -3.55281532e-01 -3.87745380e-01
-8.51911157e-02 3.68794024e-01 -1.20690846e+00 4.26669627e-01
4.82761472e-01 1.09000230e+00 -3.97024274e-01 1.26935855e-01
-3.44184309e-01 6.96236253e-01 -4.80737351e-02 -1.70857638e-01
1.91336721e-01 1.31179377e-01 3.70273501e-01 1.54065835e+00
2.21474245e-01 2.64490724e-01 -5.40566981e-01 1.58091629e+00
-3.53063732e-01 -3.42248142e-01 -8.61768603e-01 -1.10653147e-01
2.76679605e-01 9.97044265e-01 -4.16635990e-01 -6.53919399e-01
-1.98481418e-02 9.45693195e-01 5.52825749e-01 6.75753236e-01
-8.14846396e-01 -1.17547937e-01 8.27320755e-01 1.37853667e-01
-1.76490635e-01 4.79653142e-02 -7.77829111e-01 -1.00029981e+00
-3.13244015e-01 -9.08032000e-01 3.03081989e-01 -1.39264321e+00
-1.33212411e+00 1.34025764e+00 1.01507954e-01 -9.26366389e-01
-9.39362466e-01 -8.80516112e-01 -6.19928598e-01 9.51594472e-01
-9.78731453e-01 -1.49073720e+00 -3.00478399e-01 1.04085468e-01
7.31603384e-01 -1.80820018e-01 1.45324993e+00 -2.30086699e-01
-8.47827196e-01 3.16545516e-01 4.95458990e-02 1.94352075e-01
-8.86086896e-02 -1.36708748e+00 9.18326914e-01 2.91856259e-01
2.93650091e-01 1.16431904e+00 7.68606007e-01 -4.02800977e-01
-7.94340014e-01 -1.60642910e+00 1.19450831e+00 -1.18049324e+00
4.01764929e-01 -3.93694580e-01 -9.36303794e-01 1.29868007e+00
-7.75992796e-02 -2.55934596e-01 1.02131510e+00 5.16251326e-01
-1.30417362e-01 2.05486387e-01 -7.87737906e-01 9.19052362e-01
1.19320345e+00 -9.24073398e-01 -8.71786594e-01 4.00840372e-01
9.41101074e-01 -2.74918169e-01 -9.50098634e-01 4.71284628e-01
3.97664487e-01 -7.58108139e-01 8.97842050e-01 -1.48847353e+00
1.01116979e+00 -9.96795669e-03 2.40714755e-02 -1.43452883e+00
-3.37177038e-01 -6.27780080e-01 -8.41805041e-01 6.78320229e-01
9.29985464e-01 -5.23133159e-01 5.80108285e-01 1.30946112e+00
-2.72682786e-01 -1.19975495e+00 -5.14207184e-01 -2.21585482e-01
-1.36956319e-01 -8.32955658e-01 1.08503044e+00 7.25332081e-01
9.51112881e-02 5.58544397e-01 -6.49347961e-01 -9.48696584e-02
1.18987702e-01 -2.72522599e-01 3.28059673e-01 -1.49486768e+00
-3.29965293e-01 -4.73176330e-01 -3.91072810e-01 -7.94714928e-01
5.35056889e-01 -8.22160244e-01 2.19948008e-03 -2.02973700e+00
4.14776385e-01 -9.71210226e-02 -6.99549973e-01 9.38062012e-01
-1.35232389e-01 3.79286543e-03 1.16761409e-01 3.78334165e-01
-4.54618424e-01 3.71305406e-01 7.59830773e-01 -3.02765071e-01
-1.74222887e-01 1.31466761e-01 -9.14458811e-01 9.10395741e-01
8.60735178e-01 -6.53824270e-01 -4.43548828e-01 -5.55947185e-01
6.27927661e-01 1.44641697e-02 5.18364429e-01 -8.34775388e-01
1.17896214e-01 -1.52413249e-01 5.69490790e-01 -3.90796870e-01
4.40991491e-01 -8.75905156e-01 5.03842711e-01 5.59621155e-01
-1.00313890e+00 5.23844361e-01 5.84399939e-01 7.15045571e-01
-2.17694283e-01 3.27171907e-02 4.56215814e-02 -1.88429505e-01
-2.90977806e-01 4.65670936e-02 -7.85118520e-01 -1.66071996e-01
5.76225042e-01 2.17538569e-02 -2.37715334e-01 -8.90124142e-01
-9.44385946e-01 4.44343574e-02 -1.18844189e-01 5.41928172e-01
1.14147878e+00 -1.23553836e+00 -7.75021970e-01 3.91894504e-02
4.14396860e-02 1.90421551e-01 1.44217238e-02 3.35268021e-01
-7.03999996e-01 7.86265850e-01 -1.96854454e-02 -3.65553141e-01
-1.05550563e+00 3.37416977e-01 8.63893032e-01 -5.93203068e-01
-6.62043273e-01 5.22929847e-01 2.49375775e-01 -6.83878660e-01
1.90668210e-01 -1.21736753e+00 -2.30173901e-01 -4.55159724e-01
5.81163824e-01 4.21191543e-01 -2.80690938e-01 -3.10200274e-01
-1.32075340e-01 -3.08412053e-02 1.16104178e-01 1.43074557e-01
1.58465242e+00 8.60069469e-02 -2.84844249e-01 8.84130120e-01
9.41566646e-01 -7.56790042e-01 -7.08597839e-01 5.35765067e-02
2.61985153e-01 4.26434994e-01 -1.33048952e-01 -9.23150837e-01
-6.77257061e-01 9.91689205e-01 1.93271458e-01 3.34024370e-01
1.13514125e+00 7.59452656e-02 7.52080083e-01 9.95105743e-01
-1.87097073e-01 -7.60453522e-01 -1.77067712e-01 6.50374532e-01
9.81345057e-01 -1.01710761e+00 -1.49789348e-01 -5.28422091e-03
-7.24745631e-01 1.48420393e+00 3.87242168e-01 1.38761759e-01
5.06926358e-01 1.34034827e-01 8.37244019e-02 -4.17323798e-01
-1.16818929e+00 2.88036764e-01 8.13688517e-01 4.36110914e-01
6.47946477e-01 6.40133560e-01 6.81569278e-01 1.03264046e+00
-3.59759688e-01 4.29624841e-02 4.75502640e-01 4.35303569e-01
-1.81798171e-02 -6.56121850e-01 -2.04832673e-01 9.45104003e-01
-1.14061005e-01 -6.16201937e-01 -6.34327233e-01 6.63326859e-01
-1.08902425e-01 8.15847039e-01 4.22426432e-01 -7.37688541e-01
4.00476515e-01 2.22674385e-01 -2.31822222e-01 -8.86685073e-01
-8.91139150e-01 -6.91604853e-01 4.37750429e-01 -5.27952313e-01
-4.51415479e-01 -1.70315951e-01 -1.65297329e+00 -5.24183929e-01
1.25292122e-01 5.07538803e-02 4.01505291e-01 1.27263772e+00
7.02716827e-01 1.41059935e+00 1.36437103e-01 -5.29664993e-01
-3.43072891e-01 -1.25334251e+00 -1.71061754e-01 5.07361472e-01
5.61668456e-01 -7.56614879e-02 -2.89051265e-01 8.82866010e-02] | [8.341506004333496, 6.055195331573486] |
4b297c59-01a0-45a8-bc6a-b968da6c804f | bounded-projection-matrix-approximation-with | 2305.15430 | null | https://arxiv.org/abs/2305.15430v1 | https://arxiv.org/pdf/2305.15430v1.pdf | Bounded Projection Matrix Approximation with Applications to Community Detection | Community detection is an important problem in unsupervised learning. This paper proposes to solve a projection matrix approximation problem with an additional entrywise bounded constraint. Algorithmically, we introduce a new differentiable convex penalty and derive an alternating direction method of multipliers (ADMM) algorithm. Theoretically, we establish the convergence properties of the proposed algorithm. Numerical experiments demonstrate the superiority of our algorithm over its competitors, such as the semi-definite relaxation method and spectral clustering. | ['Qiang Sun', 'Hengchao Chen', 'Zheng Zhai'] | 2023-05-21 | null | null | null | null | ['community-detection'] | ['graphs'] | [ 2.47199863e-01 -2.90992409e-01 -2.00956255e-01 -9.39884968e-03
-7.06561744e-01 -2.99286962e-01 1.68056667e-01 -1.36689812e-01
-6.25282347e-01 7.93575704e-01 5.27495332e-02 -1.18578792e-01
-3.24735135e-01 -2.12041721e-01 -3.13455194e-01 -9.44399774e-01
-4.70094569e-02 5.72937846e-01 -2.15184018e-01 2.34322950e-01
4.54516262e-01 2.75623947e-01 -6.26251161e-01 -2.32390672e-01
1.22416139e+00 4.27877903e-01 2.10161075e-01 2.68962830e-01
3.17110121e-02 4.27113682e-01 -7.14001432e-02 2.60716602e-02
4.56887484e-01 -4.23339933e-01 -7.81340301e-01 1.00959325e+00
9.66815278e-02 -8.92030746e-02 -2.08052397e-01 1.38675499e+00
2.95588523e-01 3.11096162e-01 6.87749565e-01 -1.49951279e+00
-5.32987535e-01 5.41983843e-01 -1.45895147e+00 2.96174020e-01
1.71271458e-01 -2.62308389e-01 1.02635789e+00 -1.58206415e+00
6.89740300e-01 1.08201146e+00 7.70181775e-01 2.72369891e-01
-1.41660869e+00 -4.87764835e-01 4.08514947e-01 2.71846980e-01
-1.85234237e+00 -1.11312315e-01 1.03426325e+00 -6.37426019e-01
3.98709029e-01 2.24655569e-01 4.14481908e-01 5.56331873e-01
-4.87428695e-01 1.19052339e+00 1.12213981e+00 -4.01429445e-01
2.23698109e-01 2.03733489e-01 2.87306666e-01 9.84507918e-01
7.09418058e-01 -4.05515850e-01 -2.94700205e-01 -7.60658264e-01
8.55790377e-01 8.12066048e-02 -5.29269338e-01 -6.33992672e-01
-1.35549283e+00 1.21398747e+00 1.41351968e-01 -8.39054137e-02
-4.57121670e-01 -9.05892774e-02 1.09325655e-01 2.02127360e-02
4.27576870e-01 -5.78412414e-02 7.83507228e-02 1.37651309e-01
-1.09962535e+00 -1.55895576e-01 8.14625978e-01 9.83223140e-01
5.60866356e-01 1.85098857e-01 5.50263971e-02 8.92379224e-01
5.70627809e-01 3.96934330e-01 2.84182161e-01 -9.90480721e-01
6.20799959e-01 5.19271672e-01 1.49399161e-01 -1.39450705e+00
-2.60957688e-01 -5.30390918e-01 -1.31163752e+00 -3.69368702e-01
1.96846560e-01 -3.73451829e-01 -1.90010414e-01 1.24679255e+00
5.05905271e-01 5.60977161e-01 -1.84870884e-01 1.20855308e+00
3.49899769e-01 8.16666186e-01 -3.91301632e-01 -1.03543603e+00
7.66209424e-01 -1.25024140e+00 -9.60213959e-01 -4.24556322e-02
2.91684031e-01 -6.04201078e-01 6.62603736e-01 4.82890993e-01
-9.81374443e-01 -1.11005284e-01 -8.69966865e-01 4.70780104e-01
3.00964594e-01 3.68374586e-01 8.18655908e-01 6.55828536e-01
-8.08793068e-01 2.72956669e-01 -8.93653333e-01 -2.28478923e-01
2.41912246e-01 4.75599766e-01 -3.23764712e-01 1.05023971e-02
-5.87829232e-01 1.64399996e-01 6.45813763e-01 4.48400974e-01
-6.66938186e-01 -3.87560368e-01 -7.16274321e-01 -1.37933418e-01
3.93560290e-01 -4.72308278e-01 6.90422475e-01 -6.57349467e-01
-1.33208990e+00 7.89067864e-01 -2.84137309e-01 -2.34152347e-01
5.10890484e-01 -8.99429545e-02 -7.44187459e-02 3.55364323e-01
2.37747833e-01 2.96762045e-02 8.05117786e-01 -1.35252631e+00
-5.98426461e-01 -1.40929639e-01 -2.38110140e-01 4.23952579e-01
-9.49112475e-01 1.34891212e-01 -5.22315681e-01 -6.33484364e-01
4.78901595e-01 -9.87040937e-01 -9.67021763e-01 -4.64697927e-03
-4.54895943e-01 -1.39466494e-01 7.24858344e-01 -6.79068983e-01
1.47851729e+00 -2.22365379e+00 5.94233215e-01 6.63964748e-01
4.85083550e-01 -2.87379194e-02 -1.64339207e-02 6.02086842e-01
-1.29732266e-01 -2.07735702e-01 -9.01571333e-01 -3.31362844e-01
-1.22874856e-01 1.76833466e-01 8.03140774e-02 1.02068615e+00
-4.50085960e-02 3.31636488e-01 -1.20001590e+00 -7.66173303e-01
-1.36627063e-01 3.35476398e-01 -7.13578820e-01 -1.66398704e-01
4.18903440e-01 2.35427365e-01 -4.63901252e-01 4.95532721e-01
1.03893495e+00 -5.96220315e-01 5.25095940e-01 -9.02172476e-02
-2.07622021e-01 -3.02447289e-01 -2.11624742e+00 1.43134058e+00
9.90574062e-02 3.37111712e-01 7.26309657e-01 -1.52006805e+00
7.72432685e-01 3.81591707e-01 7.59139657e-01 1.90579981e-01
3.30586731e-02 2.71244079e-01 -2.74962604e-01 -2.95371115e-01
3.26083720e-01 -1.42696396e-01 3.08174044e-01 6.65211439e-01
-2.75491685e-01 4.56271648e-01 3.45021099e-01 6.34004951e-01
8.61968935e-01 -4.74576563e-01 4.02172357e-01 -6.24715090e-01
1.15435147e+00 5.36483005e-02 9.58529830e-01 4.90245551e-01
-3.93089890e-01 4.59925473e-01 1.93149909e-01 -1.63309738e-01
-7.25093305e-01 -1.03900385e+00 -2.52371758e-01 7.21155107e-01
1.23906709e-01 -4.21765000e-01 -7.10775435e-01 -6.72469079e-01
-2.16624495e-02 -5.72211146e-02 -2.23485991e-01 1.25158370e-01
-6.94056571e-01 -1.31892395e+00 3.52162495e-02 4.83966440e-01
5.50143659e-01 -2.97903180e-01 9.32105184e-02 3.18669677e-01
-5.84751785e-01 -8.62272203e-01 -8.84194851e-01 -1.70492172e-01
-1.23861361e+00 -9.55910087e-01 -8.31378281e-01 -1.34393966e+00
1.36010778e+00 7.81783998e-01 7.02658832e-01 1.72440290e-01
-1.86599344e-01 5.54266274e-01 -8.28313008e-02 2.64208257e-01
4.78320979e-02 -8.88691023e-02 4.97819275e-01 5.30861735e-01
3.18578988e-01 -5.53955019e-01 -5.66630721e-01 2.65608788e-01
-7.73866296e-01 -5.40075526e-02 6.49289608e-01 1.16980493e+00
9.65514481e-01 5.52913129e-01 2.59906441e-01 -1.17981267e+00
1.14451790e+00 -4.53920662e-01 -8.30221534e-01 1.93932846e-01
-8.58560026e-01 1.82536505e-02 4.01450098e-01 -4.38136727e-01
-1.03644240e+00 6.90079927e-01 5.82640588e-01 -5.67118168e-01
4.61284637e-01 1.01370072e+00 -2.48948559e-01 -2.53202766e-01
2.36620054e-01 6.77703619e-01 4.54095937e-02 -6.85179412e-01
4.92017448e-01 7.63281226e-01 4.87931222e-01 -8.66710901e-01
1.38751090e+00 1.00101960e+00 -9.97215286e-02 -9.54345047e-01
-6.32725954e-01 -1.19990790e+00 -7.44620264e-01 -2.91601866e-01
3.72952938e-01 -9.86351430e-01 -6.29024565e-01 5.07251732e-02
-9.55922365e-01 1.72949627e-01 1.67701110e-01 8.02758634e-01
-4.65344310e-01 1.04403400e+00 -8.25392008e-01 -1.09718227e+00
-3.40584874e-01 -7.19232857e-01 4.02224571e-01 -1.16022890e-02
3.09746057e-01 -1.13938951e+00 4.29785758e-01 5.92616081e-01
-1.97304770e-01 3.13732386e-01 2.15842783e-01 -3.38511199e-01
-2.78298736e-01 -8.99711102e-02 -4.38621730e-01 2.82347500e-01
1.32926702e-01 8.18071291e-02 -2.77670085e-01 -8.08927596e-01
2.65159518e-01 2.04226255e-01 9.12745476e-01 2.47469202e-01
9.46042180e-01 -6.19828403e-01 -5.84062994e-01 6.60646319e-01
1.70127988e+00 -2.06572846e-01 1.65263563e-01 2.42414936e-01
9.03020144e-01 4.32072967e-01 7.79484808e-01 8.10160995e-01
2.34930590e-01 1.16332099e-01 2.00660393e-01 -1.05259735e-02
6.56364024e-01 3.04730181e-02 2.98433572e-01 1.64222205e+00
-2.87178814e-01 4.80093896e-01 -1.01727605e+00 6.59663141e-01
-2.40716243e+00 -1.14473712e+00 -8.51839185e-01 2.18664527e+00
9.45400238e-01 -1.46708578e-01 3.43250632e-01 4.76707876e-01
1.15848410e+00 -2.04235718e-01 -2.91977763e-01 -2.29252815e-01
-1.33857280e-01 -4.00466844e-02 6.07333958e-01 6.28584743e-01
-1.37539840e+00 6.96739495e-01 7.12914610e+00 9.33177114e-01
-5.97284377e-01 2.07946613e-01 1.75054535e-01 1.36445582e-01
-1.13620259e-01 4.32685539e-02 -6.09460831e-01 3.27294469e-01
4.33110416e-01 -4.52469707e-01 6.68677568e-01 9.87614930e-01
4.66178477e-01 3.17690194e-01 -7.67818630e-01 1.30169511e+00
3.69756669e-01 -1.22421563e+00 -5.79686761e-02 3.21857482e-01
1.26308429e+00 -2.81235695e-01 -3.97296324e-02 -1.18963145e-01
1.83448285e-01 -5.84542334e-01 3.15982491e-01 2.07559377e-01
3.90240043e-01 -9.88650024e-01 3.48699242e-01 4.76911843e-01
-1.44635618e+00 -1.81484640e-01 -7.99990833e-01 -4.24585640e-01
2.79127657e-01 8.68920565e-01 -6.33502424e-01 3.14015836e-01
2.10648999e-01 9.14633155e-01 -2.94428080e-01 1.50377572e+00
-3.00843656e-01 8.77428710e-01 -3.82798076e-01 7.77943581e-02
5.14433563e-01 -9.96078134e-01 8.83977234e-01 1.37099171e+00
-1.54098213e-01 4.47878182e-01 6.01688027e-01 7.04509199e-01
-1.90962046e-01 4.90461558e-01 -3.04283470e-01 -2.26505324e-01
4.62111056e-01 1.41965413e+00 -9.11611021e-01 -2.53240407e-01
-6.58113956e-01 1.16636753e+00 4.15822446e-01 5.47375500e-01
-7.74929047e-01 -3.28559697e-01 3.64393502e-01 -1.81065112e-01
3.96325082e-01 -4.78425890e-01 -1.65592760e-01 -1.34197223e+00
2.05908909e-01 -8.35827291e-01 7.26251900e-01 -1.00024126e-01
-1.60005140e+00 -6.09095693e-02 -2.01001436e-01 -1.35306466e+00
4.05649811e-01 -4.50360775e-01 -7.85060465e-01 6.48333132e-01
-1.38672972e+00 -7.54916310e-01 -8.10865983e-02 8.88691485e-01
3.88196170e-01 -7.71157593e-02 4.39743966e-01 5.48269629e-01
-1.08361423e+00 4.90670323e-01 8.33467603e-01 4.31446344e-01
3.26590210e-01 -1.45766640e+00 -3.95697862e-01 1.21086109e+00
3.79494391e-02 8.92174423e-01 6.73745811e-01 -4.96271759e-01
-1.34654880e+00 -1.06364512e+00 7.69253552e-01 1.94110706e-01
9.96918917e-01 -5.37464628e-03 -8.68160307e-01 6.83292985e-01
2.08139673e-01 9.04442184e-03 1.15242434e+00 4.15731147e-02
-2.24704280e-01 9.68400538e-02 -1.15558791e+00 3.09077531e-01
1.01695311e+00 -4.97177750e-01 -5.42755961e-01 9.62164640e-01
3.03898126e-01 -8.65823179e-02 -7.98805833e-01 1.83506176e-01
2.93239892e-01 -4.80707467e-01 1.21752965e+00 -6.27648234e-01
1.72012404e-01 -6.33514822e-01 -1.41322255e-01 -1.00145769e+00
-8.58183980e-01 -9.11403477e-01 -4.62543666e-01 1.24064052e+00
3.04042697e-01 -6.57362938e-01 9.77408111e-01 4.39686805e-01
3.94482076e-01 -6.28083110e-01 -7.15680182e-01 -1.12415385e+00
-6.44196495e-02 -6.44140784e-03 -1.47104353e-01 1.53030121e+00
5.88593721e-01 4.11286384e-01 -6.26117349e-01 4.40205157e-01
1.37215304e+00 2.19842747e-01 4.45817769e-01 -1.39572668e+00
-4.33924735e-01 -4.31567013e-01 -2.07465887e-01 -1.18262398e+00
1.02300964e-01 -1.05559754e+00 -2.42408589e-02 -1.60473907e+00
9.18197095e-01 -2.37728834e-01 -7.89022148e-01 -3.32102664e-02
-5.24283350e-01 2.82692730e-01 -6.48864657e-02 5.69964707e-01
-8.98464799e-01 6.40756667e-01 1.07669961e+00 -3.02061141e-01
-2.87204146e-01 -2.20520478e-02 -5.68694293e-01 8.54452491e-01
7.47174680e-01 -6.38183296e-01 -3.22216541e-01 -4.15389836e-01
2.25298628e-01 -1.12142734e-01 -9.18517560e-02 -7.96725452e-01
3.93414855e-01 -4.27984506e-01 6.31278008e-02 -8.00855994e-01
1.02259636e-01 -8.42453778e-01 6.56979159e-02 8.42422485e-01
-1.35034427e-01 -8.13259110e-02 -3.03824484e-01 9.70308423e-01
-2.45919272e-01 -4.05318379e-01 7.83647478e-01 6.93749934e-02
-5.54021478e-01 4.23559815e-01 -5.09923637e-01 1.98256403e-01
9.78348613e-01 -2.96138793e-01 2.41471156e-01 -5.92965856e-02
-7.88638055e-01 6.53198779e-01 2.05174863e-01 -1.17536336e-01
9.38724101e-01 -1.52785742e+00 -9.81537223e-01 -1.29507616e-01
-1.94143765e-02 -2.20989913e-01 -8.80218744e-02 1.33615029e+00
-5.76531947e-01 3.72325927e-01 2.51432538e-01 -6.07109010e-01
-1.45871949e+00 6.96210384e-01 -3.34245414e-02 -1.56423926e-01
-6.08388364e-01 9.51786995e-01 1.87738258e-02 -4.95408803e-01
2.57054001e-01 2.34267041e-01 -2.98086584e-01 1.56412367e-02
5.93397260e-01 7.83106387e-01 -4.76703733e-01 -6.75495327e-01
-6.57510221e-01 4.42498565e-01 4.71448936e-02 -6.19471213e-03
1.51985407e+00 -2.54340827e-01 -6.75079048e-01 2.21769199e-01
1.28287089e+00 1.09280035e-01 -1.03963900e+00 -5.78653514e-01
1.74101278e-01 -7.08244264e-01 1.47644132e-01 6.42603710e-02
-1.16235971e+00 5.50678015e-01 3.22397262e-01 -2.54246444e-02
1.04450357e+00 -3.94875258e-01 6.99329734e-01 7.16263890e-01
3.71607125e-01 -1.37541807e+00 8.10389034e-03 2.86984861e-01
5.92165232e-01 -1.29536331e+00 4.15906101e-01 -7.52400219e-01
-4.00579721e-01 8.60271811e-01 4.16112930e-01 -4.27271873e-01
7.20677853e-01 -1.00194275e-01 -2.53981560e-01 -7.47607201e-02
-5.91808975e-01 -1.92726091e-01 1.02556631e-01 5.93082011e-01
3.34222257e-01 1.50125235e-01 -1.07620788e+00 3.89678776e-01
3.06657076e-01 -2.16716975e-01 7.07111180e-01 8.54061961e-01
-4.66912627e-01 -9.85775411e-01 -7.19346046e-01 2.98817396e-01
-2.50872076e-01 -1.69825882e-01 -5.00193596e-01 3.59226495e-01
-2.09206626e-01 1.07266319e+00 -3.33395839e-01 -9.77320783e-03
-1.37127459e-01 -1.09229088e-01 4.29593056e-01 -6.55547023e-01
-1.38202935e-01 3.98664802e-01 -1.96018964e-01 -1.72530040e-01
-6.68684661e-01 -8.65811348e-01 -1.29921675e+00 -2.84704059e-01
-7.14866042e-01 6.57257318e-01 3.51266026e-01 7.78878272e-01
-5.91058508e-02 6.04020953e-02 1.05483890e+00 -6.39237463e-01
-7.21811414e-01 -5.84751785e-01 -8.00362647e-01 3.23992968e-01
-4.37256061e-02 -5.95645010e-01 -6.21875226e-01 2.71062255e-01] | [7.56760311126709, 4.466563701629639] |
693c0faa-61ed-4d80-9314-27fdd9230a09 | automatic-diagnosis-of-knee-osteoarthritis | 2307.04442 | null | https://arxiv.org/abs/2307.04442v1 | https://arxiv.org/pdf/2307.04442v1.pdf | Automatic diagnosis of knee osteoarthritis severity using Swin transformer | Knee osteoarthritis (KOA) is a widespread condition that can cause chronic pain and stiffness in the knee joint. Early detection and diagnosis are crucial for successful clinical intervention and management to prevent severe complications, such as loss of mobility. In this paper, we propose an automated approach that employs the Swin Transformer to predict the severity of KOA. Our model uses publicly available radiographic datasets with Kellgren and Lawrence scores to enable early detection and severity assessment. To improve the accuracy of our model, we employ a multi-prediction head architecture that utilizes multi-layer perceptron classifiers. Additionally, we introduce a novel training approach that reduces the data drift between multiple datasets to ensure the generalization ability of the model. The results of our experiments demonstrate the effectiveness and feasibility of our approach in predicting KOA severity accurately. | ['Rachid Jennane', 'Alessandro Bruno', 'Aladine Chetouani', 'Yassine Nasser', 'Mohamed Amine Kerkouri', 'Marouane Tliba', 'Aymen Sekhri'] | 2023-07-10 | null | null | null | null | ['management'] | ['miscellaneous'] | [-4.59491670e-01 -3.02081555e-01 -5.48070252e-01 -1.30937710e-01
-9.06306684e-01 1.88492432e-01 -2.37425920e-02 -9.88167431e-03
-4.26056176e-01 8.56284440e-01 5.14542460e-01 -5.74116930e-02
-6.06456339e-01 -5.98416388e-01 -3.44885677e-01 -4.38842297e-01
-6.05230927e-01 5.55604100e-01 6.68322265e-01 -1.95501023e-03
1.45204157e-01 3.13000917e-01 -1.41297698e+00 5.85328460e-01
6.88935578e-01 1.19298458e+00 3.35738212e-01 5.30808508e-01
4.81452942e-01 1.04173243e+00 -5.45189857e-01 1.21751130e-01
1.85796142e-01 1.10460103e-01 -8.07982862e-01 -2.83005834e-01
9.21228155e-03 -5.59322119e-01 -1.74215108e-01 1.96879491e-01
6.88194454e-01 -3.21305424e-01 6.52603924e-01 -9.70879555e-01
-8.47499594e-02 9.97693911e-02 -3.77390623e-01 2.80716896e-01
2.40455762e-01 2.17331827e-01 7.80977011e-01 -9.11213279e-01
4.37031686e-01 7.09192038e-01 1.03895676e+00 3.51847231e-01
-1.13202095e+00 -7.06756413e-01 -3.11606556e-01 5.52983224e-01
-1.17823827e+00 -9.56900716e-02 6.42945588e-01 -6.25775158e-01
8.44934165e-01 2.41730720e-01 1.21515834e+00 7.49724984e-01
1.25498450e+00 9.07560110e-01 1.30221653e+00 -2.39436686e-01
4.60829288e-01 -4.30552870e-01 2.30561405e-01 4.99921173e-01
2.12808639e-01 1.00028567e-01 -5.88377237e-01 -3.81869227e-01
7.30116963e-01 3.49532515e-01 -4.07700837e-02 -4.52102482e-01
-8.89876783e-01 7.28208482e-01 5.39218128e-01 2.64103357e-02
-5.81665397e-01 2.48254851e-01 4.56791550e-01 1.18126832e-01
6.90857619e-02 4.45572734e-01 -5.02497852e-01 -2.02016518e-01
-1.04615533e+00 4.35679674e-01 5.24614096e-01 2.78724104e-01
-1.85688302e-01 -3.21861714e-01 -8.43964070e-02 1.19239163e+00
5.09777606e-01 4.04425442e-01 1.15548694e+00 -9.57753241e-01
-1.36110842e-01 7.49782383e-01 -1.75442398e-01 -6.09582245e-01
-5.79771578e-01 -4.98079747e-01 -5.68362772e-01 6.05443358e-01
1.50542438e-01 3.18910003e-01 -1.24951792e+00 1.27700090e+00
6.92711174e-02 2.21123584e-02 -2.56022692e-01 9.90047097e-01
3.29602569e-01 1.21219531e-01 2.07756564e-01 -1.03551321e-01
1.52265668e+00 -9.93018985e-01 -5.43858647e-01 -4.50599909e-01
7.57247984e-01 -6.31928027e-01 1.03019130e+00 6.21775687e-01
-9.25768077e-01 -2.34602615e-01 -9.12030160e-01 -2.08089724e-02
-9.88729820e-02 2.58446813e-01 9.16734695e-01 2.56180465e-01
-6.49288714e-01 4.52151686e-01 -1.22214758e+00 -1.06825277e-01
1.39362752e-01 7.30182767e-01 -5.83333910e-01 -2.73598209e-02
-1.24736774e+00 1.16416633e+00 3.25393677e-01 1.79094344e-01
-5.45349538e-01 -5.81950605e-01 -4.78698075e-01 -2.24325702e-01
1.02851346e-01 -9.03124034e-01 1.28665781e+00 -3.10687780e-01
-1.42021155e+00 5.19969046e-01 -1.49968907e-01 -4.89257872e-01
3.28091413e-01 -8.97189319e-01 -1.66281119e-01 -2.34807655e-02
6.92716464e-02 2.50501454e-01 3.35007757e-01 -7.54294813e-01
-7.03911304e-01 -5.27119279e-01 -5.07003546e-01 -4.26917523e-02
-1.68758526e-01 -1.28470004e-01 -2.42834702e-01 -7.68768966e-01
1.88237131e-01 -1.09189439e+00 -6.11626089e-01 1.11109525e-01
-1.88494623e-01 -2.08594128e-01 7.07641602e-01 -1.11106217e+00
1.62364101e+00 -2.05157971e+00 -5.36981337e-02 4.43638593e-01
3.63118947e-01 4.43543606e-02 4.55767393e-01 2.73936182e-01
-8.32640007e-02 -2.30672434e-01 -8.50515142e-02 1.10604778e-01
-6.06230855e-01 5.04038334e-01 1.72993258e-01 1.83049187e-01
1.61368653e-01 7.52345324e-01 -4.72417474e-01 -6.67320549e-01
2.52164394e-01 3.57950717e-01 -7.25989044e-01 3.78114343e-01
2.03331277e-01 2.05610186e-01 -3.98230284e-01 7.83964992e-01
-8.81485119e-02 -4.95769650e-01 1.05485775e-01 -1.32675424e-01
2.27527499e-01 2.47013137e-01 -1.07874906e+00 1.34071147e+00
-5.76100826e-01 2.10318089e-01 -1.42467365e-01 -7.19603717e-01
7.60407746e-01 2.55644709e-01 8.33300769e-01 -8.52296591e-01
3.89059149e-02 7.34287143e-01 2.23950297e-01 -7.92875886e-01
5.71024138e-03 -1.05381437e-01 -1.18002690e-01 1.66475832e-01
-3.40138108e-01 2.37996802e-01 7.88874924e-02 -5.85683584e-02
1.23622441e+00 -5.60347401e-02 6.69376731e-01 -1.04420073e-01
3.35061699e-01 9.56472382e-02 7.75027990e-01 6.48148060e-01
-3.82808000e-01 6.06658280e-01 2.67011195e-01 -5.83133101e-01
-1.06262636e+00 -1.33783138e+00 -2.76560217e-01 5.42592943e-01
-3.80465180e-01 -3.65145922e-01 -2.07026094e-01 -2.36487448e-01
5.26839733e-01 1.45081043e-01 -4.72554982e-01 -5.43451726e-01
-8.93332899e-01 -7.89914429e-01 1.66994289e-01 9.65659261e-01
1.63505942e-01 -7.87765443e-01 -7.33566821e-01 4.68068480e-01
-1.91738352e-01 -4.70028639e-01 1.74149852e-02 2.50921160e-01
-1.37277412e+00 -1.23910511e+00 -9.20743704e-01 -1.09265471e+00
2.35156596e-01 -1.98951155e-01 4.82029378e-01 1.34808689e-01
-6.90376163e-01 3.18808526e-01 1.28293306e-01 -1.03107385e-01
-1.92820445e-01 2.25032330e-01 2.35155642e-01 -5.43303072e-01
2.15147838e-01 -5.57146490e-01 -1.20577133e+00 2.47485995e-01
-4.61326361e-01 3.58393826e-02 1.38295794e+00 8.58584762e-01
7.43532658e-01 -1.51748165e-01 5.19437194e-01 -4.29108053e-01
9.13209319e-01 -6.75664008e-01 -6.61492273e-02 1.46672443e-01
-1.20171201e+00 1.96674079e-01 3.12217832e-01 -4.59789753e-01
-7.14322269e-01 7.51609579e-02 -9.46479514e-02 -1.72726497e-01
2.05035880e-01 9.42225695e-01 3.78177166e-01 -4.15558107e-02
2.17769697e-01 1.15415961e-01 4.67385292e-01 -8.48510563e-01
-1.50265291e-01 1.07323539e+00 7.07186162e-01 -3.69041830e-01
1.20293267e-01 2.30704159e-01 2.21693575e-01 -5.46853840e-01
-4.84972596e-01 -7.12857068e-01 -3.42466325e-01 -5.11586607e-01
4.54467833e-01 -1.00344121e+00 -6.51282430e-01 6.43122673e-01
-5.05175412e-01 -1.55985326e-01 6.48212433e-02 1.00854993e+00
-5.12388468e-01 2.74246812e-01 -1.14292216e+00 -5.76144636e-01
-7.34599352e-01 -1.32803679e+00 1.02202952e+00 -8.88652578e-02
-7.04572439e-01 -6.90949261e-01 4.76265907e-01 5.80220044e-01
4.68423247e-01 1.11134917e-01 1.03087723e+00 -5.69226325e-01
-4.02657211e-01 -6.01054430e-01 1.30011812e-01 4.24820989e-01
4.48155068e-02 -2.58648902e-01 -2.12209389e-01 -2.32823595e-01
-1.35079883e-02 -3.26404721e-01 6.23424172e-01 6.17484391e-01
1.12181652e+00 -1.19099557e-01 -5.42280674e-01 1.23463973e-01
1.31213963e+00 1.95316836e-01 7.18621612e-01 9.68817651e-01
3.23543012e-01 2.97849715e-01 7.89656103e-01 2.60641187e-01
3.38112891e-01 7.66782880e-01 1.90678656e-01 9.92913172e-02
-4.17003006e-01 5.66294976e-02 2.06236750e-01 1.15859246e+00
-3.92403722e-01 3.73117387e-01 -1.22778702e+00 6.56352401e-01
-1.85807061e+00 -5.72499990e-01 -2.42330268e-01 2.06451416e+00
1.06675053e+00 4.93292540e-01 -1.31688882e-02 2.76932210e-01
1.51952416e-01 -3.27272147e-01 -1.30197242e-01 -2.69881040e-01
4.35285628e-01 4.86379355e-01 6.70425117e-01 3.02481711e-01
-7.94393837e-01 3.74411732e-01 7.71309614e+00 4.13405329e-01
-1.23127520e+00 7.08190799e-02 1.80602670e-01 -4.32817996e-01
2.18343481e-01 -7.87335634e-02 -3.96066457e-01 7.34318316e-01
9.26661551e-01 -4.81266491e-02 -2.40720779e-01 1.14522123e+00
5.10362864e-01 -5.63167930e-01 -7.94001281e-01 4.70459551e-01
-1.44029692e-01 -1.11650109e+00 -2.26087317e-01 1.04710095e-01
1.63364917e-01 3.95464711e-02 -4.07205410e-02 1.24581560e-01
2.19457224e-01 -8.01253319e-01 3.01787704e-01 9.55260873e-01
4.73812163e-01 -4.51085329e-01 1.14893878e+00 -8.10805336e-02
-1.04846001e+00 -3.33346605e-01 3.07826340e-01 -4.67702031e-01
2.77449489e-01 7.13473618e-01 -1.06002545e+00 1.42656136e-02
9.55955327e-01 4.11947697e-01 -6.83975756e-01 1.69887006e+00
-2.10673675e-01 6.39511526e-01 -5.96247733e-01 5.28779440e-02
-2.04758465e-01 3.23345363e-01 3.35107028e-01 5.44929624e-01
2.84543931e-01 -1.44080445e-01 4.62400556e-01 2.92600766e-02
4.46891069e-01 1.93008393e-01 -1.82773367e-01 3.88997823e-01
3.78766298e-01 5.04996300e-01 -2.41241321e-01 4.79491316e-02
-5.88774621e-01 5.37063479e-01 1.71489924e-01 4.18826751e-02
-4.37089056e-01 -9.12100598e-02 4.31206465e-01 4.98352706e-01
4.37654695e-03 -2.92019129e-01 -4.36843306e-01 -7.14134395e-01
3.55055690e-01 -6.53208554e-01 6.32936180e-01 -7.99644351e-01
-1.17951679e+00 -7.24589303e-02 -4.29429412e-02 -1.30937123e+00
-5.66694796e-01 -7.14494705e-01 -3.76918674e-01 5.21933079e-01
-1.06596601e+00 -1.32230616e+00 -1.03124892e-02 2.49452800e-01
3.65705341e-01 4.01339605e-02 6.43024504e-01 3.43895614e-01
-5.59773207e-01 7.90880397e-02 2.29208350e-01 1.76368281e-01
9.62351501e-01 -1.11800098e+00 -4.19347703e-01 3.72155726e-01
-8.65177691e-01 7.30409324e-01 7.22665012e-01 -1.17047405e+00
-1.14654291e+00 -6.82321072e-01 8.35897088e-01 -6.97563291e-02
9.23311234e-01 2.43202493e-01 -7.96356201e-01 7.13454604e-01
-4.55953002e-01 -3.55941020e-02 1.12372339e+00 2.18035638e-01
6.93918541e-02 -2.43349895e-01 -7.24156797e-01 5.09291351e-01
4.81721669e-01 -2.37171873e-01 -1.15363288e+00 5.11249490e-02
7.70093054e-02 -2.92560309e-01 -1.54878342e+00 8.50167572e-01
1.34865463e+00 -6.17973328e-01 1.10180736e+00 -6.57866895e-01
6.13107264e-01 -2.61002989e-03 3.46077047e-02 -8.90622199e-01
-4.87507045e-01 3.43255788e-01 -4.44140017e-01 4.67040956e-01
2.26461008e-01 -4.56184626e-01 9.50381458e-01 7.32158542e-01
-1.94959566e-01 -1.27696335e+00 -1.16333306e+00 -7.60330498e-01
3.54802497e-02 -4.29821074e-01 -1.44862374e-02 4.33247805e-01
2.48457551e-01 6.80364817e-02 -5.64899325e-01 -5.97185083e-02
5.48700511e-01 -1.33003891e-02 4.86182779e-01 -1.51909530e+00
-3.11905086e-01 -1.92350477e-01 -8.52025509e-01 -2.84791678e-01
-4.94212300e-01 -7.15144217e-01 -1.27111241e-01 -1.82544172e+00
4.93319184e-01 -3.08230489e-01 -6.80846334e-01 4.92056996e-01
-1.69752855e-02 3.62641901e-01 -9.01838243e-02 8.19596827e-01
-2.89398342e-01 3.36269110e-01 1.03964341e+00 6.96775168e-02
-3.89817804e-01 1.05148457e-01 -1.47876605e-01 6.49732888e-01
8.29358876e-01 -6.33433759e-01 -2.59111464e-01 -1.65402055e-01
1.29405528e-01 -3.55158485e-02 7.28027940e-01 -1.32157791e+00
1.94877625e-01 -1.38759717e-01 7.53506303e-01 -6.45967841e-01
3.30787152e-01 -7.61871636e-01 2.72027463e-01 1.46228969e+00
-6.20680973e-02 -6.57721087e-02 -5.76130562e-02 4.68204230e-01
-3.03225845e-01 9.07078758e-02 6.73292577e-01 -4.48986441e-02
-7.89467692e-01 -4.40259948e-02 -8.02369595e-01 -3.02972764e-01
1.27355456e+00 -2.03209087e-01 -3.62478793e-02 -6.32360205e-02
-1.16315651e+00 5.10565281e-01 1.37851104e-01 6.38224721e-01
8.68520081e-01 -1.40586758e+00 -3.15835446e-01 -2.69116703e-02
2.14512840e-01 -2.63707966e-01 5.63763529e-02 1.26564276e+00
-6.30368054e-01 5.72007477e-01 -4.47353035e-01 -7.12506652e-01
-1.51509476e+00 1.26008302e-01 4.25521791e-01 -6.92905009e-01
-8.43639493e-01 3.96947980e-01 -5.52368522e-01 -6.97752386e-02
4.07774091e-01 -3.77439290e-01 -2.66348153e-01 -3.10525775e-01
4.82926935e-01 4.66110319e-01 6.63926974e-02 -4.71999161e-02
-6.14283085e-01 2.83599287e-01 -5.11726081e-01 1.84514876e-02
1.40069437e+00 -7.17241615e-02 -2.47671410e-01 6.56716526e-01
9.92334068e-01 -5.92189729e-02 -7.90805221e-01 -3.17170285e-02
2.70414740e-01 -2.88574934e-01 4.12954777e-01 -1.05978680e+00
-8.45117807e-01 4.70753908e-01 1.29578888e+00 -2.69310504e-01
9.37775612e-01 2.98583686e-01 1.35755408e+00 3.38868737e-01
3.23302776e-01 -1.24824405e+00 7.17337802e-02 -6.77075014e-02
7.52581179e-01 -1.12281501e+00 2.89337009e-01 -3.03509325e-01
-2.07635105e-01 1.18013287e+00 6.18551850e-01 -3.80756825e-01
8.12181652e-01 3.66910249e-01 2.41308331e-01 -2.05973119e-01
-6.88756108e-01 9.41026509e-02 2.99780756e-01 3.14887553e-01
5.77629089e-01 2.04084426e-01 -1.09560800e+00 9.76697683e-01
-2.04618320e-01 7.75095582e-01 6.43751696e-02 1.55984986e+00
-6.58339024e-01 -1.10672271e+00 -4.79476094e-01 1.10247779e+00
-6.51204228e-01 2.46994108e-01 -1.43934011e-01 1.02781105e+00
-2.34416768e-01 3.40477884e-01 -6.97324127e-02 -3.81787300e-01
4.62439656e-01 2.65405655e-01 2.05925927e-01 -3.04634511e-01
-2.65502304e-01 2.71550179e-01 3.11675757e-01 -8.06496918e-01
-2.29778662e-01 -5.70126951e-01 -1.35891271e+00 1.13225959e-01
-1.15005799e-01 -6.43406585e-02 5.09977460e-01 1.12735975e+00
2.92471558e-01 7.40925908e-01 3.93184364e-01 -1.81267932e-01
-6.30222499e-01 -9.21734452e-01 -6.31615818e-01 1.98032379e-01
2.83524066e-01 -1.10233676e+00 -2.08975762e-01 -1.19602725e-01] | [14.551765441894531, -1.7580856084823608] |
f022e16d-a2eb-4cac-a506-00a0cdcb7847 | variational-approach-for-intensity-domain | 2207.04204 | null | https://arxiv.org/abs/2207.04204v1 | https://arxiv.org/pdf/2207.04204v1.pdf | Variational Approach for Intensity Domain Multi-exposure Image Fusion | Recent innovations shows that blending of details captured by single Low Dynamic Range (LDR) sensor overcomes the limitations of standard digital cameras to capture details from high dynamic range scene. We present a method to produce well-exposed fused image that can be displayed directly on conventional display devices. The ambition is to preserve details in poorly illuminated and brightly illuminated regions. Proposed approach does not require true radiance reconstruction and tone manipulation steps. The aforesaid objective is achieved by taking into account local information measure that select well-exposed regions across input exposures. In addition, Contrast Limited Adaptive Histogram equalization (CLAHE) is introduced to improve uniformity of input multi-exposure image prior to fusion. | ['Vinay Kumar', 'Dinesh Arora', 'Harbinder Singh'] | 2022-07-09 | null | null | null | null | ['multi-exposure-image-fusion'] | ['computer-vision'] | [ 9.94198620e-01 -5.77933371e-01 4.42293137e-01 -4.27622110e-01
-6.59065962e-01 -7.46310711e-01 5.01010954e-01 -1.58803120e-01
-4.93799090e-01 1.02185762e+00 2.16113225e-01 -1.42911859e-02
-2.14251176e-01 -7.69164920e-01 -4.78888154e-01 -8.47743571e-01
5.91892779e-01 -4.93414521e-01 4.39708054e-01 -3.56248856e-01
3.22100937e-01 6.90100312e-01 -1.86652052e+00 3.13365549e-01
6.97306693e-01 9.14803863e-01 5.00326753e-01 1.05466592e+00
9.23874006e-02 5.34804940e-01 -6.14106417e-01 -2.13186480e-02
5.86123526e-01 -7.21831679e-01 -1.10838532e-01 3.02710861e-01
9.01916027e-01 -5.54007292e-01 -4.73888852e-02 1.15026069e+00
5.15519559e-01 3.33988190e-01 6.82177424e-01 -7.07797050e-01
-7.86828816e-01 -1.44149765e-01 -1.01013899e+00 7.00688779e-01
6.66221321e-01 2.48278812e-01 1.69116095e-01 -6.73323452e-01
4.51141834e-01 8.17381859e-01 3.72583866e-01 1.45145357e-01
-1.25469649e+00 -5.84219277e-01 -3.94672662e-01 -2.47699738e-01
-1.17784989e+00 -5.45453131e-01 1.08527935e+00 -3.49971578e-02
8.15394402e-01 7.12363780e-01 4.23958510e-01 5.20831585e-01
7.03610659e-01 -3.48491520e-01 2.28185558e+00 -5.04531622e-01
-1.24128126e-01 5.07519007e-01 9.45116878e-02 5.16048849e-01
5.57174385e-01 5.26093483e-01 -5.61985552e-01 1.46792307e-01
7.98372090e-01 1.79837316e-01 -7.86813557e-01 1.36017561e-01
-9.18896258e-01 6.92049414e-02 -5.83521239e-02 6.40425980e-01
-3.24219316e-01 -3.77452195e-01 -1.69737816e-01 4.40413505e-01
1.03210732e-01 1.99670613e-01 8.82718638e-02 1.12605721e-01
-8.50593746e-01 -3.55227470e-01 4.71848361e-02 5.39743483e-01
8.56454909e-01 1.19435824e-01 -1.15178466e-01 5.51168501e-01
1.51327074e-01 9.61415410e-01 2.60084808e-01 -8.48005772e-01
1.98448211e-01 2.57104754e-01 2.50644535e-01 -8.75166714e-01
8.86251703e-02 -2.72435278e-01 -8.04068565e-01 1.12011015e+00
2.01626569e-01 -9.54238996e-02 -1.04549944e+00 1.20561981e+00
1.58068925e-01 1.29079269e-02 3.73637348e-01 1.12142813e+00
5.95919609e-01 9.11294758e-01 -2.80691147e-01 -5.50793648e-01
1.28099453e+00 -3.21424216e-01 -1.10981941e+00 1.03576824e-01
-5.56088448e-01 -1.36466134e+00 1.16208196e+00 7.64057398e-01
-1.54508555e+00 -9.26458716e-01 -1.41051090e+00 -2.39901140e-01
-2.31535122e-01 -6.00711219e-02 7.82832131e-02 9.56015468e-01
-9.60270345e-01 2.29930520e-01 -1.12797111e-01 -1.06454611e-01
-7.66897202e-02 2.95933813e-01 -5.74789405e-01 -8.51431191e-02
-7.81201184e-01 9.59439516e-01 1.59123003e-01 -1.22914217e-01
-3.88415486e-01 -6.82974339e-01 -5.33151746e-01 -8.98506716e-02
1.81983322e-01 -5.76312780e-01 5.45874596e-01 -1.03073692e+00
-1.98126268e+00 1.05564833e+00 -8.93709213e-02 -2.23837033e-01
2.53379434e-01 -1.27838761e-01 -7.70637631e-01 4.70117718e-01
-5.36216438e-01 -1.02633378e-02 9.48741257e-01 -1.48354864e+00
-5.16205490e-01 -2.16885999e-01 -4.15522516e-01 8.00701261e-01
5.49619123e-02 1.44172415e-01 -4.90588807e-02 -3.28621417e-01
7.36087114e-02 -1.38036296e-01 1.79877013e-01 3.06375921e-02
-6.98013827e-02 6.27718866e-01 1.19688427e+00 -6.88887894e-01
1.15602040e+00 -2.05722475e+00 -5.69429755e-01 8.71193781e-02
-5.20314574e-02 5.18206060e-01 3.79485369e-01 3.93402249e-01
3.74154672e-02 -3.42989564e-01 -2.55079836e-01 1.64438739e-01
-3.87134612e-01 -3.15086097e-01 -4.42063034e-01 6.37591541e-01
-9.60282311e-02 3.65436196e-01 -2.49991447e-01 -6.79384828e-01
9.52253938e-01 8.77331257e-01 7.91845545e-02 2.81086534e-01
2.39464760e-01 7.30069637e-01 3.36728096e-02 6.35849416e-01
1.25503397e+00 2.57571310e-01 -3.64179701e-01 -6.19568586e-01
-4.81568515e-01 -3.91224474e-01 -1.39353716e+00 1.24594104e+00
-4.51453686e-01 7.73269951e-01 2.40481541e-01 -2.62361556e-01
1.42045832e+00 1.30245030e-01 1.19974643e-01 -1.07164168e+00
1.99102551e-01 4.53009009e-02 -6.50673509e-01 -5.39417267e-01
6.61517620e-01 -5.35720408e-01 1.10511176e-01 2.44235858e-01
-3.77062023e-01 -1.52660370e-01 -2.70535380e-01 -2.52096832e-01
5.52851439e-01 7.74976835e-02 6.81476295e-01 -6.28618672e-02
8.63552094e-01 -2.73478925e-01 1.55806944e-01 6.28731847e-01
-1.38267756e-01 9.81483221e-01 -2.89684743e-01 -2.27729864e-02
-1.15823555e+00 -1.52668071e+00 -3.37114036e-01 7.58049130e-01
5.35226941e-01 5.12925267e-01 -1.93310112e-01 1.47731528e-01
-4.18766737e-01 5.46439707e-01 -1.78898185e-01 3.68399084e-01
-6.64511561e-01 -3.95304590e-01 1.20453574e-01 -1.46031737e-01
8.65471542e-01 -7.45713890e-01 -1.37316716e+00 -7.25251585e-02
3.06741059e-01 -9.26308095e-01 -2.55999684e-01 1.10857114e-01
-6.73566341e-01 -8.54582191e-01 -8.87020409e-01 -6.26376033e-01
4.22167629e-01 5.44408798e-01 9.90153491e-01 -3.71092379e-01
-7.73540437e-01 2.89450616e-01 -1.88444763e-01 -7.39801526e-02
-8.46198052e-02 -8.02467763e-01 -2.88686931e-01 1.87009051e-01
2.33682081e-01 -5.27530372e-01 -9.91515815e-01 3.66665214e-01
-1.03801489e+00 3.24388355e-01 8.45962644e-01 4.64499891e-01
8.43912482e-01 4.03747827e-01 3.02349180e-01 -7.17856050e-01
4.77671385e-01 1.11174464e-01 -8.82782578e-01 4.35405195e-01
-4.97326672e-01 -1.57975957e-01 7.13036537e-01 -3.11297238e-01
-1.99958098e+00 -7.30579346e-02 1.33517265e-01 -1.24354772e-01
-5.45082748e-01 -3.49644840e-01 -1.25107497e-01 -2.91426569e-01
6.49685621e-01 6.24643981e-01 -9.37753543e-02 -3.27033967e-01
2.09860325e-01 9.39724922e-01 1.20770144e+00 -1.11344449e-01
7.70661950e-01 8.43117476e-01 2.16238573e-01 -9.53173637e-01
-3.34746182e-01 -3.39232713e-01 -3.10145497e-01 -4.44343597e-01
1.04813910e+00 -8.21440637e-01 -7.35281944e-01 2.33074784e-01
-5.78923464e-01 -9.51284766e-02 -1.43155485e-01 6.33860528e-01
-4.57904011e-01 3.34865898e-01 -2.99689859e-01 -1.17903078e+00
-4.23039466e-01 -6.31295800e-01 8.74444008e-01 8.90718102e-01
2.37273052e-01 -7.93709874e-01 1.83456272e-01 3.31070721e-01
8.27476263e-01 7.39723861e-01 5.78933239e-01 3.85375589e-01
-7.86979198e-01 -4.86613810e-02 -3.62891555e-01 4.02374625e-01
4.14474159e-01 -9.03110206e-02 -1.22484612e+00 -9.89567637e-02
3.72708768e-01 -9.42495242e-02 6.47665918e-01 5.76132059e-01
7.42916882e-01 1.66642189e-01 -7.30023831e-02 6.81244016e-01
2.41528702e+00 5.51580250e-01 9.30243254e-01 1.27391577e-01
2.18968764e-01 4.26125199e-01 7.85394251e-01 1.40923098e-01
-2.29230225e-01 6.14962399e-01 -3.86313573e-02 -6.92314088e-01
-5.87462962e-01 -5.81269078e-02 3.15616399e-01 2.67960697e-01
1.76213995e-01 -5.99059820e-01 -3.26183349e-01 2.27976084e-01
-1.01271164e+00 -1.13973200e+00 -3.99631709e-01 2.25959611e+00
8.78993750e-01 3.41209397e-02 -3.54549468e-01 1.40568182e-01
9.44933891e-01 6.52084574e-02 -2.87630737e-01 -6.26864195e-01
-7.25354791e-01 5.93407750e-01 7.67375052e-01 1.04192400e+00
-6.59152925e-01 4.15266097e-01 6.30347919e+00 6.45199776e-01
-1.34668863e+00 -9.63963801e-04 5.17058432e-01 -1.57682389e-01
-5.32123029e-01 -2.29925867e-02 -7.64154911e-01 6.29239738e-01
5.96010268e-01 -2.30069876e-01 2.34162793e-01 5.76810278e-02
3.09348941e-01 -1.04871786e+00 -5.91926396e-01 1.18750978e+00
4.48450178e-01 -8.68091047e-01 -9.51308310e-02 1.68428421e-02
9.72127855e-01 -7.84351110e-01 6.94641531e-01 -5.41812897e-01
-7.47463182e-02 -1.02740824e+00 2.52846610e-02 1.24378121e+00
1.19073796e+00 -9.23892975e-01 4.89858031e-01 -9.49135795e-02
-1.02998912e+00 1.17604040e-01 -3.91997814e-01 -6.29492253e-02
4.77254152e-01 4.99421358e-01 -4.98429984e-01 4.83132213e-01
6.51209772e-01 9.81350541e-02 -4.99260336e-01 9.27383840e-01
3.47290367e-01 8.94073173e-02 -3.09717774e-01 3.58152717e-01
-1.29246793e-03 -4.87912744e-01 5.84399402e-01 1.10357857e+00
5.11497319e-01 5.39856911e-01 -2.80725420e-01 7.96062648e-01
4.55747545e-01 -7.79146105e-02 -1.03910184e+00 4.55648601e-01
2.15927780e-01 1.14081049e+00 -6.69700503e-01 -3.58699441e-01
-5.18735647e-01 1.22122478e+00 -5.28430283e-01 4.53025311e-01
-6.29432797e-01 -7.89804101e-01 1.00529857e-01 3.69083107e-01
2.51439184e-01 -1.62445739e-01 -5.47180057e-01 -6.56595528e-01
-2.09056750e-01 -6.29675627e-01 3.24039161e-01 -1.34502089e+00
-8.68909717e-01 7.16696084e-01 1.54713944e-01 -1.27245009e+00
3.08510542e-01 -3.59934807e-01 -6.30557895e-01 1.39399147e+00
-1.78217757e+00 -1.00083077e+00 -7.57163346e-01 9.63871777e-01
4.92853284e-01 1.09199844e-02 4.03075188e-01 2.43788913e-01
-7.40657151e-02 3.28689307e-01 2.73989856e-01 -6.10081017e-01
1.15300453e+00 -1.02965939e+00 -4.60656077e-01 1.38165045e+00
-3.08236480e-01 4.56888676e-01 1.08496916e+00 -6.38050735e-01
-1.14983690e+00 -7.00349331e-01 4.97020453e-01 -2.43838578e-01
-1.40806273e-01 -7.73683935e-03 -8.12077165e-01 2.72682041e-01
9.08315122e-01 -1.25624850e-01 6.54651642e-01 -8.16593826e-01
8.20237398e-02 -4.63947147e-01 -1.52995884e+00 1.99434519e-01
3.37979227e-01 -5.36402106e-01 -6.17188334e-01 -2.58734405e-01
1.92882836e-01 -4.34040844e-01 -7.60741532e-01 4.48458403e-01
6.36978209e-01 -1.60588920e+00 9.50288177e-01 4.76211220e-01
1.64026320e-01 -8.01378667e-01 -4.41156983e-01 -7.42677867e-01
1.03789426e-01 -9.91765678e-01 3.76853496e-01 1.37841606e+00
4.18393686e-02 -7.89353490e-01 4.81488049e-01 3.35121542e-01
-1.24142863e-01 -2.79155046e-01 -4.46376860e-01 -3.64884675e-01
-6.26321018e-01 3.21484149e-01 2.39024848e-01 5.96962869e-01
-4.15760994e-01 1.64502725e-01 -7.27480948e-01 4.18477803e-01
1.05786085e+00 3.52812737e-01 4.87813354e-01 -7.67186046e-01
-3.85625988e-01 1.10869221e-02 -2.15752706e-01 -5.46304107e-01
-6.02490008e-01 -2.84959793e-01 -1.38309941e-01 -1.38481867e+00
2.64459044e-01 -2.11860433e-01 -3.84513438e-01 -2.26368278e-01
-7.96086192e-02 1.02470481e+00 4.01141718e-02 -1.27818324e-02
-5.33298135e-01 1.89016014e-01 1.30875218e+00 4.65280384e-01
-3.39095414e-01 -3.20769399e-01 -5.90659201e-01 4.68193769e-01
7.16733336e-01 -1.19005395e-02 -6.34146214e-01 -2.16814697e-01
1.58651173e-01 3.52318883e-01 4.98497128e-01 -1.32024026e+00
4.07984495e-01 -1.01761028e-01 1.05592549e+00 -9.41244483e-01
4.42003727e-01 -1.10861981e+00 8.30711365e-01 1.14897646e-01
-3.35438430e-01 2.03672707e-01 8.10477883e-02 5.80617368e-01
-1.43064395e-01 -6.90687150e-02 1.36111510e+00 -1.33499712e-01
-7.01756418e-01 -4.23767209e-01 -3.01361054e-01 -4.01087463e-01
1.40328348e+00 -1.07933891e+00 -6.60488307e-01 -3.66112262e-01
-3.63983929e-01 -5.01477897e-01 1.01099229e+00 -2.14597240e-01
7.48319328e-01 -1.07662129e+00 -3.60290319e-01 5.65496325e-01
-2.19287395e-01 -4.81537730e-01 9.32508826e-01 6.52456641e-01
-8.97408247e-01 2.20226288e-01 -9.28474844e-01 -3.88600707e-01
-1.91949511e+00 5.34450591e-01 3.40070009e-01 9.33851898e-02
-9.90232527e-01 5.19216597e-01 1.25614941e-01 6.59862757e-01
-2.75523029e-02 -1.26638144e-01 -3.36021632e-01 -4.41979080e-01
7.63943255e-01 5.28674185e-01 -1.29875854e-01 -6.44324303e-01
-6.96062148e-02 1.29650652e+00 8.74416754e-02 -4.60116416e-01
1.13668561e+00 -9.92468178e-01 2.21212000e-01 5.37916243e-01
1.22175252e+00 4.19459760e-01 -1.56113982e+00 -1.16337843e-01
-7.95872509e-01 -1.07385051e+00 4.53606278e-01 -1.18973947e+00
-5.39336920e-01 8.02805007e-01 1.33492339e+00 1.46562597e-02
1.77501738e+00 -4.21549648e-01 6.44886494e-01 -5.70740625e-02
1.08268373e-01 -1.11630285e+00 2.60447800e-01 -3.12866330e-01
5.81121862e-01 -1.09348905e+00 3.06620896e-01 -3.46691698e-01
-8.94751370e-01 1.27310026e+00 4.24547136e-01 -2.24667743e-01
3.06103230e-01 6.64402604e-01 2.41534859e-01 -2.45710045e-01
-6.49460077e-01 -3.72123480e-01 2.85917938e-01 8.03428054e-01
3.10736179e-01 -4.27497983e-01 -3.23468626e-01 -1.99682757e-01
1.19653732e-01 -5.48855662e-02 8.15130532e-01 1.08366132e+00
-9.34439838e-01 -4.96951401e-01 -8.31681371e-01 2.60619015e-01
-9.11141098e-01 9.56013799e-02 -1.05042696e-01 5.71056902e-01
2.88817406e-01 1.12315941e+00 7.70747587e-02 -7.06978142e-02
1.02190152e-01 -1.15813091e-01 7.22401202e-01 -1.35525107e-01
-4.81715828e-01 4.52102691e-01 -2.01429397e-01 -3.73451561e-01
-6.69766665e-01 -2.69994080e-01 -7.00531244e-01 -1.07060470e-01
-1.71690226e-01 -1.90229833e-01 6.91499412e-01 3.20828736e-01
-2.22483440e-03 6.26339257e-01 6.52491570e-01 -4.39173043e-01
-2.39942204e-02 -6.58050299e-01 -8.61411452e-01 3.48763168e-01
8.51491332e-01 -3.44424397e-01 -5.41438162e-01 3.56899977e-01] | [10.850545883178711, -2.466388702392578] |
0a46711d-a9c4-41d8-948c-010dabaa8652 | multilingual-zero-shot-constituency-parsing | 2004.13805 | null | https://arxiv.org/abs/2004.13805v4 | https://arxiv.org/pdf/2004.13805v4.pdf | Multilingual Chart-based Constituency Parse Extraction from Pre-trained Language Models | As it has been unveiled that pre-trained language models (PLMs) are to some extent capable of recognizing syntactic concepts in natural language, much effort has been made to develop a method for extracting complete (binary) parses from PLMs without training separate parsers. We improve upon this paradigm by proposing a novel chart-based method and an effective top-K ensemble technique. Moreover, we demonstrate that we can broaden the scope of application of the approach into multilingual settings. Specifically, we show that by applying our method on multilingual PLMs, it becomes possible to induce non-trivial parses for sentences from nine languages in an integrated and language-agnostic manner, attaining performance superior or comparable to that of unsupervised PCFGs. We also verify that our approach is robust to cross-lingual transfer. Finally, we provide analyses on the inner workings of our method. For instance, we discover universal attention heads which are consistently sensitive to syntactic information irrespective of the input language. | ['Sang-goo Lee', 'Taeuk Kim', 'Bowen Li'] | 2020-04-08 | null | https://aclanthology.org/2021.findings-emnlp.41 | https://aclanthology.org/2021.findings-emnlp.41.pdf | findings-emnlp-2021-11 | ['constituency-parsing'] | ['natural-language-processing'] | [ 2.07678705e-01 4.04872239e-01 1.03456862e-02 -5.40002167e-01
-1.18625379e+00 -9.89365578e-01 6.95643127e-01 2.55009890e-01
-4.97157037e-01 7.51642764e-01 2.48424917e-01 -7.57844508e-01
9.26380083e-02 -6.29094839e-01 -8.86651039e-01 -4.00794029e-01
6.47872360e-03 4.24410343e-01 3.52948084e-02 -2.55614877e-01
1.60292983e-01 5.36479414e-01 -1.30767405e+00 3.29973549e-01
1.07356787e+00 3.84215266e-01 4.08611238e-01 6.41894162e-01
-3.16716015e-01 5.80710888e-01 -5.36108673e-01 -6.75939620e-01
7.24252826e-03 -3.01784933e-01 -1.09861255e+00 1.26496153e-02
5.34853816e-01 1.38812259e-01 1.21501654e-01 9.77718294e-01
1.22119710e-01 -2.15052888e-01 7.12521076e-01 -6.22138441e-01
-6.66675866e-01 8.79176080e-01 -1.54408470e-01 2.01319098e-01
3.21339726e-01 -1.16519101e-01 1.38404381e+00 -7.31650949e-01
7.07137406e-01 1.26545370e+00 4.32727873e-01 5.17148554e-01
-1.40197313e+00 -3.02558184e-01 4.43499595e-01 -9.09371376e-02
-1.08272040e+00 -6.35286629e-01 6.92568064e-01 -3.58322620e-01
1.31798255e+00 4.63105552e-02 9.92775261e-02 9.89767075e-01
2.45471850e-01 9.41661239e-01 1.34197974e+00 -1.13378513e+00
8.25907513e-02 4.02887732e-01 2.07590982e-01 7.04968333e-01
3.19940418e-01 -1.77065283e-01 -3.06840241e-01 4.07537818e-02
4.12929565e-01 -6.83471978e-01 -9.23357904e-02 -2.62338042e-01
-1.14471662e+00 9.25338686e-01 -1.19132705e-01 7.52564907e-01
-1.45386666e-01 -1.88003942e-01 5.55758119e-01 2.49451682e-01
4.67754155e-01 5.95519900e-01 -8.27925742e-01 1.90802198e-02
-7.35449135e-01 -3.80589962e-02 9.73324239e-01 9.75299597e-01
7.40654051e-01 -9.47458968e-02 1.28665760e-01 7.55808830e-01
-1.51174446e-03 5.39588511e-01 4.70201731e-01 -7.57321179e-01
6.77511096e-01 4.79313701e-01 -1.36597902e-01 -6.85009718e-01
-3.75679046e-01 -3.00368845e-01 -2.97646821e-01 -1.85329184e-01
5.08121669e-01 -2.81232417e-01 -6.10017717e-01 2.12785459e+00
6.04668595e-02 -4.60449278e-01 3.83726716e-01 3.55950296e-01
3.26794475e-01 4.57219988e-01 4.75400001e-01 -3.02973688e-01
1.41464424e+00 -6.76014543e-01 -5.10388017e-01 -3.34654301e-01
9.29403067e-01 -6.47409856e-01 1.20850074e+00 4.27186131e-01
-1.00083888e+00 -4.86665338e-01 -9.19533789e-01 -1.30129784e-01
-4.57291752e-01 2.99505025e-01 9.46647108e-01 9.20523286e-01
-1.04613340e+00 5.47674716e-01 -8.14610779e-01 -4.32090789e-01
7.64991641e-02 3.69482130e-01 -5.88413656e-01 2.94917263e-03
-1.10420096e+00 1.06009626e+00 5.83091497e-01 9.26893651e-02
-6.48581982e-01 -1.77919939e-01 -1.04700935e+00 -3.61994803e-02
1.90855294e-01 -3.64802808e-01 1.31388092e+00 -1.16223252e+00
-1.52806199e+00 9.61944878e-01 -4.72930551e-01 -3.30187500e-01
-5.39136827e-02 -1.85772404e-01 -4.66365904e-01 2.39275455e-01
3.54345769e-01 4.58744019e-01 5.33422947e-01 -1.14965832e+00
-7.11115539e-01 -3.45433682e-01 1.44862294e-01 5.09304516e-02
-4.92234379e-01 4.76412326e-01 -2.23322645e-01 -4.26472843e-01
-6.18508495e-02 -8.87956619e-01 -2.36107022e-01 -8.15369666e-01
-3.61413956e-01 -5.01794875e-01 2.39655271e-01 -8.61900985e-01
1.19631410e+00 -2.06606269e+00 2.40238473e-01 7.39811584e-02
-6.10816777e-02 4.15460944e-01 -2.41753131e-01 5.81119001e-01
-1.20651022e-01 3.47328007e-01 -5.18543065e-01 -4.07468438e-01
1.63882628e-01 5.73435068e-01 -3.46015513e-01 2.03250155e-01
5.62047541e-01 1.00642490e+00 -9.02442157e-01 -5.86312711e-01
4.77296486e-02 2.61567414e-01 -4.97578025e-01 3.24765518e-02
-3.05040717e-01 5.17159879e-01 -4.37322259e-01 4.10393715e-01
2.71374464e-01 -4.95193005e-02 6.67542100e-01 2.54996270e-01
-2.56718218e-01 8.78498793e-01 -8.16091120e-01 1.69190788e+00
-7.86537409e-01 4.55604494e-01 6.81003779e-02 -1.28038347e+00
6.80811763e-01 5.13433218e-01 -4.42124382e-02 -4.96760845e-01
4.76496853e-02 4.58125502e-01 1.56354696e-01 -3.93605679e-01
4.31981683e-01 -3.12364608e-01 -4.74557966e-01 4.65269506e-01
5.50732195e-01 7.85956383e-02 5.06662369e-01 1.43986046e-01
9.34496701e-01 3.10990483e-01 5.50934494e-01 -5.21506488e-01
7.86897838e-01 9.22140256e-02 4.26846027e-01 7.52768517e-01
-2.70988327e-02 1.93063915e-01 5.27815819e-01 -1.57343119e-01
-9.47181761e-01 -1.01133657e+00 -3.78150016e-01 1.26095903e+00
-5.05812705e-01 -5.34876287e-01 -9.90701675e-01 -9.67639923e-01
-2.57498145e-01 9.73346531e-01 -4.16716188e-01 2.51488984e-01
-9.77587223e-01 -8.58458877e-01 6.99056864e-01 5.22874594e-01
1.12569764e-01 -1.23301888e+00 -2.89426476e-01 4.47911859e-01
-1.40356004e-01 -1.55019915e+00 3.43427099e-02 3.82765681e-01
-9.02196527e-01 -9.38430667e-01 -3.33800912e-01 -9.31157410e-01
6.41946793e-01 -1.37990400e-01 1.27245891e+00 -2.19307050e-01
3.50889750e-02 5.42685091e-01 -4.63981479e-01 -3.40939254e-01
-8.27660441e-01 4.48660165e-01 8.45099390e-02 -8.40850472e-02
6.72412395e-01 -6.14926517e-01 -3.61311622e-02 -3.63886088e-01
-7.62930512e-01 -1.05528854e-01 7.69480765e-01 4.80751961e-01
3.37604344e-01 -1.76732212e-01 7.55905867e-01 -1.28521192e+00
7.77155280e-01 -3.67369115e-01 -5.68705916e-01 3.48058313e-01
-4.82907563e-01 4.48881954e-01 1.06938541e+00 -3.51833850e-02
-1.21132934e+00 2.75737435e-01 -4.41710621e-01 3.64097089e-01
-6.50644124e-01 4.58868712e-01 -4.06639725e-01 -5.48101868e-03
5.20325005e-01 3.62149388e-01 -3.75580639e-01 -5.61322451e-01
6.68633163e-01 6.84110343e-01 5.64497769e-01 -1.09159446e+00
7.80498087e-01 2.79300630e-01 -2.76504040e-01 -1.01957023e+00
-8.21150184e-01 -3.08954656e-01 -9.33800697e-01 1.91597894e-01
1.01113760e+00 -7.85696864e-01 -3.08133096e-01 1.54634714e-01
-1.30576384e+00 -1.82847187e-01 1.44353509e-01 4.56764102e-01
-5.27304471e-01 6.83891952e-01 -7.76443243e-01 -7.57157207e-01
-2.52213001e-01 -1.01541328e+00 1.02773654e+00 -6.16819598e-02
-3.52893353e-01 -1.27525306e+00 1.49731338e-01 2.60357767e-01
1.08543023e-01 -8.43610615e-02 1.27907860e+00 -1.15968812e+00
-4.33753759e-01 1.19492533e-02 -3.38456035e-02 5.00311553e-01
9.66845453e-02 -6.00705743e-02 -1.09884536e+00 -1.87705219e-01
3.09069380e-02 -2.98102796e-01 6.56291664e-01 1.61427200e-01
6.63722575e-01 -1.65566027e-01 -2.81285197e-01 2.52251834e-01
1.46708333e+00 -2.45139804e-02 3.69104087e-01 4.76532996e-01
5.96122205e-01 8.55989158e-01 1.25979394e-01 -2.72847235e-01
6.29282236e-01 5.51577449e-01 -4.44009565e-02 1.32316723e-01
1.32920533e-01 -2.70498633e-01 6.81412280e-01 1.22262955e+00
3.63639719e-03 -3.14303562e-02 -8.52952838e-01 9.22992229e-01
-1.57401061e+00 -5.92923701e-01 -2.07506016e-01 2.10895896e+00
1.03414607e+00 1.63307443e-01 -6.56411648e-02 8.99982527e-02
4.74473983e-01 -5.18954843e-02 1.17622495e-01 -9.48438525e-01
-1.65598318e-01 7.19633520e-01 4.03473198e-01 7.38702118e-01
-1.22960734e+00 1.38971210e+00 6.68623543e+00 5.52374601e-01
-9.56852674e-01 7.58446753e-02 3.25373650e-01 4.88597065e-01
-4.63035613e-01 1.84183836e-01 -1.09931684e+00 1.31295130e-01
1.31412828e+00 -5.58602810e-02 3.08194488e-01 8.52430344e-01
-1.99714359e-02 -8.53634179e-02 -1.22000754e+00 2.92847008e-01
7.25930557e-02 -8.71354699e-01 -4.39897962e-02 -5.88430911e-02
5.70012510e-01 1.80840194e-01 -3.24413180e-01 4.96422380e-01
6.50235415e-01 -9.89954889e-01 5.00339508e-01 -9.24536735e-02
5.77643812e-01 -6.79743707e-01 6.02871001e-01 5.29993653e-01
-1.06827366e+00 9.49669778e-02 -3.22142065e-01 -3.52216512e-01
1.99591517e-01 4.96946096e-01 -9.71011341e-01 8.31693769e-01
3.32419574e-01 2.51995295e-01 -7.19820023e-01 5.79345345e-01
-8.45944524e-01 8.77270341e-01 -2.35405818e-01 -9.55650136e-02
4.09170896e-01 -1.85684245e-02 3.99257272e-01 1.73099303e+00
2.21915822e-02 -9.25279409e-02 1.58458844e-01 6.38835907e-01
9.81417373e-02 6.29076004e-01 -8.04433167e-01 -1.79972127e-01
1.05764732e-01 1.37529492e+00 -7.49995589e-01 -3.95128787e-01
-7.04151034e-01 8.30008924e-01 8.67066562e-01 3.55809629e-01
-2.95450538e-01 -4.33368951e-01 3.61120224e-01 -4.07386452e-01
5.21536052e-01 -5.66864908e-01 -2.49793574e-01 -1.40343273e+00
2.64924556e-01 -8.89063418e-01 2.72056580e-01 -3.69623184e-01
-1.21536291e+00 8.08166206e-01 -7.83723872e-03 -6.17076576e-01
-6.27952993e-01 -1.02313316e+00 -3.19883347e-01 1.00805771e+00
-1.72642529e+00 -1.13652551e+00 5.29163301e-01 3.28853220e-01
4.59609181e-01 -7.52260163e-02 1.28437221e+00 1.88973516e-01
-4.54890341e-01 4.82014567e-01 -4.30948175e-02 4.51567560e-01
4.80692744e-01 -1.47518241e+00 5.88084102e-01 1.31706023e+00
7.12976515e-01 9.96895790e-01 6.64207101e-01 -3.61902684e-01
-1.29825377e+00 -9.82354462e-01 1.75556612e+00 -7.45143294e-01
8.99255395e-01 -7.48150110e-01 -9.80394900e-01 1.06037045e+00
4.93592411e-01 -3.78212452e-01 7.70107269e-01 5.96322119e-01
-5.06493866e-01 6.94922879e-02 -7.24206686e-01 5.48183799e-01
7.93831229e-01 -6.71803415e-01 -1.13188982e+00 3.72189134e-01
7.14234293e-01 -1.10651270e-01 -8.00813615e-01 3.05450648e-01
2.25122407e-01 -8.80871952e-01 5.41888177e-01 -7.68944204e-01
5.36937118e-01 -1.43942922e-01 -2.03088447e-01 -1.29271841e+00
-2.85878837e-01 -5.05444944e-01 2.64186621e-01 1.40174890e+00
7.79543817e-01 -7.34232962e-01 3.57842505e-01 5.69922090e-01
-4.16408420e-01 -4.25785929e-01 -8.61093819e-01 -7.40231395e-01
6.30297720e-01 -6.47751689e-01 2.33776629e-01 9.19470072e-01
3.75814348e-01 6.66726291e-01 -2.47210022e-02 3.54562163e-01
4.09078896e-01 2.08982736e-01 6.29486203e-01 -1.16853738e+00
-6.22870028e-01 -4.29966480e-01 -1.53539866e-01 -1.07908154e+00
7.55266309e-01 -1.18761528e+00 1.04801655e-01 -1.37057209e+00
2.57665843e-01 -4.17237073e-01 -3.83643091e-01 5.63202083e-01
-2.25803882e-01 3.56890410e-02 1.77876920e-01 6.41073147e-03
-5.60618579e-01 9.60429013e-02 9.58961844e-01 2.23467752e-01
3.43421996e-02 -2.28594139e-01 -1.04668331e+00 8.14786434e-01
8.63012493e-01 -5.29885709e-01 -2.21695095e-01 -5.98751307e-01
2.49156132e-01 -8.80335271e-02 7.18376264e-02 -7.03051925e-01
-3.79238315e-02 3.80627103e-02 5.77338338e-02 -1.61700621e-01
-2.01895744e-01 -5.43938398e-01 -4.41586524e-01 5.34091406e-02
-2.45682314e-01 7.96294585e-02 4.71119493e-01 2.45671585e-01
-1.92375422e-01 -5.53275108e-01 4.17235762e-01 -4.33490664e-01
-7.06281245e-01 -1.57925859e-01 -5.58482945e-01 2.48192310e-01
5.89297652e-01 2.02475473e-01 -1.84039265e-01 -3.37628871e-02
-6.25853658e-01 -3.99925038e-02 4.10488069e-01 3.55663717e-01
-5.42383045e-02 -9.03371096e-01 -6.81613505e-01 2.06586853e-01
9.26353112e-02 -2.58063793e-01 -1.92031026e-01 7.16250598e-01
-3.67147833e-01 1.05231714e+00 6.49436638e-02 -3.60066116e-01
-1.00595653e+00 6.00948453e-01 6.77818507e-02 -4.40496087e-01
-5.89446425e-01 6.80505157e-01 3.09839517e-01 -7.79968858e-01
-1.05880626e-01 -3.81080061e-01 -1.80359110e-01 -2.27284148e-01
4.27713752e-01 -2.75963336e-01 3.16171288e-01 -8.01737547e-01
-3.91274840e-01 4.28826630e-01 -2.18691424e-01 -2.17417091e-01
1.30294788e+00 -7.17217401e-02 -2.73133218e-01 5.04769504e-01
1.04011226e+00 5.94124794e-01 -8.43263507e-01 -4.86330837e-02
5.01671612e-01 1.97154760e-01 -3.22158515e-01 -7.89900601e-01
-5.60424924e-01 1.00023890e+00 -1.20698623e-01 3.74078512e-01
1.08554721e+00 2.00275660e-01 6.37464702e-01 5.58130205e-01
4.98563290e-01 -1.09636116e+00 -5.75207174e-01 6.97962642e-01
3.76083374e-01 -1.23916805e+00 -3.87094796e-01 -3.85498047e-01
-4.46055561e-01 1.10088921e+00 1.52265206e-01 -1.03696905e-01
2.12133303e-01 3.36369216e-01 1.72287285e-01 1.31110819e-02
-8.19160104e-01 -3.10097545e-01 1.07469156e-01 6.89037144e-01
8.85335922e-01 2.57875174e-01 -6.86525822e-01 7.13188529e-01
-3.91326547e-01 -3.23070109e-01 5.15441179e-01 9.11268234e-01
-4.89255816e-01 -1.66987872e+00 -2.32222617e-01 1.55572155e-02
-9.16310430e-01 -5.12036443e-01 -4.04763043e-01 1.01722085e+00
-1.49764493e-01 1.03637004e+00 -2.91080743e-01 -2.02982128e-03
1.16478309e-01 7.55158305e-01 6.95264637e-01 -9.36792910e-01
-4.82241362e-01 -3.04729547e-02 4.70394015e-01 -1.98843285e-01
-5.42153418e-01 -7.92200625e-01 -1.12846828e+00 2.22858712e-01
-1.63028255e-01 4.09941822e-01 6.13799393e-01 1.41176724e+00
1.91880539e-01 3.26344252e-01 3.66108805e-01 -5.95562279e-01
-4.36502248e-01 -9.05903041e-01 -3.67727995e-01 2.44591087e-01
2.27266476e-02 -3.97999942e-01 -4.65066493e-01 2.03054592e-01] | [10.400900840759277, 9.607809066772461] |
ecb96c3f-428c-4920-9241-bc0c65d30948 | contact-and-human-dynamics-from-monocular | 2007.11678 | null | https://arxiv.org/abs/2007.11678v2 | https://arxiv.org/pdf/2007.11678v2.pdf | Contact and Human Dynamics from Monocular Video | Existing deep models predict 2D and 3D kinematic poses from video that are approximately accurate, but contain visible errors that violate physical constraints, such as feet penetrating the ground and bodies leaning at extreme angles. In this paper, we present a physics-based method for inferring 3D human motion from video sequences that takes initial 2D and 3D pose estimates as input. We first estimate ground contact timings with a novel prediction network which is trained without hand-labeled data. A physics-based trajectory optimization then solves for a physically-plausible motion, based on the inputs. We show this process produces motions that are significantly more realistic than those from purely kinematic methods, substantially improving quantitative measures of both kinematic and dynamic plausibility. We demonstrate our method on character animation and pose estimation tasks on dynamic motions of dancing and sports with complex contact patterns. | ['Leonidas J. Guibas', 'Davis Rempe', 'Aaron Hertzmann', 'Ruben Villegas', 'Jimei Yang', 'Bryan Russell'] | 2020-07-22 | null | https://www.ecva.net/papers/eccv_2020/papers_ECCV/html/2918_ECCV_2020_paper.php | https://www.ecva.net/papers/eccv_2020/papers_ECCV/papers/123500069.pdf | eccv-2020-8 | ['human-dynamics'] | ['computer-vision'] | [-9.98419076e-02 1.65712699e-01 -2.66054392e-01 -1.56909630e-01
-5.08472085e-01 -6.34353161e-01 4.43360299e-01 -6.93329573e-02
-2.94969559e-01 6.61520898e-01 4.02415484e-01 -1.37331024e-01
1.56455085e-01 -7.11209953e-01 -1.07750988e+00 -1.33772671e-01
-5.42019725e-01 8.19955647e-01 5.32952011e-01 -4.29187000e-01
2.25106776e-01 6.93673491e-01 -1.13829994e+00 7.03621376e-03
5.82524121e-01 4.81008530e-01 -2.06022695e-01 1.34031630e+00
3.68249565e-01 6.32543802e-01 -3.68762553e-01 -4.83395457e-01
3.77112150e-01 -4.84091610e-01 -9.03506279e-01 1.16399020e-01
6.92936063e-01 -8.07717741e-01 -6.56488061e-01 4.99184936e-01
1.58034593e-01 5.73545516e-01 7.15519130e-01 -1.15587258e+00
-1.02922641e-01 2.99097240e-01 -5.49693465e-01 -5.39703593e-02
8.55590463e-01 5.86001277e-01 8.84738564e-01 -5.10268092e-01
9.32025135e-01 1.37261176e+00 1.29367185e+00 7.24418044e-01
-1.16215575e+00 -1.80036575e-01 1.39892980e-01 2.79684011e-02
-1.08354664e+00 -1.54680058e-01 5.58814347e-01 -6.45151436e-01
1.09107888e+00 2.49368638e-01 1.19352460e+00 1.31394911e+00
4.60087478e-01 7.64032841e-01 3.72812182e-01 -1.87000632e-01
1.30033955e-01 -7.87048519e-01 -2.21984223e-01 1.04189384e+00
1.46698415e-01 2.50963837e-01 -5.98796606e-01 -3.50009620e-01
1.22666776e+00 -3.95254850e-01 -2.27337658e-01 -3.55339885e-01
-1.47547877e+00 3.89962524e-01 2.11320192e-01 -4.98187870e-01
-4.79382694e-01 8.60077143e-01 2.36609101e-01 -4.81028467e-01
3.15877587e-01 4.34435040e-01 -4.92617875e-01 -6.66732430e-01
-1.05528557e+00 1.06406069e+00 9.59009528e-01 9.84690905e-01
2.17500433e-01 2.79347032e-01 1.69691816e-01 8.68655145e-02
2.52334952e-01 5.87368727e-01 -1.14965014e-01 -1.74257302e+00
4.19942051e-01 9.62945819e-02 5.34147561e-01 -1.22217774e+00
-6.91875041e-01 2.55022854e-01 -4.30748552e-01 4.86777693e-01
9.15890336e-01 -3.99758220e-01 -8.74712646e-01 1.59037805e+00
5.59202254e-01 6.04211807e-01 -1.78015485e-01 1.32562900e+00
4.04792517e-01 6.32026434e-01 1.56957969e-01 3.34415138e-01
9.26666975e-01 -8.98636580e-01 -4.36297953e-01 -1.55323774e-01
4.60404962e-01 -5.59235096e-01 9.79088604e-01 3.39012176e-01
-1.33696508e+00 -4.62138474e-01 -9.29719627e-01 -1.40123695e-01
4.69728351e-01 -7.89521076e-03 6.04604423e-01 3.67462158e-01
-7.95418382e-01 1.46660566e+00 -1.52252376e+00 -3.13297063e-01
3.54591459e-02 1.99079007e-01 -3.04283798e-01 4.59565103e-01
-9.60203648e-01 1.02512610e+00 1.23139814e-01 2.13163719e-01
-9.19280231e-01 -6.61829472e-01 -9.78471458e-01 -3.09320956e-01
3.00668806e-01 -1.03458285e+00 1.39665532e+00 -5.33776999e-01
-1.84839618e+00 6.15904689e-01 -1.04904920e-01 -4.23184186e-01
1.11886442e+00 -8.63250732e-01 9.06989351e-02 3.33167195e-01
-1.56610802e-01 7.82344699e-01 4.45386738e-01 -1.28765881e+00
-1.97287485e-01 7.66201243e-02 1.16616324e-01 3.80510300e-01
4.27634537e-01 -3.69721353e-01 -6.38633847e-01 -6.91514492e-01
1.99751019e-01 -1.33156657e+00 -5.04060686e-01 5.55822253e-01
-8.41348410e-01 2.55314708e-01 4.81131434e-01 -1.02664983e+00
8.72985303e-01 -1.35863221e+00 6.71386957e-01 1.87061399e-01
1.72048584e-01 2.09417287e-02 9.37068686e-02 3.38464707e-01
3.19671661e-01 1.45283744e-01 -1.99096009e-01 -3.71063083e-01
-5.46619482e-03 2.52333015e-01 -3.02924067e-01 5.28803051e-01
2.89951831e-01 8.84472489e-01 -1.10324669e+00 -5.59440374e-01
2.07335532e-01 3.03922653e-01 -9.36857462e-01 2.63289839e-01
-5.86325765e-01 7.28037238e-01 -4.48651135e-01 6.43351614e-01
3.11472416e-01 -4.22329493e-02 2.84398586e-01 -3.12236637e-01
-1.37961879e-01 1.93301201e-01 -1.21268857e+00 1.91884613e+00
-1.29970610e-01 5.76562285e-01 -1.17374718e-01 -3.80841970e-01
5.00377536e-01 1.56769216e-01 6.82376862e-01 2.02755108e-01
1.47023052e-01 -5.20898812e-02 -2.32786104e-01 -7.22417653e-01
8.32239985e-01 -1.82776943e-01 -2.22518057e-01 4.29256231e-01
-1.10224918e-01 -4.40151632e-01 -2.05902338e-01 1.38467789e-01
1.17401814e+00 1.26184380e+00 1.03200704e-01 -4.54914682e-02
-1.01226784e-01 7.07507968e-01 6.99602842e-01 5.28473496e-01
-1.97370693e-01 1.01283669e+00 4.64266151e-01 -6.62244380e-01
-1.58396792e+00 -1.42039931e+00 4.41994727e-01 6.40304387e-01
5.17880023e-01 -5.87570965e-01 -9.09469187e-01 -2.56669074e-01
2.42775589e-01 4.48831469e-01 -4.44900870e-01 -1.93763778e-01
-1.29571545e+00 -4.02919978e-01 1.00501013e+00 9.25688088e-01
1.64998680e-01 -6.07347012e-01 -9.55806196e-01 2.93774426e-01
-4.92642581e-01 -1.33049679e+00 -4.94567573e-01 -4.00010437e-01
-9.67897356e-01 -1.21904516e+00 -6.93925798e-01 -2.58725703e-01
3.17330599e-01 -2.36083344e-01 1.25586343e+00 3.78982812e-01
-3.45672667e-01 2.73377359e-01 -1.52257398e-01 -1.10012993e-01
-5.62797904e-01 -2.77477473e-01 4.98355836e-01 -5.13121903e-01
-1.72555059e-01 -4.03942943e-01 -5.75821817e-01 2.31033653e-01
-2.25466460e-01 4.69018757e-01 -7.52166510e-02 5.79316497e-01
5.96224785e-01 -4.95928138e-01 -2.12506816e-01 -2.81690627e-01
2.04832986e-01 -2.38879517e-01 -3.28002542e-01 -4.27621491e-02
2.26430610e-01 6.95288852e-02 6.32552505e-01 -6.42194450e-01
-1.03621233e+00 3.47915858e-01 -2.99431145e-01 -7.11815834e-01
-2.61303693e-01 8.12282935e-02 1.26106054e-01 2.41873506e-02
9.06774402e-01 -1.62923828e-01 -1.27206653e-01 -4.19780254e-01
4.29709643e-01 -1.97719559e-01 9.57460701e-01 -1.06771958e+00
9.17533934e-01 5.91866314e-01 4.31813419e-01 -8.55477333e-01
-5.79550028e-01 1.56658471e-01 -1.12638724e+00 -5.57274818e-01
1.15249932e+00 -8.67888808e-01 -1.20589983e+00 7.56841302e-01
-1.32934308e+00 -7.69127190e-01 -1.12457186e-01 8.84168506e-01
-9.70126152e-01 7.30191946e-01 -1.07913256e+00 -7.71891296e-01
6.06208593e-02 -9.44468617e-01 1.27515090e+00 6.21017814e-02
-9.39529598e-01 -1.05001605e+00 3.35501790e-01 8.32571611e-02
-3.04073960e-01 1.12989330e+00 7.05451429e-01 6.02059625e-02
-6.24344826e-01 -9.62128639e-02 3.74983221e-01 -1.96888879e-01
-1.52717397e-01 6.98011279e-01 -5.36182582e-01 -4.14008135e-03
-6.91822052e-01 -3.05861771e-01 5.29239655e-01 6.79621935e-01
8.82385612e-01 -2.70946980e-01 -4.85536724e-01 9.18755054e-01
1.06567681e+00 -3.71944308e-01 3.75004381e-01 2.57988811e-01
1.07423031e+00 6.69644892e-01 6.98685348e-01 5.84213793e-01
4.78487343e-01 7.79171586e-01 4.45168823e-01 1.38236731e-01
-1.65587887e-01 -7.68297315e-01 3.41294080e-01 5.50398886e-01
-8.05288911e-01 -3.78076196e-01 -1.17985725e+00 4.35438424e-01
-2.01558781e+00 -1.05360186e+00 -4.68950242e-01 1.99917305e+00
7.61222899e-01 3.93818110e-01 4.98021305e-01 -2.47101292e-01
6.33426607e-01 -8.53629112e-02 -7.64980018e-01 -2.73003042e-01
1.80249870e-01 1.01136953e-01 6.42935693e-01 8.07892382e-01
-1.02852583e+00 1.17798185e+00 7.55776787e+00 1.92888543e-01
-8.74145508e-01 -2.92733669e-01 2.74332076e-01 -3.53522122e-01
-3.64394128e-01 1.78529724e-01 -6.02974772e-01 1.05725788e-01
7.71670520e-01 7.01202229e-02 1.04284145e-01 6.27129972e-01
5.68280756e-01 -1.84221074e-01 -1.27970660e+00 6.07746720e-01
-3.01635027e-01 -1.64961362e+00 4.67588082e-02 -4.87706773e-02
5.81526399e-01 -1.66128471e-01 -3.58581007e-01 -1.93791419e-01
7.26914704e-01 -1.04937708e+00 1.31049585e+00 8.55609000e-01
6.86291873e-01 -7.02542126e-01 9.26804766e-02 6.23491645e-01
-1.24343264e+00 4.08831596e-01 -1.07474834e-01 -4.38481838e-01
8.14860761e-01 1.21904030e-01 -5.56226492e-01 3.64143968e-01
5.74908495e-01 6.85217977e-01 -6.96773157e-02 9.18330073e-01
-3.20276439e-01 6.52162313e-01 -7.27739930e-01 -2.89131850e-02
2.03205302e-01 -2.58141756e-01 8.93299997e-01 1.16288853e+00
3.03641170e-01 4.88077581e-01 3.61004919e-01 1.04360414e+00
2.61521757e-01 -3.99291724e-01 -7.14241982e-01 2.46934921e-01
3.93387377e-01 8.89628291e-01 -8.31823230e-01 -3.19918931e-01
2.38650858e-01 1.03512847e+00 1.29332915e-01 3.25952083e-01
-1.23682201e+00 -5.46967564e-03 1.03359556e+00 2.83526540e-01
-2.06563622e-01 -1.04401386e+00 -2.75019228e-01 -1.33038175e+00
-4.65461891e-03 -4.06589925e-01 -9.18113738e-02 -8.86286676e-01
-8.38242471e-01 2.62983888e-01 4.36601251e-01 -1.32812452e+00
-6.10951185e-01 -5.15836239e-01 -8.30476463e-01 7.32551336e-01
-6.62947953e-01 -9.88375664e-01 -3.45918447e-01 1.71468303e-01
4.53608513e-01 4.81462091e-01 5.66922843e-01 -1.89962149e-01
-3.96597534e-01 3.18053126e-01 -4.68774080e-01 2.73504794e-01
3.71907383e-01 -1.25059140e+00 1.24053514e+00 8.08619618e-01
-4.22749557e-02 3.57449502e-01 1.14923561e+00 -1.10117567e+00
-1.51564980e+00 -8.52266431e-01 4.69510704e-01 -9.35717463e-01
6.15630746e-01 -9.62884501e-02 -8.64483833e-01 8.72020900e-01
-3.39221120e-01 -3.83390225e-02 4.10941273e-01 -2.11406484e-01
6.00050483e-03 7.07547903e-01 -9.37811971e-01 1.01404583e+00
1.60186052e+00 -2.30954438e-01 -6.40490055e-01 3.17872256e-01
5.42595029e-01 -1.16124952e+00 -9.10056770e-01 2.79785872e-01
9.64282453e-01 -7.66437531e-01 1.35394478e+00 -1.18276262e+00
6.41972661e-01 -2.26824000e-01 -3.23426239e-02 -1.19192588e+00
-8.55987146e-02 -8.88648927e-01 -3.22168350e-01 4.09507632e-01
6.97015151e-02 2.86689550e-01 1.17642629e+00 8.52357268e-01
-2.14137793e-01 -7.11127698e-01 -7.57666409e-01 -7.79805124e-01
2.54375339e-01 -5.73927462e-01 2.61518300e-01 7.59019196e-01
-1.58419251e-03 -2.78031975e-01 -7.31385231e-01 4.04591739e-01
8.72506320e-01 1.16709515e-03 1.06756067e+00 -8.41395617e-01
-5.26291251e-01 -2.04025537e-01 -5.75068772e-01 -1.42089093e+00
2.56710947e-01 -3.71790171e-01 4.41155225e-01 -1.42084932e+00
-1.59292996e-01 -1.52195558e-01 7.36711085e-01 2.72100180e-01
-2.32354730e-01 1.92711547e-01 2.72377431e-01 2.27806315e-01
-2.08456069e-01 4.69742715e-01 1.37315893e+00 2.91065514e-01
-1.60870731e-01 -4.99170758e-02 1.85882673e-01 1.50425243e+00
6.26584768e-01 -3.16339165e-01 -1.65838614e-01 -5.95513165e-01
2.39483967e-01 5.79260468e-01 8.61438751e-01 -1.16826749e+00
1.69271491e-02 -6.41295552e-01 7.66205788e-01 -6.06161416e-01
7.21237123e-01 -2.61781693e-01 3.75871122e-01 8.09800506e-01
-3.90224218e-01 3.04051757e-01 3.46535683e-01 8.19172382e-01
3.71418029e-01 2.22674627e-02 5.42339683e-01 -3.27990174e-01
-8.07653069e-01 3.60343009e-01 -6.13309264e-01 1.39386058e-01
9.22503710e-01 -5.09737492e-01 3.64597850e-02 -6.03985608e-01
-1.19905460e+00 1.18521839e-01 8.06461811e-01 3.48601818e-01
7.67568111e-01 -1.36958325e+00 -7.24186242e-01 -3.31345797e-01
-4.15131003e-01 1.06396995e-01 1.83626026e-01 5.65595031e-01
-1.57379258e+00 -1.85986966e-01 -3.75622749e-01 -8.91611457e-01
-1.12330711e+00 8.50706473e-02 5.43394923e-01 9.33242962e-02
-1.01685929e+00 8.92110705e-01 -2.40358412e-01 -5.71792364e-01
3.74255818e-03 -6.69698775e-01 2.35560283e-01 -6.28928721e-01
1.39665917e-01 7.20298886e-01 -4.83309031e-01 -8.57533932e-01
-4.22759295e-01 9.84289587e-01 6.52670145e-01 -4.03843254e-01
1.08272481e+00 -6.02403395e-02 4.35484856e-01 3.96997809e-01
9.03844178e-01 1.06725879e-01 -2.15662432e+00 3.06315333e-01
-4.50501479e-02 -5.40639341e-01 -4.74115014e-01 -4.08681095e-01
-6.72414780e-01 8.81731451e-01 1.37964329e-02 -3.87716830e-01
3.86110634e-01 -1.89429924e-01 1.29678154e+00 5.19161642e-01
5.19256949e-01 -1.08905256e+00 2.92301744e-01 6.25466585e-01
6.84341133e-01 -8.90961647e-01 2.54000932e-01 -5.97523868e-01
-7.08507717e-01 1.43710828e+00 8.64391148e-01 -5.56984484e-01
5.08076131e-01 5.88538826e-01 -8.19621831e-02 -5.58988228e-02
-6.99423432e-01 1.30282134e-01 3.90699506e-01 6.08869314e-01
3.63034397e-01 1.15650602e-01 5.53678609e-02 3.05500060e-01
-4.87746656e-01 -1.41777731e-02 8.26344728e-01 1.12801814e+00
-5.42057753e-01 -8.34094226e-01 -5.48523247e-01 -7.79589191e-02
-2.16905877e-01 2.49317706e-01 -4.70292509e-01 1.00186121e+00
3.53208184e-02 5.63694417e-01 2.41997004e-01 -7.37566113e-01
3.85520577e-01 -1.53095379e-01 9.59025979e-01 -4.46736604e-01
-4.23406690e-01 -5.98256625e-02 4.91421610e-01 -1.02002323e+00
-1.78304672e-01 -7.96867967e-01 -1.88305485e+00 -8.04818213e-01
-1.28066447e-02 -2.41560861e-01 3.79503071e-01 8.67985487e-01
2.31974065e-01 3.77106160e-01 -1.93733498e-02 -1.73468983e+00
-5.05700231e-01 -5.58843911e-01 -1.65904909e-01 6.26179636e-01
3.48687530e-01 -7.33537436e-01 -8.25571921e-03 5.09191751e-01] | [7.099533557891846, -0.5573651790618896] |
8d363d3c-38e0-4bb9-acba-a0ee98b33d0b | learning-with-batch-wise-optimal-transport | 1903.08923 | null | http://arxiv.org/abs/1903.08923v1 | http://arxiv.org/pdf/1903.08923v1.pdf | Learning with Batch-wise Optimal Transport Loss for 3D Shape Recognition | Deep metric learning is essential for visual recognition. The widely used
pair-wise (or triplet) based loss objectives cannot make full use of semantical
information in training samples or give enough attention to those hard samples
during optimization. Thus, they often suffer from a slow convergence rate and
inferior performance. In this paper, we show how to learn an importance-driven
distance metric via optimal transport programming from batches of samples. It
can automatically emphasize hard examples and lead to significant improvements
in convergence. We propose a new batch-wise optimal transport loss and combine
it in an end-to-end deep metric learning manner. We use it to learn the
distance metric and deep feature representation jointly for recognition.
Empirical results on visual retrieval and classification tasks with six
benchmark datasets, i.e., MNIST, CIFAR10, SHREC13, SHREC14, ModelNet10, and
ModelNet40, demonstrate the superiority of the proposed method. It can
accelerate the convergence rate significantly while achieving a
state-of-the-art recognition performance. For example, in 3D shape recognition
experiments, we show that our method can achieve better recognition performance
within only 5 epochs than what can be obtained by mainstream 3D shape
recognition approaches after 200 epochs. | ['Lin Xu', 'Yuai Liu', 'Han Sun'] | 2019-03-21 | learning-with-batch-wise-optimal-transport-1 | http://openaccess.thecvf.com/content_CVPR_2019/html/Xu_Learning_With_Batch-Wise_Optimal_Transport_Loss_for_3D_Shape_Recognition_CVPR_2019_paper.html | http://openaccess.thecvf.com/content_CVPR_2019/papers/Xu_Learning_With_Batch-Wise_Optimal_Transport_Loss_for_3D_Shape_Recognition_CVPR_2019_paper.pdf | cvpr-2019-6 | ['3d-shape-recognition'] | ['computer-vision'] | [-9.13311467e-02 -6.67379975e-01 1.61053296e-02 -9.10788000e-01
-9.44608152e-01 -2.96393216e-01 4.15842116e-01 -1.08275317e-01
-6.77490592e-01 3.57669592e-01 -1.49991199e-01 -2.64262974e-01
-3.80523264e-01 -5.01289666e-01 -5.85874498e-01 -6.09930217e-01
1.50133178e-01 5.40334463e-01 -3.23714130e-02 3.32387179e-01
3.54997993e-01 7.63216019e-01 -1.19333780e+00 1.69439778e-01
7.65899062e-01 1.71217024e+00 8.64337757e-02 3.72411937e-01
-9.01546851e-02 6.31703794e-01 -3.69863749e-01 -3.24902445e-01
5.02765417e-01 -6.23271130e-02 -5.47149420e-01 1.60660580e-01
7.41897762e-01 -4.37932134e-01 -6.17401838e-01 9.32104588e-01
8.47554922e-01 1.91487998e-01 8.61907125e-01 -1.17489636e+00
-1.14341080e+00 1.95085302e-01 -6.94795072e-01 2.39401668e-01
-1.86615035e-01 1.59712553e-01 1.10329652e+00 -1.44992518e+00
2.21532702e-01 1.17641962e+00 5.85924387e-01 6.31221533e-01
-1.23074389e+00 -6.43075228e-01 1.99796796e-01 4.88053232e-01
-1.69916701e+00 -3.86305839e-01 7.25550950e-01 -2.17444330e-01
1.05398250e+00 7.31786117e-02 2.38328323e-01 8.67962301e-01
-2.67867893e-01 1.17947757e+00 8.69770825e-01 -1.75752267e-01
1.05011389e-01 -2.02976659e-01 9.81745422e-02 8.27401936e-01
2.99672168e-02 1.66091666e-01 -4.07667190e-01 1.01747386e-01
6.49343252e-01 1.93804190e-01 -2.42938831e-01 -4.42076892e-01
-1.06170464e+00 6.06879890e-01 7.11761355e-01 1.04228362e-01
-2.26898074e-01 1.72006950e-01 3.32324296e-01 5.77411652e-01
5.92259169e-01 1.04429513e-01 -3.28978270e-01 -1.62668273e-01
-8.10262740e-01 1.06389031e-01 5.36081374e-01 8.01296353e-01
5.72079062e-01 1.35774329e-01 -4.03200597e-01 1.29741669e+00
3.42714846e-01 7.53699362e-01 4.15349185e-01 -6.71169996e-01
5.33177912e-01 6.49655819e-01 -1.15176775e-02 -1.11020029e+00
-2.49909282e-01 -5.19811928e-01 -9.38840330e-01 2.43336946e-01
3.91927153e-01 1.86102435e-01 -1.19315910e+00 1.66500151e+00
3.58194560e-01 3.03875715e-01 -2.29885846e-01 1.30337393e+00
7.66281247e-01 5.23819029e-01 -1.79472581e-01 2.95326829e-01
9.07846689e-01 -1.03054929e+00 -1.64552167e-01 4.77222390e-02
5.81631958e-01 -8.09183300e-01 1.18528342e+00 3.71379554e-01
-9.94660258e-01 -5.59848666e-01 -1.20894456e+00 -2.00272918e-01
-2.22539201e-01 3.33091348e-01 4.54951704e-01 4.24446434e-01
-7.17683434e-01 7.37165809e-01 -9.32338357e-01 -7.21104071e-02
9.47224677e-01 2.82692254e-01 -2.42635876e-01 -4.64502424e-01
-6.48179412e-01 8.31596076e-01 1.06255785e-01 2.71138519e-01
-1.10214770e+00 -9.63535190e-01 -7.19037294e-01 1.15187354e-01
2.47859538e-01 -5.02492666e-01 1.03585088e+00 -6.63272858e-01
-1.50945449e+00 9.79959488e-01 -8.67536888e-02 -4.24528033e-01
6.45178318e-01 -2.08711401e-01 -2.07375586e-01 5.87008754e-03
-3.98644537e-01 7.20460951e-01 8.32398772e-01 -8.31981599e-01
-4.98495936e-01 -5.28158426e-01 -4.43612002e-02 1.48481429e-01
-6.75396681e-01 1.72493290e-02 -4.02094156e-01 -7.01728582e-01
1.74549669e-01 -6.95106924e-01 4.19135913e-02 6.78454280e-01
-2.64136106e-01 -5.19297898e-01 7.81175137e-01 -4.43554550e-01
8.72946322e-01 -2.23118019e+00 7.56709874e-02 2.91869015e-01
2.70029396e-01 5.94557464e-01 -5.83577931e-01 -9.76985767e-02
2.25703657e-01 -3.85075994e-02 -5.17353415e-01 -5.50266147e-01
3.89424592e-01 2.50016630e-01 -1.08551919e-01 6.74608469e-01
2.85831928e-01 1.01488769e+00 -6.79949999e-01 -1.21449068e-01
1.95672810e-01 7.68091261e-01 -4.94175732e-01 2.05939963e-01
-3.35064903e-02 1.37838144e-02 -3.58265638e-01 7.31391549e-01
8.16720009e-01 -4.40486044e-01 -3.82705241e-01 -1.86943188e-01
2.79783636e-01 2.49904200e-01 -9.63875890e-01 1.84502244e+00
-5.16606450e-01 7.34893203e-01 -1.51729971e-01 -1.37533486e+00
1.37808871e+00 -1.67073622e-01 3.05877268e-01 -1.05935955e+00
8.78037736e-02 3.82481456e-01 -4.70351651e-02 -1.84123218e-01
-7.39731314e-03 8.44250992e-02 3.16340357e-01 5.63514173e-01
-6.50006309e-02 8.01338106e-02 -4.91178147e-02 -2.52067000e-01
8.92015934e-01 -1.51355088e-01 -1.18926600e-01 -1.81305513e-01
7.67727315e-01 -5.00303924e-01 6.68411076e-01 6.55499697e-01
-2.58018464e-01 8.36047173e-01 2.70010889e-01 -6.87980950e-01
-1.15836668e+00 -1.05527341e+00 -2.34376281e-01 7.89934874e-01
1.37010276e-01 -1.32233977e-01 -3.99978012e-01 -9.16216612e-01
4.06737208e-01 3.93464714e-01 -5.49828827e-01 -5.40315688e-01
-4.96478945e-01 -7.36731708e-01 5.82415044e-01 7.29934990e-01
7.03531206e-01 -7.78562129e-01 -1.64296433e-01 2.33941171e-02
1.74487323e-01 -1.13398325e+00 -8.33447814e-01 -2.24820912e-01
-8.70661914e-01 -1.00473464e+00 -1.01739562e+00 -9.35012639e-01
9.46337581e-01 4.57193583e-01 8.29197705e-01 1.35332420e-01
-4.57558870e-01 3.23379546e-01 -1.10486992e-01 -1.96564078e-01
1.64201081e-01 1.00554079e-01 3.24404351e-02 2.93019980e-01
6.00283325e-01 -4.27760392e-01 -7.38802016e-01 7.00981915e-01
-6.02381945e-01 -2.65150189e-01 5.47108352e-01 1.23456001e+00
7.70089328e-01 -3.69233012e-01 5.05016208e-01 -3.86807352e-01
5.59720695e-01 -1.83410987e-01 -7.12828040e-01 4.53241795e-01
-8.87530565e-01 2.06899256e-01 6.11017287e-01 -4.44081634e-01
-7.14831889e-01 -2.26413384e-01 -5.46710864e-02 -9.48323548e-01
8.56546909e-02 5.23963213e-01 -1.68982252e-01 -5.31862438e-01
5.14482260e-01 4.16221410e-01 1.79937452e-01 -8.33734453e-01
1.62809193e-01 6.31547928e-01 3.18457574e-01 -6.77028775e-01
7.62669683e-01 3.91902953e-01 1.24053452e-02 -6.25162780e-01
-1.00343668e+00 -2.42666334e-01 -3.57678503e-01 7.09408820e-02
4.61678535e-01 -7.25388467e-01 -7.61234522e-01 8.12506795e-01
-1.05235016e+00 -4.44112092e-01 -1.28216878e-01 5.92832148e-01
-3.83632839e-01 3.93910795e-01 -5.22009194e-01 -5.63525081e-01
-5.87251604e-01 -9.94404435e-01 8.54183316e-01 1.71259403e-01
3.47399533e-01 -9.28865612e-01 -2.00649112e-01 1.74807757e-01
6.29153728e-01 7.29690418e-02 7.93706119e-01 -6.31646097e-01
-5.64072967e-01 -3.06629092e-01 -9.11575496e-01 7.47488022e-01
8.30413625e-02 -4.92645465e-02 -7.95946419e-01 -4.64673340e-01
-2.29021311e-01 -6.55604720e-01 1.26967466e+00 1.07552394e-01
1.77110350e+00 -2.90849775e-01 1.22488379e-01 1.06861246e+00
1.31670594e+00 1.85423270e-01 4.69969332e-01 -2.87376251e-02
9.00843859e-01 3.65186542e-01 5.15067339e-01 5.19352317e-01
4.15781587e-01 7.83849537e-01 3.79405677e-01 2.78401207e-02
-3.08843046e-01 -1.47909328e-01 1.79672554e-01 8.18018556e-01
1.97186366e-01 1.11496197e-02 -9.16754901e-01 4.45859283e-01
-1.83774745e+00 -6.95333958e-01 3.15581530e-01 2.39096999e+00
7.45760083e-01 8.07841346e-02 -8.29018727e-02 1.01340763e-01
5.21010578e-01 1.56659663e-01 -9.44748461e-01 -3.25423896e-01
-1.85504377e-01 2.16871724e-01 2.87801713e-01 5.47570407e-01
-1.04849672e+00 8.01404655e-01 5.07092619e+00 1.08594823e+00
-1.36946344e+00 5.20498231e-02 8.97915363e-01 -3.48036408e-01
-2.38637388e-01 -3.74026686e-01 -7.07856417e-01 3.99382651e-01
6.25950694e-01 -1.31841004e-01 5.85598588e-01 8.50218952e-01
-3.85009055e-03 4.31450903e-01 -1.31509876e+00 1.61030924e+00
3.09715927e-01 -1.23510373e+00 1.33055538e-01 -3.79465870e-03
7.68254697e-01 2.31929213e-01 3.35794717e-01 2.48058558e-01
4.23608944e-02 -1.28378403e+00 6.66867971e-01 6.36588395e-01
8.66487324e-01 -8.92190933e-01 5.20592630e-01 2.15010211e-01
-9.93457496e-01 -1.73891876e-02 -9.72716808e-01 1.30110547e-01
-1.05532050e-01 9.12835181e-01 -5.11588871e-01 2.88499177e-01
6.00537896e-01 1.05470324e+00 -4.13437814e-01 1.51411450e+00
-2.49143094e-01 5.22989094e-01 -5.07262588e-01 -3.44723433e-01
6.10624015e-01 -1.51371017e-01 5.07200718e-01 1.07060516e+00
3.83126885e-01 9.16118398e-02 8.29011425e-02 9.88296390e-01
-5.01793921e-01 4.24413569e-02 -4.26206529e-01 -3.07101440e-02
4.06891912e-01 1.00956643e+00 -3.73303324e-01 -1.63813084e-01
-3.45084131e-01 1.01890504e+00 5.97356856e-01 4.07623470e-01
-8.07386458e-01 -7.30227053e-01 9.67853367e-01 -2.77845204e-01
7.43946910e-01 -3.84540170e-01 -5.21142244e-01 -1.17720509e+00
4.16672796e-01 -7.17253685e-01 3.74959260e-01 -3.18557054e-01
-1.72767484e+00 4.89074618e-01 -4.63269114e-01 -1.16692781e+00
2.42858827e-01 -9.81562257e-01 -4.49638426e-01 7.89184093e-01
-1.70891774e+00 -8.69287670e-01 -4.02187586e-01 6.12877727e-01
3.51819336e-01 -3.49875599e-01 4.80368614e-01 7.65902698e-01
-6.09329998e-01 1.25995755e+00 3.45209867e-01 3.41998607e-01
6.87699139e-01 -1.19075859e+00 5.11227548e-01 5.31346917e-01
3.72672319e-01 2.31317148e-01 -3.84877063e-02 -3.98327634e-02
-1.38274872e+00 -1.19822061e+00 9.65134203e-01 -3.16859126e-01
3.82540017e-01 -4.04833138e-01 -1.03990769e+00 1.22140169e-01
-2.42412210e-01 6.49342597e-01 5.83974183e-01 -4.37149592e-02
-8.60237896e-01 -7.51928389e-01 -1.19819677e+00 4.20411527e-01
1.46536732e+00 -6.29714787e-01 -3.64394844e-01 3.64176542e-01
5.36025643e-01 -4.30606544e-01 -7.92206526e-01 4.89362389e-01
7.69473910e-01 -7.18858957e-01 1.14450169e+00 -9.20015633e-01
3.79502565e-01 -1.84510708e-01 -4.74553764e-01 -1.24454331e+00
-2.50393778e-01 -3.89470398e-01 -2.57583350e-01 9.93485332e-01
5.45148373e-01 -7.21406460e-01 7.59322703e-01 7.49136448e-01
-2.03518897e-01 -1.22029102e+00 -1.19876409e+00 -1.16796434e+00
3.10318410e-01 -5.49442410e-01 6.94121540e-01 7.71591544e-01
-5.34382164e-01 2.73152981e-02 -3.57777774e-01 -1.30602047e-01
1.00340855e+00 3.65105629e-01 6.68475270e-01 -9.86443102e-01
-1.65392250e-01 -9.22589183e-01 -7.62418509e-01 -1.56056595e+00
2.24425986e-01 -1.16415489e+00 -1.28622681e-01 -1.39936078e+00
2.57538736e-01 -7.62875497e-01 -9.21415508e-01 7.07015812e-01
-8.41414407e-02 3.92780781e-01 3.46533298e-01 1.07489727e-01
-6.56602859e-01 1.00432348e+00 1.32902002e+00 -5.33794105e-01
1.60519078e-01 -1.21310391e-01 -5.12646198e-01 4.41908032e-01
6.34128928e-01 -5.34511387e-01 -3.64597112e-01 -1.08973873e+00
-6.62544891e-02 -3.99248719e-01 4.78592336e-01 -7.77607143e-01
4.14905787e-01 -1.64686199e-02 5.71113110e-01 -4.55960065e-01
2.79952645e-01 -7.34765589e-01 -4.67409968e-01 3.72742921e-01
-5.21562696e-01 -9.31762680e-02 9.98378620e-02 5.25190234e-01
-3.37391078e-01 -1.02073193e-01 9.64197874e-01 2.00348660e-01
-5.91902971e-01 8.90268743e-01 3.34949464e-01 2.74229795e-01
7.80979395e-01 -2.42014140e-01 -2.11416289e-01 -1.16548233e-01
-5.27984917e-01 4.95682240e-01 1.94456145e-01 5.32844782e-01
1.16187787e+00 -1.85608864e+00 -9.70358908e-01 2.36139685e-01
1.33201838e-01 -9.65707004e-02 2.18629047e-01 9.46208775e-01
-3.45680088e-01 3.77844006e-01 -1.07372239e-01 -8.29471469e-01
-1.01300311e+00 3.08115393e-01 5.80689669e-01 -2.09908426e-01
-6.28120005e-01 1.10543811e+00 1.60775855e-01 -5.15910983e-01
7.03783393e-01 -4.13554251e-01 2.60339826e-01 -2.22179547e-01
7.20632851e-01 3.71330619e-01 1.92429960e-01 -2.38117203e-01
-6.43498778e-01 7.97532558e-01 -3.13534349e-01 1.25746444e-01
1.45328403e+00 9.07852426e-02 2.35250890e-01 1.88428104e-01
1.82961905e+00 -6.76973760e-01 -1.61833441e+00 -6.54693782e-01
-7.00000580e-03 -8.88009965e-01 2.19964147e-01 -9.17088687e-01
-1.60082757e+00 1.20272517e+00 7.21270204e-01 -2.80927777e-01
1.07776594e+00 -2.73500383e-01 1.12595093e+00 7.83545315e-01
2.10910395e-01 -1.01042891e+00 1.42980188e-01 6.01354778e-01
1.06663799e+00 -1.50922465e+00 -1.77503884e-01 1.01558171e-01
-3.22807819e-01 1.11954367e+00 5.56059361e-01 -2.86999285e-01
7.52452970e-01 -3.08140703e-02 -6.54040501e-02 6.40694872e-02
-8.24049652e-01 -1.42923951e-01 7.92636752e-01 4.90996122e-01
1.06916755e-01 2.92040836e-02 -2.18234316e-01 4.13236797e-01
1.97391838e-01 -2.24338785e-01 -1.22981504e-01 5.85237086e-01
-3.50571036e-01 -9.14926469e-01 6.54337481e-02 5.26455283e-01
-3.21603864e-01 -1.61074996e-02 -4.76477802e-01 4.49083745e-01
-3.62960696e-01 6.46542609e-01 1.63277388e-01 -3.85033876e-01
4.60659593e-01 -7.75862113e-02 7.93316483e-01 -1.83897331e-01
-1.56795919e-01 -3.20481747e-01 -1.15354136e-01 -6.59493208e-01
-1.63523883e-01 -5.65104961e-01 -1.14374483e+00 -5.05703866e-01
-1.48074880e-01 -1.47305861e-01 6.65983379e-01 9.04333472e-01
6.16805017e-01 4.27371748e-02 1.07416904e+00 -4.90751266e-01
-1.14644146e+00 -7.63613343e-01 -3.59282166e-01 5.52894473e-01
2.61083186e-01 -7.89260566e-01 -3.98133725e-01 -3.72379780e-01] | [9.571362495422363, 2.976148843765259] |
9b0c23ef-514d-42dd-ae68-a9068c2550b9 | grobid-combining-automatic-bibliographic-data | null | null | https://link.springer.com/chapter/10.1007/978-3-642-04346-8_62 | https://link.springer.com/content/pdf/10.1007/978-3-642-04346-8_62.pdf | GROBID: Combining Automatic Bibliographic Data Recognition and Term Extraction for Scholarship Publications | Based on state of the art machine learning techniques, GROBID (GeneRation Of BIbliographic Data) performs reliable bibliographic data extractions from scholar articles combined with multi-level term extractions. These two types of extraction present synergies and correspond to complementary descriptions of an article. This tool is viewed as a component for enhancing the existing and the future large repositories of technical and scientific publications. | ['Patrice Lopez'] | 2009-09-01 | null | null | null | research-and-advanced-technology-for-digital | ['term-extraction'] | ['natural-language-processing'] | [-3.18987936e-01 2.83477958e-02 -1.07953298e+00 2.46792838e-01
-9.99517322e-01 -5.95909238e-01 1.15370142e+00 8.45156848e-01
-4.38647598e-01 1.04839730e+00 4.45282042e-01 -6.70669556e-01
-5.00486910e-01 -6.12197161e-01 -4.83723700e-01 -1.48557112e-01
9.85020176e-02 5.59925318e-01 -1.43075109e-01 1.79448754e-01
1.24295032e+00 8.74309897e-01 -1.69737840e+00 4.60374832e-01
6.35588050e-01 5.02357602e-01 2.48369023e-01 3.65084231e-01
-8.61355424e-01 6.69374168e-01 -7.99986959e-01 -4.43389177e-01
-3.35704327e-01 -2.80500919e-01 -9.79645848e-01 -5.38817585e-01
3.13318253e-01 2.08201528e-01 -1.33204967e-01 9.95191932e-01
1.76141426e-01 -5.99221051e-01 1.02419496e+00 -1.05577219e+00
-6.94978893e-01 9.06229675e-01 -5.28604627e-01 4.00991380e-01
8.96565795e-01 -4.52473462e-01 9.61685181e-01 -1.07429826e+00
1.13378882e+00 1.05557561e+00 6.17773652e-01 -3.62069681e-02
-7.20189452e-01 -8.47985625e-01 -5.65333962e-01 5.48330665e-01
-1.30867934e+00 -4.78839099e-01 8.59036565e-01 -9.54412580e-01
1.03472388e+00 1.52587742e-01 8.31750274e-01 1.21124232e+00
7.28704691e-01 3.20650756e-01 1.35685027e+00 -1.14690089e+00
2.99692452e-01 3.84266138e-01 3.97554904e-01 4.65774834e-01
1.09967887e+00 -2.21776485e-01 -6.05257630e-01 -6.79943740e-01
5.62729239e-01 -2.07988650e-01 -5.01936898e-02 3.71596426e-01
-1.10574818e+00 6.16282403e-01 -8.79116654e-02 8.78637314e-01
-8.92534256e-01 -2.53884315e-01 4.86953497e-01 2.82072335e-01
3.08429331e-01 9.07584071e-01 -2.95156389e-01 -3.39421570e-01
-1.62780929e+00 5.28271675e-01 1.36402488e+00 1.20625162e+00
3.84172916e-01 -2.84778893e-01 -1.16306499e-01 3.47449690e-01
5.09577990e-01 2.14920551e-01 5.99409699e-01 -7.29470849e-01
3.04842323e-01 1.09343398e+00 -2.69938703e-03 -1.07351458e+00
-3.76704603e-01 -3.49286020e-01 -4.39580530e-01 -1.26349419e-01
-1.08632110e-01 1.34366393e-01 -7.62664795e-01 7.61466742e-01
-9.97443199e-02 -3.94129932e-01 3.87501568e-02 3.18943262e-01
1.67367017e+00 8.07788074e-01 3.92395109e-01 -4.21299279e-01
1.44731045e+00 -4.99876261e-01 -1.29562640e+00 2.48823836e-01
5.49882293e-01 -1.17659950e+00 2.77073175e-01 3.91341031e-01
-1.42805123e+00 -1.07498609e-01 -1.08670878e+00 -1.52408928e-01
-1.07914412e+00 5.53825907e-02 6.77240252e-01 3.99222672e-01
-9.45523620e-01 5.60478091e-01 -4.08963025e-01 -1.69012979e-01
6.34697497e-01 2.38120668e-02 -3.69972229e-01 3.02034970e-02
-1.08548141e+00 1.53800035e+00 4.54385579e-01 -2.53928602e-01
-3.18398811e-02 -1.13800144e+00 -4.77489889e-01 3.53792496e-02
3.02309662e-01 -7.37356722e-01 8.45027566e-01 -3.48469794e-01
-8.84781659e-01 1.49033201e+00 -2.80352980e-01 -2.14238971e-01
2.79847205e-01 -1.15496203e-01 -6.79705441e-01 2.90878683e-01
5.73925972e-01 -2.98459474e-02 2.71109283e-01 -1.09748054e+00
-7.95140386e-01 -5.16448140e-01 -7.62954056e-01 -8.16782936e-02
-4.73012447e-01 6.28761649e-01 -8.52502823e-01 -7.70614803e-01
-1.29009500e-01 -5.50759196e-01 1.00464679e-01 -6.13776326e-01
-4.13862646e-01 -7.59079635e-01 5.69345176e-01 -1.06144142e+00
1.58708727e+00 -1.14595544e+00 3.18918884e-01 3.95544231e-01
3.95638078e-01 -4.17665318e-02 5.03736854e-01 8.25682521e-01
-1.82883181e-02 7.42489934e-01 3.97213876e-01 -3.05954516e-02
-5.69446012e-02 -2.61125267e-01 1.71548594e-02 4.15191233e-01
-2.41923168e-01 8.02094281e-01 -7.62032390e-01 -9.83202577e-01
1.07494831e-01 8.66576657e-02 2.92297423e-01 -1.22321760e-02
2.97454242e-02 -4.14219983e-02 -8.46257567e-01 9.63070691e-01
2.59738028e-01 -1.46213964e-01 1.17106676e-01 -1.89350933e-01
-7.07970619e-01 7.10357785e-01 -9.67754364e-01 1.66853106e+00
-2.73886800e-01 7.63294101e-01 -1.91618949e-01 -8.40668082e-01
8.91805828e-01 6.62055612e-01 6.82542145e-01 -7.03160465e-01
3.22011799e-01 7.06418931e-01 -4.74388510e-01 -6.66557789e-01
5.25281787e-01 9.62714255e-02 -9.93273631e-02 3.25036526e-01
4.49831009e-01 -7.08401650e-02 6.03740096e-01 3.29688877e-01
9.25937057e-01 3.34058702e-01 8.42558980e-01 -6.92609191e-01
6.96483672e-01 5.44701695e-01 1.82732686e-01 8.28772724e-01
3.02043557e-01 1.22395553e-01 3.24316502e-01 -2.93748200e-01
-1.55650818e+00 -3.81683916e-01 -7.40183949e-01 2.85255671e-01
-4.40740049e-01 -5.31349480e-01 -6.51700616e-01 -6.44822493e-02
1.22848257e-01 6.20306909e-01 -4.46461201e-01 3.45903158e-01
-4.63527590e-01 -5.08722901e-01 5.07304847e-01 7.05143288e-02
-1.26284495e-01 -1.19972348e+00 -6.18391216e-01 3.25571224e-02
1.49437219e-01 -1.02139056e+00 4.08797115e-01 2.71888316e-01
-7.06787467e-01 -1.17746127e+00 -9.96944606e-01 -9.32398200e-01
1.35890409e-01 -2.06114724e-01 1.27108407e+00 2.66818911e-01
-4.46048647e-01 1.72230601e-02 -4.53621089e-01 -6.72237098e-01
-6.85485542e-01 5.89505553e-01 -2.13780731e-01 -1.15739191e+00
8.84536564e-01 -5.88007450e-01 1.29815415e-01 -5.05173564e-01
-6.63323104e-01 4.55187485e-02 7.62770057e-01 5.20088494e-01
5.64755321e-01 -8.35796446e-02 1.03595412e+00 -7.17120409e-01
1.02051210e+00 -9.55684662e-01 -8.95057678e-01 2.30923697e-01
-1.27051699e+00 1.01199843e-01 2.96780586e-01 1.67218521e-01
-1.11431015e+00 -6.32416308e-01 -4.79643047e-02 -4.21859734e-02
-5.42594373e-01 1.17144489e+00 5.07433414e-02 -1.10946849e-01
4.92543548e-01 -2.77003348e-01 -2.09245399e-01 -8.49275589e-01
3.27581674e-01 1.22191751e+00 4.77072537e-01 -5.24885178e-01
6.16008461e-01 -5.17126694e-02 2.83455849e-01 -8.45436275e-01
-3.44044268e-01 -8.76196861e-01 -8.68752778e-01 -2.23518208e-01
5.03605008e-01 -7.15063691e-01 -5.25946736e-01 -2.00631529e-01
-1.23871768e+00 7.07963288e-01 -3.40757077e-03 6.64949596e-01
3.19779143e-02 -1.07598998e-01 -3.91084015e-01 -6.45020187e-01
-6.16804957e-01 -1.00193059e+00 9.09758747e-01 4.81700122e-01
-6.48414314e-01 -9.80633557e-01 4.70957339e-01 3.64950091e-01
1.34790778e-01 3.98379654e-01 1.06003416e+00 -9.20162320e-01
-6.44811243e-02 -5.93556225e-01 -1.21849284e-01 -4.77834880e-01
8.68176073e-02 4.11808401e-01 -6.43938899e-01 2.62244999e-01
-5.57023287e-02 8.24821368e-02 6.76476479e-01 4.66878027e-01
9.75922585e-01 -7.33142734e-01 -8.51375937e-01 3.15363258e-01
1.73622060e+00 5.84287703e-01 5.45847714e-01 1.26694894e+00
6.42839909e-01 7.84281909e-01 1.09422542e-01 2.58362353e-01
1.01894140e-01 6.22598290e-01 -1.16705716e-01 2.35697269e-01
-5.56311682e-02 3.47520262e-02 -2.78502762e-01 9.12095070e-01
-4.22219843e-01 2.90810950e-02 -1.52684343e+00 8.33042324e-01
-1.53779066e+00 -9.64901149e-01 -6.63251162e-01 1.78682780e+00
1.12117863e+00 3.13970149e-01 1.13364324e-01 1.91775158e-01
5.26582599e-01 -2.53221899e-01 1.23713113e-01 -6.10757709e-01
-1.07037023e-01 6.28078163e-01 8.38131070e-01 5.05222864e-02
-6.82983875e-01 8.64652455e-01 7.55136728e+00 8.96674037e-01
-6.99666858e-01 2.62488611e-02 2.47241417e-03 6.03465457e-03
-4.81495231e-01 9.68221277e-02 -8.47779870e-01 3.68527263e-01
1.54567087e+00 -8.93482208e-01 -1.53919652e-01 9.18958306e-01
9.24641639e-02 -8.67125988e-02 -9.44883645e-01 6.03022456e-01
-1.53412148e-01 -2.43604255e+00 3.13628048e-01 2.71760166e-01
7.42246926e-01 2.56975051e-02 -5.19120336e-01 6.65168231e-03
1.18293367e-01 -1.14193022e+00 7.01755583e-01 7.20404685e-01
6.14542663e-01 -8.97251964e-01 1.09784508e+00 8.98544043e-02
-7.69210935e-01 1.22908786e-01 1.47511531e-02 -5.75655289e-02
3.52585912e-02 7.00430036e-01 -4.52086359e-01 1.17987502e+00
9.28039432e-01 7.48778164e-01 -7.41830826e-01 1.28893340e+00
-1.53871104e-01 6.30492687e-01 1.37902558e-01 -2.59195477e-01
2.89620697e-01 -2.13354871e-01 6.96598411e-01 1.77605057e+00
1.62589043e-01 4.38830033e-02 -2.25851372e-01 1.16786504e+00
-4.78827089e-01 3.86312187e-01 -7.18623936e-01 -3.66644621e-01
1.02916479e+00 1.24402237e+00 -6.89466834e-01 -5.62039673e-01
-6.15393877e-01 2.49942631e-01 -3.52782421e-02 6.78702444e-02
-8.41850862e-02 -6.74363017e-01 -2.50912905e-01 7.77449459e-02
-2.07796007e-01 8.09303299e-02 -8.68161082e-01 -8.15372646e-01
1.09187007e-01 -9.51562405e-01 2.27745339e-01 -7.02655613e-01
-1.09235263e+00 4.01828766e-01 3.28833759e-01 -9.12984490e-01
-4.04989332e-01 -3.97333384e-01 -6.29749000e-01 1.24526632e+00
-1.28025603e+00 -1.18338466e+00 5.56790940e-02 5.93010560e-02
3.00550878e-01 -5.42813182e-01 1.05820274e+00 4.26791817e-01
-5.34114957e-01 1.33898586e-01 4.52696651e-01 -2.12395992e-02
4.47958320e-01 -1.28490746e+00 1.52938306e-01 3.44750464e-01
-1.87971594e-03 8.31540883e-01 5.80559075e-01 -1.16137326e+00
-1.67008471e+00 -5.41303992e-01 1.57630467e+00 -5.51885605e-01
1.08995211e+00 2.87135243e-01 -1.15526927e+00 3.60258400e-01
7.70799279e-01 -9.05895829e-01 8.83518219e-01 2.14303493e-01
5.14440313e-02 3.06498826e-01 -8.48780453e-01 5.99820495e-01
5.78868032e-01 -3.62700939e-01 -1.48695529e+00 4.56085026e-01
3.41278255e-01 -7.69257173e-02 -1.42501903e+00 1.16630174e-01
8.14339995e-01 -2.13450640e-01 9.27967668e-01 -5.50252914e-01
9.94276047e-01 2.81281263e-01 5.02549529e-01 -7.26548553e-01
-2.90182173e-01 -5.92555642e-01 -2.07870051e-01 1.52587605e+00
6.20820820e-01 -9.89744887e-02 4.82724190e-01 5.06267428e-01
-1.95032209e-01 -3.98018330e-01 -1.05662346e+00 -6.80624366e-01
4.49758679e-01 -1.23548910e-01 3.21307033e-01 1.41283619e+00
7.06494451e-01 3.68632883e-01 6.35625273e-02 -6.55018449e-01
7.36162782e-01 7.85409063e-02 3.22403848e-01 -1.85046864e+00
7.44788945e-01 -1.33726597e+00 -3.18147123e-01 1.94515035e-01
3.75410169e-01 -9.70826507e-01 -4.75793540e-01 -2.03163290e+00
3.62424850e-01 -1.59372687e-01 -2.41594225e-01 4.07776833e-01
-3.16811213e-03 -2.20544815e-01 -4.62593943e-01 1.12591195e+00
-9.12682712e-02 -1.28901154e-01 6.11322522e-01 1.40733957e-01
-4.30788517e-01 -2.49459952e-01 -8.93803596e-01 8.63021076e-01
5.80730796e-01 -9.26964939e-01 4.42021191e-01 2.11136624e-01
6.32596791e-01 -3.22452225e-02 8.82747918e-02 -7.23136306e-01
4.55412090e-01 -4.18784767e-01 4.92575020e-01 -9.68999565e-01
-5.34366012e-01 -6.67008758e-01 3.35932285e-01 3.30695093e-01
-5.33229351e-01 3.46025705e-01 4.99870688e-01 1.29036903e-01
-4.13543314e-01 -7.29813099e-01 2.55228192e-01 -5.15161693e-01
-4.75026906e-01 -2.43592858e-01 -7.12945759e-01 -4.28158998e-01
7.33322263e-01 -1.63866490e-01 -4.72793728e-01 2.56960571e-01
-2.19081879e-01 7.71195441e-02 5.98228931e-01 5.83803475e-01
2.82398969e-01 -1.17945707e+00 -7.84312367e-01 -5.36247790e-01
7.13035241e-02 -4.43070561e-01 -4.87408340e-01 7.21333504e-01
-6.84016824e-01 1.02311265e+00 -6.30658448e-01 1.65480465e-01
-1.18991983e+00 5.80166817e-01 -4.17901427e-01 -4.72252220e-01
-8.84871006e-01 1.19280830e-01 -8.46395612e-01 2.97348827e-01
4.51849550e-01 2.91772604e-01 -1.23280025e+00 5.49807012e-01
6.10385954e-01 6.33637607e-01 3.67951512e-01 -7.50016570e-01
-7.19997227e-01 3.36419761e-01 -1.47395641e-01 -3.51517737e-01
1.55233037e+00 1.16260409e-01 -5.05384743e-01 8.15734088e-01
1.09294486e+00 -3.87706165e-03 2.46895358e-01 9.57830250e-03
8.96278203e-01 1.30077198e-01 6.35728061e-01 -9.23966885e-01
-6.09608293e-01 4.04514551e-01 1.86868519e-01 2.38481954e-01
6.43310189e-01 6.62360936e-02 1.28878355e-01 3.48467410e-01
-8.06076154e-02 -1.41555142e+00 -5.96774518e-01 7.46967420e-02
9.45958734e-01 -1.02009666e+00 9.84886050e-01 -2.46891268e-02
-2.74092564e-03 1.64283907e+00 -3.93326767e-02 1.71962515e-01
6.95286751e-01 5.66547394e-01 -3.24665457e-01 -7.91243613e-01
-5.62664866e-01 1.56732365e-01 1.00973797e+00 2.58637279e-01
9.06307817e-01 -3.74034584e-01 -1.45667720e+00 8.49522769e-01
-5.23823127e-02 4.59054768e-01 5.41832030e-01 1.41960514e+00
-3.98571908e-01 -1.16498995e+00 -8.65231514e-01 7.12956011e-01
-1.11811769e+00 -3.95733953e-01 -7.25503564e-01 1.10194218e+00
-1.98615998e-01 7.68212438e-01 -2.80236542e-01 5.23101985e-02
2.17756271e-01 2.22505331e-01 2.95393914e-01 -4.71670151e-01
-1.02469492e+00 4.53872502e-01 3.83769751e-01 -1.43931061e-01
-9.41654384e-01 -6.50985420e-01 -1.17267895e+00 -3.60821694e-01
-2.93340504e-01 6.10205412e-01 1.40252388e+00 1.38022304e+00
4.55885738e-01 9.82599616e-01 1.56470746e-01 -9.10244942e-01
6.17307350e-02 -1.21007395e+00 -1.84863329e-01 -8.53092074e-02
-2.65692435e-02 -6.48603678e-01 -3.83023411e-01 4.19042975e-01] | [9.611987113952637, 8.413188934326172] |
5924c4f8-f2c2-4948-a88a-decec670f4e5 | robust-and-fast-heart-rate-variability-1 | 1902.06151 | null | http://arxiv.org/abs/1902.06151v1 | http://arxiv.org/pdf/1902.06151v1.pdf | Robust and fast heart rate variability analysis of long and noisy electrocardiograms using neural networks and images | Heart rate variability studies depend on the robust calculation of the
tachogram, the heart rate times series, usually by the detection of R peaks in
the electrocardiogram (ECG). ECGs however are subject to a number of sources of
noise which are difficult to filter and therefore reduce the tachogram
accuracy. We describe a pipeline for fast calculation of tachograms from noisy
ECGs of several hours' length. The pipeline consists of three stages. A neural
network (NN) trained to detect R peaks and distinguish these from noise; a
measure to robustly detect false positives (FPs) and negatives (FNs) produced
by the NN; a simple "alarm" algorithm for automatically removing FPs and
interpolating FNs. In addition, we introduce the approach of encoding ECGs,
tachograms and other cardiac time series in the form of raster images, which
greatly speeds and eases their visual inspection and analysis. | [] | 2019-02-16 | robust-and-fast-heart-rate-variability | https://arxiv.org/abs/1902.06151 | https://arxiv.org/pdf/1902.06151 | arxiv190206151-search-help-advanced-search | ['heart-rate-variability'] | ['medical'] | [ 7.65633762e-01 -2.26011112e-01 5.11186182e-01 -4.65594769e-01
-7.92356431e-01 -4.05121714e-01 6.06615543e-02 6.22031033e-01
-2.91074008e-01 6.19645059e-01 -1.85784444e-01 -3.69132340e-01
-3.91474590e-02 -6.88466012e-01 -2.59047985e-01 -5.60594678e-01
-5.19668758e-01 1.10251844e-01 1.65466905e-01 2.25251585e-01
2.87631691e-01 5.69213688e-01 -1.31286705e+00 1.82038084e-01
4.77278918e-01 1.15615928e+00 -3.70965660e-01 1.19461274e+00
4.67217535e-01 5.27771413e-01 -1.10695195e+00 1.70736209e-01
2.01033980e-01 -7.74556577e-01 -3.70302320e-01 -2.44684264e-01
6.24243245e-02 -4.46147114e-01 1.99073628e-01 6.83523238e-01
9.74948645e-01 -3.54109019e-01 3.84258926e-01 -8.50395501e-01
1.53365046e-01 3.98172528e-01 -5.92837334e-01 6.12914741e-01
4.33780015e-01 3.80185723e-01 2.16573894e-01 -6.46482289e-01
6.09452367e-01 7.53515422e-01 1.42373013e+00 1.60251204e-02
-1.41743457e+00 -2.96668559e-01 -7.87467539e-01 -2.74409235e-01
-1.50820041e+00 -3.75016481e-01 6.98286831e-01 -4.50418770e-01
9.95714903e-01 6.85223758e-01 9.13286746e-01 6.66783094e-01
4.85526353e-01 -4.15650532e-02 1.11144745e+00 -5.73045373e-01
2.34381840e-01 -1.80981487e-01 1.49264887e-01 3.10862005e-01
9.38640684e-02 2.18366191e-01 -3.44865441e-01 -2.72956073e-01
1.12174416e+00 1.13598451e-01 -4.40396786e-01 2.89600104e-01
-1.29775703e+00 2.91808873e-01 -1.01294182e-01 2.24564254e-01
-7.29039729e-01 1.04937449e-01 7.26553559e-01 4.75368500e-01
5.25982752e-02 5.15883982e-01 -4.11559194e-01 -3.91507983e-01
-1.29516149e+00 2.07910109e-02 6.28126442e-01 4.16256517e-01
4.54778910e-01 1.27448305e-01 -2.32466429e-01 6.76042855e-01
-3.98202054e-02 3.86902750e-01 5.13531685e-01 -8.88708889e-01
-7.09721670e-02 5.67172825e-01 1.68440547e-02 -9.52123702e-01
-8.07497203e-01 -2.95129359e-01 -9.41247106e-01 1.33614868e-01
6.24015152e-01 -4.53212529e-01 -7.87642777e-01 1.04277217e+00
2.58702546e-01 3.28960307e-02 -3.85680288e-01 8.62537622e-01
7.49508619e-01 3.58061165e-01 -8.87162834e-02 -6.72814727e-01
1.35629570e+00 8.02951455e-02 -7.81648040e-01 1.50897592e-01
5.07357597e-01 -7.20557034e-01 8.11987638e-01 4.72589374e-01
-1.19233608e+00 -6.78683043e-01 -1.11246097e+00 2.57397443e-01
-2.32734025e-01 1.82269245e-01 2.19226882e-01 7.22524524e-01
-9.26830471e-01 1.21134007e+00 -9.75679755e-01 -7.43523687e-02
3.01141053e-01 3.19918454e-01 -1.05353408e-01 5.70403337e-01
-1.12204921e+00 7.62046993e-01 1.08223379e-01 3.48201841e-01
-2.23942429e-01 -8.35650206e-01 -9.53336358e-01 -4.44412753e-02
-2.83386204e-02 -1.05021194e-01 1.10916436e+00 -8.49627733e-01
-1.18367052e+00 9.62744832e-01 7.34771043e-02 -5.93020976e-01
6.20569646e-01 -2.54316907e-02 -4.76885140e-01 4.29263383e-01
-6.55369759e-02 2.25552350e-01 1.14469504e+00 -7.08948195e-01
-4.19625133e-01 -3.29245955e-01 -6.79159343e-01 -2.68565506e-01
3.80228519e-01 1.70526519e-01 -2.29279637e-01 -6.31658733e-01
3.30230236e-01 -6.04247928e-01 -3.35098505e-01 -1.78592458e-01
-3.03853661e-01 3.09828222e-01 5.63382506e-01 -1.02431071e+00
1.51038420e+00 -2.17755079e+00 -7.21478581e-01 5.57406247e-01
3.18410218e-01 4.21833158e-01 3.50266665e-01 2.77959317e-01
-4.22474653e-01 1.66944303e-02 -3.13315451e-01 1.70949772e-01
-3.00770998e-01 4.20848764e-02 -2.14729965e-01 6.36829555e-01
3.34200144e-01 9.16057348e-01 -7.79217958e-01 -3.64950538e-01
5.05291700e-01 6.19945347e-01 1.90246657e-01 2.26937234e-01
3.46745223e-01 4.96224314e-01 2.23693773e-01 4.61148411e-01
4.18905139e-01 -6.46288972e-03 3.01009983e-01 -4.14914489e-01
-5.12399256e-01 4.35415149e-01 -1.20930469e+00 1.09412086e+00
1.85910970e-01 7.10948050e-01 -1.65456697e-01 -8.05974603e-01
1.38061988e+00 5.59019208e-01 5.35209775e-01 -6.23239160e-01
3.00234050e-01 2.57998765e-01 -3.63720134e-02 -5.51095545e-01
2.72438116e-02 -5.86240292e-02 2.48947069e-01 6.56457841e-01
-2.21518859e-01 -2.56952971e-01 3.05639267e-01 -2.06284240e-01
1.16611362e+00 1.74595147e-01 4.99485046e-01 -2.12369844e-01
2.44546115e-01 -1.30732015e-01 5.94840765e-01 8.19092214e-01
-3.08435589e-01 1.04506898e+00 8.56520414e-01 -1.15494096e+00
-8.53912294e-01 -7.96091676e-01 -3.58308107e-01 2.45272577e-01
-3.95122439e-01 -6.37143672e-01 -6.54616714e-01 -2.85806686e-01
-1.46684855e-01 2.34283268e-01 -5.39773881e-01 -9.90368426e-02
-6.98103964e-01 -1.00880837e+00 6.40167832e-01 6.59079134e-01
3.16832177e-02 -1.14632499e+00 -1.72617579e+00 5.32006204e-01
-8.01255926e-02 -7.78506279e-01 -2.39307836e-01 6.62180722e-01
-1.17645383e+00 -1.23032618e+00 -5.36955416e-01 -4.01229382e-01
6.14918172e-01 -2.57096946e-01 1.20111191e+00 2.29229420e-01
-1.22962272e+00 1.63023487e-01 -8.31941813e-02 -9.06745076e-01
-5.52540958e-01 -6.17690802e-01 -1.69652402e-01 -1.19175464e-01
2.99362928e-01 -8.04516554e-01 -8.54106665e-01 2.41231963e-01
-6.78106368e-01 -8.03971663e-02 2.95986086e-01 4.26636159e-01
1.02023041e+00 -1.16647415e-01 7.63378739e-01 -9.07085478e-01
7.06700087e-01 -5.83328269e-02 -8.07921171e-01 -7.39711225e-02
-6.52107120e-01 -4.65790004e-01 7.42007613e-01 -2.05167159e-01
-1.96081325e-01 3.26956809e-01 -1.76389530e-01 -3.04485440e-01
-2.94975787e-01 2.72029012e-01 6.00726664e-01 3.48577760e-02
1.05111563e+00 2.97287498e-02 2.83185482e-01 -2.04511240e-01
-1.91023275e-01 6.89484537e-01 1.06154311e+00 -3.83616164e-02
3.57879966e-01 2.81451136e-01 1.56669632e-01 -1.07356524e+00
-1.71809703e-01 -5.90340555e-01 -9.12054181e-01 -5.17283678e-01
7.25818217e-01 -5.91227531e-01 -6.93443000e-01 4.12238985e-01
-1.04491460e+00 -3.23827982e-01 -6.30784214e-01 4.19307113e-01
-3.18190545e-01 4.62994099e-01 -7.47935653e-01 -1.02927244e+00
-9.05793369e-01 -6.43755615e-01 8.97965074e-01 2.95191914e-01
-8.55963171e-01 -5.98864555e-01 1.52737781e-01 -1.93979338e-01
4.67569441e-01 1.08479965e+00 6.73593104e-01 -2.48163506e-01
-9.19601694e-02 -5.94207704e-01 -5.02444170e-02 3.29283983e-01
1.99912295e-01 4.17071819e-01 -1.26093662e+00 -1.17175449e-02
3.79146338e-01 5.81123456e-02 4.12678570e-01 8.01779509e-01
9.07637596e-01 -5.14517948e-02 -1.73068091e-01 5.04435122e-01
1.32242608e+00 4.68297362e-01 1.10854089e+00 -5.19680157e-02
2.72387654e-01 3.17745626e-01 3.06045055e-01 6.60864830e-01
-1.36959448e-01 3.09870243e-01 1.63154870e-01 -6.51026130e-01
1.38836816e-01 1.77605525e-01 6.82505369e-02 5.41724861e-01
-6.15887105e-01 4.95095432e-01 -9.92760777e-01 4.12659138e-01
-1.51490510e+00 -8.54899943e-01 -8.32323015e-01 2.57825494e+00
8.75049531e-01 2.18366951e-01 4.13060039e-01 9.71722066e-01
7.85333693e-01 -1.66136637e-01 -3.27884376e-01 -5.76456666e-01
1.10888965e-01 6.82997286e-01 3.79088044e-01 1.30726188e-01
-1.07892752e+00 -9.86838862e-02 7.32578754e+00 -2.29790911e-01
-1.54203081e+00 -4.18832421e-01 8.15911472e-01 6.66345283e-02
2.78409690e-01 -3.40403467e-01 -2.72539616e-01 4.33538020e-01
1.38324285e+00 2.68369410e-02 1.24378100e-01 5.90795875e-01
5.35015047e-01 -3.14465612e-01 -8.99958372e-01 1.25225890e+00
-2.97735870e-01 -1.43809688e+00 -5.19642711e-01 -4.57484961e-01
1.58452109e-01 -7.23973513e-02 -6.13524675e-01 -7.22683668e-02
-4.31109816e-01 -9.17764068e-01 5.45856059e-01 6.72913432e-01
1.28102291e+00 -7.09972322e-01 7.95896113e-01 -1.10199682e-01
-1.03819668e+00 1.36099860e-01 -2.08707795e-01 -9.59978625e-02
1.50459856e-01 1.13824880e+00 -9.81204391e-01 2.85017818e-01
7.67666578e-01 4.87287641e-01 -5.34534037e-01 1.10077035e+00
-1.14250265e-01 6.10564351e-01 -4.10027564e-01 3.80541831e-01
-3.86185259e-01 -2.37386525e-02 3.84052932e-01 1.44463587e+00
1.47448540e-01 2.95149833e-01 -1.31970748e-01 8.97109032e-01
4.57649678e-01 2.44695228e-02 -3.69205058e-01 -1.24901459e-02
4.57599342e-01 1.44814575e+00 -1.20776117e+00 -3.40410352e-01
-3.00816268e-01 5.97417712e-01 -4.33479160e-01 8.89977589e-02
-7.23299444e-01 -1.07227468e+00 -4.35488373e-02 5.28262854e-01
-1.75223798e-02 1.77787334e-01 -6.20129049e-01 -5.88921428e-01
2.57129341e-01 -1.01865804e+00 6.50645912e-01 -7.04323947e-01
-8.42533529e-01 7.12292314e-01 -4.07619804e-01 -1.22399414e+00
-4.27038550e-01 -4.14973229e-01 -1.01389861e+00 1.16791594e+00
-9.47199225e-01 -3.54105920e-01 -5.34049988e-01 4.76165980e-01
1.80144608e-01 6.24221981e-01 1.24497461e+00 3.47798407e-01
-3.25173467e-01 1.79943293e-01 -4.95493799e-01 2.79969901e-01
5.78793287e-01 -1.52383113e+00 6.60002947e-01 7.13127077e-01
-2.13076547e-01 6.14046335e-01 6.66994512e-01 -7.03618646e-01
-1.08398581e+00 -9.81705666e-01 1.08501482e+00 -3.31157297e-01
1.89718619e-01 -1.09592803e-01 -9.09345746e-01 3.94829154e-01
-3.18194360e-01 2.45474607e-01 5.61227858e-01 -2.46181935e-01
1.44768283e-01 -2.76312917e-01 -1.10638011e+00 2.66976267e-01
1.13727435e-01 -4.93787974e-01 -5.14763772e-01 9.85578597e-02
-5.90027347e-02 -5.49625218e-01 -1.17656696e+00 4.13615704e-01
9.09826458e-01 -1.26999784e+00 8.48558724e-01 3.29701579e-03
1.68922544e-01 -4.66422170e-01 6.19411170e-01 -8.88887048e-01
3.99591923e-02 -1.28953171e+00 1.37647301e-01 8.86879921e-01
2.27572203e-01 -4.63939786e-01 3.66964459e-01 5.86602211e-01
-9.28024650e-02 -5.44927955e-01 -6.66805506e-01 -3.92785579e-01
-8.08539450e-01 -5.16202331e-01 2.47246742e-01 9.75510299e-01
3.59792113e-01 2.44458482e-01 -1.68998316e-01 -7.05402419e-02
5.85484266e-01 -7.02794567e-02 3.61811727e-01 -1.44097996e+00
-1.89338446e-01 -1.32475853e-01 -5.93516350e-01 -6.91474825e-02
-1.09357285e+00 -3.68095189e-01 2.84533411e-01 -1.14628816e+00
-3.70352328e-01 -2.38644361e-01 -2.68282920e-01 5.45074642e-01
-1.68790698e-01 7.58360326e-01 -2.87981242e-01 7.29681551e-02
-1.36391804e-01 -4.67210621e-01 6.93428516e-01 4.27044481e-01
-7.88370311e-01 1.91452801e-01 -1.76290810e-01 8.44268739e-01
7.84257472e-01 -6.32450283e-01 -1.96546853e-01 2.00563729e-01
2.39677384e-01 5.56738019e-01 5.12621999e-01 -1.19617462e+00
-2.71193027e-01 4.80743736e-01 9.30836380e-01 -9.03175533e-01
-1.29647613e-01 -4.16593343e-01 6.37459576e-01 7.97502100e-01
-1.78726360e-01 6.12519026e-01 2.91493148e-01 1.31403118e-01
-5.66329211e-02 -9.70605910e-02 1.07836306e+00 -3.53395611e-01
1.08492315e-01 -1.21837027e-01 -6.44432724e-01 -2.42508780e-02
6.17270231e-01 -5.73627770e-01 4.99924691e-03 -1.77210301e-01
-9.43225265e-01 -1.35392845e-01 1.01015598e-01 -9.47591066e-02
5.87516189e-01 -9.51360166e-01 -5.34773290e-01 6.39765739e-01
-1.89070627e-01 -7.41930753e-02 3.05195332e-01 1.45772862e+00
-1.05306566e+00 -5.20342477e-02 -4.70804065e-01 -8.60628963e-01
-1.49269152e+00 2.09973678e-01 5.99187076e-01 -1.61235869e-01
-1.09398210e+00 4.92679834e-01 -4.85806525e-01 3.07156354e-01
3.31728518e-01 -6.62512720e-01 -3.67192984e-01 1.85042650e-01
1.08970678e+00 6.58960283e-01 4.73707795e-01 -2.23076656e-01
-4.75390643e-01 3.23989302e-01 3.38443339e-01 3.38200592e-02
1.31349242e+00 -9.18314233e-02 -1.87096998e-01 5.98638833e-01
6.93736434e-01 -1.36034518e-01 -1.10941172e+00 4.02700871e-01
8.77937451e-02 -1.39079556e-01 -2.22718865e-01 -7.60393918e-01
-7.71713078e-01 9.06657517e-01 8.07427943e-01 6.22466743e-01
1.42685795e+00 -6.58073485e-01 6.98425293e-01 -1.71223909e-01
-7.86361173e-02 -1.24339032e+00 -3.38535219e-01 5.78541309e-02
6.33773863e-01 -5.22596240e-01 7.49856904e-02 -1.96536437e-01
-3.53090733e-01 1.66544485e+00 -5.40616363e-02 -1.21013857e-01
5.20551145e-01 5.84354758e-01 7.74963558e-01 -3.43892604e-01
-5.24272323e-01 8.05308968e-02 2.79463101e-02 8.33706141e-01
6.88596010e-01 -8.71801600e-02 -5.30483305e-01 6.03208482e-01
-1.21643379e-01 4.71956223e-01 6.20774627e-01 1.08402371e+00
-4.20031935e-01 -6.14452720e-01 -5.86164892e-01 6.76443338e-01
-9.26587045e-01 -4.96487273e-03 -3.13623607e-01 4.96148527e-01
2.21430555e-01 8.28344464e-01 1.04239278e-01 -1.80578396e-01
4.90659118e-01 3.20636958e-01 3.03036720e-01 -4.00944412e-01
-1.07559347e+00 4.93264467e-01 1.14745453e-01 -7.57100880e-01
-1.86597332e-01 -4.60913122e-01 -1.33244765e+00 1.57219872e-01
-1.89637974e-01 -5.36047900e-03 7.34935343e-01 7.56640971e-01
1.97102532e-01 7.09594905e-01 4.68041599e-01 -7.17212260e-01
-3.07681143e-01 -7.72091329e-01 -7.57614613e-01 5.19198895e-01
5.11272490e-01 1.73723623e-01 -2.69469112e-01 6.69956088e-01] | [14.195388793945312, 3.165226697921753] |
302f5b10-be38-4b4d-9f9a-a908d45cf296 | a-large-scale-comparative-study-of-accurate | 2304.04811 | null | https://arxiv.org/abs/2304.04811v2 | https://arxiv.org/pdf/2304.04811v2.pdf | A Large-Scale Comparative Study of Accurate COVID-19 Information versus Misinformation | The COVID-19 pandemic led to an infodemic where an overwhelming amount of COVID-19 related content was being disseminated at high velocity through social media. This made it challenging for citizens to differentiate between accurate and inaccurate information about COVID-19. This motivated us to carry out a comparative study of the characteristics of COVID-19 misinformation versus those of accurate COVID-19 information through a large-scale computational analysis of over 242 million tweets. The study makes comparisons alongside four key aspects: 1) the distribution of topics, 2) the live status of tweets, 3) language analysis and 4) the spreading power over time. An added contribution of this study is the creation of a COVID-19 misinformation classification dataset. Finally, we demonstrate that this new dataset helps improve misinformation classification by more than 9\% based on average F1 measure. | ['Xingyi Song', 'Kalina Bontcheva', 'Carolina Scarton', 'Iknoor Singh', 'Freddy Heppell', 'Ye Jiang', 'Yida Mu'] | 2023-04-10 | null | null | null | null | ['misinformation'] | ['miscellaneous'] | [-1.92992181e-01 -4.85504940e-02 -4.94904786e-01 1.20411217e-01
-6.78000808e-01 -8.72944951e-01 1.41715741e+00 1.06478214e+00
-6.04004562e-01 7.42217362e-01 8.09259832e-01 -4.62185174e-01
1.28263384e-01 -1.02636850e+00 -3.08937103e-01 -3.25955361e-01
-2.05324262e-01 4.97271240e-01 2.32226670e-01 -5.41849554e-01
3.72056752e-01 1.09254293e-01 -1.32018042e+00 7.80547932e-02
1.00122058e+00 6.09256923e-01 -5.74541502e-02 5.77854633e-01
-3.50467026e-01 5.72735667e-01 -9.16350365e-01 -3.99884522e-01
-1.99393079e-01 -2.74219751e-01 -6.74833357e-01 -6.14509284e-01
1.42514378e-01 -3.59841764e-01 -1.11517988e-01 1.00584173e+00
3.73732150e-01 -2.36936003e-01 8.88918757e-01 -1.20511281e+00
-1.86215028e-01 5.96723497e-01 -4.91582572e-01 7.84745216e-01
6.03676975e-01 -5.56933694e-02 6.32255197e-01 -4.11239326e-01
1.05708504e+00 1.19333601e+00 9.62107897e-01 -2.68136382e-01
-9.17429805e-01 -8.58689785e-01 -6.24752566e-02 -1.82036757e-01
-1.63801074e+00 -6.82343617e-02 2.44066030e-01 -1.14052463e+00
6.74689174e-01 4.66409236e-01 8.51688504e-01 1.29475379e+00
5.29218256e-01 1.43908009e-01 1.20527017e+00 1.14307657e-01
2.41119519e-01 5.27095675e-01 3.75611126e-01 2.86705554e-01
7.97738552e-01 -1.73552372e-02 -1.72858104e-01 -6.04512155e-01
-1.24571025e-01 2.44724780e-01 1.36411814e-02 8.01164687e-01
-1.07424748e+00 1.49599552e+00 4.34523314e-01 5.82252622e-01
-4.05934334e-01 -3.70934397e-01 5.48664868e-01 3.80546570e-01
1.25841415e+00 3.70167553e-01 -1.98681235e-01 -3.59121531e-01
-8.34189296e-01 4.43944871e-01 8.72086763e-01 5.15354514e-01
7.32563972e-01 -2.34901533e-01 1.79530337e-01 5.60461819e-01
1.46959454e-01 1.03870177e+00 3.41363102e-01 -2.83764720e-01
7.33522296e-01 4.40581858e-01 3.28657866e-01 -2.12773299e+00
-7.84094453e-01 -6.14557803e-01 -7.70405352e-01 -5.05953312e-01
3.85597259e-01 -6.70216084e-01 -4.99638438e-01 1.59240162e+00
2.72819221e-01 -1.29398704e-01 -3.95053238e-01 5.66881120e-01
8.02161336e-01 8.99915814e-01 2.36028284e-01 -2.93803096e-01
1.36964154e+00 -1.25646532e-01 -8.16415131e-01 -6.46722317e-02
7.32395768e-01 -8.84045064e-01 3.09600502e-01 -8.04756507e-02
-6.53479695e-01 -5.04311547e-02 -7.98260391e-01 4.20157611e-01
-1.18162370e+00 -7.66429961e-01 3.72996926e-01 7.89604664e-01
-8.76377702e-01 1.72637347e-02 -1.72120705e-01 -6.05198503e-01
3.24406534e-01 -3.99743944e-01 3.19219828e-02 1.16168760e-01
-1.61535394e+00 9.49739337e-01 3.18187326e-01 -7.27860093e-01
-6.99065328e-01 -1.00837004e+00 -5.42222381e-01 -5.44654012e-01
1.39898006e-02 -5.02328396e-01 8.80853772e-01 -7.92473972e-01
-4.87603009e-01 9.10388887e-01 -1.85188636e-01 -4.80884373e-01
7.61704922e-01 1.51601434e-02 -8.54461014e-01 2.38767952e-01
6.33072734e-01 7.17251822e-02 6.12351477e-01 -1.10381973e+00
-8.17443848e-01 -4.45045680e-01 -1.38979614e-01 -2.40884408e-01
-2.41727769e-01 3.73898566e-01 1.40569881e-01 -9.48405504e-01
-2.94419438e-01 -7.38548398e-01 1.93566401e-02 -7.27729738e-01
-3.88147891e-01 -3.02499607e-02 7.70423412e-01 -9.11541045e-01
1.58940387e+00 -1.89612806e+00 -5.51094055e-01 3.25727701e-01
7.25262523e-01 3.63102019e-01 2.50072956e-01 1.12647557e+00
2.87864089e-01 7.90368557e-01 9.54447687e-02 -6.62540272e-02
-1.59307390e-01 -8.54885131e-02 -5.90430379e-01 9.42272663e-01
-1.78152546e-01 6.92506611e-01 -1.15637910e+00 -2.89041758e-01
-1.44772172e-01 4.38976258e-01 -4.58883196e-01 -2.35404059e-01
-1.67765394e-01 2.77670294e-01 -6.60388052e-01 4.01945233e-01
9.52169597e-01 -4.56159323e-01 -1.55843943e-01 1.91743463e-01
-7.24507689e-01 3.41781467e-01 -4.23258901e-01 6.55420721e-01
-1.89203024e-01 9.84699249e-01 -3.19211148e-02 -4.14027542e-01
6.17033362e-01 2.10040152e-01 6.18257046e-01 -5.81893981e-01
4.50624526e-01 -7.42950141e-02 -2.88272917e-01 -3.25317949e-01
8.99337828e-01 -9.90356803e-02 -4.48820651e-01 8.89890969e-01
-5.52141428e-01 2.40006372e-02 1.10960260e-01 5.19855678e-01
7.10450649e-01 -1.17700446e+00 4.16042984e-01 -6.02889836e-01
1.56235442e-01 5.14433384e-01 -3.53883766e-02 8.86843562e-01
-1.79758161e-01 2.59401917e-01 3.04090142e-01 -4.72385943e-01
-7.31440902e-01 -8.79710615e-01 -5.61709464e-01 8.31317425e-01
4.18739654e-02 -6.43013120e-01 -7.48423338e-01 -4.73063678e-01
3.17453235e-01 8.94502163e-01 -8.24593663e-01 3.08135360e-01
-8.70384425e-02 -1.19743359e+00 8.40884924e-01 -2.78472483e-01
6.49582863e-01 -2.65472949e-01 -3.14613938e-01 1.32126495e-01
-8.54473889e-01 -1.03692949e+00 -2.92098075e-01 -4.73319799e-01
-4.04223442e-01 -1.04547882e+00 -6.85571849e-01 -3.60131204e-01
4.98095930e-01 6.30378783e-01 9.92365301e-01 1.66251734e-01
-1.75213460e-02 3.72630030e-01 -5.05489349e-01 -8.89934003e-01
-6.62090361e-01 2.38630876e-01 2.62758553e-01 -1.19844444e-01
5.08620858e-01 -2.83814639e-01 -4.41931963e-01 2.64191628e-01
-7.92858124e-01 -2.83949912e-01 -9.06413887e-03 9.51894522e-02
-1.50952891e-01 4.13731843e-01 7.50003219e-01 -9.14371789e-01
8.77618909e-01 -1.48030436e+00 -3.52045745e-01 -4.20035094e-01
-7.21447587e-01 -6.03184700e-01 2.19090149e-01 -1.08202033e-01
-8.83022010e-01 -1.04574358e+00 -2.56885976e-01 7.56613135e-01
-7.03786090e-02 8.64392161e-01 8.88722956e-01 3.31971049e-01
8.95359218e-01 -5.72758541e-03 4.05521877e-02 -4.29278195e-01
1.53824925e-01 1.19969785e+00 7.77560323e-02 2.10296169e-01
9.69394207e-01 8.31314504e-01 -6.67222798e-01 -1.71027398e+00
-8.10556531e-01 -7.90318251e-01 -2.62273252e-01 -3.25474322e-01
1.03185809e+00 -1.11602020e+00 -6.23084128e-01 7.86429048e-01
-1.16097987e+00 -3.57133299e-02 3.46005797e-01 3.77174944e-01
2.10968569e-01 2.99101114e-01 -6.83105707e-01 -8.67332518e-01
-2.25145116e-01 -7.31700122e-01 3.74032289e-01 -2.18106985e-01
-7.66857445e-01 -1.48402655e+00 5.88569522e-01 5.10343432e-01
1.00308669e+00 7.45619774e-01 5.62628746e-01 -1.04185057e+00
-1.44518465e-01 -4.15327013e-01 -7.21000314e-01 -4.64624763e-01
2.57633179e-01 -1.49948135e-01 -8.94903064e-01 -3.82380992e-01
-1.72169521e-01 -1.61997616e-01 8.48811626e-01 2.99687445e-01
2.81398952e-01 -9.72615004e-01 -7.12453306e-01 1.59919634e-01
1.43222964e+00 -6.86030760e-02 1.88660011e-01 5.19472897e-01
3.26727450e-01 6.40609384e-01 3.46349269e-01 8.86963308e-01
6.97232366e-01 5.28741062e-01 3.82971644e-01 1.84434265e-01
7.22467294e-03 -5.68271697e-01 2.92336136e-01 1.05611992e+00
-1.50374258e-02 -4.44304138e-01 -1.23359656e+00 6.40256166e-01
-1.11455786e+00 -1.22542787e+00 -5.22973597e-01 1.78200614e+00
7.41978168e-01 -3.36041413e-02 6.18569136e-01 -3.06919590e-02
7.69198120e-01 5.70138156e-01 6.97150826e-02 -1.98635548e-01
-1.05485030e-01 -5.15413404e-01 7.61952579e-01 8.47501457e-01
-1.09469497e+00 3.49630624e-01 7.44145966e+00 7.06236362e-01
-1.18388057e+00 2.80701607e-01 4.71671045e-01 4.01529580e-01
-7.61040092e-01 -4.93667185e-01 -9.76018965e-01 1.02735913e+00
1.29300523e+00 -5.52593291e-01 8.87285173e-02 3.53048921e-01
5.92770636e-01 -3.23217392e-01 -7.21523315e-02 6.44670725e-01
3.64938051e-01 -1.62636352e+00 5.22270836e-02 3.51000220e-01
8.75341296e-01 7.60527730e-01 4.02670875e-02 -4.61697988e-02
4.91625279e-01 -8.22906077e-01 6.70932710e-01 3.26265246e-01
8.21627498e-01 -7.10558236e-01 9.49753463e-01 7.12152302e-01
-8.51892591e-01 2.44209263e-03 -6.80289268e-02 -2.52974004e-01
4.50808257e-01 1.08197510e+00 -1.17324996e+00 3.62730145e-01
5.36851764e-01 1.06114209e+00 -5.80781698e-01 8.03870857e-01
1.90873727e-01 8.66542518e-01 -1.65648535e-01 -5.17026365e-01
4.97968435e-01 -1.39953107e-01 9.90770459e-01 1.76550329e+00
2.97570497e-01 -4.95012924e-02 -7.96493050e-03 6.25262499e-01
1.59540072e-01 1.16017498e-01 -1.01497149e+00 -5.18447340e-01
5.35602748e-01 7.33728766e-01 -5.29338360e-01 -4.05520916e-01
-2.37412721e-01 2.64459014e-01 -5.50498962e-02 3.15038711e-01
-5.93075395e-01 -1.97353169e-01 7.68582642e-01 5.03161907e-01
-2.56380793e-02 -3.17091286e-01 -2.26280913e-01 -9.01470780e-01
-5.78162551e-01 -6.49645329e-01 6.14396989e-01 -3.12651664e-01
-1.46406090e+00 5.45879066e-01 2.72826135e-01 -8.73310447e-01
-3.52294207e-01 -2.35863272e-02 -5.17391205e-01 7.57541597e-01
-1.58907449e+00 -7.02462971e-01 -4.17534560e-01 2.08729237e-01
1.66552052e-01 -6.98039541e-03 6.53696775e-01 3.64520103e-01
-2.86737144e-01 3.21041346e-01 3.07665467e-01 1.02098882e-01
4.25277233e-01 -7.84789801e-01 4.30559218e-01 4.63166475e-01
-6.15741313e-01 4.66416299e-01 9.75118279e-01 -1.06870961e+00
-8.23821187e-01 -1.37508404e+00 1.51531732e+00 -8.11769962e-01
1.10878730e+00 -2.57871777e-01 -4.05138791e-01 5.90935111e-01
4.36125547e-01 -9.31763351e-01 9.15842056e-01 -1.09826133e-01
-4.78193104e-01 1.83334813e-01 -1.43609107e+00 5.74630201e-01
7.61594474e-01 -5.60987055e-01 -6.81611538e-01 5.37266374e-01
9.18395936e-01 1.97188735e-01 -8.58318150e-01 -2.12004423e-01
5.37291288e-01 -8.40155065e-01 8.74692202e-01 -5.12456238e-01
3.66269201e-01 3.79192308e-02 -1.41526446e-01 -1.48959780e+00
-4.03173864e-01 -7.63157845e-01 5.18362880e-01 1.21135759e+00
6.21579289e-01 -1.26114357e+00 1.90678641e-01 2.11054727e-01
2.47690678e-01 -2.80320533e-02 -6.21531665e-01 -5.36034107e-01
2.86957949e-01 -5.78987896e-01 4.42952245e-01 1.38561308e+00
2.07442850e-01 1.25430718e-01 -2.18163684e-01 6.34601247e-03
5.86488008e-01 -1.17355369e-01 7.25989819e-01 -1.36651099e+00
3.12408328e-01 -4.94703323e-01 -3.26439917e-01 -7.45741665e-01
-2.91775882e-01 -8.97398353e-01 -4.69093800e-01 -1.34330177e+00
4.67428297e-01 -6.02644145e-01 3.51020932e-01 -1.60095051e-01
2.12159097e-01 4.36112314e-01 5.98404929e-03 3.60271692e-01
-4.37527269e-01 4.54886146e-02 1.00978041e+00 -3.74321282e-01
-1.85500398e-01 -2.68269386e-02 -8.24486732e-01 7.70786881e-01
9.87869024e-01 -4.76604223e-01 -3.21650088e-01 -1.39238223e-01
1.02353573e+00 -1.64264530e-01 2.05450177e-01 -6.37430191e-01
-1.35162822e-03 -1.31570548e-01 1.25323996e-01 -9.96841252e-01
1.28459126e-01 -4.63831991e-01 3.13899070e-01 8.44495714e-01
-1.83814868e-01 5.47900677e-01 2.51341850e-01 6.90515995e-01
-1.20678402e-01 1.29738182e-01 5.08103251e-01 -7.92975351e-02
-2.18968436e-01 1.62268013e-01 -1.27353966e+00 8.60763490e-01
9.73106027e-01 1.56830445e-01 -1.04444778e+00 -8.14245462e-01
-1.47015676e-01 -3.29945572e-02 5.58049977e-01 3.30685019e-01
2.31611371e-01 -1.06051862e+00 -1.09659278e+00 7.74455443e-02
1.70403719e-01 -7.09872961e-01 9.45674554e-02 1.09573889e+00
-6.38304830e-01 9.04127479e-01 4.94879559e-02 -3.97654861e-01
-1.05484092e+00 4.86119568e-01 -1.54532744e-02 -3.76003347e-02
-4.83160764e-01 2.82889903e-01 -3.45194712e-02 -2.47778505e-01
-1.76993683e-02 -8.26319307e-02 -4.60139960e-01 9.43540931e-01
1.05959737e+00 8.02140236e-01 -1.67148650e-01 -1.26823831e+00
-5.05615294e-01 3.22636724e-01 5.72220236e-03 -1.94174975e-01
1.14469445e+00 -5.51319420e-01 -1.44726440e-01 5.55473566e-01
1.65909731e+00 4.21911776e-01 -3.24153721e-01 -2.07228027e-02
7.35380175e-03 -6.34229183e-01 9.45783034e-02 -8.54031503e-01
-4.47527409e-01 3.10006440e-01 7.32833073e-02 1.05454278e+00
3.97165656e-01 1.46577016e-01 1.06338978e+00 -1.25978291e-01
2.24165916e-01 -9.34268057e-01 -4.77792211e-02 6.38705730e-01
9.47973251e-01 -1.11628008e+00 6.50802925e-02 -2.97700971e-01
-4.34225321e-01 5.31561852e-01 -4.06884700e-01 2.14586347e-01
1.33063102e+00 9.47593823e-02 2.48164669e-01 -6.34176195e-01
-3.94009709e-01 -5.43414056e-02 1.89044937e-01 6.66341424e-01
1.98876828e-01 3.26215774e-01 -7.45786369e-01 3.68035942e-01
-5.77713847e-01 -2.57569909e-01 6.99716270e-01 5.67168832e-01
-5.93923926e-01 -5.02099216e-01 -3.57148379e-01 5.92201948e-01
-9.56420064e-01 5.82376234e-02 -5.65874040e-01 8.70742977e-01
4.33908813e-02 1.48992395e+00 2.72936374e-01 -5.79875588e-01
-3.42128575e-01 -2.85916686e-01 -3.67214471e-01 -2.27971971e-01
-5.79674423e-01 -4.85877424e-01 6.41445577e-01 -2.62970448e-01
-3.73793095e-01 -9.33303773e-01 -7.98450828e-01 -1.00127566e+00
-1.28478989e-01 3.27890843e-01 8.15508544e-01 8.95716131e-01
4.21050429e-01 -2.95391195e-02 8.78130198e-01 -4.37355846e-01
-3.23162556e-01 -8.58072877e-01 -5.00499010e-01 4.47058529e-01
8.15558434e-01 -4.53507751e-01 -9.45328951e-01 -3.70923758e-01] | [8.458965301513672, 9.811025619506836] |
793dd14b-fad4-4338-a07c-53369dc376d6 | nflat-non-flat-lattice-transformer-for | 2205.05832 | null | https://arxiv.org/abs/2205.05832v3 | https://arxiv.org/pdf/2205.05832v3.pdf | NFLAT: Non-Flat-Lattice Transformer for Chinese Named Entity Recognition | Recently, Flat-LAttice Transformer (FLAT) has achieved great success in Chinese Named Entity Recognition (NER). FLAT performs lexical enhancement by constructing flat lattices, which mitigates the difficulties posed by blurred word boundaries and the lack of word semantics. In FLAT, the positions of starting and ending characters are used to connect a matching word. However, this method is likely to match more words when dealing with long texts, resulting in long input sequences. Therefore, it significantly increases the memory and computational costs of the self-attention module. To deal with this issue, we advocate a novel lexical enhancement method, InterFormer, that effectively reduces the amount of computational and memory costs by constructing non-flat lattices. Furthermore, with InterFormer as the backbone, we implement NFLAT for Chinese NER. NFLAT decouples lexicon fusion and context feature encoding. Compared with FLAT, it reduces unnecessary attention calculations in "word-character" and "word-word". This reduces the memory usage by about 50% and can use more extensive lexicons or higher batches for network training. The experimental results obtained on several well-known benchmarks demonstrate the superiority of the proposed method over the state-of-the-art hybrid (character-word) models. | ['Xiao-Jun Wu', 'ZhenHua Feng', 'Xiaoning Song', 'Shuang Wu'] | 2022-05-12 | null | null | null | null | ['chinese-named-entity-recognition'] | ['natural-language-processing'] | [-2.19604447e-02 -2.29275599e-01 7.34208673e-02 -2.05423534e-01
-5.06765902e-01 -3.88671875e-01 2.23202109e-01 3.26918721e-01
-1.15563023e+00 7.28428960e-01 3.45651448e-01 -4.50664610e-01
4.35275108e-01 -1.07269776e+00 -3.43500257e-01 -6.72835886e-01
4.90538865e-01 1.80994317e-01 3.22795242e-01 -1.77140489e-01
1.46269128e-01 2.54657686e-01 -1.14673328e+00 2.66042501e-01
1.14104784e+00 7.85583019e-01 6.61010623e-01 1.14839710e-02
-9.26085711e-01 4.10628885e-01 -6.72541916e-01 -8.00558865e-01
2.13070493e-02 -1.67138785e-01 -6.77043915e-01 -2.74845839e-01
3.61116268e-02 -1.48759604e-01 -1.98461771e-01 1.24065387e+00
8.15550625e-01 3.15383852e-01 1.84815362e-01 -5.22452474e-01
-7.89251089e-01 1.03737175e+00 -5.59487939e-01 3.69429320e-01
1.33350110e-02 -2.83859193e-01 1.02242482e+00 -1.23354781e+00
3.87018204e-01 1.15543282e+00 8.22794676e-01 4.32936758e-01
-9.01430547e-01 -7.56117761e-01 3.84861737e-01 2.58164257e-01
-1.64454782e+00 -3.44198704e-01 4.19916451e-01 7.40838051e-02
1.34764373e+00 1.79336175e-01 3.00929278e-01 7.38877594e-01
1.19801583e-02 8.25446129e-01 7.46948481e-01 -7.06384957e-01
6.33538216e-02 1.07052885e-01 4.26308632e-01 5.43923199e-01
2.52646536e-01 -3.81434590e-01 -1.64000213e-01 -3.89440134e-02
7.30907857e-01 6.54354244e-02 -3.47690165e-01 2.29818419e-01
-9.86550748e-01 7.30627239e-01 2.43899360e-01 5.70183158e-01
-4.22433943e-01 -2.76375473e-01 7.18977094e-01 5.67163490e-02
4.63266343e-01 3.16631854e-01 -4.65416461e-01 -1.78000122e-01
-8.03674579e-01 -2.97991991e-01 6.48188233e-01 1.11851883e+00
7.37477243e-01 1.55234858e-01 -1.71274796e-01 1.15890896e+00
-2.81440821e-02 3.58327538e-01 8.33276868e-01 -1.26352906e-01
7.24888742e-01 7.41162062e-01 -1.41012788e-01 -8.29286575e-01
-5.10058641e-01 -5.32063365e-01 -1.08295178e+00 -5.14952302e-01
1.63959920e-01 -3.20128143e-01 -9.08377588e-01 1.70049989e+00
1.19748555e-01 2.41793811e-01 1.11413352e-01 6.56624079e-01
8.60446215e-01 9.76046503e-01 3.91022116e-01 -2.80316025e-01
1.65605175e+00 -1.12339711e+00 -9.61820781e-01 -4.05460685e-01
8.06372583e-01 -9.80531514e-01 1.39804578e+00 -8.89549702e-02
-1.00975180e+00 -6.14038825e-01 -8.75659049e-01 -2.27765948e-01
-6.64536059e-01 4.07386690e-01 4.46908414e-01 7.06231296e-01
-7.72404373e-01 4.72830325e-01 -7.07516074e-01 -2.83086538e-01
6.31272495e-02 2.93799490e-01 -2.94884980e-01 -4.71815951e-02
-1.69928765e+00 8.59816134e-01 8.15300822e-01 3.74593168e-01
5.25419740e-03 -5.63566267e-01 -1.04842615e+00 5.26023448e-01
3.58714461e-01 -2.21095622e-01 9.53819990e-01 -5.99528134e-01
-1.32991493e+00 4.14917052e-01 -3.49863350e-01 -2.53719300e-01
1.08789898e-01 -4.23472852e-01 -7.24237680e-01 -2.70266771e-01
1.33102825e-02 6.22333288e-01 4.31587934e-01 -6.23806417e-01
-6.49782896e-01 -1.55991226e-01 -1.40263408e-01 3.21867198e-01
-9.48548973e-01 1.53469399e-01 -8.79047275e-01 -9.94629622e-01
-2.59692725e-02 -6.04650736e-01 -3.65127265e-01 -6.05930865e-01
-2.26328254e-01 -3.24220747e-01 6.41514838e-01 -7.27099299e-01
1.95848191e+00 -2.20190430e+00 -2.45322898e-01 9.22514498e-02
-7.81677961e-02 8.20656955e-01 -2.44526401e-01 5.10692656e-01
-1.50290122e-02 2.39640221e-01 -8.79172608e-02 -3.54108632e-01
-1.59771562e-01 2.83516258e-01 -7.58571327e-02 8.18723515e-02
3.36309701e-01 9.16184962e-01 -6.75492585e-01 -6.04296148e-01
5.84132746e-02 5.79051077e-01 -5.48042655e-01 6.56788424e-02
1.02930635e-01 -6.48910478e-02 -2.76860505e-01 3.72771293e-01
8.67520273e-01 -1.20200187e-01 4.05762434e-01 -2.70637304e-01
-3.75348449e-01 7.48809695e-01 -1.34459794e+00 1.48346817e+00
-7.93499708e-01 1.80299401e-01 -2.04715297e-01 -7.40941882e-01
1.14058709e+00 4.63619769e-01 -7.63152316e-02 -9.18327034e-01
3.10709506e-01 2.47186512e-01 -1.79049015e-01 -2.20903471e-01
6.99362755e-01 -4.20297123e-02 -2.00076640e-01 1.40327111e-01
1.97296799e-03 4.91366625e-01 4.00652647e-01 -5.64491004e-02
9.79083359e-01 -1.14017148e-02 5.20671785e-01 -3.52383792e-01
8.84692013e-01 -1.84736773e-01 1.02178252e+00 3.41679871e-01
3.17064933e-02 2.66229451e-01 2.65552998e-01 -4.61401045e-01
-1.00530422e+00 -8.05837333e-01 -5.39656244e-02 1.32152510e+00
1.60646856e-01 -6.88544989e-01 -8.95107448e-01 -6.02273643e-01
-4.10408586e-01 8.70800853e-01 -2.63393760e-01 -1.14280030e-01
-1.06235242e+00 -8.81525457e-01 9.04550254e-01 8.82800937e-01
6.52907968e-01 -1.41124809e+00 -3.61615360e-01 7.63642132e-01
-5.01794815e-01 -1.25058365e+00 -1.02203715e+00 3.56810153e-01
-8.41064811e-01 -4.87977535e-01 -7.35155642e-01 -1.32371688e+00
7.26295829e-01 2.60837346e-01 1.04157078e+00 1.40587315e-01
-8.67859125e-02 -4.62532133e-01 -6.43481731e-01 -1.47265315e-01
-9.42253470e-02 4.71110702e-01 -1.34312227e-01 -8.75529498e-02
7.70623326e-01 -3.23663652e-01 -3.15786302e-01 3.53773564e-01
-1.09990036e+00 9.25510153e-02 7.82078207e-01 1.15651107e+00
5.92523217e-01 2.10234731e-01 6.28971636e-01 -9.38377082e-01
7.14365780e-01 -3.24856281e-01 -5.71765780e-01 4.22731638e-01
-6.61905289e-01 2.67284513e-01 1.09083104e+00 -6.18616402e-01
-1.44999838e+00 -7.97205605e-03 -4.50283468e-01 -4.59479950e-02
-4.13916510e-04 7.17344940e-01 -4.50689971e-01 4.14319456e-01
2.29358450e-01 3.96198988e-01 -3.53781462e-01 -9.70911682e-01
2.84140557e-01 9.33566988e-01 5.13255537e-01 -4.67177480e-01
3.94441098e-01 9.48652625e-02 -5.65362036e-01 -7.60416925e-01
-5.68236887e-01 -6.30684555e-01 -6.35637283e-01 2.26628453e-01
7.98647702e-01 -9.34759974e-01 -4.67529088e-01 4.44316804e-01
-1.41106474e+00 1.43886030e-01 -1.10839583e-01 5.25417149e-01
1.65177882e-01 5.90293229e-01 -1.01470435e+00 -6.90234184e-01
-8.21026087e-01 -1.02412391e+00 7.52619326e-01 4.88370597e-01
-1.50234476e-01 -9.95918334e-01 -2.40238681e-01 -1.90098025e-02
5.31435311e-01 -4.84922081e-01 1.19491208e+00 -7.30899274e-01
-9.02396888e-02 -1.53415933e-01 -4.42992866e-01 4.44359988e-01
1.04889527e-01 -4.79928702e-01 -7.85999179e-01 -2.34341025e-01
-1.26259744e-01 2.25031003e-02 7.46483505e-01 -8.01640227e-02
9.54415321e-01 -2.73072749e-01 -2.32388943e-01 4.14703757e-01
1.49439347e+00 4.73988861e-01 7.62533367e-01 4.05357927e-01
8.16882908e-01 4.56873000e-01 6.54701650e-01 3.73103440e-01
4.11264151e-01 5.09051859e-01 8.68055671e-02 -4.48531032e-01
2.90392693e-02 -2.96383679e-01 4.89373833e-01 1.49391270e+00
2.22612172e-01 -4.95724916e-01 -9.36045527e-01 7.07258463e-01
-1.68405795e+00 -6.96164072e-01 -7.75985718e-02 2.08727598e+00
1.14149415e+00 2.23804921e-01 -3.49285781e-01 1.59893841e-01
1.07648516e+00 4.37294468e-02 -2.78079331e-01 -5.35178959e-01
-3.57942939e-01 5.27174234e-01 6.04951143e-01 3.91407818e-01
-1.09116364e+00 1.37374115e+00 5.10702419e+00 1.40077055e+00
-1.10310423e+00 2.56145537e-01 5.07450640e-01 2.54319310e-01
-2.29641467e-01 -1.25601083e-01 -1.38199699e+00 4.98429030e-01
8.97287607e-01 7.49442307e-03 2.27666914e-01 7.83262312e-01
6.75343443e-03 1.56893879e-01 -7.05187261e-01 8.38934422e-01
-2.01242580e-03 -1.23603463e+00 2.15337873e-01 -2.80564398e-01
4.41775918e-01 1.55571122e-02 -5.03955722e-01 4.92053479e-01
-1.23171732e-02 -6.45955086e-01 6.54176295e-01 9.78598446e-02
8.50678563e-01 -1.04301858e+00 1.25707769e+00 3.18409860e-01
-1.74810624e+00 1.13285430e-01 -6.81265891e-01 8.48407857e-03
3.79484802e-01 7.51320541e-01 -5.90659082e-01 4.72219020e-01
5.23835301e-01 3.21399808e-01 -4.24669772e-01 8.80869091e-01
-2.16675922e-01 5.83579659e-01 -3.29088986e-01 -2.36997470e-01
4.77726370e-01 -2.63792098e-01 2.33870342e-01 1.77304232e+00
2.86044121e-01 8.77032280e-02 2.66791999e-01 5.56533158e-01
-1.50251314e-01 6.71203136e-01 -1.28982976e-01 -1.76144063e-01
7.36529231e-01 1.30634058e+00 -8.72834146e-01 -4.56427187e-01
-6.37675881e-01 1.09210575e+00 5.39145529e-01 1.58920497e-01
-8.92072678e-01 -1.00063181e+00 5.03077924e-01 -3.36956643e-02
5.20685554e-01 -2.14789063e-01 -3.24119925e-01 -1.11313248e+00
1.06450051e-01 -7.15142787e-01 3.43510300e-01 -2.86051393e-01
-1.22145188e+00 1.09442163e+00 -3.90967816e-01 -1.00683868e+00
1.91259474e-01 -5.19431651e-01 -5.36217988e-01 1.02524304e+00
-1.60169256e+00 -7.91673183e-01 -6.13643378e-02 4.94605750e-01
7.21159935e-01 1.48178175e-01 9.92913127e-01 9.36321497e-01
-9.37631011e-01 9.71863210e-01 5.87052144e-02 5.30176580e-01
6.33172393e-01 -1.07082772e+00 9.26067114e-01 9.36308444e-01
1.36673614e-01 8.85995090e-01 3.90140489e-02 -7.68921971e-01
-1.01162291e+00 -1.15112042e+00 1.45455837e+00 1.78139046e-01
4.87116665e-01 -5.14215708e-01 -1.10488510e+00 3.53257328e-01
1.85715079e-01 -1.38004437e-01 7.79495835e-01 4.11185995e-02
-3.03272754e-01 -3.17057036e-02 -8.16934109e-01 8.51537466e-01
9.66539621e-01 -4.72770184e-01 -8.19335520e-01 -1.01480126e-01
9.43114400e-01 -2.08708525e-01 -8.19922209e-01 3.40912074e-01
3.54175091e-01 -6.77956998e-01 7.86047041e-01 -2.22997293e-01
-4.80231792e-02 -3.64040554e-01 1.63287204e-02 -1.25679326e+00
-5.84181070e-01 -5.00878870e-01 2.21762970e-01 1.66693962e+00
7.29114413e-01 -6.54561698e-01 5.09781897e-01 2.30574101e-01
-3.06434244e-01 -8.06251109e-01 -8.45034420e-01 -8.19408953e-01
2.35250089e-02 -3.57075036e-01 8.80729616e-01 9.99125481e-01
2.01066688e-01 6.47004604e-01 -2.42968917e-01 -2.53421478e-02
1.78722814e-01 -1.76768750e-01 5.95383309e-02 -1.01048601e+00
-4.72117998e-02 -3.32217991e-01 -1.36549532e-01 -1.31678545e+00
1.57724604e-01 -7.70412445e-01 1.91159561e-01 -1.32342815e+00
-1.38980607e-02 -6.35039806e-01 -4.89901364e-01 7.37211108e-01
-6.41715527e-01 2.16323480e-01 2.41786301e-01 -3.96432467e-02
-4.90103334e-01 6.25523627e-01 1.00741351e+00 1.53587341e-01
-2.68186599e-01 -3.80752563e-01 -5.50820351e-01 7.53101408e-01
8.63369167e-01 -5.57190895e-01 -1.82262555e-01 -7.82107234e-01
2.35453442e-01 -2.42461979e-01 -3.30391377e-01 -8.40430439e-01
3.83590937e-01 1.81341127e-01 3.00174385e-01 -8.34740043e-01
6.25497028e-02 -8.01133573e-01 -4.94114459e-02 4.02327508e-01
-4.93624434e-02 6.05369031e-01 4.38767791e-01 2.11392716e-01
-3.62814814e-01 -5.27450442e-01 8.47163856e-01 -1.38469666e-01
-8.08783233e-01 2.55222674e-02 -5.12604713e-01 2.23558679e-01
7.46417522e-01 -2.43417114e-01 -1.56111330e-01 4.82145958e-02
-4.29239839e-01 1.93075970e-01 1.48470536e-01 4.68861431e-01
4.47178870e-01 -1.34266210e+00 -4.96063650e-01 5.37295759e-01
3.34574399e-03 1.07613523e-02 2.25996375e-01 5.97934663e-01
-6.31757021e-01 5.51985383e-01 3.76368351e-02 -1.05642132e-01
-1.12719750e+00 5.43510914e-01 9.05650202e-03 -6.40953898e-01
-7.06632733e-01 9.73647952e-01 2.39103213e-01 -4.76156205e-01
3.27531546e-01 -3.47984523e-01 -4.48835999e-01 2.38686144e-01
7.38297164e-01 4.67669338e-01 3.88051033e-01 -6.13727331e-01
-3.65690559e-01 4.88053411e-01 -5.17451048e-01 9.94749814e-02
1.15156782e+00 -3.12016577e-01 -2.97738284e-01 7.30790868e-02
1.01152873e+00 2.67987996e-01 -7.74533868e-01 -5.34677684e-01
3.78279656e-01 -2.99858544e-02 1.56750828e-01 -5.42319357e-01
-1.09685290e+00 8.85622621e-01 3.72219652e-01 -2.91886665e-02
1.31063223e+00 -5.17190754e-01 1.44822979e+00 4.37403649e-01
3.10861528e-01 -1.27008247e+00 -3.74748975e-01 9.53511119e-01
3.22657794e-01 -8.60760808e-01 -3.71928692e-01 -5.23555696e-01
-6.60139322e-01 1.08467019e+00 8.15575182e-01 8.16984847e-02
6.77015841e-01 6.91829503e-01 2.31252417e-01 2.36684829e-01
-5.39755523e-01 -2.73877531e-01 1.22927487e-01 1.98155999e-01
6.16150796e-01 8.99364799e-03 -8.73407304e-01 8.99590671e-01
-6.64385632e-02 -2.84528077e-01 3.01048279e-01 9.48142946e-01
-4.51206535e-01 -1.48064291e+00 -2.97308087e-01 2.96726435e-01
-6.84825003e-01 -8.95017684e-01 -1.62091758e-02 5.44055820e-01
3.52572560e-01 9.43407834e-01 2.39102885e-01 -2.84570217e-01
4.52937722e-01 3.07160884e-01 -6.53393716e-02 -7.29947031e-01
-8.99158955e-01 5.04779518e-01 1.19746380e-01 -3.14941525e-01
-1.13594010e-01 -3.36126029e-01 -1.66784680e+00 -2.82457769e-01
-8.99860680e-01 4.72726762e-01 5.05967617e-01 8.54952395e-01
4.27032322e-01 6.12945199e-01 3.65488797e-01 -4.51806486e-01
-5.34938395e-01 -1.04156351e+00 -4.69266295e-01 2.43768886e-01
-3.05418283e-01 -4.71575558e-01 -2.41167378e-02 -1.51656106e-01] | [9.833178520202637, 9.838254928588867] |
7aee1a52-e089-420e-8027-c1c8f62c11e8 | the-go-transformer-natural-language-modeling | 2007.03500 | null | https://arxiv.org/abs/2007.03500v3 | https://arxiv.org/pdf/2007.03500v3.pdf | The Go Transformer: Natural Language Modeling for Game Play | This work applies natural language modeling to generate plausible strategic moves in the ancient game of Go. We train the Generative Pretrained Transformer (GPT-2) to mimic the style of Go champions as archived in Smart Game Format (SGF), which offers a text description of move sequences. The trained model further generates valid but previously unseen strategies for Go. Because GPT-2 preserves punctuation and spacing, the raw output of the text generator provides inputs to game visualization and creative patterns, such as the Sabaki project's game engine using auto-replays. Results demonstrate that language modeling can capture both the sequencing format of championship Go games and their strategic formations. Compared to random game boards, the GPT-2 fine-tuning shows efficient opening move sequences favoring corner play over less advantageous center and side play. Game generation as a language modeling task offers novel approaches to more than 40 other board games where historical text annotation provides training data (e.g., Amazons & Connect 4/6). | ['David Noever', 'Matthew Ciolino', 'Josh Kalin'] | 2020-07-07 | null | null | null | null | ['text-annotation', 'game-of-go', 'board-games'] | ['natural-language-processing', 'playing-games', 'playing-games'] | [ 3.68661694e-02 6.06518149e-01 1.44861013e-01 1.92152321e-01
-7.59495258e-01 -1.04287267e+00 7.75030613e-01 -5.22830129e-01
-6.84068426e-02 6.54899538e-01 6.97878361e-01 -7.05589235e-01
-1.14959544e-02 -1.28059494e+00 -5.24168670e-01 -2.76606470e-01
5.48699275e-02 8.73368740e-01 2.44781420e-01 -1.15087581e+00
3.44886541e-01 -1.56410113e-01 -1.42387414e+00 7.48098314e-01
4.06765521e-01 1.78938940e-01 6.13228619e-01 1.23024213e+00
-5.16992211e-01 1.28920627e+00 -8.35108161e-01 -6.86756074e-01
5.79897761e-01 -1.02297461e+00 -1.01914966e+00 -5.24739444e-01
-1.25128537e-01 -2.47249380e-02 -6.00668669e-01 6.99546993e-01
4.55706060e-01 4.28685844e-01 3.63822758e-01 -1.16414368e+00
-5.80863655e-01 1.52760267e+00 -1.43572330e-01 2.48534530e-01
7.06601262e-01 6.18886769e-01 1.36295867e+00 -3.05513889e-01
1.08605015e+00 9.98805106e-01 8.03730309e-01 9.07630444e-01
-1.25072706e+00 -5.15127659e-01 -1.56679943e-01 -1.05218627e-02
-1.24756455e+00 -2.47052833e-01 4.82980907e-01 -4.93626744e-01
1.23475814e+00 4.51940745e-01 1.26841545e+00 1.35167825e+00
2.91754276e-01 8.92565846e-01 6.36645854e-01 -3.76789540e-01
1.96369916e-01 -6.11272871e-01 -4.77878004e-01 5.85051596e-01
-2.87029833e-01 2.38189727e-01 -7.22169340e-01 -1.10296242e-01
1.21442080e+00 -5.78079045e-01 1.03426017e-01 1.84596390e-01
-8.77640128e-01 8.34748626e-01 1.62772626e-01 5.44805303e-02
-1.45214170e-01 5.87844551e-01 4.98532057e-01 1.34821758e-01
1.71315849e-01 9.64911520e-01 -1.47482410e-01 -1.22771263e+00
-1.19331384e+00 1.13349402e+00 8.51358712e-01 1.44290709e+00
4.45480198e-01 4.24036980e-01 -2.53263712e-01 5.57867467e-01
1.59166947e-01 -4.14227955e-02 9.29094791e-01 -8.75197411e-01
5.28236687e-01 4.55166221e-01 -1.48189828e-01 -8.77568185e-01
-4.69610363e-01 -4.71525937e-01 -3.08877200e-01 1.23775139e-01
4.71807629e-01 -6.23012334e-02 -7.43759751e-01 1.67162907e+00
-3.01764607e-01 -4.90499996e-02 -9.46713835e-02 5.98681867e-01
1.03703392e+00 7.12733388e-01 7.88946748e-02 5.30150056e-01
1.27381229e+00 -9.97887075e-01 -2.55410820e-01 -8.14657927e-01
8.64562511e-01 -6.30360186e-01 1.73967338e+00 3.67232889e-01
-1.45759320e+00 -4.73050028e-01 -8.47469389e-01 -1.75602809e-01
-3.71797159e-02 -8.99450928e-02 9.86536384e-01 7.11423159e-01
-1.27004099e+00 5.80551565e-01 -9.45909977e-01 -2.81085163e-01
2.90169448e-01 -4.37176228e-02 4.17978242e-02 2.98524976e-01
-1.18702042e+00 5.64500332e-01 8.22169125e-01 -1.71860427e-01
-9.44212079e-01 -8.09596777e-01 -9.28407431e-01 1.41206294e-01
4.29730624e-01 -9.61410224e-01 1.76840091e+00 -5.70636451e-01
-1.84113860e+00 9.96816933e-01 4.43344623e-01 -6.45096898e-01
7.09906101e-01 4.18033870e-03 -5.79800978e-02 -4.41406965e-01
3.14565212e-01 6.83037639e-01 2.44932279e-01 -5.92770636e-01
-7.21837759e-01 1.85681179e-01 5.18298447e-01 4.50458258e-01
5.65151155e-01 1.03473440e-02 -6.08110487e-01 -9.77261245e-01
-1.13189571e-01 -8.50590348e-01 -5.47506392e-01 -7.90213764e-01
-6.37030602e-01 1.15287110e-01 -5.88406399e-02 -8.42129290e-01
1.73954213e+00 -1.63108730e+00 1.64878160e-01 1.68097034e-01
2.40001872e-01 -1.93937287e-01 -9.23774764e-02 8.63179266e-01
-1.56280085e-01 5.34197211e-01 1.00050829e-01 -1.44190401e-01
4.41025347e-01 2.41540447e-01 -5.50154448e-01 -2.76973516e-01
-1.63139343e-01 1.50649023e+00 -9.84913468e-01 -5.23069128e-02
3.96702737e-02 -2.16976479e-01 -1.15932822e+00 1.12592066e-02
-7.93082416e-01 2.37585261e-01 -2.84603596e-01 2.25447282e-01
-3.21994461e-02 -1.00805454e-01 4.57897782e-01 5.34554124e-01
-1.56974778e-01 1.02801061e+00 -9.32417333e-01 2.29256845e+00
-4.92954224e-01 7.81821012e-01 -5.23827076e-01 -2.40689173e-01
9.47774172e-01 -1.20457321e-01 -1.28907815e-01 -8.64100218e-01
1.52069423e-02 5.98587207e-02 3.41735631e-01 -2.64713228e-01
1.39980805e+00 -2.96125382e-01 -8.46475005e-01 7.46866405e-01
9.14479867e-02 -6.90159202e-01 5.51123619e-01 6.62714958e-01
1.45409858e+00 9.21547234e-01 4.47390109e-01 -2.11969569e-01
-1.53162569e-01 5.58285892e-01 4.52926278e-01 1.22416592e+00
4.42382425e-01 8.18962336e-01 7.68905580e-01 -5.03483534e-01
-1.44093156e+00 -9.38083470e-01 8.41429770e-01 1.39147329e+00
-2.60030389e-01 -1.60946131e+00 -9.83417988e-01 -1.15018435e-01
-6.10175788e-01 1.12342799e+00 -6.33403480e-01 -2.57414609e-01
-9.17986691e-01 -6.45206571e-01 1.13435125e+00 5.74360073e-01
3.35427403e-01 -1.38326168e+00 -7.35557854e-01 5.56496859e-01
-4.34130698e-01 -5.28727949e-01 -4.17671651e-01 1.76072288e-02
-4.06498760e-01 -1.00423598e+00 -1.48532525e-01 -5.33101797e-01
3.67270573e-03 -3.27873051e-01 1.75425744e+00 7.37797050e-03
-3.09325367e-01 1.00769982e-01 -5.12517393e-01 -3.46400470e-01
-8.62058580e-01 5.60003281e-01 -3.56789529e-01 -9.97297764e-01
1.36878744e-01 -7.60782778e-01 -2.25012794e-01 1.51052415e-01
-8.21133852e-01 9.51881886e-01 5.71416579e-02 8.37212026e-01
4.55895454e-01 -3.44350887e-03 -1.56009020e-02 -1.04394233e+00
1.10556483e+00 -3.78113091e-01 -5.40131092e-01 -2.55494583e-02
-1.40771836e-01 7.36801177e-02 5.99998116e-01 -1.42728031e-01
-9.44454849e-01 -3.75152647e-01 -4.35717136e-01 2.47997358e-01
1.55832484e-01 6.43853366e-01 -2.48744369e-01 5.29794931e-01
1.11074054e+00 4.93812263e-01 -4.63634014e-01 -2.81717300e-01
6.56018317e-01 1.63919777e-01 9.86770988e-01 -8.63943756e-01
7.55085647e-01 -7.14152008e-02 -3.67519170e-01 -4.79292601e-01
-2.26255476e-01 2.14807555e-01 -2.61683226e-01 -2.94883221e-01
6.37612998e-01 -8.56644630e-01 -7.70323575e-01 3.55742902e-01
-9.91954625e-01 -1.10704863e+00 -8.33782375e-01 -1.45999506e-01
-1.19110811e+00 -2.50603884e-01 -8.36054265e-01 -7.51915216e-01
-2.91281700e-01 -9.68686283e-01 9.52956915e-01 4.34679359e-01
-1.01818669e+00 -1.01323926e+00 3.42812777e-01 2.29521617e-01
3.20221335e-01 3.64312410e-01 1.01761425e+00 -6.53741658e-01
-5.89290082e-01 1.37021825e-01 5.66780329e-01 -5.03537118e-01
3.44485454e-02 2.81919185e-02 -3.48714739e-01 2.58555591e-01
-3.37902874e-01 -4.68291417e-02 3.35347176e-01 4.41058189e-01
7.09397674e-01 -4.56560612e-01 -9.46787223e-02 8.80346775e-01
1.02424967e+00 5.18277824e-01 1.15840292e+00 9.67948020e-01
6.10230505e-01 3.40926051e-01 5.05971491e-01 4.82119530e-01
5.15291095e-01 7.33605206e-01 2.85731524e-01 2.44412333e-01
-1.75047919e-01 -1.21708822e+00 4.44610536e-01 5.53790331e-01
-2.25229159e-01 -4.25890148e-01 -1.14388788e+00 5.13245523e-01
-1.82730627e+00 -1.46399963e+00 -1.67314589e-01 1.75028956e+00
8.67990494e-01 5.36058962e-01 4.95377779e-01 -5.17335013e-02
3.14335108e-01 3.85118604e-01 -2.57302582e-01 -5.57042718e-01
-4.13998216e-01 8.13960373e-01 3.86725396e-01 5.69094539e-01
-5.94011784e-01 1.65589106e+00 6.84078312e+00 1.37798321e+00
-6.49568081e-01 6.50601834e-02 5.41699350e-01 -6.25233233e-01
-8.27474296e-01 2.23781213e-01 -5.35939515e-01 5.12490094e-01
9.68664944e-01 -7.67239630e-01 9.86017764e-01 7.73950934e-01
5.26005268e-01 3.14347334e-02 -7.47473180e-01 1.10767663e+00
-1.41022399e-01 -2.12386775e+00 1.59456134e-01 1.80188507e-01
4.86219436e-01 -2.19491646e-01 -5.33624440e-02 7.33972192e-01
1.33285594e+00 -1.44255424e+00 1.60535157e+00 4.92981344e-01
8.34411144e-01 -9.11997020e-01 2.00864524e-01 3.62108409e-01
-1.24990141e+00 -1.05496123e-01 -3.05780381e-01 -5.87852538e-01
5.91911614e-01 -2.46648714e-01 -1.17630196e+00 6.45267308e-01
5.63674271e-01 5.57901382e-01 -6.52124286e-01 6.86941803e-01
-7.11529076e-01 7.03251183e-01 -2.39506245e-01 -3.85152549e-02
4.16390002e-01 -3.72568220e-01 7.25746930e-01 9.82590377e-01
5.49785435e-01 3.64238024e-01 -1.35389015e-01 1.30647027e+00
7.77522996e-02 -3.02932542e-02 -6.81933343e-01 -5.57272792e-01
4.98349696e-01 9.26424086e-01 -9.30714190e-01 -1.00705735e-01
3.20651084e-01 1.07841730e+00 3.05569824e-02 1.14395872e-01
-9.12319362e-01 -2.24675745e-01 7.99583912e-01 6.62157655e-01
9.75104719e-02 -4.71300095e-01 -4.23021555e-01 -9.49947000e-01
-5.19738078e-01 -1.20438552e+00 3.97884816e-01 -1.33593130e+00
-7.22785234e-01 8.79479706e-01 4.99064140e-02 -9.94116306e-01
-1.01037180e+00 -4.34129655e-01 -1.00907409e+00 1.02864468e+00
-8.72206390e-02 -1.32150853e+00 -2.17132866e-02 1.73046529e-01
1.06094098e+00 -4.66683120e-01 6.92947447e-01 -1.66590944e-01
-3.03162128e-01 5.27555645e-01 -1.48621559e-01 2.58241773e-01
2.88938917e-02 -1.53389335e+00 1.69347858e+00 9.33546484e-01
6.12679720e-01 5.50896347e-01 9.86673355e-01 -9.62183535e-01
-1.02233541e+00 -8.46452773e-01 3.78520370e-01 -6.28577709e-01
8.33438337e-01 -6.94595337e-01 -3.08927059e-01 1.09071434e+00
9.50548425e-02 -9.99367356e-01 7.36608446e-01 -9.19172093e-02
1.17880329e-01 6.13281667e-01 -5.51940560e-01 1.22345984e+00
1.74876893e+00 -5.89187980e-01 -6.70112073e-01 -2.30978224e-02
6.09242201e-01 -1.19211948e+00 -2.47605607e-01 -4.57312644e-01
5.36374271e-01 -9.78181660e-01 7.80669928e-01 -1.16942954e+00
1.00543618e+00 -3.25119734e-01 3.21612731e-02 -1.65042293e+00
-5.02242327e-01 -1.45138741e+00 3.67324352e-01 1.13442934e+00
6.28542423e-01 -5.48100956e-02 1.22506261e+00 6.02708697e-01
-5.55330455e-01 -3.72031271e-01 -6.59442782e-01 -5.11248708e-01
2.13033363e-01 -1.04358792e+00 8.94712090e-01 6.31831765e-01
3.13201964e-01 1.40594646e-01 -4.78074342e-01 -4.75288391e-01
1.81763507e-02 -8.85969773e-03 1.13963234e+00 -7.90355206e-01
-1.01148105e+00 -7.22357273e-01 -5.06892323e-01 -1.29470026e+00
-6.97907507e-02 -1.18118107e+00 1.16551802e-01 -1.66951478e+00
-2.77840555e-01 -2.92908072e-01 6.27210259e-01 5.48240960e-01
3.45205329e-02 2.22982287e-01 3.97667676e-01 1.20372847e-01
-3.57097983e-01 4.90841985e-01 1.01738966e+00 -3.32334116e-02
-6.17948711e-01 -8.73306245e-02 -1.27673316e+00 7.22862959e-01
7.35587299e-01 -5.15887022e-01 -5.57303011e-01 -4.77319002e-01
1.14137578e+00 1.24112643e-01 1.85971484e-01 -8.63596320e-01
2.40840852e-01 -4.45004731e-01 -5.77888675e-02 -1.98585659e-01
-1.03171021e-02 1.78495690e-01 9.30337787e-01 2.23219961e-01
-3.57290447e-01 4.43128765e-01 5.66458225e-01 2.84320116e-01
7.79584125e-02 -1.86046690e-01 9.60470662e-02 -7.66743064e-01
-8.33282948e-01 -5.97065762e-02 -1.07345462e+00 4.78524983e-01
6.12695396e-01 -7.18879282e-01 -4.99418199e-01 -9.13004458e-01
-1.04132056e+00 5.01059294e-02 7.37726986e-01 6.28931999e-01
3.05903584e-01 -1.27686334e+00 -6.41102672e-01 7.61172250e-02
-1.04182377e-01 2.30900720e-01 4.96034592e-01 1.37358859e-01
-1.32268214e+00 3.88112254e-02 -5.39738357e-01 -2.95365825e-02
-9.47049677e-01 -5.40243760e-02 5.68033099e-01 -7.78285146e-01
-8.32618117e-01 1.04728472e+00 2.63030797e-01 -4.61827755e-01
-6.29259706e-01 -3.68074834e-01 4.84235957e-02 -2.99805015e-01
6.41022325e-01 2.40187049e-01 1.76639517e-03 -3.79671097e-01
-1.42260604e-02 -8.02712217e-02 9.76417065e-02 -7.49556422e-01
1.58520424e+00 1.30535215e-01 1.49809569e-01 3.50138515e-01
1.62880048e-01 2.47977287e-01 -1.21755624e+00 2.59442776e-01
2.38716438e-01 -2.93560743e-01 -4.75789011e-01 -8.23285699e-01
-7.25470543e-01 5.35954773e-01 -1.66496947e-01 1.48869470e-01
7.57085025e-01 -5.30932732e-02 6.16993248e-01 9.59591046e-02
8.53956044e-01 -9.68977869e-01 4.06575669e-03 1.03696513e+00
9.06595767e-01 -1.15523532e-01 -5.10314226e-01 -1.65602654e-01
-8.93089473e-01 1.00351632e+00 7.31033087e-01 -1.66728199e-01
-1.16172865e-01 6.00128591e-01 -2.21343800e-01 -4.25294638e-01
-8.85992110e-01 -1.66485935e-01 -2.26379082e-01 7.44335294e-01
3.45594287e-01 1.24491856e-01 -4.04946208e-01 1.42733061e+00
-1.72215700e+00 -1.08666517e-01 1.02659345e+00 1.03775632e+00
6.03974313e-02 -1.28090465e+00 -1.49257690e-01 4.11365598e-01
-2.25456342e-01 -6.90710008e-01 -6.02510035e-01 9.76053178e-01
1.58508062e-01 6.00538254e-01 4.45481867e-01 -7.81975985e-01
4.13396478e-01 2.09081247e-01 4.95307386e-01 -1.15990949e+00
-1.16731989e+00 -4.11951505e-02 4.98522818e-01 -5.37569761e-01
3.06084573e-01 -4.00009632e-01 -1.30587316e+00 -7.53009439e-01
2.60908455e-01 3.30919683e-01 1.44475192e-01 6.60434067e-01
2.66337603e-01 8.92778575e-01 -3.08524936e-01 -8.96298587e-01
5.58844246e-02 -8.93003941e-01 -7.67872572e-01 2.42405817e-01
-6.15244508e-01 -2.60523766e-01 2.06357509e-01 1.84279487e-01] | [3.7042086124420166, 1.3343197107315063] |
01b5db73-04c5-4e3d-9c04-ce885cd84bba | crowd-level-abnormal-behavior-detection-via | 2212.00501 | null | https://arxiv.org/abs/2212.00501v1 | https://arxiv.org/pdf/2212.00501v1.pdf | Crowd-level Abnormal Behavior Detection via Multi-scale Motion Consistency Learning | Detecting abnormal crowd motion emerging from complex interactions of individuals is paramount to ensure the safety of crowds. Crowd-level abnormal behaviors (CABs), e.g., counter flow and crowd turbulence, are proven to be the crucial causes of many crowd disasters. In the recent decade, video anomaly detection (VAD) techniques have achieved remarkable success in detecting individual-level abnormal behaviors (e.g., sudden running, fighting and stealing), but research on VAD for CABs is rather limited. Unlike individual-level anomaly, CABs usually do not exhibit salient difference from the normal behaviors when observed locally, and the scale of CABs could vary from one scenario to another. In this paper, we present a systematic study to tackle the important problem of VAD for CABs with a novel crowd motion learning framework, multi-scale motion consistency network (MSMC-Net). MSMC-Net first captures the spatial and temporal crowd motion consistency information in a graph representation. Then, it simultaneously trains multiple feature graphs constructed at different scales to capture rich crowd patterns. An attention network is used to adaptively fuse the multi-scale features for better CAB detection. For the empirical study, we consider three large-scale crowd event datasets, UMN, Hajj and Love Parade. Experimental results show that MSMC-Net could substantially improve the state-of-the-art performance on all the datasets. | ['Wentong Cai', 'Ruimin Hu', 'Shangwei Xie', 'Haiyan Yin', 'Yuanjing Li', 'Linbo Luo'] | 2022-12-01 | null | null | null | null | ['video-anomaly-detection'] | ['computer-vision'] | [-4.07492042e-01 -8.32923591e-01 3.96333963e-01 1.28148168e-01
-1.75121501e-02 -1.13751225e-01 7.03595042e-01 4.56325114e-01
-3.18752229e-01 5.72084725e-01 3.96997303e-01 -6.56027123e-02
3.26955840e-02 -7.82482564e-01 -4.13502276e-01 -7.76101589e-01
-6.65725708e-01 2.27438241e-01 9.77211535e-01 -6.18946075e-01
1.24790020e-01 6.75519407e-01 -1.56723189e+00 1.50170609e-01
7.77255654e-01 6.44980788e-01 -1.52172476e-01 1.08597422e+00
-3.49103302e-01 1.24993181e+00 -9.61397588e-01 -2.53049970e-01
1.20585866e-01 -4.55954432e-01 -4.32598531e-01 2.90659070e-01
3.02453846e-01 -4.20714229e-01 -7.41074741e-01 9.17309046e-01
5.85764766e-01 4.16226506e-01 4.93832737e-01 -1.74467123e+00
-4.61173594e-01 -1.78334087e-01 -9.68730927e-01 1.32056808e+00
5.90178192e-01 7.18039215e-01 4.29385573e-01 -8.43084574e-01
4.28532124e-01 1.63332689e+00 7.02694595e-01 4.34274942e-01
-4.69360501e-01 -5.57185352e-01 4.79798585e-01 4.45247382e-01
-1.34788036e+00 -1.54268026e-01 9.32973444e-01 -7.27935553e-01
7.99064040e-01 2.25682497e-01 1.17083502e+00 1.10967720e+00
3.23662847e-01 8.58016014e-01 6.36307061e-01 1.29246145e-01
1.81319103e-01 -6.57171130e-01 4.89567891e-02 7.73427725e-01
6.90571487e-01 -3.36135745e-01 -6.65372312e-01 -4.09649968e-01
6.18238270e-01 1.62452430e-01 -2.33840361e-01 2.02781126e-01
-1.06776321e+00 7.31037438e-01 2.80033737e-01 4.34902489e-01
-4.00769979e-01 -2.19777189e-02 8.22471976e-01 2.11905912e-01
5.98666906e-01 -3.35431956e-02 1.43900156e-01 -3.60082150e-01
-4.46602404e-01 5.74792266e-01 4.29129452e-01 6.33215785e-01
5.94133735e-01 4.13432598e-01 -4.55384254e-01 4.18930233e-01
9.77854207e-02 6.19158089e-01 5.52269518e-01 -6.43148541e-01
6.35395706e-01 9.13844347e-01 1.41349271e-01 -2.07789087e+00
-7.32725561e-01 1.31839097e-01 -1.33051252e+00 -7.24851564e-02
5.80773890e-01 -4.07174319e-01 -6.41072035e-01 1.36843824e+00
7.03318298e-01 9.83780980e-01 -1.07417479e-01 9.55342412e-01
1.04262948e+00 7.83195555e-01 1.27279405e-02 -4.09844518e-01
1.06197214e+00 -8.22669208e-01 -9.92816925e-01 -2.43298352e-01
7.40525365e-01 -3.08216453e-01 8.24675202e-01 -7.93619156e-02
-8.33676159e-01 -4.35468316e-01 -6.48030400e-01 5.33534527e-01
-4.25271392e-01 -5.25107861e-01 4.36538786e-01 3.03369999e-01
-9.56197500e-01 1.37404025e-01 -8.81457150e-01 -5.06920040e-01
5.90551972e-01 5.32188304e-02 -4.30358112e-01 -2.09952462e-02
-1.19661999e+00 4.38719630e-01 1.05604425e-01 4.26609457e-01
-7.61538267e-01 -2.85193443e-01 -9.94931102e-01 -5.03096342e-01
4.53113556e-01 -5.17392576e-01 7.87342906e-01 -6.74756706e-01
-7.77818382e-01 6.85438514e-01 -4.42456961e-01 -3.63197178e-01
7.30238795e-01 -5.97525388e-02 -1.00330794e+00 1.95842013e-01
4.92913187e-01 5.66421486e-02 6.56829596e-01 -1.05525446e+00
-9.07556534e-01 -2.36076251e-01 -1.49804622e-01 1.52286530e-01
-3.30107629e-01 2.79853493e-01 -4.82674599e-01 -8.32329571e-01
-2.58409530e-01 -7.06135392e-01 -2.62224197e-01 -2.44539872e-01
-3.70289832e-01 -5.63562155e-01 1.44332421e+00 -6.70099139e-01
1.64765954e+00 -2.04697514e+00 -3.99974734e-02 1.80573478e-01
5.52409708e-01 6.45629883e-01 -1.31205499e-01 2.78512210e-01
4.03153777e-01 5.23845106e-02 -3.10275733e-01 -2.56753862e-01
-5.16841650e-01 3.80327851e-01 -2.32674759e-02 7.86071539e-01
3.39402139e-01 9.57450867e-01 -1.27546990e+00 -6.74943388e-01
1.18258633e-01 1.52149767e-01 -3.04235488e-01 4.17807639e-01
2.28343546e-01 8.34411383e-01 -5.04957795e-01 9.22713995e-01
7.48169601e-01 -1.72377720e-01 -4.19095635e-01 2.85431713e-01
-5.83925582e-02 -7.25754261e-01 -1.33750701e+00 9.67251062e-01
2.79205143e-01 8.16596687e-01 1.03724837e-01 -7.88177848e-01
7.08587527e-01 2.78387547e-01 5.99396348e-01 -5.81904411e-01
3.13684270e-02 1.49195269e-01 -6.03507608e-02 -1.09119630e+00
5.27963698e-01 3.74242276e-01 -1.00950755e-01 2.04178169e-01
-4.25492853e-01 4.34775829e-01 4.27687377e-01 3.30243111e-01
1.40598810e+00 -6.81018174e-01 3.69868487e-01 -4.55345586e-02
9.99835551e-01 -5.41484728e-02 9.55845952e-01 8.24027717e-01
-8.53209734e-01 4.11896616e-01 5.77593088e-01 -1.10406065e+00
-7.28482962e-01 -7.15967238e-01 5.53996742e-01 8.48178566e-01
5.55756867e-01 -1.96525291e-01 -5.84046245e-01 -8.48177254e-01
1.02102771e-01 1.15467653e-01 -6.63982630e-01 -1.68467209e-01
-1.20356464e+00 -8.98540914e-01 7.18217611e-01 4.60537702e-01
9.82869565e-01 -1.24151099e+00 -3.73502463e-01 1.74807280e-01
-3.46625060e-01 -1.33290398e+00 -8.06830525e-01 -9.32853460e-01
-4.80313301e-01 -1.29096544e+00 -5.94991207e-01 -6.70079470e-01
6.26792789e-01 9.13736880e-01 1.11824596e+00 7.07945108e-01
-3.45342159e-01 6.56659484e-01 -3.72050524e-01 -4.95950282e-01
-1.73408553e-01 -2.96593398e-01 5.33462346e-01 5.96240401e-01
6.48113966e-01 -5.51020682e-01 -7.09816694e-01 4.90872264e-01
-1.05000353e+00 -6.19016826e-01 8.67432207e-02 5.81575751e-01
3.26775670e-01 3.71989608e-01 5.75879335e-01 -3.73800874e-01
1.03521597e+00 -9.81048584e-01 -2.55839050e-01 -1.07342690e-01
1.56749949e-01 -6.89044535e-01 6.46252334e-01 -5.52785575e-01
-8.13992798e-01 -3.86758208e-01 1.46672726e-01 -5.97385049e-01
-7.09368467e-01 2.02274323e-01 -1.12991827e-03 -7.99855664e-02
4.61377770e-01 3.89013708e-01 -1.86889946e-01 4.83958237e-03
-2.54461914e-01 3.68566394e-01 7.42110550e-01 -1.92910895e-01
1.13287270e+00 8.52487326e-01 2.53887922e-01 -1.50093508e+00
-4.85361934e-01 -8.39871287e-01 -5.93253791e-01 -7.38186181e-01
1.24669600e+00 -9.55462337e-01 -6.99964046e-01 1.05162680e+00
-1.26989746e+00 -1.10136472e-01 -8.32002684e-02 2.25620687e-01
1.35881484e-01 8.22149992e-01 -6.10512912e-01 -1.07453215e+00
-2.94672325e-02 -8.26491296e-01 1.14276862e+00 4.99048978e-01
-3.77110690e-02 -1.34330320e+00 3.61186147e-01 1.31892979e-01
3.13199997e-01 7.79764295e-01 2.95951843e-01 -8.29017997e-01
-4.73482609e-01 -2.77216494e-01 -3.75829674e-02 -1.87500983e-01
3.93981427e-01 1.14791259e-01 -6.24737978e-01 -4.33379799e-01
-2.25093976e-01 5.87133765e-02 6.68868721e-01 4.34455991e-01
9.51186717e-01 -4.68964517e-01 -3.82013649e-01 4.97741073e-01
8.06159973e-01 1.60723761e-01 5.58699965e-01 4.88457561e-01
1.27049196e+00 5.46062350e-01 5.22821963e-01 7.07406223e-01
6.29396856e-01 5.10115862e-01 5.55546224e-01 3.52182984e-02
3.91718298e-02 -6.75737783e-02 5.69632232e-01 8.45631897e-01
-5.80887914e-01 -5.15430033e-01 -1.10178149e+00 7.82208681e-01
-2.10002446e+00 -1.48416626e+00 -6.24751031e-01 1.70612776e+00
-1.12098657e-01 9.19115096e-02 7.11393118e-01 9.74140503e-03
1.12504554e+00 6.35383844e-01 -3.32216859e-01 2.15206649e-02
-6.44308507e-01 -7.74795413e-01 9.44716334e-02 4.53290939e-01
-1.34989023e+00 7.59950221e-01 5.46416807e+00 7.50542760e-01
-8.24103355e-01 4.48123850e-02 4.65856761e-01 -1.92114674e-02
9.79008600e-02 -4.86583501e-01 -7.01416373e-01 8.99861455e-01
5.08853674e-01 -8.27389956e-02 1.39448673e-01 4.59274173e-01
5.39173543e-01 -3.26405354e-02 -3.99355203e-01 1.17047632e+00
2.86439627e-01 -1.25077903e+00 2.54353672e-01 -3.96281555e-02
7.83999324e-01 -1.70823887e-01 -1.68459550e-01 2.76929647e-01
1.30625457e-01 -9.22633052e-01 3.69558960e-01 5.50602674e-01
1.43553942e-01 -8.05660248e-01 1.13184214e+00 4.40496653e-01
-1.92508638e+00 -2.60127336e-01 -3.11426401e-01 -3.56063783e-01
5.64537287e-01 5.66933632e-01 -3.25774044e-01 4.95266378e-01
1.02009368e+00 9.99849617e-01 -9.85020638e-01 1.11500072e+00
2.74711668e-01 5.86424589e-01 -2.56247878e-01 -2.05065131e-01
5.12007415e-01 -2.39357367e-01 1.42755568e+00 1.35511994e+00
4.48368698e-01 1.97452664e-01 4.72243667e-01 1.19670898e-01
1.77763820e-01 3.56545337e-02 -1.22472715e+00 3.71281505e-01
4.26348388e-01 9.51585114e-01 -6.64157212e-01 -3.57481211e-01
-5.47837496e-01 1.00308096e+00 2.83564895e-01 4.75726187e-01
-8.81362677e-01 -1.56474948e-01 8.24187279e-01 2.33313769e-01
2.41846979e-01 -3.82359743e-01 3.65172952e-01 -1.26795816e+00
3.55224907e-01 -7.34788656e-01 7.14842081e-01 -2.77433038e-01
-1.68596029e+00 6.10689342e-01 8.27085972e-02 -1.44116914e+00
-1.32107198e-01 -3.91922534e-01 -1.58877480e+00 3.92942101e-01
-1.30815744e+00 -1.00835109e+00 -9.47473228e-01 1.08571482e+00
7.38063574e-01 -6.22123599e-01 1.60204321e-01 3.12652558e-01
-1.02135742e+00 3.36490124e-01 -3.53040367e-01 5.86076677e-01
2.90278465e-01 -1.02576590e+00 5.52034378e-01 1.33157432e+00
-3.98707718e-01 9.11518633e-02 6.25119925e-01 -1.18997741e+00
-1.27785218e+00 -1.54505193e+00 5.43893635e-01 -5.37910581e-01
8.33921850e-01 -2.24388316e-02 -1.20293009e+00 5.61999619e-01
-7.68515375e-03 7.06715643e-01 3.62680614e-01 -5.11169732e-01
1.39520958e-01 1.15811571e-01 -9.33479071e-01 7.46752858e-01
1.35917711e+00 -2.13700742e-01 -4.08283085e-01 3.38476837e-01
5.88548660e-01 -2.36504138e-01 -3.82878512e-01 4.02626067e-01
-4.58443947e-02 -1.27245140e+00 9.54337656e-01 -9.56410408e-01
-1.44452304e-01 -7.13212609e-01 -1.75385468e-03 -1.34495211e+00
-4.78643745e-01 -8.04363132e-01 -7.71146357e-01 1.21312249e+00
-2.81792253e-01 -7.46051550e-01 6.06962502e-01 1.34591997e-01
-2.43308991e-01 -5.15555382e-01 -1.23235190e+00 -9.62582588e-01
-3.00571501e-01 -2.26499557e-01 8.72114480e-01 9.65188563e-01
-4.03523773e-01 4.93861772e-02 -7.07978547e-01 6.05943024e-01
5.90529144e-01 -4.04652476e-01 1.20779347e+00 -1.22118175e+00
1.77073747e-01 -6.66660011e-01 -1.14502001e+00 -8.21615934e-01
1.17718749e-01 -3.55047911e-01 -1.39881834e-01 -1.39390457e+00
-2.28180154e-03 1.53411150e-01 8.89319852e-02 -2.27734461e-01
-8.00878227e-01 1.96347550e-01 6.36872463e-03 3.16543192e-01
-1.15936852e+00 6.52858973e-01 1.22223330e+00 -2.99722731e-01
-4.67638850e-01 3.23019363e-02 -1.39329642e-01 1.26984596e+00
9.16527510e-01 -1.29493326e-01 -1.59524128e-01 -4.40672040e-01
1.04964778e-01 -4.47598100e-02 5.57378650e-01 -1.41277742e+00
5.99972725e-01 -3.26747537e-01 2.08062276e-01 -5.99971950e-01
7.37608969e-02 -5.51374137e-01 -1.51163369e-01 5.75937152e-01
2.42458507e-01 9.07910049e-01 2.32168168e-01 1.30285585e+00
-4.61239278e-01 3.81819457e-01 4.12436485e-01 -1.19460210e-01
-1.16860723e+00 8.55726242e-01 -8.73700857e-01 6.05157912e-01
1.38782084e+00 -3.02930325e-01 -5.56915283e-01 -7.18307793e-01
-3.95286083e-01 6.58083856e-01 3.07178676e-01 5.72265089e-01
8.97217870e-01 -1.68331635e+00 -1.06837237e+00 1.04525395e-01
1.99661911e-01 2.73080200e-01 7.18424499e-01 1.09660697e+00
-7.65355647e-01 -5.37155867e-02 -6.84895143e-02 -8.25920761e-01
-1.20342302e+00 4.85121369e-01 4.66865420e-01 -3.76096308e-01
-7.49686003e-01 7.11886406e-01 2.11662054e-02 -2.13072270e-01
9.30715799e-02 3.76919508e-02 -6.05322003e-01 1.39014823e-02
9.49492693e-01 1.05565429e+00 -4.18323606e-01 -1.29593349e+00
-5.72938800e-01 7.02028573e-01 3.30831110e-01 4.09649014e-01
9.97811615e-01 -2.65761971e-01 -2.59208202e-01 5.27325690e-01
8.74976575e-01 1.83587790e-01 -1.31738579e+00 -1.89314857e-01
-9.03660059e-02 -7.28860438e-01 -5.49221337e-01 3.87067199e-01
-1.26293695e+00 6.52013898e-01 4.10577208e-01 6.37542188e-01
9.70016241e-01 4.21820655e-02 1.09052682e+00 4.41568673e-01
2.62226105e-01 -1.02777171e+00 6.98976636e-01 6.74451828e-01
8.10448706e-01 -1.53508174e+00 -1.44042909e-01 -2.49339953e-01
-9.64899719e-01 8.70672345e-01 1.15571177e+00 -1.67757854e-01
6.56164587e-01 3.85893248e-02 -2.16022264e-02 -6.27251863e-01
-3.77803952e-01 -2.03747794e-01 2.20370695e-01 9.00758088e-01
-1.12284437e-01 -2.89489161e-02 2.05643654e-01 3.41454417e-01
1.13467932e-01 -4.26543832e-01 5.96254468e-01 1.06991816e+00
-6.44588232e-01 -2.57472217e-01 -8.05477440e-01 5.78914106e-01
-3.38841289e-01 4.45271373e-01 -4.45386678e-01 8.73151720e-01
3.95520538e-01 1.24971390e+00 4.32841420e-01 -5.26053667e-01
5.86491525e-01 -2.96566844e-01 -1.70178548e-01 -7.45892003e-02
-2.31693670e-01 -3.35885644e-01 1.23973340e-02 -7.33916402e-01
-6.08345628e-01 -7.66854525e-01 -1.18247545e+00 -6.95783079e-01
8.50689039e-02 1.00970462e-01 -4.35193539e-01 1.09420562e+00
2.94314414e-01 5.18725991e-01 6.92590952e-01 -1.02605224e+00
2.10987300e-01 -8.17390442e-01 -5.48646331e-01 9.73408580e-01
7.67342269e-01 -9.09718513e-01 -6.39604032e-01 -2.25162536e-01] | [7.855021953582764, 1.542382001876831] |
0f5f9cc1-7b39-4452-abf1-a1dc6acf318d | bipartite-flat-graph-network-for-nested-named | 2005.00436 | null | https://arxiv.org/abs/2005.00436v1 | https://arxiv.org/pdf/2005.00436v1.pdf | Bipartite Flat-Graph Network for Nested Named Entity Recognition | In this paper, we propose a novel bipartite flat-graph network (BiFlaG) for nested named entity recognition (NER), which contains two subgraph modules: a flat NER module for outermost entities and a graph module for all the entities located in inner layers. Bidirectional LSTM (BiLSTM) and graph convolutional network (GCN) are adopted to jointly learn flat entities and their inner dependencies. Different from previous models, which only consider the unidirectional delivery of information from innermost layers to outer ones (or outside-to-inside), our model effectively captures the bidirectional interaction between them. We first use the entities recognized by the flat NER module to construct an entity graph, which is fed to the next graph module. The richer representation learned from graph module carries the dependencies of inner entities and can be exploited to improve outermost entity predictions. Experimental results on three standard nested NER datasets demonstrate that our BiFlaG outperforms previous state-of-the-art models. | ['Hai Zhao', 'Ying Luo'] | 2020-05-01 | bipartite-flat-graph-network-for-nested-named-1 | https://aclanthology.org/2020.acl-main.571 | https://aclanthology.org/2020.acl-main.571.pdf | acl-2020-6 | ['nested-named-entity-recognition', 'nested-mention-recognition'] | ['natural-language-processing', 'natural-language-processing'] | [-3.23908657e-01 7.43485570e-01 -2.17562988e-02 -3.93054187e-01
-2.53216803e-01 -7.71673739e-01 3.77109081e-01 4.34072822e-01
-4.09112304e-01 6.09932125e-01 3.75715017e-01 -4.13521647e-01
2.39465222e-01 -1.30290163e+00 -9.58184183e-01 -3.73279393e-01
-4.31606829e-01 3.30727756e-01 5.03789902e-01 -2.09611475e-01
-4.18246627e-01 4.23456043e-01 -7.31593668e-01 3.81564945e-01
8.58992994e-01 7.20290303e-01 1.40029863e-01 6.25295103e-01
-5.83155394e-01 1.44427896e+00 -2.89708257e-01 -8.84938657e-01
-9.04093459e-02 -4.13123704e-02 -1.19526207e+00 -6.77097023e-01
1.09888703e-01 -2.26525098e-01 -7.73313522e-01 8.76091540e-01
4.54779625e-01 2.91693598e-01 2.50993401e-01 -8.70455563e-01
-1.09291399e+00 1.30103469e+00 -1.70088917e-01 7.64768124e-02
3.49430919e-01 -3.65243405e-01 1.54929352e+00 -7.24476039e-01
7.42047608e-01 1.04518020e+00 9.79042888e-01 5.74831426e-01
-9.65182960e-01 -5.42879999e-01 4.34284687e-01 -4.70263660e-02
-1.24927771e+00 -1.03028558e-01 5.77247798e-01 -3.01849544e-01
1.35201228e+00 5.06717302e-02 5.04391432e-01 6.77357495e-01
1.60950989e-01 8.38098228e-01 2.80134708e-01 -6.88454285e-02
-2.77315974e-01 -2.87430018e-01 5.81290483e-01 1.06290507e+00
3.70840132e-01 -1.16513588e-01 6.74319714e-02 -1.65251672e-01
7.81678557e-01 1.03529066e-01 -1.54089645e-01 -3.91500384e-01
-9.47853982e-01 7.01005697e-01 1.50259912e+00 6.49118066e-01
-5.62807262e-01 1.85524166e-01 2.99763501e-01 1.46686345e-01
3.15349311e-01 2.14699671e-01 -7.24138796e-01 4.14111495e-01
-4.53816652e-01 -5.04414439e-01 1.24311864e+00 1.18067348e+00
1.09575856e+00 -3.28757577e-02 -3.07270497e-01 7.62648046e-01
5.65025926e-01 4.70374189e-02 2.36092120e-01 -9.33784693e-02
8.71930063e-01 1.33318734e+00 -2.20043942e-01 -1.09623504e+00
-7.29392290e-01 -4.65407789e-01 -1.03823113e+00 -7.40181923e-01
2.47328840e-02 -4.24072802e-01 -1.03305256e+00 1.72445226e+00
4.45063114e-01 3.20169985e-01 1.45672694e-01 6.17635846e-01
1.69328976e+00 5.96224904e-01 4.50028479e-01 4.70653683e-01
1.26687360e+00 -1.24221539e+00 -5.04877806e-01 -3.92572582e-01
1.29559231e+00 -1.62416682e-01 2.95249611e-01 -5.42456508e-01
-7.97177017e-01 -4.78983998e-01 -6.47141218e-01 -6.10641778e-01
-1.00011814e+00 2.72226781e-01 6.37579918e-01 1.80764318e-01
-1.36094785e+00 6.55382454e-01 -7.96999156e-01 -3.80202472e-01
1.90181360e-01 5.84393144e-01 -8.26781154e-01 1.97614506e-01
-1.78828120e+00 7.97543108e-01 8.66679370e-01 6.64802134e-01
-7.19552040e-01 -3.98243666e-01 -1.59613824e+00 4.97564465e-01
3.65586132e-02 -8.49179089e-01 9.12655473e-01 -4.71014082e-01
-9.19217050e-01 8.11197579e-01 -1.35282785e-01 -5.24546385e-01
-1.63871627e-02 -1.10353880e-01 -5.57075083e-01 -2.74662197e-01
1.01588771e-01 5.98546684e-01 1.26908012e-02 -1.23427463e+00
-5.79311848e-01 -3.52321893e-01 4.77839530e-01 1.10615209e-01
-3.80538434e-01 -1.86277807e-01 -7.27164030e-01 -3.27330887e-01
1.88971590e-02 -7.12341070e-01 -2.83347249e-01 -8.48988712e-01
-1.04557621e+00 -5.65425992e-01 4.46766853e-01 -9.60773885e-01
1.87185335e+00 -1.88227820e+00 2.05314264e-01 1.03953220e-01
5.48201203e-01 2.83232510e-01 -2.26515353e-01 8.92252445e-01
-3.63454431e-01 3.71700585e-01 2.19705701e-02 -5.21420836e-01
1.41694441e-01 3.90943199e-01 -1.56700671e-01 2.08421171e-01
4.24166471e-01 1.42013502e+00 -1.14835048e+00 -5.22259831e-01
-8.87620673e-02 6.29686534e-01 -2.67522931e-01 4.68997031e-01
1.58033937e-01 9.93809029e-02 -5.17280936e-01 3.88566941e-01
7.01257944e-01 -7.58363366e-01 4.93753433e-01 -4.39011335e-01
-3.35111916e-02 7.28351951e-01 -9.21989858e-01 1.33130836e+00
-3.90443623e-01 1.40207142e-01 -2.29967162e-02 -6.94305599e-01
9.43080902e-01 3.48901600e-01 1.67279616e-02 -4.46755797e-01
1.74654685e-02 1.56624138e-03 -1.00791968e-01 -3.19957256e-01
5.21924913e-01 2.37505913e-01 -3.16274703e-01 -7.80758485e-02
5.69739878e-01 6.23155892e-01 4.55734491e-01 5.22647619e-01
1.32352436e+00 8.29129741e-02 4.08008754e-01 3.67377363e-02
6.17137194e-01 -3.64850491e-01 9.52294230e-01 6.20829821e-01
-9.75252613e-02 1.26673445e-01 6.19938493e-01 -6.15413964e-01
-6.68032527e-01 -1.10085523e+00 3.16212147e-01 1.36955881e+00
1.75422505e-01 -5.54384053e-01 -6.36555314e-01 -1.20092916e+00
-1.26843512e-01 4.68523026e-01 -9.02426243e-01 -7.62740672e-02
-7.98895001e-01 -2.97022998e-01 8.97268295e-01 9.88748431e-01
6.30862474e-01 -1.30605471e+00 1.62166774e-01 1.98149025e-01
-1.37143597e-01 -1.30811822e+00 -7.02722967e-01 3.18684995e-01
-8.28945637e-01 -1.04796362e+00 -4.35787022e-01 -1.27919364e+00
8.04935932e-01 -8.71712491e-02 1.38848722e+00 3.69100720e-01
2.98820615e-01 1.64770260e-01 -3.54798615e-01 9.71878022e-02
-2.09529743e-01 6.15624547e-01 -4.19530749e-01 4.86102840e-03
3.06559235e-01 -5.20429552e-01 -4.37020808e-01 1.09531842e-01
-7.12059557e-01 2.52255082e-01 7.65449584e-01 1.00800300e+00
4.56059933e-01 -2.55186766e-01 3.19851696e-01 -1.46026123e+00
3.17603469e-01 -7.92863607e-01 -4.40236419e-01 6.00047588e-01
-2.19218105e-01 2.20385507e-01 9.74141777e-01 -4.46100757e-02
-1.10230458e+00 2.00762618e-02 -2.08859727e-01 -1.98584810e-01
-1.25321284e-01 9.19494569e-01 -4.16336775e-01 1.22597277e-01
2.58407205e-01 1.19615093e-01 -1.03639615e+00 -5.95242798e-01
4.85792667e-01 3.67148340e-01 4.66059297e-01 -3.10980707e-01
7.34523058e-01 1.94227487e-01 -1.89021379e-01 -5.63274264e-01
-9.78959918e-01 -5.26162744e-01 -9.18513119e-01 4.74186754e-03
1.11750877e+00 -1.04946041e+00 -7.15271413e-01 4.33830172e-01
-1.46780884e+00 -4.54867095e-01 -1.05529256e-01 1.47208363e-01
2.46853307e-01 3.05556417e-01 -1.33300185e+00 -6.78691924e-01
-6.99594557e-01 -7.99855649e-01 1.17842638e+00 6.14925683e-01
2.55193412e-01 -1.62192810e+00 2.18109712e-01 1.33671075e-01
9.09610465e-02 1.00090891e-01 9.02025759e-01 -1.20723081e+00
-4.48912561e-01 -3.40056121e-01 -6.10155165e-01 3.99726368e-02
2.73666829e-02 -2.31150746e-01 -8.32512796e-01 -3.26221943e-01
-7.09126234e-01 -4.81099933e-02 1.10809970e+00 -6.56481013e-02
4.96065080e-01 -3.40327233e-01 -8.23423207e-01 7.33677328e-01
1.22052646e+00 -9.46305096e-02 6.89763010e-01 2.57772177e-01
1.48318028e+00 4.84058678e-01 -2.39443649e-02 1.31548708e-02
1.36650026e+00 4.86710779e-02 5.42083681e-01 -5.00535131e-01
5.74677065e-03 -9.25089836e-01 4.15889889e-01 1.28566098e+00
-3.79766198e-03 -5.48162282e-01 -1.07115638e+00 7.98003376e-01
-1.92322576e+00 -7.56184518e-01 -3.70354325e-01 1.77468371e+00
4.80957985e-01 -6.68800250e-02 -1.97416395e-01 -6.56399012e-01
1.23058283e+00 3.73454154e-01 -4.25570220e-01 -4.55637306e-01
-1.36156112e-01 -1.37030995e-02 6.68459237e-01 3.60622823e-01
-1.46710539e+00 1.26421964e+00 5.40607119e+00 4.68772769e-01
-6.92458093e-01 -1.51503608e-02 4.48653400e-01 7.81286478e-01
-5.08887291e-01 3.10953021e-01 -1.00617647e+00 3.20816755e-01
9.60875511e-01 3.36287916e-02 3.24062377e-01 8.56743455e-01
-3.27082604e-01 5.78262746e-01 -1.04030955e+00 4.21609312e-01
-3.27163726e-01 -1.10751426e+00 2.81254500e-01 -6.20907219e-03
7.26210535e-01 6.02087796e-01 -6.22688353e-01 8.64321351e-01
1.07914197e+00 -8.19686949e-01 3.02535295e-01 7.53409505e-01
3.15384358e-01 -6.21321440e-01 1.09073579e+00 3.33147973e-01
-2.19133306e+00 -1.13548398e-01 -4.49221760e-01 2.83482850e-01
1.56186029e-01 5.24022758e-01 -7.62407184e-01 1.30637062e+00
6.77934945e-01 8.77908409e-01 -4.73450571e-01 8.80609751e-01
-7.42107391e-01 5.32976329e-01 -2.09473103e-01 -1.43793598e-01
4.45093393e-01 -3.11473817e-01 3.85302812e-01 1.61389601e+00
5.92551790e-02 -4.71144952e-02 2.90187627e-01 8.88285398e-01
-8.95124435e-01 1.53527662e-01 -7.63460934e-01 -3.91866982e-01
5.91323197e-01 1.62145555e+00 -5.06365240e-01 -4.59485173e-01
-7.60114133e-01 1.20746362e+00 1.39342725e+00 5.56356788e-01
-7.86266565e-01 -7.44396865e-01 2.94411182e-01 -2.84150690e-01
8.18451166e-01 -9.86501575e-02 1.87094226e-01 -1.49934256e+00
-1.50781438e-01 -1.06361084e-01 1.17820323e+00 -6.88811183e-01
-1.57951093e+00 9.40175772e-01 -5.64678371e-01 -7.73376167e-01
6.63522854e-02 -4.91985381e-01 -9.68431830e-01 9.58636582e-01
-1.53173172e+00 -1.82550931e+00 -4.73479964e-02 4.58267003e-01
-2.10298508e-01 2.69327849e-01 9.17280555e-01 4.48226213e-01
-8.62816453e-01 6.91137254e-01 -1.64568111e-01 1.24657464e+00
2.87366480e-01 -1.70118070e+00 9.21147585e-01 9.82043862e-01
3.38673979e-01 1.11959040e+00 -1.70092970e-01 -9.95677352e-01
-1.34268522e+00 -1.46899259e+00 1.46869779e+00 -3.74800384e-01
9.39058900e-01 -4.81238186e-01 -8.01213205e-01 1.40614331e+00
2.37598732e-01 2.66473442e-01 8.14072847e-01 2.96319008e-01
-6.32028580e-01 9.25656408e-02 -7.28773832e-01 4.87102985e-01
1.18843341e+00 -7.39295065e-01 -8.27252030e-01 -1.60473078e-01
1.20294034e+00 -5.42822599e-01 -1.40446401e+00 5.69570482e-01
3.06238532e-01 -6.96974695e-01 8.30763221e-01 -9.84081089e-01
1.68320090e-01 -2.73862958e-01 1.25759631e-01 -1.39661980e+00
-6.04355812e-01 -2.75747240e-01 -5.11813402e-01 1.56329513e+00
8.06905806e-01 -8.11207473e-01 5.21995664e-01 5.70141971e-01
-4.35433179e-01 -7.48986483e-01 -7.13471651e-01 -4.09996301e-01
-6.87090680e-03 -4.68464196e-02 6.90954268e-01 1.32396924e+00
1.70106813e-01 8.21807206e-01 -1.96147472e-01 6.62636936e-01
3.40027750e-01 3.17358941e-01 4.69208866e-01 -1.23409927e+00
-1.65663719e-01 -1.66517183e-01 -5.61628819e-01 -1.41454947e+00
3.73231947e-01 -1.31146419e+00 -8.45017657e-03 -2.24602294e+00
3.03229332e-01 -1.77568525e-01 -6.89031661e-01 9.71784532e-01
-4.88013864e-01 -6.19963072e-02 1.30206212e-01 -2.01223895e-01
-1.04626107e+00 5.59956074e-01 1.15633214e+00 -2.35898256e-01
-1.79361612e-01 -3.79984975e-01 -6.83063686e-01 6.90256774e-01
3.54534119e-01 -3.67105812e-01 5.54496609e-03 -8.28404367e-01
4.75770056e-01 -4.35552336e-02 2.10584819e-01 -6.63555145e-01
8.10237944e-01 4.59749430e-01 4.35722619e-01 -6.92409873e-01
-7.67051876e-02 -7.46858239e-01 1.58173740e-01 1.90675214e-01
-2.42928058e-01 8.48229155e-02 8.80142227e-02 6.48999631e-01
-3.06321830e-01 1.12417783e-03 2.92422712e-01 -1.22390248e-01
-8.08920622e-01 6.98635161e-01 1.14025682e-01 2.08336860e-01
7.45221496e-01 1.91043645e-01 -6.40588105e-01 -2.60334462e-01
-1.21446943e+00 8.00445378e-01 1.39308989e-01 4.35062081e-01
4.27796513e-01 -1.45748854e+00 -4.58975405e-01 -3.94406691e-02
4.49756719e-02 2.68430889e-01 4.93863761e-01 8.15550804e-01
-3.92234772e-01 4.23424900e-01 3.72187912e-01 -1.76087260e-01
-1.04789627e+00 6.39169097e-01 6.72205865e-01 -1.05101335e+00
-4.15247917e-01 1.06862652e+00 5.55564225e-01 -1.28439438e+00
1.21023078e-02 -4.00696069e-01 -5.76154411e-01 -2.35196739e-03
1.38260812e-01 2.39188597e-01 -4.29725982e-02 -1.07927001e+00
-4.26237285e-01 2.57316947e-01 -1.64976761e-01 5.48101664e-01
1.46743965e+00 -8.87556225e-02 -6.42239153e-01 4.65096742e-01
1.24599814e+00 -9.86229852e-02 -6.32043242e-01 -2.70639718e-01
2.79412359e-01 3.24671298e-01 -1.90200761e-01 -4.43760753e-01
-1.19377267e+00 8.83987606e-01 -1.18146529e-02 6.04939640e-01
8.41658056e-01 2.68230706e-01 1.14536476e+00 6.31350815e-01
3.79162997e-01 -7.69551754e-01 -5.85182667e-01 1.06546712e+00
4.57957566e-01 -8.46952319e-01 -5.44148922e-01 -4.36995685e-01
-4.29648131e-01 1.11833882e+00 8.78631830e-01 -1.12281077e-01
7.11745501e-01 1.92642018e-01 -1.38815612e-01 -3.88716012e-01
-6.93483114e-01 -5.39843142e-01 5.31327069e-01 2.61041969e-01
7.18368769e-01 1.65093258e-01 -2.53282860e-02 1.10227907e+00
5.23211695e-02 -3.10829699e-01 5.84361739e-02 5.94450831e-01
-2.04819277e-01 -1.07009387e+00 2.11603329e-01 4.92382973e-01
-4.88588393e-01 -5.79691708e-01 -7.59158254e-01 8.74514699e-01
2.26139754e-01 8.84721339e-01 1.74253602e-02 -7.77498543e-01
5.54256678e-01 1.15629353e-01 -8.45273361e-02 -7.52958417e-01
-1.02289784e+00 -3.94191831e-01 5.73516309e-01 -5.35072327e-01
-1.63697988e-01 -1.01188056e-01 -1.82850373e+00 -2.19172373e-01
-7.25618660e-01 5.41082144e-01 7.97066316e-02 9.28142369e-01
7.81319082e-01 8.59967828e-01 4.86407548e-01 -5.24749696e-01
1.43727109e-01 -1.01654923e+00 -7.01031089e-01 5.79589307e-01
1.85940921e-01 -3.77085298e-01 -1.89662516e-01 -2.42593184e-01] | [9.439289093017578, 9.392099380493164] |
a30ebc67-aebd-44ca-a823-297334ebf52c | waymo-open-dataset-panoramic-video-panoptic | 2206.07704 | null | https://arxiv.org/abs/2206.07704v1 | https://arxiv.org/pdf/2206.07704v1.pdf | Waymo Open Dataset: Panoramic Video Panoptic Segmentation | Panoptic image segmentation is the computer vision task of finding groups of pixels in an image and assigning semantic classes and object instance identifiers to them. Research in image segmentation has become increasingly popular due to its critical applications in robotics and autonomous driving. The research community thereby relies on publicly available benchmark dataset to advance the state-of-the-art in computer vision. Due to the high costs of densely labeling the images, however, there is a shortage of publicly available ground truth labels that are suitable for panoptic segmentation. The high labeling costs also make it challenging to extend existing datasets to the video domain and to multi-camera setups. We therefore present the Waymo Open Dataset: Panoramic Video Panoptic Segmentation Dataset, a large-scale dataset that offers high-quality panoptic segmentation labels for autonomous driving. We generate our dataset using the publicly available Waymo Open Dataset, leveraging the diverse set of camera images. Our labels are consistent over time for video processing and consistent across multiple cameras mounted on the vehicles for full panoramic scene understanding. Specifically, we offer labels for 28 semantic categories and 2,860 temporal sequences that were captured by five cameras mounted on autonomous vehicles driving in three different geographical locations, leading to a total of 100k labeled camera images. To the best of our knowledge, this makes our dataset an order of magnitude larger than existing datasets that offer video panoptic segmentation labels. We further propose a new benchmark for Panoramic Video Panoptic Segmentation and establish a number of strong baselines based on the DeepLab family of models. We will make the benchmark and the code publicly available. Find the dataset at https://waymo.com/open. | ['Dragomir Anguelov', 'Henrik Kretzschmar', 'Liang-Chieh Chen', 'Yukun Zhu', 'Siyuan Qiao', 'Hang Yan', 'Xinchen Yan', 'Alex Zihao Zhu', 'Jieru Mei'] | 2022-06-15 | null | null | null | null | ['temporal-sequences'] | ['reasoning'] | [ 4.06078398e-01 -4.00672197e-01 -5.35562932e-01 -5.47208786e-01
-6.84227526e-01 -1.00346506e+00 6.72701359e-01 -4.50421363e-01
-4.99621540e-01 1.22051522e-01 -2.39246055e-01 -3.08850914e-01
1.43451199e-01 -8.47143173e-01 -1.06675553e+00 -5.64839423e-01
9.47320908e-02 4.39545035e-01 6.46117210e-01 -4.49367724e-02
1.41037881e-01 3.49781632e-01 -2.11438799e+00 1.15810707e-01
7.11762190e-01 1.19274557e+00 3.96043360e-01 7.54564881e-01
2.74821460e-01 5.41062534e-01 -1.93088889e-01 -3.75670880e-01
9.51271236e-01 -1.44062549e-01 -7.77112007e-01 5.56561112e-01
1.32554376e+00 -4.98511583e-01 -6.41012549e-01 1.18157876e+00
-2.21234038e-01 2.62938887e-01 4.05340493e-01 -1.73598778e+00
-3.55899394e-01 3.35844159e-01 -6.58014655e-01 4.12120402e-01
-2.89426208e-01 4.12646890e-01 1.26463521e+00 -3.56006473e-01
7.94337034e-01 1.05466270e+00 5.23414373e-01 2.66577542e-01
-7.91911781e-01 -8.69672775e-01 8.05830359e-02 2.47526824e-01
-1.15439355e+00 -3.93403262e-01 4.29462016e-01 -7.19034255e-01
6.61073446e-01 2.22660154e-01 7.26453125e-01 8.25304449e-01
7.51230568e-02 7.31886387e-01 1.12047434e+00 1.59815788e-01
5.58647048e-03 -1.35461599e-01 2.09849387e-01 6.85005248e-01
1.90314621e-01 1.91046908e-01 -2.00885102e-01 2.63020426e-01
6.37798727e-01 3.57456475e-01 -1.22471573e-02 -5.87677002e-01
-1.36606467e+00 8.86530280e-01 7.12869465e-01 -1.20890044e-01
-9.06683207e-02 3.76237869e-01 2.57404506e-01 1.63738385e-01
2.51466632e-01 3.65136147e-01 -5.71344137e-01 -4.77452837e-02
-1.07963181e+00 3.21379572e-01 5.88842690e-01 1.12596154e+00
1.29660594e+00 -2.00975388e-01 3.37569386e-01 8.63593519e-01
2.06059709e-01 9.52913702e-01 8.26751292e-02 -1.52445102e+00
5.04922509e-01 5.48991680e-01 -3.35362926e-02 -9.79416430e-01
-3.86293918e-01 1.33259490e-01 -2.48319745e-01 1.90459937e-01
4.32592243e-01 -1.50298148e-01 -1.36319280e+00 1.51156688e+00
2.76310116e-01 5.89021325e-01 5.24522327e-02 1.18344283e+00
8.20564747e-01 9.37036157e-01 -6.69040382e-02 2.66495913e-01
1.55706501e+00 -1.45074606e+00 -3.23287606e-01 -6.79575861e-01
4.20120060e-01 -8.23555112e-01 8.84522319e-01 1.56186804e-01
-3.37817848e-01 -6.22175932e-01 -1.10222399e+00 -1.41306192e-01
-7.17035174e-01 -3.22835743e-02 9.06370640e-01 5.46434581e-01
-1.00122452e+00 8.14862773e-02 -7.30011284e-01 -6.43851876e-01
5.22366047e-01 1.38660848e-01 -4.91662115e-01 -4.11052704e-01
-1.08972442e+00 7.18871236e-01 4.22364414e-01 -6.82534277e-02
-1.31393540e+00 -7.11703598e-01 -1.07884789e+00 -4.25765008e-01
6.47155941e-01 -2.62359023e-01 1.36797464e+00 -1.08687115e+00
-1.02386963e+00 1.18723524e+00 1.60822615e-01 -6.97450280e-01
4.33976740e-01 -1.51494130e-01 -5.65381765e-01 5.34791112e-01
4.07962948e-01 1.53976655e+00 7.38854885e-01 -1.08505917e+00
-1.22114098e+00 -1.33673638e-01 3.36271256e-01 1.70464844e-01
3.43687162e-02 -1.52464304e-02 -1.04981875e+00 -2.47304663e-01
-5.97914755e-02 -1.56774449e+00 -3.00657064e-01 -1.71902657e-01
-2.58553892e-01 7.85882324e-02 1.18041420e+00 -4.94365841e-01
5.99717081e-01 -2.14847732e+00 -8.88252556e-02 -2.41536334e-01
2.21463680e-01 1.56918779e-01 -2.13045299e-01 -4.87110242e-02
1.69320673e-01 4.41813283e-02 -7.23311543e-01 -1.01887666e-01
-1.44859135e-01 3.63648444e-01 -5.45122862e-01 5.74619949e-01
2.97828820e-02 8.77600431e-01 -7.76931763e-01 -4.82917249e-01
6.22293174e-01 8.44000354e-02 -5.02228498e-01 3.79066616e-02
-5.40051639e-01 2.47837737e-01 -3.22458625e-01 9.18018401e-01
9.16051507e-01 -1.31156910e-02 -2.86734819e-01 -1.48878440e-01
-2.74289608e-01 -1.00659272e-02 -8.50802541e-01 1.68096852e+00
-2.45366693e-01 1.28258276e+00 7.11858943e-02 -9.17493045e-01
7.78537869e-01 -6.55569881e-02 6.77982748e-01 -4.71877903e-01
2.38839537e-01 2.36132592e-01 -1.60707638e-01 -5.57363331e-01
1.09514046e+00 2.04799294e-01 -5.20881593e-01 3.04951787e-01
-2.67608222e-02 -8.01668584e-01 6.34365857e-01 2.12679625e-01
9.05574858e-01 -1.44726951e-02 -2.46984005e-01 -2.94883490e-01
2.40672618e-01 7.30787933e-01 6.25690460e-01 7.02280223e-01
-5.77903092e-01 8.81049395e-01 2.02408910e-01 -5.61773360e-01
-1.17607105e+00 -9.35522556e-01 -5.67409694e-01 9.10214067e-01
6.72301769e-01 -3.46291184e-01 -7.72009790e-01 -5.16209304e-01
1.08100221e-01 3.05540562e-01 -5.97773671e-01 3.40291023e-01
-4.13155496e-01 -8.92729640e-01 6.41229868e-01 3.30513626e-01
9.80684698e-01 -9.49634075e-01 -8.05551410e-01 -1.48437589e-01
-3.25012714e-01 -1.83354092e+00 -4.91358191e-01 -8.20782185e-02
-6.50776744e-01 -1.52544343e+00 -3.82493287e-01 -7.82035708e-01
3.13492537e-01 9.19990361e-01 1.04406822e+00 -2.26452008e-01
-3.98922235e-01 5.72043002e-01 -2.38618508e-01 -1.76317081e-01
-2.36667633e-01 1.28235266e-01 -1.27307445e-01 1.14571854e-01
6.19274795e-01 -1.19907796e-01 -6.74188673e-01 7.05986381e-01
-1.11573315e+00 3.74627709e-01 4.34827030e-01 8.09979364e-02
6.92923129e-01 8.48967731e-02 2.20378283e-02 -1.00775754e+00
-3.09605151e-01 -8.02301407e-01 -1.27515364e+00 -3.75754498e-02
-1.84844688e-01 -4.65094179e-01 2.67160296e-01 -3.07004992e-02
-7.49161243e-01 3.92535836e-01 4.12951149e-02 -5.90604782e-01
-4.39492166e-01 1.78460956e-01 1.45008013e-01 -8.18339437e-02
3.32132131e-01 -2.38528792e-02 -1.23116732e-01 -8.51321667e-02
8.07138681e-01 8.23464870e-01 1.02324986e+00 -2.26023257e-01
7.75083184e-01 8.39220107e-01 -1.91223189e-01 -9.50020730e-01
-9.92721081e-01 -9.44866121e-01 -6.13468468e-01 -3.83806616e-01
1.33776855e+00 -1.42247427e+00 -2.33755454e-01 8.74521673e-01
-5.97985029e-01 -6.23615682e-01 9.77500379e-02 4.21826482e-01
-6.96163237e-01 2.95720071e-01 -4.31271106e-01 -2.30798587e-01
-2.36039832e-02 -1.63934720e+00 1.18180645e+00 3.64916176e-01
2.71263391e-01 -7.58134305e-01 -9.56759751e-02 9.18357670e-01
1.06354430e-01 3.16122025e-01 1.40194997e-01 -1.38477415e-01
-1.21873701e+00 -1.09854639e-01 -4.64729041e-01 3.96158725e-01
-4.08752412e-02 3.48675907e-01 -9.92935359e-01 -2.93041784e-02
-2.45761648e-01 -3.99432242e-01 1.52425468e+00 5.60613215e-01
1.05010974e+00 1.57494694e-01 -4.70196575e-01 1.01492548e+00
1.45375848e+00 7.61242956e-02 7.22683012e-01 5.67034304e-01
1.07285535e+00 7.45498538e-01 8.87748241e-01 -1.42974943e-01
7.26756215e-01 5.78873873e-01 6.51589811e-01 -5.97707406e-02
-1.02638826e-01 -1.59285575e-01 4.27682221e-01 4.44578379e-01
4.08714294e-01 -3.35960567e-01 -1.20662165e+00 9.71059978e-01
-1.81216609e+00 -8.05755734e-01 -3.46173763e-01 1.97227097e+00
4.95397419e-01 2.05561984e-02 4.07905281e-02 -3.10015768e-01
8.63866448e-01 5.47609389e-01 -5.93213141e-01 -6.73741028e-02
-2.76589662e-01 -2.66762912e-01 1.28297007e+00 4.27803934e-01
-1.88083780e+00 1.42679107e+00 5.96873331e+00 6.97552145e-01
-1.32824767e+00 1.08318202e-01 7.91414678e-01 -1.64119340e-02
7.46833608e-02 2.51694798e-01 -1.26040792e+00 4.77654576e-01
1.16283059e+00 8.70173648e-02 4.52555120e-01 9.71002281e-01
-4.36176248e-02 -5.07222950e-01 -8.38662207e-01 1.13433027e+00
2.40466684e-01 -1.53254676e+00 -9.18724909e-02 1.72384560e-01
1.33491838e+00 1.19320929e+00 3.34128216e-02 -1.94718819e-02
5.52905083e-01 -7.49275386e-01 8.50967765e-01 2.19441075e-02
7.21157849e-01 -3.86177957e-01 4.13898796e-01 6.34494936e-04
-1.31153238e+00 -2.59864897e-01 -4.29131716e-01 3.58673893e-02
2.92746782e-01 3.64226162e-01 -4.17806149e-01 1.28820524e-01
1.10497224e+00 1.33130884e+00 -8.37174892e-01 1.18525457e+00
-1.84224740e-01 7.08568037e-01 -5.94475627e-01 5.04846931e-01
7.59043813e-01 -4.84893352e-01 4.05532658e-01 1.00337887e+00
2.41464183e-01 -1.36219412e-01 4.94704217e-01 6.68764412e-01
-3.68917137e-01 -2.10517794e-01 -8.92590940e-01 -1.52358249e-01
5.18482745e-01 1.65658712e+00 -1.11162925e+00 -5.71747899e-01
-7.18445003e-01 6.18999839e-01 -2.45178446e-01 2.64625460e-01
-9.34691489e-01 -1.89422339e-01 1.12299168e+00 -1.46672368e-01
4.37253267e-01 -4.25201148e-01 -1.64599225e-01 -1.21799636e+00
-3.17065746e-01 -6.91008687e-01 5.35780132e-01 -8.52851748e-01
-9.73518252e-01 5.44062018e-01 3.90874892e-01 -1.21606863e+00
6.66972548e-02 -7.13585794e-01 -3.71946663e-01 2.27177471e-01
-1.76845312e+00 -1.25131011e+00 -6.58189058e-01 4.67649490e-01
8.41631711e-01 -9.77246985e-02 2.42576420e-01 4.45672661e-01
-6.23000085e-01 -8.43769833e-02 1.57168284e-01 2.61025906e-01
7.23767161e-01 -9.86212134e-01 7.28103161e-01 1.23725224e+00
2.87500709e-01 -7.92092551e-03 5.20391226e-01 -2.95851529e-01
-1.24178123e+00 -1.61366487e+00 4.21597481e-01 -9.48901474e-01
9.17768121e-01 -5.02177238e-01 -5.97710192e-01 1.01414621e+00
2.23387465e-01 1.17156364e-01 2.15081826e-01 -4.74464744e-01
-2.91097790e-01 -3.71392995e-01 -7.40972757e-01 4.97615367e-01
9.15399015e-01 -4.76388991e-01 -5.34973145e-01 6.22294068e-01
9.06122804e-01 -5.38870156e-01 -5.90425491e-01 5.33609569e-01
3.32061499e-01 -8.60926390e-01 9.31089699e-01 2.36792982e-01
5.57195961e-01 -6.44430697e-01 -3.38699937e-01 -1.03448248e+00
2.24253133e-01 -3.70193899e-01 5.62087953e-01 1.08997488e+00
3.51988524e-01 -7.33871162e-01 6.05403900e-01 5.27483344e-01
-4.77307498e-01 -1.45711586e-01 -7.14731693e-01 -6.17355347e-01
8.14770982e-02 -8.03267598e-01 6.06991470e-01 8.95984769e-01
-5.81365049e-01 8.50988925e-02 -3.87812048e-01 3.60905886e-01
5.23763418e-01 6.12093210e-01 1.12785435e+00 -1.11668992e+00
1.18118882e-01 -4.36920524e-01 -5.72263122e-01 -1.29359794e+00
3.46371740e-01 -9.88782346e-01 3.99786055e-01 -1.37648177e+00
2.08925962e-01 -7.12811708e-01 -1.92798227e-02 5.41014969e-01
6.16488159e-02 1.04150164e+00 1.88188359e-01 5.26386559e-01
-8.39483976e-01 2.65891582e-01 1.04187143e+00 -3.87192100e-01
8.76633599e-02 -2.52960116e-01 -4.73854065e-01 7.68631279e-01
1.01060951e+00 -5.03048062e-01 -4.47009712e-01 -7.25224912e-01
-8.20636898e-02 -2.82808721e-01 5.25612772e-01 -1.31693935e+00
1.81562871e-01 -3.31771314e-01 -1.57867491e-01 -8.31946850e-01
4.41205949e-01 -8.52514803e-01 1.79882482e-01 2.46790588e-01
-4.69703898e-02 -3.80270071e-02 4.32701528e-01 5.69543540e-01
-3.81000102e-01 -1.50276404e-02 9.18173194e-01 -1.32101312e-01
-1.79194927e+00 7.19093442e-01 -5.61026514e-01 2.74975598e-01
1.34011447e+00 -1.04203284e-01 -6.99875891e-01 -1.25024185e-01
-6.72637438e-03 6.14182353e-01 9.30053413e-01 1.06634510e+00
1.46861464e-01 -8.58487844e-01 -4.91358668e-01 2.50466168e-01
5.25751829e-01 1.44226789e-01 2.32187793e-01 7.48588443e-01
-1.18759978e+00 6.42980158e-01 -4.02867228e-01 -1.01529539e+00
-1.06598425e+00 3.92459869e-01 3.04960400e-01 5.56091249e-01
-8.97088826e-01 6.63832247e-01 5.66807449e-01 -6.27358913e-01
-1.89646482e-01 -6.33713007e-01 -1.86013475e-01 1.08799562e-01
4.29492950e-01 1.47668526e-01 -1.21169962e-01 -1.20637453e+00
-2.90577799e-01 9.07445550e-01 1.31445110e-01 -2.65517514e-02
1.16293097e+00 -3.47445220e-01 -2.18810737e-01 3.36304307e-01
1.28571200e+00 -2.27911443e-01 -1.89778030e+00 -1.77082986e-01
-1.68365896e-01 -6.81889415e-01 2.54725546e-01 -3.71266663e-01
-1.51599693e+00 7.82609761e-01 6.08037591e-01 -5.89653067e-02
9.62703407e-01 1.61692619e-01 1.11488247e+00 4.79642361e-01
4.24683660e-01 -1.12983465e+00 -2.33846635e-01 5.56122601e-01
3.57131660e-01 -1.72815013e+00 -9.00136903e-02 -6.12306952e-01
-7.80489028e-01 8.47892642e-01 7.18877077e-01 -2.64200717e-01
5.54644525e-01 2.04523094e-02 6.10017598e-01 -2.85820156e-01
-4.98219937e-01 -5.01627088e-01 1.61282450e-01 3.94540668e-01
-2.62759805e-01 3.43631357e-01 1.65475592e-01 2.36641914e-02
-4.24116969e-01 -1.41205460e-01 7.63555527e-01 5.86615503e-01
-5.87152719e-01 -7.55299985e-01 -3.63844305e-01 4.81624007e-01
-1.87439308e-01 -1.93555504e-02 -2.36993074e-01 8.75985026e-01
2.96096981e-01 1.28387737e+00 4.99688357e-01 -3.12661260e-01
-4.21309024e-02 -3.08036447e-01 4.43849154e-02 -4.04070646e-01
2.40852069e-02 -3.62507910e-01 6.45230711e-02 -5.89024365e-01
-8.02713573e-01 -8.54170024e-01 -1.08173203e+00 -3.92180502e-01
-3.89897376e-02 -4.86172512e-02 8.23277712e-01 7.64177203e-01
2.10976526e-01 1.20647565e-01 5.34017861e-01 -1.10365546e+00
2.35559508e-01 -6.79047644e-01 -5.60436368e-01 5.91522992e-01
3.17420512e-01 -7.70627439e-01 -3.88789356e-01 4.23119575e-01] | [8.370272636413574, -1.6246753931045532] |
7345f624-2515-41ff-bfad-7a219cd6c505 | progressive-refinement-a-method-of-coarse-to | 1804.08256 | null | http://arxiv.org/abs/1804.08256v1 | http://arxiv.org/pdf/1804.08256v1.pdf | Progressive refinement: a method of coarse-to-fine image parsing using stacked network | To parse images into fine-grained semantic parts, the complex fine-grained
elements will put it in trouble when using off-the-shelf semantic segmentation
networks. In this paper, for image parsing task, we propose to parse images
from coarse to fine with progressively refined semantic classes. It is achieved
by stacking the segmentation layers in a segmentation network several times.
The former segmentation module parses images at a coarser-grained level, and
the result will be feed to the following one to provide effective contextual
clues for the finer-grained parsing. To recover the details of small
structures, we add skip connections from shallow layers of the network to
fine-grained parsing modules. As for the network training, we merge classes in
groundtruth to get coarse-to-fine label maps, and train the stacked network
with these hierarchical supervision end-to-end. Our coarse-to-fine stacked
framework can be injected into many advanced neural networks to improve the
parsing results. Extensive evaluations on several public datasets including
face parsing and human parsing well demonstrate the superiority of our method. | ['Zhengxing Sun', 'Yunhan Sun', 'Jiagao Hu', 'Jinlong Shi'] | 2018-04-23 | null | null | null | null | ['face-parsing', 'human-parsing'] | ['computer-vision', 'computer-vision'] | [ 4.00423080e-01 6.96129799e-01 -5.84688634e-02 -9.08723235e-01
-8.11600029e-01 -6.12561464e-01 6.07380690e-03 -1.36036932e-01
-1.95164055e-01 4.87906516e-01 1.46649688e-01 -2.04076767e-01
3.53957027e-01 -1.02653253e+00 -8.32887352e-01 -4.94906455e-01
2.73927003e-01 5.59170902e-01 6.03935838e-01 7.86303263e-03
4.25279401e-02 2.81937838e-01 -1.39038444e+00 7.05221653e-01
9.02592003e-01 1.36402631e+00 3.44460040e-01 5.99857807e-01
-6.98647678e-01 5.75688899e-01 -5.00700593e-01 -6.96505606e-01
1.00706905e-01 -1.69523388e-01 -1.20879436e+00 3.96164298e-01
5.39545238e-01 -4.20145214e-01 3.70833993e-01 1.46158063e+00
-8.49167407e-02 -3.56480479e-01 7.56496713e-02 -8.62625659e-01
-6.09999478e-01 7.77957439e-01 -7.28573620e-01 -1.34713084e-01
1.23971559e-01 -2.45673746e-01 9.12254751e-01 -8.38667572e-01
3.71622115e-01 1.79698157e+00 5.59842587e-01 7.10709333e-01
-7.72350907e-01 -7.93587446e-01 6.89283311e-01 -2.86787838e-01
-7.98899293e-01 -3.20227176e-01 6.29659593e-01 -4.13982153e-01
3.79271865e-01 -1.57579616e-01 4.25349355e-01 7.16374278e-01
-2.13997379e-01 6.26180470e-01 9.39384460e-01 -1.41117424e-01
4.35397169e-03 -1.90654755e-01 7.07885861e-01 1.25830555e+00
3.46641988e-01 -4.23259676e-01 -8.20218846e-02 1.46338776e-01
9.66798961e-01 1.29608765e-01 2.24501833e-01 7.40437657e-02
-5.85916996e-01 9.77406740e-01 8.20396781e-01 7.18110949e-02
-2.70934284e-01 1.29421189e-01 2.10202828e-01 -2.13926658e-01
6.09595418e-01 8.40361491e-02 -8.06733429e-01 3.00116599e-01
-9.38196480e-01 -1.63290426e-01 7.07680941e-01 9.29001212e-01
1.23058581e+00 -4.29903805e-01 -2.44528517e-01 1.07236493e+00
4.15431291e-01 3.61350439e-02 2.25000590e-01 -1.23293304e+00
7.77726471e-01 9.05601501e-01 -2.23883778e-01 -5.69928646e-01
-4.71532881e-01 -2.37825185e-01 -8.71585667e-01 -2.15955265e-02
4.67017561e-01 -2.85094708e-01 -1.62165439e+00 1.57731152e+00
5.70343554e-01 9.32340473e-02 -1.72154516e-01 9.85742867e-01
1.10027730e+00 5.94854534e-01 7.37880349e-01 1.85030833e-01
1.91994762e+00 -1.44806027e+00 -3.68987352e-01 -5.84554791e-01
3.46686602e-01 -6.68926597e-01 1.22702885e+00 1.27505690e-01
-1.05732715e+00 -8.85660410e-01 -8.61759782e-01 -4.34366673e-01
-3.69950265e-01 2.65840113e-01 6.85191274e-01 4.90117788e-01
-1.20032251e+00 7.32264161e-01 -9.52413797e-01 -2.75571883e-01
1.12006712e+00 2.64519811e-01 -3.15706939e-01 -2.64115185e-01
-1.10789406e+00 2.94430375e-01 5.81687808e-01 2.96178877e-01
-6.99573994e-01 -4.60684627e-01 -1.11209893e+00 2.81510830e-01
4.09157842e-01 -9.31101084e-01 1.17684960e+00 -9.78832066e-01
-1.33371150e+00 1.09680271e+00 -2.92270958e-01 7.51324045e-03
1.65800720e-01 -1.66059911e-01 1.67107835e-01 4.07704949e-01
7.13393271e-01 1.27096331e+00 8.69465351e-01 -1.22035062e+00
-1.00079000e+00 -6.52822196e-01 4.21160877e-01 3.44302319e-02
4.94900383e-02 6.02994151e-02 -1.09552455e+00 -5.16860068e-01
4.48069960e-01 -6.48109972e-01 -3.72606367e-01 -3.37710232e-01
-7.10750282e-01 -5.17547965e-01 6.24913871e-01 -8.86941552e-01
8.32879841e-01 -1.98891950e+00 -1.81577392e-02 -1.87634230e-01
2.02349529e-01 3.94627526e-02 -3.40450764e-01 -2.32922763e-01
-1.01529747e-01 3.63974631e-01 -3.48655552e-01 -6.77151263e-01
-1.01254456e-01 3.13688278e-01 -1.87527135e-01 -7.43390247e-02
7.60429442e-01 1.05211246e+00 -6.58181190e-01 -7.55731344e-01
-6.81039914e-02 2.14937493e-01 -7.52434731e-01 4.46065038e-01
-3.60803813e-01 4.71064448e-01 -1.00224316e+00 9.36932504e-01
8.65088999e-01 -6.59564018e-01 -9.20630991e-02 -3.98529887e-01
2.34443128e-01 1.41224608e-01 -7.52408087e-01 1.67746651e+00
-4.11167502e-01 -1.53466444e-02 3.34867328e-01 -9.62283611e-01
6.21551514e-01 -1.68915242e-01 5.98004311e-02 -7.02869475e-01
2.18887776e-01 -1.31862551e-01 -3.66034269e-01 -4.59451765e-01
3.21918339e-01 -2.19187781e-01 -5.94640672e-01 1.51057109e-01
2.59277314e-01 -2.02047117e-02 2.33115390e-01 1.16541050e-01
6.74165606e-01 3.33551407e-01 9.98879317e-03 -1.97363868e-01
5.78769028e-01 1.33862138e-01 8.22316945e-01 4.86263007e-01
-1.40914083e-01 7.56712258e-01 7.52172589e-01 -5.82516551e-01
-8.04564059e-01 -8.59558105e-01 -8.85655135e-02 1.75869715e+00
3.13258082e-01 -2.54538447e-01 -1.70706427e+00 -9.58911002e-01
-1.82605848e-01 4.22204398e-02 -7.74434984e-01 1.57078654e-01
-7.13372052e-01 -7.98129678e-01 4.94338810e-01 8.46511960e-01
1.06183732e+00 -1.37790000e+00 -2.49949440e-01 2.09085807e-01
-2.86440045e-01 -1.59589469e+00 -4.06071693e-01 2.51826525e-01
-9.69929338e-01 -1.16745412e+00 -4.50281948e-01 -1.22744083e+00
9.74668860e-01 6.64626285e-02 1.35431659e+00 4.08030480e-01
-2.59779304e-01 -2.19259486e-01 -3.45284373e-01 -1.28463313e-01
-1.70824394e-01 2.26487592e-01 -6.28655255e-01 -1.23047128e-01
1.81199059e-01 -2.24051014e-01 -6.34829521e-01 2.40060166e-01
-7.38759398e-01 3.01985174e-01 6.67708158e-01 7.15494871e-01
8.81261289e-01 1.60577327e-01 3.95523489e-01 -1.59292364e+00
3.34710598e-01 -1.24252841e-01 -7.41389632e-01 2.97436982e-01
-1.60790563e-01 2.85215557e-01 7.57723331e-01 7.28666112e-02
-1.18033254e+00 3.64382029e-01 -6.04170620e-01 -1.36827603e-01
-5.78681350e-01 2.00475037e-01 -6.95053160e-01 6.33105114e-02
2.28114668e-02 -1.80460423e-01 -5.34165561e-01 -7.73567677e-01
6.08276367e-01 4.79817629e-01 7.39224076e-01 -8.02663684e-01
5.99169075e-01 3.16695809e-01 -3.59137982e-01 -3.58694047e-01
-1.77370894e+00 -1.44845411e-01 -8.86986613e-01 3.73467535e-01
1.63085318e+00 -1.13023686e+00 -4.30565149e-01 6.38848066e-01
-1.25149345e+00 -6.10664666e-01 1.82559702e-03 -3.80944610e-01
-3.18389624e-01 2.02766553e-01 -1.06663752e+00 -1.54318795e-01
-3.42577279e-01 -1.37377918e+00 1.76985276e+00 8.24453115e-01
2.78515726e-01 -8.40878725e-01 -3.82802367e-01 8.74632776e-01
-1.19328775e-01 3.90319116e-02 9.42801714e-01 -4.57060248e-01
-7.06177652e-01 4.51926701e-02 -8.28615546e-01 3.71988267e-01
5.17652184e-03 -9.94387344e-02 -1.23790288e+00 1.42829828e-02
-2.22825985e-02 -3.92689735e-01 1.26679599e+00 5.15918255e-01
1.82092905e+00 -2.62083054e-01 -4.39030796e-01 1.06283760e+00
1.24257994e+00 -9.68569592e-02 4.71711725e-01 2.48178512e-01
1.19953775e+00 7.87002146e-01 7.99896598e-01 3.68914008e-02
7.63997853e-01 8.94040167e-02 4.85243499e-01 -4.88273352e-01
-2.78097481e-01 -4.17973876e-01 1.70354098e-02 4.23211277e-01
2.82716244e-01 -5.08892648e-02 -5.64814627e-01 3.44635695e-01
-1.56705713e+00 -5.23906589e-01 1.16599433e-01 1.42221975e+00
8.24109435e-01 3.93876165e-01 -1.36003476e-02 -4.36175287e-01
1.05923641e+00 2.87594318e-01 -5.95057428e-01 -4.39610630e-01
2.56561428e-01 3.14968675e-01 3.98308158e-01 6.57991230e-01
-1.44880712e+00 1.76564217e+00 5.46829939e+00 7.52867520e-01
-9.68904138e-01 1.35189101e-01 1.32325041e+00 4.87516195e-01
-4.98722307e-02 6.07248321e-02 -1.23181832e+00 4.58560586e-01
6.34289205e-01 6.93036139e-01 2.40679204e-01 1.03493667e+00
-9.48302895e-02 -1.12924717e-01 -1.03142583e+00 8.06207716e-01
-3.50309312e-01 -1.20834935e+00 1.94514066e-01 -1.72463119e-01
6.24474704e-01 1.11178178e-02 -3.22391212e-01 4.33956921e-01
6.78064942e-01 -1.20888448e+00 6.32239401e-01 8.32223222e-02
7.39492357e-01 -6.05233788e-01 7.33706713e-01 3.73829871e-01
-1.54570568e+00 6.08928837e-02 -8.01282823e-01 1.15547985e-01
2.07210049e-01 7.40467370e-01 -5.27971923e-01 2.93803245e-01
9.99563158e-01 4.98261780e-01 -7.49833286e-01 4.57891703e-01
-5.88211477e-01 4.31428701e-01 -8.58925804e-02 4.04792577e-01
4.44608510e-01 -3.76365900e-01 -3.86911690e-01 1.26649249e+00
-7.41733005e-03 5.18803239e-01 4.96959478e-01 7.64813185e-01
-5.30228496e-01 -3.96889627e-01 1.51589718e-02 -2.27561351e-02
4.92468655e-01 1.84286082e+00 -1.18482018e+00 -6.75431252e-01
-3.48659724e-01 1.28275645e+00 7.46234000e-01 3.15747499e-01
-7.68920600e-01 -3.11753839e-01 6.59813464e-01 -7.41479024e-02
5.01793504e-01 1.60698697e-01 -5.99733412e-01 -1.05232227e+00
4.08184491e-02 -5.87732434e-01 7.22386539e-01 -7.89031088e-01
-1.43509245e+00 9.60091054e-01 -3.09245646e-01 -3.96874309e-01
-2.45751627e-02 -8.06420505e-01 -6.11590564e-01 1.06077051e+00
-1.77932072e+00 -1.41021407e+00 -6.66275978e-01 5.80753386e-01
7.23713815e-01 3.33664626e-01 6.43982232e-01 2.54401118e-01
-8.30428123e-01 4.91922408e-01 -7.46830285e-01 7.97669590e-01
3.54461372e-01 -1.37276471e+00 7.77185321e-01 9.01159465e-01
-2.37262800e-01 4.68157172e-01 1.12565480e-01 -7.37448096e-01
-6.35843337e-01 -1.55721629e+00 7.63412356e-01 -3.78641784e-01
5.73055983e-01 -6.06616080e-01 -9.03951049e-01 6.88997805e-01
-5.89898601e-02 2.70467043e-01 2.57909209e-01 1.96376801e-01
-5.24798393e-01 -5.67233004e-02 -1.11393118e+00 2.43752077e-01
1.12742209e+00 -5.81930220e-01 -7.25737333e-01 2.31816694e-01
1.26032770e+00 -5.18573165e-01 -8.00298810e-01 4.23751354e-01
2.70193875e-01 -9.64412212e-01 1.02678943e+00 -6.39324963e-01
7.66954064e-01 -1.82216406e-01 -1.28409620e-02 -1.09585941e+00
-4.28132325e-01 -9.50790718e-02 5.40456712e-01 1.47087669e+00
4.53769624e-01 -4.26405668e-01 1.20254743e+00 6.08850241e-01
-3.17238778e-01 -7.91879475e-01 -4.77591574e-01 -7.17399344e-02
3.10719997e-01 -1.56336010e-01 8.47714245e-01 6.30688071e-01
-4.93412882e-01 6.37620687e-01 1.34914532e-01 5.30908644e-01
7.84187436e-01 6.00365639e-01 4.72593099e-01 -1.41984296e+00
-1.42158091e-01 -3.58426750e-01 -2.13525034e-02 -1.40617025e+00
3.28535587e-01 -6.65294468e-01 3.85165989e-01 -1.79007876e+00
5.16518116e-01 -5.45882583e-01 -4.22797278e-02 8.07165205e-01
-8.06620955e-01 6.41529024e-01 2.43776560e-01 7.33075216e-02
-8.25606763e-01 2.04620868e-01 1.68218255e+00 -9.92770791e-02
2.86690164e-02 -1.16993807e-01 -1.28016591e+00 1.22962213e+00
6.29169166e-01 -4.53328311e-01 -3.41399133e-01 -8.10397327e-01
-1.84092507e-01 2.33469352e-01 3.58224422e-01 -7.67435253e-01
7.91285262e-02 5.36852144e-02 7.49627173e-01 -6.92967594e-01
1.79644436e-01 -6.21083498e-01 -3.95941645e-01 1.14145562e-01
-2.20996261e-01 -3.03371698e-01 3.99092883e-02 2.63810605e-01
-4.11212981e-01 -1.87393039e-01 1.11663592e+00 -5.25679946e-01
-8.87275577e-01 7.76439250e-01 3.30757290e-01 2.93328583e-01
7.14840114e-01 -2.95092076e-01 -5.36141694e-01 1.79396376e-01
-1.22418714e+00 5.47165453e-01 4.75368351e-01 3.85926068e-01
2.64989227e-01 -9.49306667e-01 -3.58582169e-01 2.99379081e-01
-1.48267895e-01 8.06955099e-01 3.10227007e-01 2.69704252e-01
-5.96446216e-01 3.38258386e-01 -2.52679467e-01 -7.68558204e-01
-1.12264562e+00 4.64621931e-01 4.55452532e-01 -3.95939976e-01
-4.35017109e-01 1.31663156e+00 9.66287255e-01 -6.10125959e-01
1.99319616e-01 -5.68566024e-01 -3.82749379e-01 1.50098488e-01
5.11060774e-01 -2.77754098e-01 -3.68069788e-03 -4.44986850e-01
-3.98513645e-01 1.12637508e+00 -1.81449682e-01 2.82226056e-01
1.43863618e+00 -4.30603653e-01 -4.79501814e-01 -1.78366348e-01
1.29150081e+00 -2.38974214e-01 -1.74522293e+00 -3.48889343e-02
8.89531970e-02 -4.44874242e-02 -2.28494540e-01 -8.69519889e-01
-1.47300076e+00 1.15843618e+00 3.36893529e-01 2.49829262e-01
1.42787683e+00 5.56894481e-01 9.11507905e-01 1.19564012e-01
3.48578006e-01 -9.25998569e-01 -6.47217557e-02 6.76310182e-01
3.33404124e-01 -1.37219524e+00 -3.40754181e-01 -1.09823084e+00
-5.33898771e-01 1.19786942e+00 8.49886239e-01 -2.44565368e-01
6.26870990e-01 3.02757055e-01 2.00769380e-01 -3.19003969e-01
-3.48384082e-01 -4.00477797e-01 1.04782358e-01 5.69039524e-01
3.79259408e-01 1.41815215e-01 6.79268241e-02 1.24891174e+00
-3.70417506e-01 -7.35529140e-02 1.96525887e-01 4.90577549e-01
-9.18224454e-01 -1.32389665e+00 -2.51715124e-01 3.08730781e-01
-6.52599454e-01 -2.55015910e-01 -4.55633551e-01 6.37865245e-01
5.51458299e-01 9.65193450e-01 2.04282448e-01 -9.27082896e-02
3.60830694e-01 2.08101962e-02 2.64971375e-01 -1.00912786e+00
-5.73294878e-01 2.10548818e-01 2.20495656e-01 -1.07961035e+00
-2.95614749e-01 -3.71200919e-01 -1.75606763e+00 -5.64747415e-02
1.05830364e-01 1.83575973e-01 4.80735987e-01 1.27766919e+00
2.89427668e-01 8.59858513e-01 5.42896688e-01 -9.76399422e-01
-1.32591531e-01 -8.87140632e-01 -3.51256639e-01 4.69051421e-01
2.23080963e-01 -3.98841798e-01 1.56874545e-02 2.61986285e-01] | [9.449613571166992, 0.36703452467918396] |
f79e532e-88b2-4400-b345-afb83c7c9b51 | ganerf-leveraging-discriminators-to-optimize | 2306.06044 | null | https://arxiv.org/abs/2306.06044v1 | https://arxiv.org/pdf/2306.06044v1.pdf | GANeRF: Leveraging Discriminators to Optimize Neural Radiance Fields | Neural Radiance Fields (NeRF) have shown impressive novel view synthesis results; nonetheless, even thorough recordings yield imperfections in reconstructions, for instance due to poorly observed areas or minor lighting changes. Our goal is to mitigate these imperfections from various sources with a joint solution: we take advantage of the ability of generative adversarial networks (GANs) to produce realistic images and use them to enhance realism in 3D scene reconstruction with NeRFs. To this end, we learn the patch distribution of a scene using an adversarial discriminator, which provides feedback to the radiance field reconstruction, thus improving realism in a 3D-consistent fashion. Thereby, rendering artifacts are repaired directly in the underlying 3D representation by imposing multi-view path rendering constraints. In addition, we condition a generator with multi-resolution NeRF renderings which is adversarially trained to further improve rendering quality. We demonstrate that our approach significantly improves rendering quality, e.g., nearly halving LPIPS scores compared to Nerfacto while at the same time improving PSNR by 1.4dB on the advanced indoor scenes of Tanks and Temples. | ['Matthias Nießner', 'Peter Kontschieder', 'Samuel Rota Bulò', 'Lorenzo Porzi', 'Norman Müller', 'Barbara Roessle'] | 2023-06-09 | null | null | null | null | ['3d-scene-reconstruction', 'novel-view-synthesis'] | ['computer-vision', 'computer-vision'] | [ 5.24501562e-01 9.26280953e-03 6.96464360e-01 -2.24510700e-01
-1.05228138e+00 -1.02016413e+00 5.34702063e-01 -4.51292545e-01
2.21174166e-01 7.26169586e-01 4.81729507e-01 -1.62738934e-01
2.48434007e-01 -1.06974173e+00 -1.19531381e+00 -7.43608415e-01
1.83196068e-01 -1.10226765e-01 -3.41788203e-01 -4.23019171e-01
-7.16703832e-02 7.83820510e-01 -1.31492710e+00 2.96042442e-01
1.06925249e+00 8.57643485e-01 3.40780523e-03 1.00075948e+00
3.31317246e-01 1.07929218e+00 -8.51273298e-01 -5.19883692e-01
7.55368233e-01 -4.73545313e-01 -1.30014226e-01 2.34016582e-01
7.55250216e-01 -9.69411373e-01 -7.52945006e-01 9.35943842e-01
5.05454063e-01 1.96760163e-01 5.40926158e-01 -8.65816057e-01
-7.50733137e-01 -4.96266447e-02 -4.72985804e-01 -2.53362119e-01
5.73252022e-01 5.82939029e-01 7.41626799e-01 -6.77331507e-01
3.86897862e-01 1.28364551e+00 5.75751543e-01 4.21819389e-01
-1.50426269e+00 -6.76430762e-01 -1.04291081e-01 -4.45226401e-01
-1.23380351e+00 -5.95223963e-01 9.17911530e-01 -2.05000073e-01
5.45144379e-01 5.63467681e-01 4.48817551e-01 1.35927022e+00
3.79813552e-01 2.66435057e-01 1.20717752e+00 -9.47266445e-02
2.09543124e-01 -6.82586804e-02 -9.22032893e-01 5.64603567e-01
-1.34895340e-01 3.83970112e-01 -3.82632881e-01 -8.26032385e-02
1.37852407e+00 -3.14956717e-02 -6.77474618e-01 -1.05009638e-01
-1.02617097e+00 5.43865144e-01 8.52488577e-01 -2.21673787e-01
-5.36519766e-01 3.77847463e-01 -1.39963776e-01 1.79074988e-01
4.72884864e-01 7.67983913e-01 -1.07839227e-01 3.37077498e-01
-6.04480207e-01 4.34913009e-01 4.55221623e-01 7.84565747e-01
5.89211285e-01 7.42917776e-01 -1.86531588e-01 7.97648489e-01
9.95478928e-02 1.01070166e+00 -6.04781844e-02 -1.43070376e+00
5.28876483e-01 3.06161642e-02 3.29885930e-01 -9.79148448e-01
8.69565234e-02 -5.69232047e-01 -1.04915977e+00 8.87213767e-01
2.04414934e-01 -2.24708945e-01 -1.14326155e+00 1.98350179e+00
1.75430834e-01 1.73260257e-01 2.94932604e-01 1.00726402e+00
6.04992330e-01 9.87678051e-01 -1.29549995e-01 2.05901027e-01
9.65154946e-01 -6.30854428e-01 -6.73876405e-01 -3.10421646e-01
-1.32288083e-01 -9.30222273e-01 1.12825286e+00 4.96365279e-01
-1.32347012e+00 -6.19610012e-01 -1.12649286e+00 2.62497831e-02
1.95165247e-01 -2.83133239e-01 5.01631916e-01 6.24972165e-01
-9.16966379e-01 5.71206331e-01 -5.91640353e-01 3.65930647e-01
5.05925059e-01 -1.53797537e-01 -3.33988160e-01 -4.08532888e-01
-1.00218356e+00 6.34078860e-01 -2.94276983e-01 4.36586849e-02
-1.32878840e+00 -1.16862094e+00 -9.67170537e-01 9.89049324e-04
2.39394158e-01 -1.03255069e+00 9.52213347e-01 -1.03613758e+00
-1.62636995e+00 3.39244306e-01 1.18107259e-01 -2.49911159e-01
7.81646967e-01 -2.34048635e-01 -3.73638034e-01 2.20161617e-01
7.23628178e-02 6.11787260e-01 8.82902861e-01 -1.85981774e+00
-1.33929178e-01 -1.85898259e-01 4.58175004e-01 4.77952719e-01
1.16954900e-01 -5.53143919e-01 -2.13942245e-01 -1.02238286e+00
1.46794409e-01 -7.06304133e-01 -3.90601069e-01 2.55736113e-01
-4.96993154e-01 8.43632460e-01 5.10501564e-01 -1.06908154e+00
4.36326802e-01 -2.06624365e+00 1.12450130e-01 9.27508548e-02
2.93715119e-01 -8.62275064e-02 -3.19786251e-01 3.07781637e-01
-1.38522699e-01 1.17775463e-01 -3.57421339e-01 -4.34866160e-01
-2.49028623e-01 3.71264935e-01 -6.55492961e-01 3.62664521e-01
2.67603308e-01 9.26112831e-01 -8.11315715e-01 2.32482404e-01
4.68151212e-01 9.08880949e-01 -7.97513485e-01 6.13152385e-01
-3.11615884e-01 1.07283306e+00 -2.84437150e-01 4.97545540e-01
9.03695583e-01 2.99489703e-02 -4.40208614e-02 -4.12107587e-01
8.92191082e-02 1.65543780e-01 -1.03848934e+00 1.80342174e+00
-1.03640485e+00 6.38884842e-01 1.61978066e-01 -4.71283972e-01
8.42929661e-01 1.82709366e-01 3.84796739e-01 -9.32470858e-01
-2.51106005e-02 -1.48411351e-03 -4.10540640e-01 -1.56943172e-01
5.16806304e-01 -4.50387478e-01 -3.02893855e-02 1.24029271e-01
-2.42053926e-01 -6.85398757e-01 -6.41424537e-01 1.77428663e-01
1.21210861e+00 3.12242746e-01 -1.36509627e-01 2.59544313e-01
9.95519459e-02 -4.25635129e-01 4.49831158e-01 6.28619492e-01
3.92656654e-01 1.28025651e+00 1.43022060e-01 -2.68134683e-01
-1.52974260e+00 -1.63575768e+00 1.07859768e-01 3.55875999e-01
3.37323882e-02 -6.25384375e-02 -6.58068419e-01 -3.74245375e-01
-1.52584389e-01 1.04088664e+00 -4.77138311e-01 -2.55853266e-01
-6.70947254e-01 -3.53109270e-01 5.72821915e-01 4.99071181e-01
7.52437294e-01 -5.80642879e-01 -5.15882671e-01 1.96575075e-01
-3.67805690e-01 -1.24841154e+00 -3.25155973e-01 2.74435561e-02
-6.80649579e-01 -6.77173436e-01 -9.71095204e-01 -1.19068854e-01
6.38021171e-01 3.30498010e-01 1.36551726e+00 -2.02155262e-01
-2.83881336e-01 2.66322494e-01 -1.79362416e-01 -1.88610926e-01
-7.10801184e-01 -7.54560113e-01 -1.98715150e-01 -1.38286063e-02
-8.06160629e-01 -1.04694951e+00 -9.40022707e-01 3.43417376e-01
-1.12935734e+00 1.55133948e-01 3.90414447e-01 7.65387177e-01
5.53362370e-01 3.12794089e-01 2.00026095e-01 -8.52278352e-01
1.67940900e-01 -2.65621960e-01 -7.69852281e-01 -2.62916312e-02
-2.23788366e-01 -2.69186124e-02 1.14284492e+00 -2.76119411e-01
-1.43682873e+00 -1.52693152e-01 -4.47589487e-01 -7.78282344e-01
-2.02848107e-01 -2.01536715e-01 -5.04110456e-01 -3.01671058e-01
8.45426261e-01 2.00314447e-01 -3.81043583e-01 -2.72191674e-01
4.74651635e-01 2.40418911e-01 9.34964359e-01 -5.30132174e-01
1.37155032e+00 6.46076322e-01 2.36024082e-01 -5.53118408e-01
-8.17319930e-01 2.76398331e-01 -3.36889215e-02 -3.59892070e-01
7.77832091e-01 -1.37503171e+00 -5.43467522e-01 5.56642652e-01
-1.09341860e+00 -6.60516858e-01 -4.97769088e-01 2.67985344e-01
-5.87218881e-01 1.11959308e-01 -5.13042867e-01 -6.89367831e-01
-1.01951867e-01 -1.06391537e+00 1.13912237e+00 2.17526987e-01
2.38149747e-01 -8.33249569e-01 -1.09493732e-01 5.47856688e-01
5.04558504e-01 1.00164604e+00 7.41560638e-01 5.26799858e-01
-1.04397762e+00 2.37187278e-02 -1.17382802e-01 6.64872885e-01
3.16496015e-01 -1.84232593e-01 -1.34561193e+00 -2.70678461e-01
2.25821689e-01 -2.07303479e-01 6.29875839e-01 3.21343958e-01
1.09408796e+00 -4.33744133e-01 2.22135887e-01 1.13720608e+00
1.66165578e+00 2.77578443e-01 1.06859422e+00 -1.74525697e-02
9.11982059e-01 4.27273184e-01 1.78988755e-01 5.50801516e-01
1.38784096e-01 6.25460505e-01 8.23438883e-01 -4.11355704e-01
-5.80752850e-01 -6.79626584e-01 1.05484672e-01 2.13050902e-01
-1.53385207e-01 -7.18186677e-01 -5.31144321e-01 2.33269647e-01
-1.13995481e+00 -8.11385632e-01 3.08120623e-02 2.13699746e+00
5.53418100e-01 -5.25677912e-02 -4.12300140e-01 -1.46549894e-02
3.75138402e-01 4.41473544e-01 -7.39921451e-01 -1.37909502e-01
-3.61906290e-01 3.79171312e-01 7.18590140e-01 5.80165923e-01
-5.69422841e-01 5.58295727e-01 5.77115393e+00 3.72965157e-01
-1.14668870e+00 -8.81835818e-03 8.13751757e-01 -1.77252918e-01
-9.50942636e-01 -1.37351081e-01 -8.09599832e-02 2.31918231e-01
6.29024506e-01 1.71171501e-01 9.75962937e-01 5.33733308e-01
3.58550638e-01 7.02841729e-02 -7.27437675e-01 9.52424467e-01
1.99478820e-01 -1.27662742e+00 2.38251701e-01 1.85646027e-01
1.14647686e+00 -2.81400383e-01 4.73840743e-01 -1.53812720e-02
6.43874645e-01 -1.23636460e+00 9.03906822e-01 6.75109088e-01
1.15021670e+00 -9.34214711e-01 3.91193509e-01 2.20391348e-01
-8.22756529e-01 1.74667031e-01 -3.28229159e-01 1.24382026e-01
5.44407666e-01 7.09773421e-01 -6.14470124e-01 8.84428442e-01
5.26838124e-01 2.52687067e-01 -1.70461759e-01 7.10428774e-01
-6.17048264e-01 4.15222377e-01 -2.01937616e-01 6.76393211e-01
2.26067573e-01 -1.86116740e-01 6.89420223e-01 5.87075233e-01
5.18110514e-01 4.49896812e-01 -1.19596340e-01 1.15308928e+00
-4.20319676e-01 -3.88287812e-01 -8.94347191e-01 3.04456890e-01
1.98197037e-01 1.08935654e+00 -1.61673635e-01 5.06686419e-02
-1.48746669e-01 1.42808700e+00 1.77492350e-02 6.33936465e-01
-1.03438294e+00 -2.39880070e-01 9.71564531e-01 2.81153768e-01
1.94532469e-01 -1.78340852e-01 -4.17393595e-01 -1.20405734e+00
1.20419666e-01 -1.01649976e+00 -2.82998323e-01 -1.34770453e+00
-1.18628252e+00 7.76948333e-01 -4.13523436e-01 -1.22440672e+00
-2.32003704e-01 -1.70212492e-01 -6.19380236e-01 1.22291422e+00
-1.60774255e+00 -1.18618870e+00 -7.11621523e-01 5.58654547e-01
4.45618272e-01 9.26212147e-02 8.45957100e-01 3.94660920e-01
-1.27474397e-01 4.98613507e-01 8.89056697e-02 -2.16812398e-02
6.75774634e-01 -1.25884235e+00 7.79378593e-01 1.13546717e+00
8.72766748e-02 3.12680662e-01 8.94400835e-01 -4.62964684e-01
-1.50188529e+00 -1.35429621e+00 1.88236777e-02 -4.38694239e-01
1.39146894e-01 -3.82274568e-01 -8.69765162e-01 5.64382613e-01
2.29427949e-01 1.97378665e-01 3.63398790e-01 -5.39899051e-01
-6.94032490e-01 -1.26185253e-01 -1.52611637e+00 8.06934714e-01
1.10731518e+00 -5.70414186e-01 -7.99865946e-02 2.70047754e-01
9.52929139e-01 -8.21105540e-01 -8.68376136e-01 4.65626538e-01
4.04923826e-01 -1.31982827e+00 1.31961095e+00 -2.70384133e-01
1.00646949e+00 -5.39126456e-01 -7.02633381e-01 -1.70521307e+00
-1.82150155e-01 -8.16369951e-01 9.79765207e-02 1.24819767e+00
1.63868293e-01 -6.05935395e-01 5.79539716e-01 5.21198452e-01
-2.72423446e-01 -3.34435731e-01 -6.05270445e-01 -6.15779161e-01
7.28129894e-02 -3.57037455e-01 7.71461189e-01 7.44382024e-01
-9.85624611e-01 2.82375246e-01 -7.47713327e-01 7.54142344e-01
9.00021613e-01 6.10501170e-02 1.05059063e+00 -5.01649618e-01
-7.60367692e-01 8.38384852e-02 -8.05915743e-02 -1.38326907e+00
-5.77161275e-02 -4.39975977e-01 2.50364870e-01 -1.43290102e+00
-2.71824121e-01 -6.16383910e-01 2.08476022e-01 5.74092269e-02
-1.85373604e-01 6.60038948e-01 3.08720291e-01 -4.95020486e-03
9.33225453e-02 7.22935438e-01 1.73010111e+00 -8.77635647e-03
2.28030626e-02 -7.51367956e-03 -8.54981124e-01 5.39631546e-01
5.80703676e-01 -1.98058262e-01 -5.63392341e-01 -9.09026980e-01
2.02225313e-01 5.19442201e-01 8.46137583e-01 -1.24172258e+00
-3.03160757e-01 4.98239836e-03 9.26072240e-01 -2.71625042e-01
7.21711755e-01 -8.59719872e-01 6.62436247e-01 1.02769151e-01
-1.63621426e-01 -1.57180250e-01 2.40882203e-01 6.16064191e-01
-9.66219828e-02 2.60347843e-01 1.05145383e+00 -2.88369328e-01
-2.37576917e-01 2.55764157e-01 1.59181952e-02 5.89063801e-02
6.30086303e-01 9.45441574e-02 -2.89013565e-01 -9.71956074e-01
-2.82601058e-01 -2.32011825e-01 9.30458426e-01 8.51029307e-02
6.27806842e-01 -1.40878487e+00 -6.96337640e-01 4.88318861e-01
-1.03325285e-01 3.18710327e-01 6.08731091e-01 1.16656706e-01
-7.53818810e-01 -1.48940489e-01 -2.59020180e-01 -4.75259930e-01
-8.67340207e-01 3.95085692e-01 4.15961444e-01 -8.93459544e-02
-8.63902271e-01 7.76394784e-01 6.26086950e-01 -4.96097445e-01
-5.07232770e-02 -6.04390763e-02 2.81772107e-01 -6.11414909e-01
5.42915702e-01 7.07534924e-02 4.49182354e-02 -3.54874194e-01
8.29126090e-02 5.61446965e-01 4.70534682e-01 -2.69479930e-01
1.45985031e+00 -1.94660977e-01 4.05289441e-01 -4.86471951e-02
1.27147949e+00 5.97965598e-01 -1.93851149e+00 -1.41835669e-02
-1.14619124e+00 -9.72479463e-01 2.38659665e-01 -1.11072910e+00
-1.31032109e+00 6.59040213e-01 5.47087967e-01 6.89893886e-02
1.46232104e+00 -3.14201713e-01 1.00617695e+00 -1.42192221e-06
4.28586960e-01 -2.43011177e-01 2.59898841e-01 1.41801104e-01
1.05684960e+00 -9.40252841e-01 -1.89801872e-01 -4.53089148e-01
-6.15360200e-01 9.48569834e-01 2.49021232e-01 -3.13769490e-01
5.23764081e-02 6.50481045e-01 2.88967311e-01 7.43121281e-02
-4.07349437e-01 2.02762693e-01 1.32609427e-01 7.13320673e-01
8.15538913e-02 5.16758636e-02 6.50309622e-01 -2.82030948e-03
-3.84127766e-01 -4.03796613e-01 6.37876391e-01 5.06416261e-01
-4.11331728e-02 -7.59899437e-01 -6.88103437e-01 5.51391579e-02
-4.38872963e-01 -1.90509275e-01 3.67994457e-02 5.11339366e-01
2.91408841e-02 9.77366924e-01 -3.45162232e-03 -4.63362902e-01
6.51420891e-01 -5.25664032e-01 5.83071530e-01 -1.98996127e-01
-4.65590417e-01 2.01848492e-01 1.16111144e-01 -8.92995536e-01
-5.08132167e-02 -4.18582886e-01 -8.01124275e-01 -6.35410607e-01
1.33496365e-02 -2.20550507e-01 6.91769540e-01 4.34004068e-01
1.99975818e-01 1.11276710e+00 1.17993176e+00 -9.21668231e-01
-4.88361478e-01 -5.57793438e-01 -4.20845866e-01 4.01668221e-01
5.43333173e-01 -3.77072126e-01 -5.79709113e-01 9.06588882e-02] | [9.37026309967041, -3.1976981163024902] |
5444960a-a7f5-463d-996e-21d60da2c00f | localized-semantic-feature-mixers-for | null | null | http://openaccess.thecvf.com//content/CVPR2023/html/Khan_Localized_Semantic_Feature_Mixers_for_Efficient_Pedestrian_Detection_in_Autonomous_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Khan_Localized_Semantic_Feature_Mixers_for_Efficient_Pedestrian_Detection_in_Autonomous_CVPR_2023_paper.pdf | Localized Semantic Feature Mixers for Efficient Pedestrian Detection in Autonomous Driving | Autonomous driving systems rely heavily on the underlying perception module which needs to be both performant and efficient to allow precise decisions in real-time. Avoiding collisions with pedestrians is of topmost priority in any autonomous driving system. Therefore, pedestrian detection is one of the core parts of such systems' perception modules. Current state-of-the-art pedestrian detectors have two major issues. Firstly, they have long inference times which affect the efficiency of the whole perception module, and secondly, their performance in the case of small and heavily occluded pedestrians is poor. We propose Localized Semantic Feature Mixers (LSFM), a novel, anchor-free pedestrian detection architecture. It uses our novel Super Pixel Pyramid Pooling module instead of the, computationally costly, Feature Pyramid Networks for feature encoding. Moreover, our MLPMixer-based Dense Focal Detection Network is used as a light detection head, reducing computational effort and inference time compared to existing approaches. To boost the performance of the proposed architecture, we adapt and use mixup augmentation which improves the performance, especially in small and heavily occluded cases. We benchmark LSFM against the state-of-the-art on well-established traffic scene pedestrian datasets. The proposed LSFM achieves state-of-the-art performance in Caltech, City Persons, Euro City Persons, and TJU-Traffic-Pedestrian datasets while reducing the inference time on average by 55%. Further, LSFM beats the human baseline for the first time in the history of pedestrian detection. Finally, we conducted a cross-dataset evaluation which proved that our proposed LSFM generalizes well to unseen data. | ['Andreas Dengel', 'Mohammed Shariq Nawaz', 'Abdul Hannan Khan'] | 2023-01-01 | null | null | null | cvpr-2023-1 | ['pedestrian-detection'] | ['computer-vision'] | [-1.48938209e-01 -3.25744778e-01 1.69843867e-01 -3.01607281e-01
-4.63657647e-01 -1.32805407e-01 6.39969945e-01 9.01882648e-02
-9.38567340e-01 5.60901344e-01 -2.49578372e-01 -2.28116035e-01
6.30591631e-01 -1.06673038e+00 -8.18175495e-01 -7.31652796e-01
2.80636877e-01 3.27365845e-02 1.16597426e+00 -3.39842826e-01
-8.74744579e-02 4.09641325e-01 -1.99550176e+00 3.34920406e-01
8.39153826e-01 1.09468842e+00 5.04440010e-01 6.91096008e-01
1.03613742e-01 5.87432027e-01 -4.13986921e-01 -7.19661474e-01
3.57520729e-01 1.92793697e-01 -1.43794939e-01 3.20861824e-02
6.02013052e-01 -2.82046407e-01 -6.04562700e-01 1.05493939e+00
7.32983708e-01 1.19958572e-01 5.20367503e-01 -1.40668440e+00
-1.23296119e-01 1.73850983e-01 -8.18036616e-01 3.88783693e-01
1.93066318e-02 7.18554795e-01 5.31549811e-01 -1.04396117e+00
8.63596648e-02 1.44366300e+00 7.61350751e-01 2.54429668e-01
-9.47748125e-01 -7.52058089e-01 3.43978882e-01 6.19385839e-01
-1.36386108e+00 -5.30048192e-01 5.21036088e-01 -4.85286564e-01
9.35656428e-01 -3.29710660e-03 5.73395550e-01 1.10531545e+00
2.06115305e-01 8.47221911e-01 9.24569249e-01 -1.97485194e-01
2.50739664e-01 2.96314299e-01 2.94864267e-01 8.90962541e-01
6.46967053e-01 3.03993493e-01 -2.87602186e-01 3.27681065e-01
4.00415540e-01 1.18827214e-02 2.21779659e-01 -1.77974015e-01
-8.34648728e-01 6.86557531e-01 7.19240010e-01 4.01629979e-04
-2.85965860e-01 2.02687502e-01 4.47794199e-01 -2.53233820e-01
5.44208698e-02 -3.54281843e-01 -2.02930912e-01 1.12708718e-01
-8.58165145e-01 2.60426611e-01 3.49468410e-01 8.37563932e-01
7.48828173e-01 -9.89500992e-03 -4.00679797e-01 6.68707252e-01
4.35278207e-01 7.51510501e-01 1.14951782e-01 -6.04623556e-01
6.08970881e-01 7.46226907e-01 2.15642631e-01 -9.21424568e-01
-5.06452858e-01 -7.35097706e-01 -8.88517439e-01 4.68916923e-01
6.31509364e-01 -6.38738461e-03 -9.83550489e-01 1.45021009e+00
3.85380119e-01 9.56440419e-02 5.39024686e-03 1.00533855e+00
9.50511515e-01 6.22725487e-01 6.24666572e-01 3.16333115e-01
1.84621239e+00 -1.09942400e+00 -1.94777548e-01 -6.86824024e-01
1.67309701e-01 -8.20197940e-01 8.80649686e-01 1.89704046e-01
-7.58871734e-01 -1.20989132e+00 -1.15892208e+00 -2.88910627e-01
-6.14069343e-01 5.64715326e-01 6.22277439e-01 9.11881149e-01
-6.39776886e-01 1.50540426e-01 -7.77258635e-01 -5.86609840e-01
6.73596799e-01 3.25636446e-01 -2.57814825e-01 -2.22748876e-01
-1.03812885e+00 9.36306834e-01 4.96385872e-01 2.17409983e-01
-8.93509567e-01 -3.75200063e-01 -8.00023198e-01 3.28143924e-01
3.45747530e-01 -8.27332675e-01 9.73741293e-01 -3.48918945e-01
-1.18016684e+00 7.04727530e-01 -3.85054559e-01 -8.14586222e-01
7.93826818e-01 -2.76198626e-01 -4.74580765e-01 -7.03961402e-02
3.61542672e-01 1.02732384e+00 8.26072454e-01 -1.09979653e+00
-1.39286506e+00 -2.07286626e-01 7.07092956e-02 -7.13497922e-02
-1.10994242e-01 -1.25140026e-01 -6.44105852e-01 -1.87403902e-01
-2.90228993e-01 -8.09018016e-01 -4.21963930e-01 8.31844732e-02
-4.21681315e-01 -4.38612729e-01 9.77585673e-01 -3.80897045e-01
8.45905304e-01 -2.27720451e+00 -5.02153337e-01 -4.02515195e-02
-1.67783573e-02 6.89951420e-01 1.76278148e-02 -1.19726576e-01
3.34243566e-01 -4.45151806e-01 -1.63817480e-01 -5.88082552e-01
1.66822821e-01 7.69154131e-02 -4.18670513e-02 4.99058485e-01
3.33836794e-01 7.95362055e-01 -7.11082578e-01 -7.37720132e-01
8.38125527e-01 6.20810747e-01 -4.99554604e-01 6.56469613e-02
1.11264706e-01 2.83915222e-01 -3.10310692e-01 6.48055911e-01
9.53560650e-01 1.43215418e-01 -4.61653113e-01 -3.39110345e-01
-6.34339392e-01 -1.79406151e-01 -1.39322054e+00 1.39215970e+00
-3.28337342e-01 6.39920592e-01 -1.24217942e-01 -9.22486842e-01
8.66014719e-01 6.03347011e-02 -1.10545186e-02 -8.94613266e-01
3.45199704e-01 2.67126739e-01 3.69781666e-02 -4.20456201e-01
5.24102151e-01 3.05897892e-01 -6.22790977e-02 -4.90708709e-01
-4.81244735e-02 3.19785684e-01 6.42612398e-01 5.50916195e-02
9.55622673e-01 -1.02323142e-03 2.20836014e-01 -2.75698453e-01
1.02451670e+00 -6.50395900e-02 6.67218983e-01 8.38873208e-01
-6.17299795e-01 3.89887065e-01 2.22495735e-01 -5.05480349e-01
-9.25090849e-01 -1.28296459e+00 -1.99958548e-01 1.03270006e+00
3.72422189e-01 -1.92798704e-01 -8.23893368e-01 -5.44156492e-01
1.04158744e-01 7.06661880e-01 -3.40507567e-01 -1.98816545e-02
-6.68807089e-01 -9.95441556e-01 5.81803381e-01 8.11421990e-01
1.17179394e+00 -8.15036714e-01 -1.14413297e+00 3.62456232e-01
-2.20291436e-01 -1.76359284e+00 -1.72100395e-01 1.49869639e-02
-1.63489223e-01 -1.20006800e+00 -6.25859201e-01 -6.80595458e-01
3.48088264e-01 5.93517780e-01 8.32410395e-01 -2.19728291e-01
-5.71543038e-01 -1.20645747e-01 4.18159701e-02 -5.43231606e-01
-2.33690860e-03 8.95581543e-02 3.51590626e-02 2.42615432e-01
5.95772326e-01 -1.98785961e-01 -1.03087378e+00 5.14421403e-01
-3.00131202e-01 8.96053463e-02 8.88160288e-01 6.70890331e-01
3.69005442e-01 2.54874259e-01 4.25153941e-01 -3.47151995e-01
-8.24777707e-02 -1.84120417e-01 -9.31985795e-01 -7.74043053e-02
-1.60695016e-01 -1.81882530e-01 7.80519605e-01 -1.24604262e-01
-1.30850780e+00 4.34477270e-01 -4.16462541e-01 -3.69923674e-02
-5.12774765e-01 -4.71042365e-01 -5.09612620e-01 -5.92503697e-02
7.51623333e-01 1.11349396e-01 -3.21724534e-01 -3.77892464e-01
4.37494993e-01 6.48889720e-01 9.09708858e-01 -3.65361392e-01
7.78871059e-01 8.05660307e-01 3.52509230e-01 -1.09785652e+00
-6.05992496e-01 -7.28967607e-01 -7.08172560e-01 -2.54639059e-01
1.05632126e+00 -1.29951656e+00 -1.03138983e+00 6.41774774e-01
-1.27620447e+00 9.58567858e-02 -1.32943302e-01 4.00012821e-01
-1.57640696e-01 4.25588220e-01 -3.91613305e-01 -1.08572423e+00
-2.70158887e-01 -1.43323874e+00 1.08073747e+00 5.08834124e-01
3.15134645e-01 -3.20272774e-01 -5.35138488e-01 4.83375221e-01
5.06635487e-01 2.05960467e-01 3.33849430e-01 -1.16692431e-01
-8.47953379e-01 -3.27358007e-01 -8.08849692e-01 3.68673414e-01
-2.97816962e-01 -8.56161714e-02 -1.36529171e+00 -1.05977401e-01
-5.90913951e-01 8.39433521e-02 1.42720222e+00 5.36796451e-01
7.91697323e-01 1.19864859e-01 -6.06550872e-01 3.41478407e-01
1.25577807e+00 1.97427757e-02 6.45798087e-01 4.82242227e-01
6.66470528e-01 4.94573683e-01 5.30934453e-01 3.10829014e-01
7.08745122e-01 7.90508449e-01 4.12077725e-01 -4.15140450e-01
-5.85858822e-01 -1.64457947e-01 5.48809350e-01 1.47742435e-01
-9.98246148e-02 -2.15834558e-01 -6.47601187e-01 5.56980610e-01
-2.13218594e+00 -1.07342923e+00 -4.91394311e-01 2.12856817e+00
2.45260790e-01 6.77937627e-01 4.47161913e-01 4.51457113e-01
8.02951455e-01 -1.21160164e-01 -3.62178802e-01 -1.35979384e-01
-1.62826106e-01 1.77264154e-01 8.90842199e-01 2.90972829e-01
-1.65591538e+00 1.06914032e+00 4.68738747e+00 9.09857452e-01
-9.35899973e-01 4.41903234e-01 5.20584643e-01 -4.51577036e-03
7.60451734e-01 -3.34224969e-01 -1.31284392e+00 7.01908767e-01
7.57054985e-01 4.27414924e-01 -8.87185112e-02 1.12238479e+00
3.73265594e-01 -4.86086369e-01 -8.27719450e-01 1.08634114e+00
-2.47654215e-01 -9.48591948e-01 -3.66107255e-01 -1.69690445e-01
3.47551614e-01 1.97497293e-01 9.88569693e-04 5.47875762e-01
2.04506949e-01 -7.04626441e-01 8.72750521e-01 2.73107827e-01
4.47110981e-01 -9.80113983e-01 1.04607463e+00 4.06309068e-01
-1.74771190e+00 -3.35295528e-01 -7.00210035e-01 -1.05652183e-01
3.45238447e-01 8.10757399e-01 -5.33232689e-01 3.93818676e-01
9.32516932e-01 3.72914493e-01 -9.99783218e-01 1.35255277e+00
-4.25962776e-01 3.87136877e-01 -6.06564462e-01 9.83610563e-03
3.34697992e-01 2.22534791e-01 4.53681469e-01 1.74912679e+00
1.06237821e-01 -1.38224036e-01 5.22215605e-01 7.13971913e-01
2.01521873e-01 -1.40117317e-01 -3.81525576e-01 6.89638197e-01
2.35271424e-01 1.43355775e+00 -7.82886028e-01 -4.81064588e-01
-6.25129580e-01 7.78128624e-01 1.11826189e-01 2.89499044e-01
-1.09484935e+00 -4.31252420e-01 8.10290635e-01 3.72621536e-01
5.62849820e-01 -2.08622590e-01 -4.04227942e-01 -9.56377745e-01
2.01156378e-01 -3.26552480e-01 1.77205086e-01 -2.43272796e-01
-1.05117142e+00 6.03796482e-01 -1.87431425e-01 -1.01866877e+00
1.58107117e-01 -8.52034092e-01 -5.66378474e-01 6.97539210e-01
-1.96560431e+00 -1.43330491e+00 -6.34805322e-01 6.57513320e-01
7.31405735e-01 -7.82917440e-02 2.94476092e-01 8.86155486e-01
-9.60937858e-01 6.81612730e-01 -4.84316319e-01 2.88073868e-01
5.46583891e-01 -9.37034011e-01 7.63570964e-01 1.41906929e+00
-2.78189361e-01 1.46095470e-01 6.26237154e-01 -3.95729005e-01
-1.07835805e+00 -1.53833997e+00 7.50487804e-01 -2.58969069e-01
1.84789509e-01 -6.18300021e-01 -6.13262594e-01 1.28986433e-01
7.43312165e-02 3.21711779e-01 9.59388614e-02 -2.52532631e-01
-3.16170335e-01 -5.14586747e-01 -1.27114022e+00 5.47369540e-01
1.08915389e+00 -4.30798046e-02 -4.48003471e-01 2.53292114e-01
3.50398719e-01 -1.74897447e-01 -1.43156424e-01 4.61470485e-01
6.44987047e-01 -1.27168000e+00 1.41406357e+00 -8.70135576e-02
3.17929089e-02 -8.92950237e-01 -3.30155551e-01 -8.42961013e-01
-5.31635940e-01 -9.61336493e-02 -9.25523639e-02 1.20472062e+00
1.52303189e-01 -7.39035547e-01 7.84159958e-01 2.99952179e-01
-2.19691575e-01 -4.47697461e-01 -1.11589170e+00 -8.98509562e-01
-3.64089340e-01 -5.56601226e-01 3.84163111e-01 9.62790996e-02
-5.28834403e-01 4.54807431e-01 -1.37568906e-01 5.40640295e-01
1.19305611e+00 -2.48327687e-01 1.08118653e+00 -1.22877181e+00
-1.87691972e-01 -4.73586649e-01 -8.97839129e-01 -1.06510258e+00
-1.52884200e-01 -5.58982790e-01 2.04540372e-01 -1.65755343e+00
8.63991231e-02 -3.93337369e-01 -2.26662338e-01 3.00119698e-01
-4.36402202e-01 4.45473045e-01 4.14276630e-01 -1.37253821e-01
-6.70017898e-01 4.90603715e-01 8.10331762e-01 -2.48193145e-01
9.59051251e-02 1.06537648e-01 -4.01997387e-01 8.41802239e-01
6.18290961e-01 -2.37227410e-01 -7.76950791e-02 -2.44116828e-01
-3.94072115e-01 -5.83688974e-01 9.06598806e-01 -1.83959639e+00
5.86137891e-01 1.83888271e-01 8.11296165e-01 -9.95492578e-01
6.51582301e-01 -7.40431547e-01 -1.96980208e-01 8.24291229e-01
3.84393275e-01 -2.84211412e-02 3.99965733e-01 5.05136013e-01
-1.14353132e-02 1.64006680e-01 1.33203673e+00 -8.53879601e-02
-1.27205598e+00 2.64209151e-01 -3.53460848e-01 -1.17344245e-01
1.27787578e+00 -3.50312322e-01 -5.19841790e-01 2.81930834e-01
-2.60646939e-01 3.54374200e-01 1.31718829e-01 5.06900311e-01
5.09865880e-01 -1.03437102e+00 -7.83502281e-01 4.33751255e-01
1.37862056e-01 9.72298905e-02 3.23231190e-01 8.05333436e-01
-3.51310909e-01 6.76589608e-01 -2.38551840e-01 -9.14999008e-01
-1.09239507e+00 6.78384066e-01 1.85696274e-01 -3.47393788e-02
-7.16084182e-01 7.63239503e-01 4.47901756e-01 -8.98034032e-03
2.77929842e-01 -3.98022413e-01 -2.23802045e-01 -7.62160420e-02
6.37266397e-01 7.55282938e-01 1.21149369e-01 -8.68439257e-01
-5.93850374e-01 6.74850821e-01 2.72802282e-02 2.12531388e-01
8.11640918e-01 -8.53603333e-02 4.34339195e-01 -1.43785581e-01
8.88446391e-01 -2.42645234e-01 -1.45546067e+00 -1.60957545e-01
-2.57520884e-01 -4.66941744e-01 1.24384321e-01 -5.45164526e-01
-9.86307323e-01 1.11231971e+00 1.04851758e+00 -2.64309317e-01
9.13252592e-01 -2.27813929e-01 1.03089261e+00 3.09187382e-01
4.00417715e-01 -1.14711750e+00 -2.50079423e-01 3.94305706e-01
3.65474731e-01 -1.57496464e+00 -7.16326982e-02 -7.01169908e-01
-4.84757841e-01 8.86440992e-01 8.48074019e-01 -3.41969132e-02
5.17150700e-01 3.02480906e-01 -1.64183781e-01 1.30443335e-01
-4.03898895e-01 -8.71461272e-01 2.53003091e-01 6.54217005e-01
2.60967482e-02 2.79154003e-01 7.87388161e-03 5.10833263e-01
-9.94472131e-02 -1.13789262e-02 2.89441701e-02 6.66384459e-01
-8.70817423e-01 -8.26469719e-01 -5.04666567e-01 1.91426396e-01
-2.23946363e-01 2.92117018e-02 2.28003711e-01 7.13702619e-01
8.39043617e-01 1.33001769e+00 1.93892866e-02 -2.48061672e-01
7.66076565e-01 -5.80478497e-02 3.58054072e-01 -2.32238725e-01
-4.31011975e-01 -1.63127527e-01 3.37944329e-02 -6.36550128e-01
-2.85997033e-01 -6.50849819e-01 -1.16238225e+00 -3.95021021e-01
-2.93267131e-01 -3.38930786e-01 7.68085897e-01 9.68144894e-01
3.93340588e-01 6.48183465e-01 4.07105327e-01 -1.00038540e+00
-2.59890139e-01 -8.58386636e-01 -1.33104131e-01 2.97697783e-01
2.38725856e-01 -1.09426212e+00 1.40243247e-01 -4.67813350e-02] | [8.054718017578125, -0.7494316697120667] |
7cab6894-1e54-43df-a685-700c5b5c2f04 | point-cloud-transformers-applied-to-collider | 2102.05073 | null | https://arxiv.org/abs/2102.05073v2 | https://arxiv.org/pdf/2102.05073v2.pdf | Point Cloud Transformers applied to Collider Physics | Methods for processing point cloud information have seen a great success in collider physics applications. One recent breakthrough in machine learning is the usage of Transformer networks to learn semantic relationships between sequences in language processing. In this work, we apply a modified Transformer network called Point Cloud Transformer as a method to incorporate the advantages of the Transformer architecture to an unordered set of particles resulting from collision events. To compare the performance with other strategies, we study jet-tagging applications for highly-boosted particles. | ['Florencia Canelli', 'Vinicius Mikuni'] | 2021-02-09 | null | null | null | null | ['jet-tagging'] | ['graphs'] | [-1.07241794e-01 -4.13705230e-01 -9.05120149e-02 -7.08103776e-01
-3.45408440e-01 -6.01513028e-01 8.70035887e-01 4.95176792e-01
-7.84032345e-01 7.11780429e-01 1.27627388e-01 -5.95307708e-01
-1.68269277e-01 -1.02103043e+00 -3.79398733e-01 -4.31762606e-01
-1.39429510e-01 1.22158062e+00 5.74447155e-01 -7.58744001e-01
2.85453320e-01 1.03597999e+00 -1.02984297e+00 7.62728631e-01
1.71891913e-01 5.89260161e-01 1.85028926e-01 6.20701671e-01
-8.11061502e-01 6.52696550e-01 -5.99765182e-01 -6.49314940e-01
2.08019435e-01 -2.74421573e-01 -7.97165096e-01 -8.20256293e-01
-3.15902591e-01 5.65752268e-01 -3.36266518e-01 1.10517156e+00
3.99954915e-01 4.41409260e-01 3.82128268e-01 -1.27990341e+00
-1.49959981e-01 9.45373356e-01 -2.93295495e-02 9.36360836e-01
1.63293302e-01 -3.27280611e-01 9.06089306e-01 -7.22997189e-01
8.57465982e-01 1.39776671e+00 7.71064043e-01 2.40560055e-01
-9.83254492e-01 -7.55259812e-01 -3.16452652e-01 4.53190655e-01
-1.02443087e+00 -1.02849342e-01 7.16128588e-01 -2.71357983e-01
1.62724948e+00 1.85081214e-01 7.22362816e-01 7.78260350e-01
1.06443751e+00 4.97367829e-01 9.74330783e-01 -7.08187044e-01
1.47477359e-01 8.60664323e-02 4.58945870e-01 5.26854217e-01
-1.54278710e-01 6.83226287e-01 -5.79475164e-01 -2.37614006e-01
5.27229249e-01 -1.67533264e-01 6.43664956e-01 -4.87460673e-01
-1.14086747e+00 1.24500048e+00 4.38896954e-01 7.87116468e-01
1.90652832e-02 4.37904000e-02 9.50172663e-01 4.06300098e-01
9.24817145e-01 7.59887934e-01 -6.30826592e-01 -2.25339338e-01
-7.61436701e-01 7.36117840e-01 7.78784215e-01 7.77033627e-01
2.43107885e-01 -4.19456027e-02 -3.75114352e-01 5.36610901e-01
1.75685927e-01 2.81841367e-01 4.01534736e-01 -3.48911554e-01
2.65366495e-01 7.73318112e-02 -2.24626541e-01 -5.51594198e-01
-7.44947016e-01 -6.33412361e-01 -5.05713046e-01 2.83504874e-01
8.78730044e-02 1.25131458e-01 -1.05887198e+00 1.38739657e+00
4.93371069e-01 -1.28629476e-01 2.86615998e-01 7.77815938e-01
7.72106349e-01 7.14401543e-01 3.64528090e-01 -3.98688857e-03
1.29507756e+00 -9.04673934e-01 -5.96604466e-01 2.03152671e-01
6.18517101e-01 -1.09157836e+00 2.45169744e-01 4.06065613e-01
-1.04140353e+00 -7.52204835e-01 -8.27086031e-01 -1.93342581e-01
-6.91673815e-01 -5.16496897e-01 1.25767851e+00 4.75721270e-01
-1.05335748e+00 1.18349099e+00 -8.51167560e-01 -4.66439247e-01
-3.25536251e-01 5.58036745e-01 -9.32439193e-02 2.98956364e-01
-1.49163222e+00 1.04172778e+00 9.65451419e-01 -9.42108780e-02
-2.88149446e-01 -7.44756401e-01 -6.69223070e-01 -2.01022532e-02
8.77847597e-02 -7.64918983e-01 1.90288532e+00 -4.97963428e-01
-1.70414162e+00 7.35967159e-01 -3.00635546e-01 -8.80721092e-01
1.44361004e-01 -4.87229228e-02 -7.43984342e-01 5.88700175e-02
1.13514543e-01 5.33413172e-01 3.40828717e-01 -7.14578807e-01
-8.72916162e-01 -8.49878117e-02 8.34626257e-02 3.98375206e-02
4.19910699e-01 9.98195529e-01 -2.45152041e-01 -6.82919741e-01
6.35763332e-02 -9.23004627e-01 -4.95162606e-01 -7.45751798e-01
-2.07894295e-01 -7.66560256e-01 6.24736607e-01 -3.30733329e-01
4.33402210e-01 -1.96255922e+00 3.90350044e-01 2.60395736e-01
2.42762014e-01 4.39437121e-01 -1.04069509e-01 7.75478005e-01
-6.06618762e-01 1.26828581e-01 1.92493945e-01 -2.37483099e-01
1.92527533e-01 4.43057060e-01 -4.47928607e-01 -1.61599711e-01
1.18639030e-01 1.02071345e+00 -1.01889110e+00 -2.92243838e-01
3.32192212e-01 3.77958596e-01 -5.78996956e-01 -1.17815092e-01
-5.48095107e-01 5.38221359e-01 -4.49105501e-01 1.22940727e-01
5.93277097e-01 2.09924176e-01 6.14238605e-02 -2.10508212e-01
-5.55961132e-01 6.46175981e-01 -3.90254170e-01 1.68267941e+00
-2.58764476e-01 4.02947307e-01 1.54194739e-02 -1.06912899e+00
8.49433064e-01 3.01584691e-01 7.07889140e-01 -9.51609552e-01
6.74573556e-02 -2.36773014e-01 3.82006794e-01 -1.06955945e-01
9.61091638e-01 -9.64083195e-01 -3.49037260e-01 2.35343426e-01
2.73251802e-01 -3.82975072e-01 4.67980862e-01 3.28018069e-01
1.08900702e+00 1.72353595e-01 2.87601441e-01 -2.79518336e-01
4.37878609e-01 4.45571780e-01 6.55984879e-01 9.33580339e-01
-1.06009524e-02 -3.74183133e-02 2.63750702e-01 -8.84237289e-01
-1.21244431e+00 -1.17115891e+00 -1.76612124e-01 1.21218562e+00
-2.12966457e-01 -9.67010140e-01 -2.19221860e-01 -6.23116374e-01
-5.86100444e-02 1.02302122e+00 3.72325853e-02 -1.13898955e-01
-8.50024581e-01 -9.79759276e-01 4.39235121e-01 6.35161042e-01
2.15476274e-01 -1.31530511e+00 1.19833991e-01 3.14881444e-01
-1.08241718e-02 -1.13427770e+00 3.05208862e-02 6.77366853e-01
-7.17373013e-01 -5.14596879e-01 1.18070871e-01 -6.46069407e-01
7.61140836e-03 1.33485973e-01 1.55214572e+00 -2.59036779e-01
-2.48905838e-01 -6.25389665e-02 -6.00146115e-01 -7.70288885e-01
-8.84111643e-01 2.29450334e-02 3.68459910e-01 -6.45350754e-01
7.33628869e-01 -3.38023335e-01 3.43611449e-01 -1.30325392e-01
-8.09556067e-01 1.16132803e-01 3.97247463e-01 7.73175001e-01
5.58625877e-01 6.15309715e-01 9.56619158e-02 -1.03751695e+00
4.67543334e-01 -4.56270128e-01 -9.57224548e-01 -2.03675643e-01
-2.72609740e-01 3.62020791e-01 5.33392668e-01 -5.03007136e-02
-9.91143823e-01 -5.78925647e-02 -7.81503439e-01 -3.50709707e-01
-1.46257684e-01 7.39721954e-01 1.89765155e-01 -3.55116189e-01
1.25190765e-01 -8.72819498e-02 -2.84294099e-01 -6.34358346e-01
4.47938412e-01 5.59431240e-02 4.08849359e-01 -7.08087206e-01
8.79334331e-01 6.43577427e-02 2.60029256e-01 -6.55991137e-01
-9.84899223e-01 -6.13866508e-01 -8.26178789e-01 1.93498239e-01
1.16716433e+00 -4.49980617e-01 -6.99051261e-01 8.73141661e-02
-1.48351276e+00 1.06047869e-01 -4.96414065e-01 7.78931141e-01
-4.96870846e-01 3.60340595e-01 -9.42970991e-01 -4.21125621e-01
-2.95089155e-01 -9.64860857e-01 9.91741121e-01 1.73523147e-02
6.81296512e-02 -1.07577240e+00 3.41513127e-01 4.40658182e-02
2.61959612e-01 -1.77257955e-01 1.32734585e+00 -1.15671694e+00
-4.59279835e-01 -1.06184043e-01 -1.26009420e-01 -5.65783540e-03
-2.22453684e-01 -4.59261090e-01 -2.88205594e-01 -2.97774911e-01
2.66519904e-01 1.21821955e-01 7.83284903e-01 1.61107242e-01
1.16018486e+00 4.58777547e-01 -6.61105633e-01 5.70192873e-01
1.26219881e+00 5.48088133e-01 3.06476831e-01 2.52213776e-01
4.89760280e-01 3.96832526e-01 6.74965858e-01 2.25224346e-02
-9.99006405e-02 1.22947001e+00 -3.77536938e-02 -9.72293019e-02
-1.85738385e-01 -3.00993830e-01 3.06547135e-01 1.37989581e+00
6.40066266e-02 -5.03734052e-01 -1.00318003e+00 -1.79046281e-02
-1.57484138e+00 -1.17990685e+00 -1.83839545e-01 1.68088138e+00
3.85394663e-01 6.59288526e-01 -5.16270325e-02 -2.76997149e-01
6.17715597e-01 -2.59379819e-02 -4.00849581e-02 -9.26837564e-01
2.65285909e-01 1.02117455e+00 5.31319380e-01 5.66938996e-01
-1.11895728e+00 1.17349958e+00 7.57009411e+00 1.02151871e+00
-8.78942311e-01 6.84381604e-01 -2.46952474e-02 -8.57372656e-02
-2.86679089e-01 8.22756216e-02 -1.13781524e+00 3.97003949e-01
1.34711790e+00 -2.38863900e-01 2.56671906e-01 5.49145639e-01
1.09840415e-01 2.18221217e-01 -1.14703882e+00 7.94433415e-01
-4.65201549e-02 -1.69796634e+00 -4.29881141e-02 -2.89553672e-01
8.67966041e-02 6.22118413e-01 -4.36497390e-01 9.14351344e-01
7.72751391e-01 -7.66002536e-01 7.12527215e-01 4.80685294e-01
3.53943795e-01 -9.92695510e-01 1.07015955e+00 3.23958665e-01
-1.30937159e+00 2.50354767e-01 -7.07570851e-01 -2.54375905e-01
7.28378773e-01 8.32789540e-01 -1.24032867e+00 1.08062375e+00
7.65886247e-01 6.26628578e-01 -2.63640970e-01 1.05180657e+00
-7.94356316e-03 5.39863467e-01 -3.43218058e-01 -1.22407809e-01
3.34645778e-01 -3.90751988e-01 8.78724992e-01 1.42602074e+00
2.29066595e-01 1.18853174e-01 5.96556127e-01 7.40081787e-01
1.69977099e-01 -9.88014489e-02 -1.06762612e+00 -3.49700242e-01
2.92044990e-02 1.27342284e+00 -1.16609526e+00 -2.96706975e-01
-3.93220484e-01 6.06374621e-01 3.55463624e-01 -8.83219242e-02
-6.78516448e-01 -3.29226941e-01 7.21232533e-01 -3.30702722e-01
4.16353196e-01 -8.29900861e-01 2.07868174e-01 -8.53920043e-01
-3.67544919e-01 -5.79298258e-01 4.08267170e-01 -8.29281509e-01
-1.51844251e+00 8.79810214e-01 4.16871399e-01 -9.11939502e-01
-4.21851993e-01 -1.05052865e+00 -5.66836596e-01 1.00266123e+00
-1.29028153e+00 -9.91252303e-01 4.61648375e-01 2.40011543e-01
7.16242790e-01 -6.13769829e-01 7.53171027e-01 3.16483021e-01
-6.36672899e-02 8.81425515e-02 2.01225668e-01 -2.65261307e-02
5.06515980e-01 -1.36975741e+00 1.03130090e+00 6.53602958e-01
5.07425070e-01 4.58208025e-01 9.27092910e-01 -9.45530593e-01
-1.34654343e+00 -1.07260478e+00 1.37494063e+00 -5.95483541e-01
8.40234339e-01 -9.51435149e-01 -5.99638104e-01 7.69996405e-01
4.63617116e-01 -1.35233581e-01 5.32882214e-01 5.15983820e-01
-1.12481609e-01 3.59577723e-02 -8.67292523e-01 1.94946334e-01
1.12711620e+00 -7.26999223e-01 -9.88176286e-01 9.56919611e-01
1.16573966e+00 -6.69246614e-01 -8.88722122e-01 6.30303681e-01
-2.42468029e-01 -6.87248051e-01 1.25400925e+00 -1.00332761e+00
-2.79654562e-02 -2.80220538e-01 4.33648638e-02 -1.62306380e+00
-8.49323332e-01 -5.02459466e-01 5.67660809e-01 9.07961130e-01
1.32673442e-01 -5.75266421e-01 8.54260445e-01 -1.03037775e-01
-4.74835724e-01 -1.48899287e-01 -1.04804254e+00 -8.63253057e-01
4.94959056e-01 -7.03824759e-01 5.25111437e-01 7.02710092e-01
6.69326559e-02 6.83651865e-01 -1.97137311e-01 1.24636680e-01
2.70328104e-01 3.73077303e-01 1.38646498e-01 -1.60594499e+00
-5.12090445e-01 -2.38249570e-01 -5.70752859e-01 -7.11312652e-01
5.37982583e-01 -1.79952204e+00 4.94115911e-02 -9.55821276e-01
1.70146748e-01 -8.14522624e-01 -5.15890539e-01 2.73747027e-01
4.56840426e-01 2.30887845e-01 4.95398819e-01 -7.09130466e-02
-6.27946436e-01 5.09354770e-01 8.84293377e-01 -1.56018078e-01
3.05197358e-01 1.63516521e-01 -1.80920362e-01 6.77338600e-01
9.06454444e-01 -7.90605962e-01 -2.85467893e-01 -5.07415593e-01
6.91762507e-01 3.90657276e-01 2.75297254e-01 -1.17375922e+00
3.84114623e-01 1.28911272e-01 5.14516294e-01 -1.17368603e+00
4.79768783e-01 -5.49359858e-01 2.47835606e-01 2.93897212e-01
-3.54508460e-01 1.00501883e+00 5.62463701e-01 4.62952763e-01
-5.67800224e-01 -6.88195348e-01 3.57297331e-01 -4.41497266e-01
-9.02400553e-01 2.73076504e-01 -5.54272413e-01 -1.87333852e-01
8.00645471e-01 4.38627750e-01 -3.29633616e-02 1.54264390e-01
-9.23583508e-01 -4.37552892e-02 8.95324126e-02 8.95729899e-01
4.83377516e-01 -1.37761557e+00 -3.97088021e-01 3.01325738e-01
-1.34931877e-01 -5.82772374e-01 -2.56448597e-01 5.38347483e-01
-5.44499040e-01 1.16380119e+00 -3.66388112e-01 -4.82777685e-01
-1.38586628e+00 1.01384902e+00 2.11220592e-01 -7.40217209e-01
-9.18925047e-01 7.42129385e-01 -8.19671303e-02 -8.52249324e-01
-3.42087090e-01 -4.25905168e-01 -2.51374036e-01 -3.71632189e-01
3.74255866e-01 -1.94357991e-01 5.79917312e-01 -5.80057144e-01
-2.97320157e-01 2.53997803e-01 -4.95905429e-01 -2.70203263e-01
1.23824382e+00 4.08974230e-01 -5.96507430e-01 6.06081665e-01
9.18850601e-01 1.86688170e-01 -2.00412735e-01 -3.45718622e-01
4.51228291e-01 -4.64382023e-02 1.78837821e-01 -8.17812979e-01
-8.51128757e-01 8.03893864e-01 6.10831261e-01 4.45503473e-01
6.00984693e-01 1.96475476e-01 8.97382855e-01 5.22273719e-01
8.26689601e-01 -7.17179418e-01 -3.84717464e-01 1.19622386e+00
5.30364096e-01 -9.00079131e-01 -3.53805840e-01 -6.33473694e-01
-3.38922977e-01 1.39353538e+00 2.71856666e-01 -6.89982902e-03
7.43661106e-01 6.25925004e-01 -1.15977049e-01 -6.46871924e-01
-8.91578436e-01 -2.67704993e-01 1.33447126e-01 3.61270785e-01
5.52467883e-01 2.08027035e-01 -6.14712059e-01 3.58675241e-01
-5.73160350e-01 6.64377213e-02 1.33879185e-01 1.03604281e+00
-3.17466348e-01 -2.14771938e+00 -3.17820549e-01 3.95480692e-01
-5.85695982e-01 -4.19588238e-01 -2.14597523e-01 8.06298077e-01
4.94235814e-01 7.82505155e-01 2.16567338e-01 -3.58328670e-01
1.68387443e-01 3.39667976e-01 6.81208134e-01 -9.43157315e-01
-9.81757164e-01 -2.60226786e-01 3.44936341e-01 -6.13111258e-01
-6.05007589e-01 -7.89138198e-01 -1.43599200e+00 -2.80357778e-01
-5.86972609e-02 9.98708725e-01 8.74580145e-01 1.24513900e+00
1.29886806e-01 9.83624220e-01 1.84391364e-01 -1.02748334e+00
-4.03084040e-01 -7.10665941e-01 -5.17910302e-01 1.35714427e-01
-8.63787308e-02 -9.00377810e-01 1.95621178e-01 -5.87161303e-01] | [15.701865196228027, 2.9186012744903564] |
59e0f40a-e9e4-431c-a7f2-fda0278bfc2c | leveraging-affinity-cycle-consistency-to | null | null | https://openreview.net/forum?id=Hr-cI3LMKb8 | https://openreview.net/pdf?id=Hr-cI3LMKb8 | Leveraging affinity cycle consistency to isolate factors of variation in learned representations | Identifying the dominant factors of variation across a dataset is a central goal of representation learning. Generative approaches lead to descriptions that are rich enough to recreate the data, but often only a partial description is needed to complete downstream tasks or to gain insights about the dataset. In this work, we operate in the setting where limited information is known about the data in the form of groupings, or set membership, and the task is to learn representations which isolate the factors of variation that are common across the groupings. Our key insight is the use of affinity cycle consistency (ACC) between the learned embeddings of images belonging to different sets. In contrast to prior work, we demonstrate that we can work with significantly fewer constraints on the factors of variation, and in some cases, with almost no constraints on the factors of variation for part of the data. By curating datasets from Shapes3D, we quantify the effectiveness of ACC through mutual information between the learned representations and the known image generative factors. In addition, we demonstrate the applicability of ACC to the tasks of digit style isolation and synthetic-to-real object pose transfer. | ['Ameesh Makadia', 'Srikumar Ramalingam', 'Varun Jampani', 'Kieran A Murphy'] | 2021-01-01 | null | null | null | null | ['pose-transfer'] | ['computer-vision'] | [ 2.49885961e-01 2.43925508e-02 -1.85325705e-02 -2.76669383e-01
-4.19441998e-01 -9.89298522e-01 9.41740155e-01 6.77627400e-02
-7.72654191e-02 4.92045403e-01 5.05648911e-01 1.21067844e-01
-2.73286879e-01 -5.29665112e-01 -9.63288426e-01 -7.86701500e-01
2.88865775e-01 5.46916246e-01 -1.51093125e-01 -1.48540258e-01
1.81376681e-01 6.41078055e-01 -1.48244798e+00 1.28966957e-01
4.74506468e-01 6.67110562e-01 2.75339335e-01 4.52903211e-01
-8.56993124e-02 3.68396461e-01 -5.84533870e-01 -8.23490247e-02
4.14058894e-01 -6.73153698e-01 -5.47334373e-01 6.27263069e-01
5.74092388e-01 -2.07528975e-02 -3.82415920e-01 8.93353581e-01
2.20624685e-01 3.75900455e-02 1.06055105e+00 -1.11432588e+00
-6.53313935e-01 3.26581270e-01 -4.88297850e-01 1.28214341e-02
1.57568142e-01 4.79767732e-02 1.03282154e+00 -8.21840346e-01
8.13646972e-01 1.20788765e+00 4.49556738e-01 5.61856747e-01
-1.81168497e+00 -3.12702745e-01 3.66874754e-01 -2.02106446e-01
-1.32049155e+00 -5.41222155e-01 7.74854541e-01 -8.45829844e-01
3.16134393e-01 1.22219034e-01 5.74591339e-01 1.09273493e+00
1.05554517e-02 4.80450958e-01 8.04392636e-01 -4.55371648e-01
3.10368806e-01 1.85035095e-01 8.04982036e-02 4.89967883e-01
3.44323605e-01 -8.14289674e-02 -6.64075255e-01 -2.19483282e-02
9.05406058e-01 3.01572662e-02 -4.44479227e-01 -9.44291532e-01
-1.19026160e+00 7.90676177e-01 4.89832401e-01 2.73466855e-01
-1.34526744e-01 1.72558144e-01 1.04666337e-01 1.61984712e-01
2.23526925e-01 6.62627578e-01 -3.08906615e-01 -2.74338722e-02
-4.53095704e-01 4.19317722e-01 7.54693568e-01 1.19705963e+00
9.79244649e-01 -6.15228117e-02 1.04653500e-01 6.81150138e-01
1.52187958e-01 2.90043890e-01 3.72813702e-01 -8.87696147e-01
2.94745356e-01 5.34056544e-01 1.88506469e-01 -9.13435102e-01
-5.19354548e-03 -4.11206305e-01 -4.98568058e-01 3.02230895e-01
7.85945594e-01 -1.21337570e-01 -1.06490731e+00 2.27011323e+00
1.32200703e-01 -7.67558962e-02 -1.25148267e-01 7.06182063e-01
1.86366126e-01 3.08342308e-01 -3.09709817e-01 8.36453363e-02
1.10336196e+00 -3.75076145e-01 -4.28896695e-01 -3.86852801e-01
4.15821463e-01 -7.37197042e-01 1.05423462e+00 1.21733725e-01
-7.71007240e-01 -5.27564228e-01 -1.08882141e+00 -5.03985211e-02
-3.02239805e-01 -3.52792703e-02 4.56823260e-01 3.99867266e-01
-8.06249321e-01 6.88673556e-01 -8.13057601e-01 -3.92041355e-01
3.60814601e-01 1.84824988e-01 -4.50793326e-01 -3.13907325e-01
-6.17711067e-01 7.88956344e-01 8.19345564e-02 -1.10361591e-01
-8.60660732e-01 -9.82495189e-01 -8.15329969e-01 -6.65906072e-02
2.97409534e-01 -6.85230315e-01 9.33135986e-01 -1.12378645e+00
-9.54523385e-01 6.70423090e-01 -2.12567732e-01 -1.85813121e-02
5.23732185e-01 -1.86274290e-01 2.61277616e-01 -1.17551260e-01
1.02943115e-01 6.91241205e-01 9.81292605e-01 -1.54756379e+00
-1.68207437e-01 -5.10426700e-01 2.20658854e-01 2.50566781e-01
-3.03193536e-02 -3.70226473e-01 -5.87724984e-01 -6.57085001e-01
1.91020280e-01 -1.20559394e+00 -6.03161640e-02 5.07256389e-01
-2.79420584e-01 8.07555467e-02 7.29587138e-01 -5.58265209e-01
6.50043070e-01 -2.48588538e+00 7.15998948e-01 3.17023277e-01
2.55523533e-01 -1.54740736e-01 -6.28936663e-02 5.27398169e-01
-1.27290949e-01 6.46808371e-02 -3.49331290e-01 -5.16633928e-01
1.68054029e-01 5.55663347e-01 -3.56267959e-01 4.39906031e-01
6.27045512e-01 6.45031750e-01 -8.15714240e-01 -7.25989714e-02
9.19401199e-02 5.59574902e-01 -5.80159009e-01 2.23353654e-01
-3.72382134e-01 7.68229306e-01 -3.79613191e-01 1.65453196e-01
3.12029362e-01 -2.62825757e-01 2.94122070e-01 -3.30297500e-01
2.05856800e-01 3.30598891e-01 -1.44969440e+00 1.77571380e+00
-4.85975474e-01 5.77964306e-01 -7.59573355e-02 -8.59988689e-01
7.15426564e-01 -1.82227455e-02 3.35961372e-01 -2.99367934e-01
-3.86452451e-02 1.21099941e-01 3.15046787e-01 -1.58502504e-01
2.25072786e-01 -3.50896597e-01 -2.96957400e-02 6.47288680e-01
2.28516266e-01 -1.76977798e-01 2.21941739e-01 3.27719212e-01
9.35707927e-01 1.93296388e-01 1.45648733e-01 -3.46703470e-01
-8.29010680e-02 -1.75874949e-01 4.71634328e-01 5.15318453e-01
2.10552856e-01 9.43866313e-01 7.82279193e-01 -1.79642066e-01
-1.39071476e+00 -1.24880660e+00 -2.95072317e-01 5.98911285e-01
8.87183473e-02 -3.84665161e-01 -4.85772520e-01 -5.59943497e-01
2.44368866e-01 5.93596458e-01 -1.01525152e+00 -2.77118832e-01
-3.63954127e-01 -4.58930403e-01 2.07893372e-01 7.02586174e-01
3.88935208e-02 -6.95142329e-01 -5.02161384e-01 -6.28576800e-02
1.82596475e-01 -1.04607451e+00 -6.94493711e-01 2.45844841e-01
-6.08938336e-01 -1.15819621e+00 -5.34018934e-01 -5.46677947e-01
1.02718771e+00 8.68417397e-02 1.10654855e+00 -2.23201349e-01
-4.80942994e-01 5.62013149e-01 -2.08397642e-01 -3.35918576e-01
-3.72696698e-01 -1.97744504e-01 -6.69072494e-02 1.60971746e-01
3.71762291e-02 -7.70261347e-01 -3.11303675e-01 2.08813980e-01
-9.97309804e-01 9.24857706e-02 4.76542056e-01 7.42126107e-01
5.41009068e-01 -1.30463392e-01 2.72033781e-01 -1.04725504e+00
4.51261371e-01 -4.53094900e-01 -2.59542763e-01 1.57022968e-01
-2.06694365e-01 5.89540184e-01 5.62751651e-01 -8.06409776e-01
-7.70040333e-01 1.69809729e-01 4.32598978e-01 -7.04918087e-01
-1.05200000e-01 3.61722171e-01 -5.58141947e-01 2.73631811e-01
7.34002590e-01 6.63807839e-02 1.47543058e-01 -7.18812406e-01
5.68058074e-01 1.20339669e-01 3.35207254e-01 -9.29162562e-01
7.38472164e-01 4.57322866e-01 1.27573252e-01 -8.50013673e-01
-6.59284055e-01 -1.76486149e-01 -9.79122818e-01 1.29633352e-01
7.81412959e-01 -8.09165001e-01 -1.03705570e-01 2.68984258e-01
-8.48249137e-01 -2.59101003e-01 -6.97585225e-01 2.99973488e-01
-5.93024790e-01 1.85085163e-01 -1.84943527e-01 -5.62076271e-01
4.52980846e-01 -1.23406029e+00 9.85825837e-01 2.21013222e-02
-4.00265068e-01 -9.33116376e-01 4.27671187e-02 -7.65410811e-02
9.29574296e-02 3.47396851e-01 1.27640808e+00 -4.95653898e-01
-6.52565956e-01 -1.51368409e-01 -1.03828549e-01 5.08068144e-01
7.47799456e-01 -4.20869514e-02 -9.26797867e-01 -3.02887380e-01
-1.01178832e-01 -1.24528632e-01 7.73912847e-01 3.37621152e-01
9.30263638e-01 -2.36954004e-01 -2.00973436e-01 5.21597683e-01
1.18269277e+00 1.16093673e-01 5.04475772e-01 -3.04057933e-02
9.68663514e-01 9.00554121e-01 1.11629941e-01 2.59546399e-01
9.60746855e-02 8.50187182e-01 6.24188371e-02 1.54141271e-02
-2.78746367e-01 -5.32896996e-01 8.64577219e-02 4.69306022e-01
1.58065945e-01 -1.33935317e-01 -7.11987793e-01 6.57425463e-01
-1.66030562e+00 -6.44884050e-01 3.73121113e-01 2.36374092e+00
7.92099833e-01 7.54799098e-02 1.53627381e-01 4.70693409e-02
7.17755497e-01 1.27846032e-01 -8.16666245e-01 -1.68966174e-01
-1.05714304e-02 1.55793428e-01 3.52039903e-01 5.03933489e-01
-9.32469666e-01 5.22992015e-01 6.22045660e+00 4.38531071e-01
-9.74092603e-01 -3.56275946e-01 4.31757867e-01 -1.15483388e-01
-5.67774355e-01 2.20713302e-01 -5.27437687e-01 4.32459772e-01
4.84080642e-01 -1.35437950e-01 6.02555871e-01 5.99931419e-01
-1.18127204e-01 1.16455913e-01 -1.74386060e+00 7.35354125e-01
1.66882023e-01 -1.03141654e+00 2.37418517e-01 3.02326739e-01
6.31704867e-01 -2.92904913e-01 1.89211234e-01 -9.59738046e-02
4.34283257e-01 -1.06249309e+00 7.48896718e-01 5.55968523e-01
7.15771675e-01 -6.20855629e-01 2.88537294e-01 2.83994496e-01
-8.23544621e-01 9.87374559e-02 -3.54071140e-01 -4.11632992e-02
-1.15241565e-01 3.63077492e-01 -8.68335605e-01 3.71041059e-01
3.03443998e-01 8.40026200e-01 -5.63878655e-01 6.15297437e-01
-3.45974684e-01 3.29821110e-01 -2.98681110e-01 1.48104623e-01
-6.11952357e-02 -3.71747822e-01 4.68087822e-01 8.92020881e-01
1.67794332e-01 -1.43341988e-01 1.51682585e-01 1.16921401e+00
-1.70802116e-01 -2.39765957e-01 -7.72675216e-01 -3.85571867e-01
3.34273070e-01 9.67206359e-01 -5.87710977e-01 -1.40497936e-02
-3.13522339e-01 8.26771140e-01 4.23965365e-01 4.02978480e-01
-3.86137396e-01 -2.71993014e-03 1.06786358e+00 3.07383537e-01
5.85632443e-01 -6.39833748e-01 -3.25722724e-01 -1.11250114e+00
2.36501336e-01 -8.60010684e-01 -2.09360779e-03 -7.77274787e-01
-1.44873345e+00 1.30598515e-01 2.39697650e-01 -1.08007848e+00
-4.45467800e-01 -6.35669351e-01 -4.95625347e-01 1.16297328e+00
-1.00237060e+00 -1.14545989e+00 -1.07135616e-01 3.35835695e-01
5.77979207e-01 4.95196469e-02 7.67471313e-01 1.06308209e-02
-3.75288218e-01 3.01470250e-01 2.54141659e-01 2.15959311e-01
6.74401820e-01 -1.33031523e+00 3.77152979e-01 8.05899084e-01
4.89444703e-01 9.60229456e-01 8.77379119e-01 -5.18801630e-01
-1.44601953e+00 -9.05495763e-01 4.59401071e-01 -6.49911582e-01
6.21517777e-01 -6.85500145e-01 -9.90310848e-01 8.70562136e-01
-1.06892549e-01 4.72210348e-02 6.01044714e-01 3.30588996e-01
-7.53083825e-01 4.39718440e-02 -8.56283724e-01 6.18682742e-01
1.20455301e+00 -6.99166596e-01 -4.92958814e-01 5.73589001e-03
6.60734594e-01 -2.71600038e-01 -6.58844590e-01 -1.09614776e-02
7.16896892e-01 -6.04648054e-01 8.68415654e-01 -8.85110974e-01
6.26817286e-01 -4.58554208e-01 -5.43880284e-01 -1.64913177e+00
-3.95503789e-01 -1.88813254e-01 1.24808863e-01 1.43159544e+00
4.62467730e-01 -3.60556543e-01 5.95306098e-01 8.36878121e-01
-9.21076611e-02 -4.78416145e-01 -6.73999250e-01 -6.40704691e-01
3.55796307e-01 -1.08318366e-01 5.67647398e-01 9.02672887e-01
-3.93564403e-01 5.14275849e-01 -1.84366703e-01 1.90084562e-01
5.25011897e-01 2.43056253e-01 8.76965225e-01 -1.11639798e+00
-7.13850141e-01 -2.48976827e-01 -6.92059934e-01 -9.25433338e-01
1.82600886e-01 -8.23211014e-01 5.10942303e-02 -1.33581841e+00
3.75221461e-01 -5.41190207e-01 -1.26727656e-01 3.96053076e-01
-7.02542886e-02 3.69313285e-02 3.13651353e-01 3.88302237e-01
-5.11990599e-02 6.04612768e-01 1.29697013e+00 -1.60793915e-01
-1.19536094e-01 -2.75307983e-01 -9.96059060e-01 5.59301019e-01
5.86458743e-01 -3.14165503e-01 -7.48612463e-01 -7.02798426e-01
9.82615948e-02 -4.28374052e-01 4.15596962e-01 -7.67802238e-01
-4.28836159e-02 -2.68557936e-01 6.33182168e-01 -1.64192468e-01
4.23767984e-01 -7.44897544e-01 3.77955258e-01 5.06794415e-02
-5.37951052e-01 -3.80241796e-02 1.40912026e-01 7.12460339e-01
9.40199420e-02 -2.31368884e-01 7.95730174e-01 -2.09164649e-01
-4.50298458e-01 1.53926641e-01 4.73703854e-02 2.95841277e-01
8.85048866e-01 -1.35990024e-01 -1.24664649e-01 -4.27256644e-01
-7.91782677e-01 -1.07926726e-01 9.40160751e-01 5.50233305e-01
4.18813497e-01 -1.34136093e+00 -6.20768130e-01 5.80521166e-01
3.71364683e-01 1.91776633e-01 -4.42890786e-02 4.05371159e-01
-1.47520885e-01 1.20478369e-01 -2.61620522e-01 -6.74228191e-01
-1.10311162e+00 5.71183562e-01 3.25248420e-01 2.14895383e-01
-6.75780177e-01 6.26816750e-01 7.74269521e-01 -3.50170195e-01
8.50039646e-02 -2.73977488e-01 1.12084724e-01 1.56441480e-01
3.02226871e-01 -5.23649752e-02 3.03448644e-02 -6.17702603e-01
-1.84575886e-01 6.72095001e-01 -1.95309073e-01 -2.80969322e-01
1.35068285e+00 -5.24333343e-02 2.42339924e-01 8.53236496e-01
1.22272170e+00 5.56106642e-02 -1.74658203e+00 -3.17838848e-01
-1.99480131e-01 -7.51914978e-01 -2.56392449e-01 -5.37184238e-01
-8.93187106e-01 9.51352537e-01 3.56838495e-01 -5.72190545e-02
8.34063113e-01 3.72754604e-01 1.32134795e-01 1.45322725e-01
2.53185093e-01 -7.32251525e-01 2.50528932e-01 2.45590851e-01
1.01288724e+00 -8.85897219e-01 -2.47122943e-02 -3.89422983e-01
-6.37795329e-01 8.86943936e-01 3.04779470e-01 -4.20329779e-01
5.35422444e-01 2.68800884e-01 -3.73107791e-02 -1.22877203e-01
-7.02622414e-01 -1.74106836e-01 3.99673402e-01 7.82339096e-01
4.09292489e-01 -5.58007397e-02 3.19172367e-02 2.38231108e-01
-3.12191159e-01 -4.00450468e-01 4.96313632e-01 1.01771939e+00
-1.44506827e-01 -1.30095410e+00 -1.23462141e-01 4.45029289e-01
-1.60363361e-01 1.06281199e-01 -7.88181484e-01 9.75082994e-01
1.95556715e-01 5.11036992e-01 3.74767751e-01 -2.86603391e-01
3.96867931e-01 1.99811354e-01 8.60314906e-01 -8.73225689e-01
4.17989271e-04 1.26634881e-01 -4.73116450e-02 -2.16969460e-01
-3.11333954e-01 -1.05030739e+00 -9.23020303e-01 -4.59550358e-02
-1.44747406e-01 -5.89004382e-02 7.65220582e-01 9.97681499e-01
3.45038444e-01 5.60053468e-01 5.48894525e-01 -8.74380767e-01
-5.33662498e-01 -8.47757876e-01 -7.71727026e-01 8.70087743e-01
5.03538668e-01 -9.98256266e-01 -5.42442143e-01 4.30878073e-01] | [9.498619079589844, 2.1326513290405273] |
337fde44-7e40-4e70-9ad6-64904d49bb3f | text2poster-laying-out-stylized-texts-on | 2301.02363 | null | https://arxiv.org/abs/2301.02363v1 | https://arxiv.org/pdf/2301.02363v1.pdf | Text2Poster: Laying out Stylized Texts on Retrieved Images | Poster generation is a significant task for a wide range of applications, which is often time-consuming and requires lots of manual editing and artistic experience. In this paper, we propose a novel data-driven framework, called \textit{Text2Poster}, to automatically generate visually-effective posters from textual information. Imitating the process of manual poster editing, our framework leverages a large-scale pretrained visual-textual model to retrieve background images from given texts, lays out the texts on the images iteratively by cascaded auto-encoders, and finally, stylizes the texts by a matching-based method. We learn the modules of the framework by weakly- and self-supervised learning strategies, mitigating the demand for labeled data. Both objective and subjective experiments demonstrate that our Text2Poster outperforms state-of-the-art methods, including academic research and commercial software, on the quality of generated posters. | ['Zhiwu Lu', 'Ruihua Song', 'Hongteng Xu', 'Chuhao Jin'] | 2023-01-06 | null | null | null | null | ['layout-design', 'multimodal-generation'] | ['computer-vision', 'natural-language-processing'] | [ 5.73083401e-01 -2.54233573e-02 1.15743214e-02 -1.86605155e-01
-9.25014675e-01 -7.44662285e-01 8.16926479e-01 -2.36207411e-01
-1.75796285e-01 5.09882271e-01 1.43510804e-01 -2.84688979e-01
4.33631510e-01 -6.71665490e-01 -1.04979169e+00 -2.77694255e-01
4.25339073e-01 3.95771742e-01 1.63280338e-01 3.22836526e-02
3.59226018e-01 2.86356837e-01 -1.49016750e+00 5.54963171e-01
9.30287123e-01 9.58870709e-01 4.61895078e-01 7.49615848e-01
-4.27973777e-01 8.63192081e-01 -8.62590134e-01 -1.00730407e+00
1.76232412e-01 -6.16159856e-01 -4.11255687e-01 7.57987201e-01
6.95276320e-01 -5.04198730e-01 -3.11317712e-01 7.67439663e-01
4.90947992e-01 -1.96286917e-01 8.48283947e-01 -1.07116330e+00
-1.09611511e+00 5.70601046e-01 -9.16284263e-01 -2.51186520e-01
2.46669486e-01 3.82070810e-01 9.29283440e-01 -1.19815195e+00
6.78600132e-01 1.35377693e+00 4.72743899e-01 5.58202088e-01
-1.47012174e+00 -7.23016679e-01 3.01252097e-01 -2.89566457e-01
-1.28792799e+00 -5.72779596e-01 9.61053848e-01 -6.76584065e-01
6.81894720e-02 1.87296048e-01 4.93777752e-01 1.43746960e+00
-1.17708303e-01 1.28655922e+00 1.02361023e+00 -5.25991857e-01
2.61946529e-01 3.63093853e-01 -5.46836972e-01 8.62039089e-01
2.71677643e-01 -3.37653011e-01 -8.93245220e-01 1.56350911e-01
1.04367912e+00 3.02427653e-02 -1.22551573e-02 -4.05684710e-01
-1.15297043e+00 6.49997532e-01 1.02489889e-01 -4.15931530e-02
-2.80543804e-01 2.85939425e-01 1.52078971e-01 -1.59122109e-01
5.97222447e-01 3.68669510e-01 -9.07764658e-02 9.29777995e-02
-1.38407242e+00 2.15003416e-01 5.47394574e-01 1.12845087e+00
7.30819702e-01 6.29369691e-02 -5.13652503e-01 7.75621414e-01
4.73948866e-01 7.88938046e-01 1.29765615e-01 -8.42956841e-01
6.45744979e-01 7.46335328e-01 3.01774085e-01 -9.85945463e-01
4.14417356e-01 -1.85207903e-01 -7.67009079e-01 2.62351960e-01
2.47332856e-01 -2.17598841e-01 -1.13680899e+00 1.26866543e+00
2.30299488e-01 3.24969436e-03 -3.33207279e-01 6.35200799e-01
7.55207121e-01 7.45193303e-01 2.05189541e-01 6.50600344e-03
1.30984175e+00 -1.40093470e+00 -7.85064518e-01 -4.55154747e-01
-6.96554482e-02 -1.11770999e+00 1.51613081e+00 4.28426653e-01
-1.36801207e+00 -6.08157814e-01 -1.03336453e+00 -1.04184799e-01
-1.13183163e-01 7.52350628e-01 3.68588299e-01 6.06037557e-01
-8.68614852e-01 5.18493652e-01 -6.47639632e-01 -7.50828087e-02
9.14062738e-01 -3.34085561e-02 6.39949515e-02 7.63231248e-04
-5.39871514e-01 4.69222784e-01 -1.01807036e-01 1.12450920e-01
-1.14019406e+00 -7.28438556e-01 -8.69243383e-01 -1.96215674e-01
3.83578032e-01 -8.28242540e-01 1.43819654e+00 -1.21571553e+00
-1.79328406e+00 1.05187237e+00 -2.69487858e-01 -2.05322430e-01
9.91766274e-01 -5.36332726e-01 -2.51975954e-01 3.06694180e-01
3.13016891e-01 8.74661744e-01 1.43762410e+00 -1.80177116e+00
-5.22510231e-01 -5.13620786e-02 -1.96157396e-01 1.15257204e-01
-4.91811901e-01 5.88387856e-03 -1.17972052e+00 -1.13446271e+00
-4.37652737e-01 -8.27974796e-01 -3.70329946e-01 4.33320165e-01
-6.35533571e-01 -1.65161882e-02 9.78306293e-01 -7.80078232e-01
1.37072527e+00 -2.23788643e+00 -1.91972423e-02 1.62049666e-01
2.09396601e-01 2.54858643e-01 -2.21469894e-01 3.08956712e-01
3.24444383e-01 2.65643150e-01 -1.11824177e-01 -8.03983510e-01
2.35798553e-01 -2.10024163e-01 -6.37518406e-01 5.87638319e-02
4.39940929e-01 1.08841574e+00 -8.94079387e-01 -9.66541708e-01
4.43416476e-01 6.60566688e-01 -2.33818248e-01 4.07157689e-01
-8.07127059e-01 3.79371196e-01 -4.36454415e-01 7.14033604e-01
5.24442971e-01 -6.79910481e-01 9.14793015e-02 -2.43463203e-01
9.63448081e-03 -5.62877692e-02 -1.09306002e+00 1.97241831e+00
-6.00637197e-01 8.63405168e-01 -6.98240474e-02 -3.25755596e-01
8.66435289e-01 5.79237789e-02 1.44732013e-01 -6.96063221e-01
1.03091918e-01 3.54361683e-02 -7.07494438e-01 -5.00662684e-01
6.33046389e-01 2.50981927e-01 -4.49850895e-02 6.99920833e-01
-1.73643932e-01 -2.35676482e-01 4.18786883e-01 6.49949729e-01
8.40547979e-01 7.08574772e-01 -2.16028392e-01 2.49059245e-01
3.40636581e-01 -1.69523463e-01 1.38460964e-01 6.45766973e-01
2.21231282e-01 9.28174734e-01 2.77686805e-01 -3.14232230e-01
-1.23822343e+00 -1.15596199e+00 3.32094818e-01 1.16253090e+00
1.94330260e-01 -5.77393949e-01 -9.80972290e-01 -7.98799455e-01
-1.63018152e-01 6.87824547e-01 -6.17493451e-01 2.86139369e-01
-4.15300041e-01 -3.23123097e-01 4.20907021e-01 5.56934655e-01
4.04109716e-01 -1.22764552e+00 -4.66766924e-01 9.39614251e-02
-1.79835632e-01 -1.15366495e+00 -1.02033913e+00 -2.99402088e-01
-5.69985688e-01 -7.03228414e-01 -1.17390215e+00 -9.37361419e-01
1.07822978e+00 3.45693439e-01 1.42654562e+00 1.33654520e-01
-5.09664714e-01 2.92280942e-01 -1.72531679e-01 -4.48274225e-01
-4.20954585e-01 -6.50382265e-02 -4.67531532e-01 4.15594906e-01
5.32570444e-02 -3.33894998e-01 -9.10038650e-01 2.66329736e-01
-1.06270480e+00 5.53200066e-01 9.92012024e-01 5.27557135e-01
7.81693697e-01 -5.03862388e-02 2.54852235e-01 -8.64611328e-01
6.35330975e-01 9.52802896e-02 -6.54262781e-01 3.42155784e-01
-2.27082804e-01 2.02175573e-01 5.34107029e-01 -6.50876939e-01
-1.18605363e+00 3.25205117e-01 2.94913799e-01 -5.87315679e-01
7.27435797e-02 -2.47830842e-02 -3.82761389e-01 8.19007382e-02
6.27345264e-01 3.03727627e-01 -3.24682236e-01 -5.17599285e-01
6.54072285e-01 7.55490065e-01 5.65998375e-01 -6.34887397e-01
1.23808396e+00 6.33796155e-01 -4.43459362e-01 -6.49366200e-01
-1.21462238e+00 1.23544708e-02 -5.34617603e-01 -5.10304153e-01
8.98272336e-01 -1.12636387e+00 -4.20971066e-01 4.56156373e-01
-1.32317495e+00 -6.94833279e-01 -2.75227278e-01 -2.49132052e-01
-3.97117019e-01 2.86601573e-01 -5.38468599e-01 -9.60646629e-01
-4.70578909e-01 -1.11178172e+00 1.64778233e+00 3.94050568e-01
-1.85868338e-01 -7.67698228e-01 -1.66505530e-01 7.14592159e-01
1.93317533e-01 3.06166679e-01 5.08728445e-01 -7.75098428e-02
-1.07966387e+00 -7.15942755e-02 -5.21684110e-01 2.47633830e-01
9.25186500e-02 2.93847978e-01 -1.02856672e+00 -3.24341320e-02
-7.62455702e-01 -5.06274879e-01 7.60658562e-01 3.03142518e-01
1.45531559e+00 -3.91086727e-01 -4.21653658e-01 3.76870960e-01
1.19367456e+00 -7.94314668e-02 6.48660362e-01 3.27618331e-01
9.09434140e-01 5.66754401e-01 4.59599614e-01 4.99220222e-01
5.00780165e-01 5.35041153e-01 1.89411566e-01 -5.12674153e-01
-4.45510775e-01 -8.71829331e-01 3.27693909e-01 4.06356663e-01
2.63827324e-01 -3.16186041e-01 -5.97861409e-01 7.26966321e-01
-1.87285054e+00 -9.02324140e-01 3.29694413e-02 2.00897050e+00
1.26381695e+00 1.84431866e-01 1.17603175e-01 -7.47972503e-02
9.73551810e-01 2.13196293e-01 -5.98794699e-01 -9.91877541e-03
-4.02385853e-02 2.24543333e-01 2.72208720e-01 2.37808898e-01
-1.15147221e+00 1.17646086e+00 6.51825953e+00 9.59337711e-01
-1.04983842e+00 -9.44140181e-02 1.10252035e+00 -1.88160494e-01
-4.15351659e-01 -1.20969392e-01 -6.27662897e-01 7.79532194e-01
5.94259262e-01 1.15552666e-02 4.51869965e-01 9.12068605e-01
3.47826362e-01 8.75660777e-02 -1.01724398e+00 1.08249509e+00
4.75238711e-01 -1.71302640e+00 3.71032327e-01 2.88316291e-02
1.14905989e+00 -6.00189984e-01 3.76251578e-01 5.11840992e-02
6.99584186e-01 -9.96025801e-01 1.15606797e+00 4.89278704e-01
1.05510163e+00 -5.40671527e-01 3.56038548e-02 -1.21504657e-01
-1.19132161e+00 1.91995040e-01 7.58255348e-02 3.27939361e-01
3.48142743e-01 6.96364939e-01 -6.18881404e-01 1.22604944e-01
6.64893568e-01 6.67897522e-01 -9.79000568e-01 8.42418373e-01
-5.90435207e-01 4.92488950e-01 1.11710250e-01 -9.07983854e-02
2.37540510e-02 -6.32814178e-03 1.30690545e-01 1.33259547e+00
1.36261538e-01 -3.30768675e-01 3.17793369e-01 1.08608592e+00
-5.72190583e-01 1.10722817e-01 -4.67900276e-01 -4.51507568e-01
4.94725615e-01 1.60594666e+00 -9.69713688e-01 -5.51024854e-01
-1.64990827e-01 1.52077854e+00 1.27482163e-02 4.51823235e-01
-9.00386751e-01 -5.71082592e-01 1.06729008e-01 4.33497399e-01
6.19886339e-01 -7.19016120e-02 -5.50783992e-01 -1.11109102e+00
2.20568404e-01 -1.06787527e+00 2.57048067e-02 -1.27760327e+00
-1.36230481e+00 5.49204171e-01 -4.60331887e-01 -1.42016757e+00
-1.54056272e-03 -5.70437491e-01 -6.55317426e-01 5.56052029e-01
-1.45934486e+00 -1.59461606e+00 -5.47280371e-01 4.60491985e-01
8.91368210e-01 -6.63858801e-02 3.96716356e-01 1.81913331e-01
-6.03474915e-01 5.91943324e-01 -9.15302932e-02 3.28483969e-01
9.99774456e-01 -1.29738700e+00 6.23673916e-01 8.89524221e-01
3.60495716e-01 5.74212730e-01 6.40628994e-01 -7.57333934e-01
-1.42330229e+00 -1.33305717e+00 7.71644771e-01 -5.75391531e-01
7.18414783e-01 -7.59759426e-01 -3.86755347e-01 4.98675108e-01
6.29064739e-01 -2.91279703e-01 7.85888970e-01 -3.16684395e-01
-4.13695604e-01 -1.35687634e-01 -6.81606352e-01 1.19574869e+00
1.04770935e+00 -4.42543000e-01 -3.93706799e-01 5.12204289e-01
7.29173839e-01 -4.19848710e-01 -3.57673973e-01 -1.36706784e-01
5.38175523e-01 -6.94112360e-01 1.06421781e+00 -1.96728423e-01
9.71015990e-01 -5.16998887e-01 2.38361120e-01 -9.83292043e-01
3.54988836e-02 -1.12769759e+00 -2.19087318e-01 1.54726875e+00
5.10150373e-01 2.29052827e-01 9.26612854e-01 6.04818523e-01
1.33886412e-01 -5.03832281e-01 -2.22344324e-01 -3.99038404e-01
-4.01807219e-01 -2.91131020e-01 3.78100961e-01 5.19107282e-01
-4.24020767e-01 6.49303138e-01 -6.27164423e-01 -1.30979344e-01
8.35873902e-01 2.97737926e-01 1.25977528e+00 -1.07960701e+00
-2.73232311e-01 -4.83562857e-01 -2.34715473e-02 -1.26812792e+00
7.28574917e-02 -3.99848968e-01 2.27448657e-01 -1.81324065e+00
2.98158854e-01 -1.55453175e-01 1.70359552e-01 3.27777356e-01
-4.08975512e-01 6.00716352e-01 2.67220706e-01 2.03749299e-01
-1.09411561e+00 5.52536309e-01 1.44464433e+00 -4.28389758e-01
-2.32673228e-01 -2.25403830e-01 -1.11827195e+00 8.03806841e-01
3.62338543e-01 -2.55151302e-01 -3.73904228e-01 -7.11329043e-01
3.59799892e-01 -3.13212276e-01 4.28876311e-01 -8.21183145e-01
2.31264219e-01 -3.09825391e-01 9.22368526e-01 -8.18225920e-01
3.24664980e-01 -6.10244572e-01 -2.09738925e-01 9.22007784e-02
-4.92951035e-01 4.78449464e-02 1.01651631e-01 7.52498925e-01
1.23237044e-01 -3.90307792e-02 5.90806425e-01 -1.51660830e-01
-2.79350102e-01 3.85755807e-01 -2.57089257e-01 2.39346221e-01
9.70559120e-01 -1.97690010e-01 -1.93230703e-01 -5.38689137e-01
-3.09954286e-01 1.06902041e-01 7.21130669e-01 4.68167305e-01
6.45082891e-01 -1.39498580e+00 -5.90209424e-01 -1.45521248e-02
1.20727696e-01 2.06901908e-01 1.50978699e-01 3.81731421e-01
-5.42088807e-01 -2.54332665e-02 1.30256161e-01 -5.58632791e-01
-1.19398916e+00 5.11503041e-01 -1.27127603e-01 -3.42842668e-01
-6.42576158e-01 8.24678421e-01 4.01821852e-01 -2.94674095e-02
4.66960192e-01 7.57754520e-02 1.53935790e-01 -4.33070846e-02
8.43638420e-01 3.45371738e-02 -2.08220080e-01 -3.51098061e-01
-9.51351598e-02 6.04902446e-01 -2.94639856e-01 -5.45635521e-01
1.30848885e+00 -1.18828572e-01 2.68624574e-01 2.30935454e-01
7.94012845e-01 3.63466352e-01 -1.84868658e+00 -2.34898031e-01
-4.29984629e-02 -5.97048640e-01 -6.34118542e-02 -1.01026320e+00
-1.12037480e+00 1.02240455e+00 1.58441052e-01 -1.36556119e-01
1.05840886e+00 -4.67994474e-02 9.05190945e-01 1.48578763e-01
3.93806882e-02 -1.39199483e+00 9.45506811e-01 -6.56518266e-02
9.17707145e-01 -1.23309195e+00 6.22400641e-02 -3.24920088e-01
-1.09986663e+00 8.20604980e-01 6.44525707e-01 -1.67549968e-01
1.57578483e-01 4.20319468e-01 2.80563861e-01 1.21538928e-02
-6.69859409e-01 -3.24072875e-02 4.60175753e-01 6.35065854e-01
5.11072934e-01 -2.13935688e-01 1.03085041e-01 6.73035085e-01
-4.44314778e-02 1.46435037e-01 3.15193534e-01 9.78944004e-01
-2.29419470e-01 -1.12698352e+00 -4.83714670e-01 2.09051967e-01
-3.62972409e-01 -2.99525708e-01 -7.85732329e-01 4.65762615e-01
9.39275604e-03 8.95227134e-01 8.18824172e-02 -2.17247352e-01
2.09592685e-01 -1.68961480e-01 4.26189691e-01 -6.01864517e-01
-4.82810766e-01 4.68055397e-01 -1.08563662e-01 -3.28073591e-01
-3.34080577e-01 -5.68249464e-01 -8.87989521e-01 -1.60242558e-01
-1.62197948e-01 -1.77086517e-01 7.95564651e-01 7.60317266e-01
6.97258234e-01 7.62575328e-01 6.76628411e-01 -1.12407970e+00
-1.15086712e-01 -7.94555962e-01 -1.18690968e-01 6.99593067e-01
-4.11728630e-03 -4.49079126e-01 -1.87223420e-01 8.77040863e-01] | [11.474652290344238, -0.06853753328323364] |
33febb23-c1ab-4852-a5b0-7acbab141eac | continual-learning-for-lidar-semantic | 2304.03980 | null | https://arxiv.org/abs/2304.03980v1 | https://arxiv.org/pdf/2304.03980v1.pdf | Continual Learning for LiDAR Semantic Segmentation: Class-Incremental and Coarse-to-Fine strategies on Sparse Data | During the last few years, continual learning (CL) strategies for image classification and segmentation have been widely investigated designing innovative solutions to tackle catastrophic forgetting, like knowledge distillation and self-inpainting. However, the application of continual learning paradigms to point clouds is still unexplored and investigation is required, especially using architectures that capture the sparsity and uneven distribution of LiDAR data. The current paper analyzes the problem of class incremental learning applied to point cloud semantic segmentation, comparing approaches and state-of-the-art architectures. To the best of our knowledge, this is the first example of class-incremental continual learning for LiDAR point cloud semantic segmentation. Different CL strategies were adapted to LiDAR point clouds and tested, tackling both classic fine-tuning scenarios and the Coarse-to-Fine learning paradigm. The framework has been evaluated through two different architectures on SemanticKITTI, obtaining results in line with state-of-the-art CL strategies and standard offline learning. | ['Simone Milani', 'Elena Camuffo'] | 2023-04-08 | null | null | null | null | ['class-incremental-learning', 'lidar-semantic-segmentation'] | ['computer-vision', 'computer-vision'] | [ 4.09195721e-01 3.21203135e-02 -3.25501323e-01 -3.06233108e-01
-6.91203594e-01 -5.19653618e-01 6.67777479e-01 5.31123102e-01
-6.12262011e-01 7.98867404e-01 -5.66294372e-01 -9.98654962e-02
-5.26111722e-01 -7.14992464e-01 -1.03355372e+00 -3.64507198e-01
-1.69076174e-01 1.23327708e+00 6.33621752e-01 -1.34274468e-01
3.76244754e-01 9.35503125e-01 -2.12662244e+00 5.54221928e-01
8.54444504e-01 8.74713957e-01 4.02475536e-01 6.81227326e-01
-8.23391557e-01 6.36008620e-01 -5.13038397e-01 -3.12205106e-01
3.24109048e-01 7.43654892e-02 -1.09944308e+00 2.89510906e-01
1.02321196e+00 1.79633021e-01 1.50412902e-01 8.10130954e-01
3.84610504e-01 -2.20764317e-02 3.86066079e-01 -1.17327058e+00
-4.03826177e-01 6.12793684e-01 -4.21122938e-01 2.17452079e-01
2.62378275e-01 2.74860471e-01 6.99968636e-01 -1.07290804e+00
7.65924692e-01 1.29258561e+00 9.09630954e-01 3.53810549e-01
-1.34364271e+00 -5.13962507e-01 4.29964364e-01 6.91143930e-01
-1.32205403e+00 -9.29654227e-04 8.02781641e-01 -4.53411788e-01
1.15950119e+00 1.26373935e-02 1.06957889e+00 8.95461261e-01
-1.11173145e-01 1.07909036e+00 1.42617762e+00 -5.55308938e-01
4.54496473e-01 3.57988000e-01 8.42828527e-02 5.89375436e-01
2.08584815e-01 2.96775728e-01 -6.55875623e-01 1.02998428e-01
4.01543230e-01 -4.88796271e-02 2.11461097e-01 -7.96057045e-01
-8.12558174e-01 7.07231641e-01 3.14327747e-01 5.35578191e-01
-2.24254146e-01 2.25776255e-01 4.74915236e-01 4.93569821e-01
5.64472675e-01 2.56118298e-01 -8.62944186e-01 -1.75977618e-01
-1.53294301e+00 4.60925162e-01 9.04431045e-01 9.86499667e-01
1.27832949e+00 2.34394655e-01 1.66404411e-01 7.38663018e-01
-9.93882567e-02 4.11875099e-01 4.10918176e-01 -9.65071559e-01
1.78018123e-01 4.43736941e-01 -2.76058823e-01 -4.52061594e-01
-3.77056003e-01 -6.97333217e-01 -5.33024132e-01 7.37483501e-01
2.91678697e-01 1.92542687e-01 -1.20554626e+00 1.17202437e+00
3.52605402e-01 7.41665006e-01 -1.85170136e-02 4.00987804e-01
5.34981668e-01 3.75803083e-01 1.67995512e-01 -2.13620260e-01
8.72876227e-01 -1.01458073e+00 -2.96464115e-01 -3.29373896e-01
3.47569168e-01 -8.34637880e-01 1.15640473e+00 9.29558873e-01
-1.11038780e+00 -9.33875859e-01 -8.98933053e-01 -5.11446819e-02
-7.69814849e-01 -1.21705845e-01 6.33973241e-01 5.62811315e-01
-1.02025175e+00 1.05280185e+00 -1.01701045e+00 -3.93838257e-01
8.31944883e-01 4.19260055e-01 -2.60985401e-02 -1.72945693e-01
-8.59897733e-01 8.82361889e-01 6.49512947e-01 -1.89863548e-01
-9.79461372e-01 -1.21257603e+00 -2.42375731e-01 -1.43925518e-01
5.82270682e-01 -8.10015917e-01 1.33711028e+00 -1.10457945e+00
-1.52091193e+00 9.97677743e-01 7.11652935e-02 -9.96935308e-01
8.88626695e-01 -7.50047147e-01 -8.01539868e-02 1.44506961e-01
2.81327572e-02 1.06283736e+00 1.26740658e+00 -1.64540589e+00
-7.04317391e-01 -3.39764446e-01 1.18589606e-02 1.28865272e-01
-8.55033696e-02 -7.36435771e-01 -1.91861451e-01 -5.73702395e-01
8.13628808e-02 -9.24518287e-01 -1.18922718e-01 8.59745741e-02
1.86662022e-02 -1.68012604e-01 1.31250429e+00 -2.27208436e-01
8.15753818e-01 -1.88027310e+00 3.61021668e-01 -2.81881075e-02
-2.68837869e-01 5.76835573e-01 -1.18919298e-01 3.60191733e-01
-2.23837152e-01 -1.20668024e-01 -6.20420277e-01 -8.31003666e-01
-1.08297303e-01 7.05033660e-01 -5.54662526e-01 2.11658955e-01
2.40059823e-01 9.36646461e-01 -9.93276298e-01 -6.56544328e-01
5.41684270e-01 4.16036814e-01 -4.99429494e-01 -8.11117142e-02
-8.92760634e-01 3.61812681e-01 -1.39969200e-01 8.79188597e-01
9.05131936e-01 1.60730273e-01 -2.13093385e-01 -5.97201847e-02
-3.28873605e-01 -4.00996000e-01 -1.16230083e+00 2.25380588e+00
-6.04083061e-01 3.38925451e-01 -1.31234348e-01 -1.06971264e+00
8.61642838e-01 -2.10851012e-03 5.99932194e-01 -7.17271447e-01
-3.36657077e-01 5.37293196e-01 -5.37568271e-01 -4.29814070e-01
6.61702335e-01 -3.81251782e-01 2.27550194e-01 2.19805706e-02
5.11371017e-01 -7.72171736e-01 2.83652872e-01 -7.99408332e-02
8.16262066e-01 6.40059590e-01 6.66582435e-02 -5.78081086e-02
7.31161475e-01 5.22833288e-01 1.03969865e-01 9.98391688e-01
1.50887564e-01 6.39083326e-01 1.19810313e-01 -7.04658270e-01
-9.29472625e-01 -1.16784883e+00 -1.29080623e-01 8.30394268e-01
-2.22703889e-02 -2.36951888e-01 -7.20210075e-01 -6.87247217e-01
3.58375400e-01 1.05628479e+00 -4.80519086e-01 -6.37224540e-02
-6.95369422e-01 -3.59881461e-01 4.18857306e-01 2.43808240e-01
5.77975750e-01 -1.33056068e+00 -8.67430508e-01 4.31084037e-01
3.43163580e-01 -1.12897003e+00 4.12443131e-01 3.28156203e-01
-1.51796103e+00 -1.16519582e+00 -4.88659859e-01 -5.73288262e-01
1.24739192e-01 2.81287760e-01 1.52877426e+00 -6.67501464e-02
-6.33216083e-01 9.79885519e-01 -4.15678442e-01 -5.83310068e-01
-3.86911571e-01 6.32704437e-01 -2.27732733e-01 -2.16427818e-01
2.23395169e-01 -9.13336039e-01 -3.14245522e-01 -5.34214750e-02
-1.18864071e+00 -2.78813958e-01 8.22068810e-01 7.52563953e-01
1.01636422e+00 2.17836976e-01 4.32698101e-01 -1.26240206e+00
1.93860486e-01 -2.74819434e-01 -7.72831321e-01 1.56763688e-01
-8.30557287e-01 -1.43006165e-02 5.69210708e-01 -4.04816836e-01
-1.05179524e+00 3.95816177e-01 -4.61779475e-01 -9.59195077e-01
-5.59358358e-01 2.13750824e-01 1.42010659e-01 -5.02755165e-01
6.13804162e-01 2.57825345e-01 -8.81100670e-02 -6.78791702e-01
7.64505267e-01 7.38601089e-02 6.74226999e-01 -7.44750202e-01
9.46317732e-01 8.19155931e-01 2.33144745e-01 -9.83624935e-01
-1.13034725e+00 -6.06848121e-01 -1.12167025e+00 -2.01888725e-01
6.68676078e-01 -7.72247493e-01 -2.59908289e-01 7.55729079e-01
-1.19543016e+00 -5.23141623e-01 -1.16553760e+00 7.38201141e-02
-1.05658817e+00 4.34304684e-01 -2.12327972e-01 -5.57360470e-01
-1.68130070e-01 -9.44772899e-01 1.21151793e+00 -8.24038312e-02
1.86321720e-01 -1.03848124e+00 2.52336413e-01 2.76340306e-01
4.48433220e-01 1.90091178e-01 7.78112411e-01 -4.11339194e-01
-1.00257301e+00 5.99887148e-02 1.06724620e-01 5.76477051e-01
-2.02701971e-01 -1.69365346e-01 -1.13200438e+00 -4.76317436e-01
2.26791948e-02 -5.64336538e-01 1.29537928e+00 1.18428648e-01
1.32490957e+00 -2.77652941e-03 -3.29382628e-01 6.36316001e-01
1.75155890e+00 -2.46800855e-01 7.38341570e-01 5.96491516e-01
6.90578103e-01 4.17742461e-01 9.32186544e-01 2.99959540e-01
1.37890369e-01 5.52188158e-01 7.51737177e-01 1.47757351e-01
-5.91979682e-01 -1.90296814e-01 -7.98585638e-02 5.33314943e-01
1.15011781e-01 1.25774056e-01 -1.04978037e+00 7.79493153e-01
-1.71878588e+00 -8.73760939e-01 -7.26100337e-03 2.01829195e+00
7.00309753e-01 5.37158668e-01 -8.99189487e-02 3.50233972e-01
3.29384357e-01 1.66237280e-01 -5.82189262e-01 -4.53940839e-01
-1.15770467e-01 8.24117661e-01 5.01382291e-01 6.19238555e-01
-1.14439583e+00 1.42621553e+00 5.75184011e+00 1.34958470e+00
-1.28787577e+00 4.65493411e-01 2.26087183e-01 -2.83600949e-02
-2.27319434e-01 2.94513136e-01 -9.69299078e-01 2.19923794e-01
7.06579506e-01 2.98579335e-01 4.43058431e-01 1.01451230e+00
-1.89167142e-01 -3.01210821e-01 -8.48229587e-01 9.50003445e-01
9.56754237e-02 -1.43034875e+00 2.02611730e-01 -3.47611785e-01
9.87429857e-01 3.48407924e-01 1.52254015e-01 5.42488813e-01
-4.73899059e-02 -6.82805836e-01 9.07539427e-01 8.88570905e-01
6.45614624e-01 -8.51144135e-01 4.19685453e-01 4.69431728e-01
-1.18917489e+00 -3.95840883e-01 -4.28535312e-01 2.10183442e-01
3.07701051e-01 1.04364920e+00 -8.24234962e-01 7.64849424e-01
9.42943215e-01 8.29434574e-01 -9.03421104e-01 1.23589921e+00
-1.29199371e-01 5.55629969e-01 -3.76551926e-01 4.29019570e-01
5.16760468e-01 -6.48205653e-02 7.91300118e-01 1.43998361e+00
3.43581617e-01 -4.93016839e-01 2.68176258e-01 7.79636264e-01
1.67744055e-01 4.86200638e-02 -5.98061204e-01 1.35284901e-01
3.78121257e-01 9.92118895e-01 -8.75043631e-01 -4.75535274e-01
-1.91506296e-01 1.01696765e+00 4.80766982e-01 4.38115373e-02
-6.00512981e-01 9.69084650e-02 3.26874673e-01 3.35578650e-01
8.49166691e-01 -4.56426978e-01 -3.91164750e-01 -7.36665964e-01
-1.05673686e-01 -5.93040586e-01 3.83164108e-01 -6.81384385e-01
-1.24669874e+00 3.84852320e-01 3.94736320e-01 -1.06518137e+00
-4.09883186e-02 -4.32560116e-01 -3.85504663e-01 2.39064261e-01
-2.10924602e+00 -1.44209731e+00 -4.55790281e-01 5.63587427e-01
9.81968582e-01 -3.69130135e-01 7.22860038e-01 3.46142590e-01
2.06237942e-01 2.60973662e-01 1.00905679e-01 -8.48907650e-01
4.76600170e-01 -1.41077995e+00 3.11612368e-01 4.22742486e-01
3.94853920e-01 2.21941590e-01 6.39834821e-01 -6.83319688e-01
-1.30451822e+00 -1.36828578e+00 6.36270404e-01 -2.41421252e-01
6.46433949e-01 -2.45252088e-01 -1.22700512e+00 4.17242646e-01
1.27298594e-01 1.12770021e-01 1.93281457e-01 -1.01535380e-01
-2.97703892e-01 -4.04853016e-01 -1.26314402e+00 1.01500541e-01
1.18515694e+00 -4.38225329e-01 -6.36357486e-01 5.29880464e-01
7.59133101e-01 -5.26244164e-01 -6.27356768e-01 5.92895985e-01
2.49585852e-01 -1.25374293e+00 1.36052358e+00 -2.70904124e-01
7.43467733e-02 -3.03247660e-01 1.34571940e-01 -1.19770968e+00
1.34352846e-02 -4.86774802e-01 -3.97712469e-01 1.15278327e+00
6.29119128e-02 -2.46168986e-01 1.14607620e+00 -4.59740072e-01
-4.06355470e-01 -7.80675709e-01 -1.25520134e+00 -9.28411603e-01
3.73576820e-01 -7.19135165e-01 4.63351011e-01 6.99392915e-01
-9.90658998e-01 3.09673436e-02 -2.12443531e-01 1.50535570e-03
7.39750028e-01 3.24999154e-01 7.98911810e-01 -1.52387118e+00
-5.03832102e-01 -4.69085008e-01 -4.93633032e-01 -8.40737879e-01
3.66394669e-01 -9.08393681e-01 -2.88977116e-01 -1.30732191e+00
-2.92789787e-01 -5.77595711e-01 3.34783085e-02 4.16434556e-01
1.02750875e-01 2.43150890e-01 4.71243799e-01 2.15244636e-01
-7.77034402e-01 5.66565096e-01 1.05894160e+00 -4.75639373e-01
-2.22309232e-01 1.08377323e-01 -1.74726531e-01 6.52786255e-01
9.11620319e-01 -6.56271815e-01 -6.06651664e-01 -4.11217213e-01
3.26784730e-01 -3.80209178e-01 6.72055304e-01 -1.64086437e+00
5.39043903e-01 1.50302052e-01 -2.51441728e-02 -1.14223278e+00
5.28766632e-01 -1.05636299e+00 2.46270731e-01 5.25043011e-01
-8.29761289e-03 -6.51947632e-02 6.26035392e-01 8.05233717e-01
-3.05739343e-01 -4.43657607e-01 9.33468819e-01 -7.66139269e-01
-1.25815821e+00 2.45591179e-01 -9.82050374e-02 7.38346651e-02
1.08892953e+00 -5.25635064e-01 1.51573315e-01 1.29224107e-01
-1.24950397e+00 8.89296159e-02 4.53891248e-01 3.41554731e-01
8.47841144e-01 -8.77716243e-01 -6.30224049e-01 1.09790526e-01
2.17048149e-03 4.21995997e-01 4.52031225e-01 6.33753061e-01
-6.41967773e-01 3.37028354e-01 -2.85983562e-01 -1.06782305e+00
-1.21868420e+00 8.15450549e-01 3.30447376e-01 -4.38022345e-01
-6.30579710e-01 8.67149711e-01 -4.10551250e-01 -6.55985534e-01
3.03603083e-01 -2.15097442e-01 -3.43196429e-02 4.27720994e-01
-3.21044736e-02 6.74398184e-01 5.79279721e-01 -2.14493036e-01
-1.12838089e-01 1.06547952e+00 -2.26745069e-01 8.57911706e-02
1.39799201e+00 -9.82387066e-02 -1.43676907e-01 7.44605482e-01
9.10019219e-01 -4.65996861e-01 -1.38531280e+00 -3.02830666e-01
3.36573243e-01 -5.50270021e-01 -1.53086232e-02 -9.52886879e-01
-1.06777704e+00 1.18322766e+00 9.72265363e-01 1.32349497e-02
9.37399864e-01 -1.22011766e-01 6.64938271e-01 6.83038294e-01
8.24893296e-01 -1.32383835e+00 3.98679942e-01 7.84636199e-01
8.27845395e-01 -1.04789400e+00 2.80051827e-01 -3.70602429e-01
-3.81445825e-01 1.26220429e+00 3.91299158e-01 -2.82835215e-01
8.16072941e-01 3.24128658e-01 -1.33951396e-01 -1.88902110e-01
-5.99512517e-01 -5.47990382e-01 -6.98045790e-02 6.92452431e-01
-4.16657142e-02 -1.62553534e-01 -1.93933368e-01 -2.03722060e-01
-1.38745621e-01 5.24973512e-01 2.38816112e-01 1.24236858e+00
-4.72908705e-01 -1.49264002e+00 -3.30707282e-01 2.26898044e-01
-5.58430254e-02 4.67071831e-02 -3.67894888e-01 1.11791527e+00
7.51928329e-01 4.25447315e-01 -6.89786300e-02 -2.07629949e-02
5.81569672e-01 3.11231375e-01 8.28483939e-01 -7.62299538e-01
-7.99797356e-01 -1.67103007e-01 -2.59327352e-01 -5.15955627e-01
-6.28503561e-01 -1.03765535e+00 -1.24946833e+00 7.21026435e-02
-2.72703111e-01 -2.20631976e-02 7.59289801e-01 8.42746019e-01
2.99134135e-01 5.56467593e-01 2.68351763e-01 -1.21363139e+00
-4.97978836e-01 -6.73007905e-01 -4.50221926e-01 3.37134212e-01
3.03802699e-01 -8.63014400e-01 -1.46392047e-01 3.29824947e-02] | [9.43087100982666, 1.8867688179016113] |
95a92423-ecad-4b1d-abfe-a97eee18df73 | investigating-math-word-problems-using | 2105.08928 | null | https://arxiv.org/abs/2105.08928v3 | https://arxiv.org/pdf/2105.08928v3.pdf | Investigating Math Word Problems using Pretrained Multilingual Language Models | In this paper, we revisit math word problems~(MWPs) from the cross-lingual and multilingual perspective. We construct our MWP solvers over pretrained multilingual language models using sequence-to-sequence model with copy mechanism. We compare how the MWP solvers perform in cross-lingual and multilingual scenarios. To facilitate the comparison of cross-lingual performance, we first adapt the large-scale English dataset MathQA as a counterpart of the Chinese dataset Math23K. Then we extend several English datasets to bilingual datasets through machine translation plus human annotation. Our experiments show that the MWP solvers may not be transferred to a different language even if the target expressions have the same operator set and constants. But for both cross-lingual and multilingual cases, it can be better generalized if problem types exist on both source language and target language. | ['Jing Jiang', 'Lingxiao Jiang', 'Lei Wang', 'Minghuan Tan'] | 2021-05-19 | null | null | null | null | ['pretrained-multilingual-language-models'] | ['natural-language-processing'] | [-3.28880012e-01 -2.57151455e-01 -1.80349603e-01 -3.59073132e-01
-1.06194532e+00 -1.28961217e+00 4.41373885e-01 -1.37007469e-02
-6.34179413e-01 1.23202360e+00 4.59296331e-02 -7.44596899e-01
1.67054012e-02 -8.84026051e-01 -1.10039961e+00 -2.42594570e-01
1.57853931e-01 7.19820857e-01 -1.02958202e-01 -6.62686884e-01
-5.17745689e-02 2.72099525e-01 -7.71154284e-01 5.97881496e-01
1.19372857e+00 3.22201401e-01 3.23118478e-01 4.64502722e-01
-6.34349227e-01 6.28494620e-01 -3.71640086e-01 -1.00302374e+00
8.39817703e-01 -1.62919432e-01 -1.14881742e+00 -4.00778651e-01
5.20892203e-01 2.79106885e-01 -1.22680113e-01 1.25895798e+00
2.83691674e-01 -1.56487569e-01 3.20544690e-01 -1.31319940e+00
-9.18408215e-01 1.15672088e+00 -4.81880069e-01 -4.21227627e-02
6.62096262e-01 1.81234285e-01 1.25743032e+00 -8.76605213e-01
9.83371139e-01 1.42773843e+00 6.43779218e-01 1.69944286e-01
-1.20965147e+00 -7.40337968e-01 3.12455118e-01 4.38796341e-01
-1.67130125e+00 1.09114908e-01 3.34774494e-01 -3.59155744e-01
1.40190518e+00 2.90794875e-02 3.87572110e-01 1.00444496e+00
3.07221293e-01 6.87999427e-01 1.38482237e+00 -5.58720827e-01
-3.30933869e-01 3.04073691e-01 1.24768808e-01 8.45727086e-01
8.15109834e-02 -1.32873461e-01 -5.87218702e-01 -4.57677618e-03
4.48073089e-01 -6.90043390e-01 -3.33278149e-01 -1.04491800e-01
-1.68434930e+00 8.42459202e-01 9.59019512e-02 5.22241652e-01
1.90798596e-01 -7.49563426e-02 5.61863542e-01 9.35379028e-01
1.00102916e-01 7.80838788e-01 -1.05992270e+00 -2.00537771e-01
-6.52724385e-01 2.88052768e-01 1.03900349e+00 1.46441412e+00
1.06713688e+00 -9.79794636e-02 2.52949953e-01 6.77881241e-01
-1.51351243e-01 6.63098991e-01 6.57625496e-01 -7.99766779e-01
1.14418316e+00 4.62220192e-01 -6.78924993e-02 -6.69090509e-01
-2.57334203e-01 -3.63133669e-01 -3.87252927e-01 -5.77502608e-01
6.36297584e-01 -2.32784554e-01 -4.16859865e-01 1.94988501e+00
7.06049129e-02 -9.42248199e-03 4.48967367e-01 6.69504881e-01
6.07543230e-01 9.74098861e-01 -1.15962811e-01 -1.94771975e-01
1.29915726e+00 -1.31361961e+00 -4.50714320e-01 -2.81467795e-01
1.23213720e+00 -9.92701828e-01 1.29164255e+00 4.48511064e-01
-1.20551693e+00 -3.66310298e-01 -9.40754116e-01 -4.48564082e-01
-7.17457592e-01 1.49193525e-01 5.01167536e-01 4.69637513e-01
-1.04420888e+00 3.06112438e-01 -5.72828412e-01 -4.39282477e-01
-4.08208907e-01 3.30432087e-01 -5.25649309e-01 -4.37816978e-01
-1.59709144e+00 1.38597345e+00 7.37477601e-01 -1.25121102e-01
-5.16723394e-01 -9.62539852e-01 -1.01579976e+00 -2.15169191e-01
4.10670221e-01 -4.79605883e-01 1.23176706e+00 -1.11695445e+00
-1.49369764e+00 9.45025444e-01 -9.95144248e-02 -3.59309047e-01
5.61000168e-01 -1.69862196e-01 -4.94159162e-01 -3.66511971e-01
3.62014800e-01 5.41685641e-01 2.40314811e-01 -8.16194773e-01
-7.90113568e-01 -3.07221059e-02 3.72411638e-01 2.48017222e-01
7.01182708e-02 4.50198114e-01 -6.50602281e-01 -7.30466008e-01
-2.51351148e-01 -9.83526289e-01 3.12389471e-02 -5.18022358e-01
-2.92259097e-01 -1.63394004e-01 1.73587874e-01 -7.87621021e-01
1.09621990e+00 -1.86217153e+00 7.68765390e-01 2.80093014e-01
-4.09948498e-01 5.62559217e-02 -6.22772276e-01 8.19874287e-01
-3.39351714e-01 1.55787945e-01 -2.86567271e-01 -1.03831008e-01
2.14871168e-01 6.26247823e-01 -3.56868416e-01 3.90581548e-01
2.79583335e-01 1.15260696e+00 -9.05937254e-01 -5.38614988e-01
-2.45688468e-01 -7.31873810e-02 -6.51930571e-01 -1.26155198e-01
-2.89357662e-01 3.09217811e-01 -2.86788672e-01 6.22271001e-01
6.35681093e-01 1.47767484e-01 6.61516666e-01 -5.84469102e-02
-3.18472743e-01 4.96457040e-01 -1.29078889e+00 2.23264003e+00
-9.34462309e-01 4.62425411e-01 4.33659106e-02 -1.14217997e+00
4.67887133e-01 1.41409695e-01 1.59725204e-01 -8.36481810e-01
-1.50072798e-01 6.41146302e-01 3.46194863e-01 -4.08229619e-01
6.47766650e-01 -2.79514462e-01 -4.37570691e-01 3.08831394e-01
3.62768233e-01 -2.33816326e-01 7.71520078e-01 1.29421830e-01
7.12140560e-01 4.11552399e-01 3.17088336e-01 -7.66059816e-01
8.16521108e-01 2.49452367e-01 7.43327916e-01 5.23446977e-01
2.55823165e-01 1.57529712e-01 6.91005111e-01 -3.50727528e-01
-9.42722142e-01 -1.10646176e+00 -3.14131528e-01 1.26461124e+00
-1.12527058e-01 -7.68036127e-01 -5.78591704e-01 -6.68236494e-01
-1.99425563e-01 8.50615442e-01 -1.78152293e-01 3.00274134e-01
-1.08467925e+00 -7.91678071e-01 9.93204415e-01 3.01091671e-01
3.19123209e-01 -5.84914744e-01 1.30762711e-01 3.25584233e-01
-4.23698395e-01 -1.55988932e+00 -4.68729049e-01 1.87474295e-01
-3.31906259e-01 -8.53494406e-01 -4.09982532e-01 -1.10137403e+00
3.44274998e-01 -3.75651389e-01 1.41274345e+00 -2.11978137e-01
-4.81572337e-02 1.98342666e-01 -2.76229173e-01 -2.69566387e-01
-5.27518213e-01 4.72123712e-01 3.43113363e-01 -4.57475394e-01
3.24314445e-01 -4.08988535e-01 1.18463397e-01 1.31317139e-01
-7.91738987e-01 1.45485684e-01 2.06145465e-01 6.54681981e-01
4.11026001e-01 -2.07568020e-01 2.81577855e-01 -9.50161278e-01
5.29219568e-01 -4.78622258e-01 -1.05551910e+00 8.44026387e-01
-2.38953665e-01 3.45354110e-01 9.74694312e-01 -4.28620934e-01
-8.92619491e-01 -3.81272621e-02 1.72955706e-03 2.10736170e-02
1.46363035e-01 8.80677402e-01 -4.77640688e-01 -1.57115385e-02
4.56333399e-01 2.53106356e-01 -7.14502037e-01 -3.62807810e-01
7.04726338e-01 2.71660447e-01 3.18503976e-01 -1.44951570e+00
9.79806483e-01 -1.88449800e-01 -1.72209918e-01 -6.34319901e-01
-5.86398780e-01 -1.50399864e-01 -8.22685957e-01 4.07216430e-01
9.12178576e-01 -1.12007010e+00 -3.84382427e-01 3.17216396e-01
-1.55723023e+00 -4.87099588e-01 -2.48481482e-02 5.30855536e-01
-3.63599956e-01 4.44169223e-01 -9.48683381e-01 -1.08928025e-01
-7.79654905e-02 -1.65316129e+00 9.04159307e-01 -3.06958526e-01
-1.28252715e-01 -1.33750272e+00 2.26887420e-01 1.99427456e-01
1.59212753e-01 -3.09022129e-01 1.55871546e+00 -7.28169918e-01
-5.15783846e-01 1.70945793e-01 -2.42191657e-01 2.36384332e-01
-2.08360776e-02 -4.77661155e-02 -2.48772576e-01 -2.97178715e-01
-2.77909756e-01 -5.28061450e-01 2.99213827e-01 -1.65910408e-01
6.70633018e-01 -3.89558434e-01 9.88072678e-02 7.34420538e-01
1.84193587e+00 -9.20981169e-02 4.71193790e-01 6.12560093e-01
7.38205373e-01 7.04511046e-01 4.69062597e-01 -2.67490119e-01
6.94816172e-01 6.26470625e-01 -2.67549515e-01 6.21092804e-02
2.71710247e-01 -2.54579842e-01 7.89317012e-01 1.56284761e+00
1.26071513e-01 -1.54913411e-01 -1.36442959e+00 6.50350571e-01
-1.75161660e+00 -4.62095141e-01 -1.58089682e-01 1.97586727e+00
1.32121515e+00 -3.72318298e-01 -1.09705217e-01 -5.93753397e-01
3.96010756e-01 -3.14003587e-01 -3.91626470e-02 -7.69302964e-01
-4.38957900e-01 6.25691891e-01 7.66363740e-01 8.31714869e-01
-6.82407677e-01 1.56860590e+00 6.33061886e+00 1.16289794e+00
-1.22295797e+00 3.01416427e-01 2.25073308e-01 -5.87970130e-02
-3.80833507e-01 1.06692187e-01 -8.55277896e-01 1.62228212e-01
1.05713248e+00 -4.38321739e-01 8.20470273e-01 5.16042054e-01
-2.55677432e-01 1.71480581e-01 -1.52618861e+00 9.91002917e-01
1.49531662e-02 -1.15994370e+00 9.84115377e-02 -2.30173230e-01
9.84750688e-01 3.74534816e-01 -1.95099846e-01 6.49147689e-01
6.83606088e-01 -1.20810318e+00 7.96050251e-01 5.25570139e-02
7.47789443e-01 -8.14667940e-01 6.94256783e-01 4.39383864e-01
-1.35111356e+00 2.60193288e-01 -5.01987815e-01 -1.00241303e-01
2.05944642e-01 1.15257846e-02 -4.12126273e-01 1.12853765e+00
5.32697082e-01 6.68809772e-01 -5.07033765e-01 3.76413882e-01
-4.51139688e-01 3.05945903e-01 -2.84439951e-01 2.30428964e-01
5.96579969e-01 -7.77332544e-01 4.26277637e-01 1.39002645e+00
5.09126365e-01 -6.77852407e-02 5.18000960e-01 7.22387552e-01
-1.15831062e-01 7.40634918e-01 -6.16366029e-01 -1.27548799e-01
1.66529834e-01 1.09122550e+00 -3.02590251e-01 -4.71657366e-01
-8.95860910e-01 1.03131258e+00 7.38141954e-01 4.95030880e-01
-1.03588092e+00 -2.22974062e-01 6.41521215e-01 -2.29887649e-01
4.77372855e-03 -5.81342876e-01 -8.69443789e-02 -1.79716516e+00
1.33484021e-01 -1.54493666e+00 3.28706920e-01 -6.28343046e-01
-1.50730371e+00 6.63682222e-01 2.80579209e-01 -8.80503356e-01
-2.20430776e-01 -1.05419683e+00 -1.54456526e-01 1.17311585e+00
-1.69281328e+00 -1.30588937e+00 5.36870062e-01 7.62402534e-01
5.83034992e-01 -2.33113393e-01 9.17907000e-01 5.86423635e-01
-5.66770077e-01 7.06957519e-01 1.68595627e-01 3.35208297e-01
1.02126694e+00 -1.25152707e+00 3.86963755e-01 1.03299248e+00
3.97430509e-01 1.14065051e+00 5.33853173e-01 -5.45307338e-01
-1.69869328e+00 -1.18036938e+00 1.24831831e+00 -5.06630361e-01
1.42622793e+00 -4.78599757e-01 -8.11667383e-01 1.24732637e+00
7.23288476e-01 -3.73454690e-01 6.05626523e-01 2.38917783e-01
-5.60649872e-01 1.20335855e-01 -8.05304170e-01 7.04881370e-01
1.03288579e+00 -7.49056041e-01 -7.60743022e-01 6.20692194e-01
8.76345098e-01 -7.06556916e-01 -1.16413856e+00 1.73870802e-01
3.09413791e-01 -2.84910798e-01 7.80336857e-01 -1.10403585e+00
6.68891788e-01 -3.94241810e-01 -6.43147707e-01 -1.52760422e+00
-1.74921490e-02 -4.56664085e-01 5.49715877e-01 1.24870253e+00
1.01759470e+00 -8.05147767e-01 7.22609460e-02 4.72093582e-01
-1.45960167e-01 -5.19603133e-01 -1.01310706e+00 -1.22732699e+00
1.07850730e+00 -8.14706922e-01 6.58229411e-01 1.31291974e+00
2.73779631e-01 5.09915471e-01 -3.48013371e-01 2.13912189e-01
1.77815005e-01 3.98662686e-01 6.82485700e-01 -6.35029852e-01
-6.89452291e-01 -2.35480964e-01 -1.44465178e-01 -8.73619497e-01
7.69741535e-01 -1.54609084e+00 -6.25860319e-02 -1.08032835e+00
1.69580519e-01 -3.97799700e-01 1.98335834e-02 6.36996150e-01
1.07574359e-01 1.46629348e-01 3.76593381e-01 -1.40495837e-01
-3.82552236e-01 2.91798919e-01 1.05281639e+00 -2.07271084e-01
4.70254235e-02 -6.60925806e-01 -5.10582685e-01 6.98333502e-01
7.38589525e-01 -4.42193687e-01 -2.33654991e-01 -1.13736749e+00
7.67603278e-01 4.64023240e-02 -2.50375837e-01 -4.56782013e-01
7.68627375e-02 -5.85305691e-01 -2.33204648e-01 -9.16880146e-02
-2.05429252e-02 -7.57990420e-01 1.23440273e-01 3.63921076e-01
-2.27505103e-01 7.21873760e-01 5.16659319e-01 1.83693334e-01
-3.96411836e-01 -2.94833928e-01 5.37450850e-01 -4.85116571e-01
-6.82188928e-01 5.44538870e-02 -4.40436572e-01 3.93753827e-01
9.63639081e-01 1.86363280e-01 -5.37879169e-01 -3.47269885e-02
-4.46197778e-01 4.52071965e-01 6.73850954e-01 5.03002346e-01
3.05826627e-02 -1.20995450e+00 -7.99620807e-01 9.63477641e-02
2.56864101e-01 -3.38285297e-01 -1.58924371e-01 9.57906961e-01
-8.17562461e-01 7.90203869e-01 -2.15590134e-01 -2.41650686e-01
-1.03807151e+00 6.86856389e-01 4.75462407e-01 -6.12991512e-01
-3.80648337e-02 6.67284846e-01 2.83177257e-01 -1.26788425e+00
-4.93955836e-02 -6.43623412e-01 2.84184039e-01 -2.34717205e-01
1.51749775e-01 1.38170451e-01 2.64792800e-01 -8.75153124e-01
-5.31692803e-01 7.76762009e-01 -6.69000745e-02 -2.45022103e-01
1.15833604e+00 -5.58561645e-03 -6.92246139e-01 4.81146425e-01
1.23919141e+00 5.49039304e-01 -1.50627121e-01 -3.92781705e-01
1.85496092e-01 -2.60032769e-02 -5.17798364e-01 -7.49452949e-01
-9.91335869e-01 8.14534307e-01 8.17384198e-02 -5.33609807e-01
8.97066534e-01 -8.19915682e-02 8.10067296e-01 7.40937293e-01
7.39818692e-01 -1.20543694e+00 -5.48557162e-01 1.09644747e+00
8.42262387e-01 -1.02749908e+00 -1.72293112e-01 -6.31675005e-01
-7.44179964e-01 9.97533619e-01 7.76419580e-01 -2.94653960e-02
3.57740641e-01 4.96380478e-01 6.01403043e-02 7.11265355e-02
-7.62381673e-01 -1.24704793e-01 2.97833025e-01 3.47979069e-01
7.69568682e-01 2.14662045e-01 -6.71824992e-01 6.52227044e-01
-7.44403303e-01 -1.84589669e-01 5.10352075e-01 9.48604405e-01
3.34899247e-01 -1.89425826e+00 -2.72029668e-01 8.11108854e-03
-5.43330610e-01 -7.78218329e-01 -4.34002429e-01 1.13799071e+00
4.12121534e-01 5.69775522e-01 -1.09667212e-01 -8.61489922e-02
2.38399655e-01 4.05471414e-01 9.09053206e-01 -7.44972289e-01
-7.78890491e-01 -1.26117453e-01 2.80930340e-01 -5.45187891e-01
-2.71478742e-01 -4.52054203e-01 -1.21875203e+00 -5.25590777e-01
-1.87559351e-01 2.52736568e-01 5.57014704e-01 1.18844211e+00
4.06471677e-02 2.22950354e-01 1.13100797e-01 -2.38460720e-01
-3.15305710e-01 -7.51934946e-01 -4.61463958e-01 1.87654391e-01
-2.63258845e-01 -1.43328950e-01 7.89221376e-03 8.48182440e-02] | [11.021782875061035, 9.94098949432373] |
fd514388-fa22-4be9-bf18-b31ad430cb6a | a-time-resolved-clustering-method-revealing | 1912.04261 | null | https://arxiv.org/abs/1912.04261v2 | https://arxiv.org/pdf/1912.04261v2.pdf | A time resolved clustering method revealing longterm structures and their short-term internal dynamics | The last decades have not only been characterized by an explosive growth of data, but also an increasing appreciation of data as a valuable resource. Their value comes with the ability to extract meaningful patterns that are of economic, societal or scientific relevance. A particular challenge is the identification of patterns across time, including those that might only become apparent when the temporal dimension is taken into account. Here, we present a novel method that aims to achieve this by detecting dynamic clusters, i.e. structural elements that can be present over prolonged durations. It is based on an adaptive identification of majority overlaps between groups at different time points and accommodates the transient decompositions in otherwise persistent dynamic clusters. Our method enables the detection of persistent structural elements with internal dynamics and can be applied to any classifiable data, ranging from social contact networks to arbitrary sets of time stamped feature vectors. It represents a unique tool to study systems with non-trivial temporal dynamics and has a broad applicability to scientific, societal and economic data. | ['Jonas I. Liechti', 'Sebastian Bonhoeffer'] | 2019-12-09 | null | null | null | null | ['time-series-clustering'] | ['time-series'] | [ 1.98344201e-01 -5.56873977e-01 -1.00447066e-01 1.47183254e-01
1.28268823e-01 -9.81647670e-01 9.61780727e-01 7.63440132e-01
-2.32476860e-01 5.04457295e-01 -8.86075348e-02 -3.88196558e-01
-8.29771221e-01 -7.41611421e-01 -1.28259152e-01 -1.01639736e+00
-8.16124618e-01 4.87898350e-01 4.46559489e-01 -2.72478491e-01
2.42937282e-01 8.64607275e-01 -1.88711989e+00 9.62805822e-02
5.95963717e-01 8.49207640e-01 2.23018005e-02 3.66444618e-01
-1.38355672e-01 3.00236702e-01 -6.53849244e-01 1.09300777e-01
-3.21283261e-03 -1.64374098e-01 -5.02388656e-01 2.13725314e-01
-2.48788103e-01 4.98304456e-01 1.57277845e-02 6.83724821e-01
1.20047525e-01 1.48131967e-01 5.30741870e-01 -1.48624074e+00
-5.77404201e-02 3.15584123e-01 -7.65774786e-01 4.94654208e-01
3.63599032e-01 -5.11886775e-02 7.89570272e-01 -6.61500514e-01
8.19656551e-01 1.14668691e+00 6.01910174e-01 -1.74053349e-02
-1.60772252e+00 -4.46780056e-01 2.06250340e-01 4.68671441e-01
-1.32120419e+00 -2.22624898e-01 7.73359179e-01 -1.03771460e+00
7.21983671e-01 6.00540400e-01 7.58909047e-01 8.79688084e-01
2.42138416e-01 1.90393850e-01 1.05419910e+00 -3.64223748e-01
2.95231879e-01 -1.93060532e-01 2.46228516e-01 1.84404179e-01
3.21858555e-01 -6.02190532e-02 -1.92267478e-01 -4.67601776e-01
2.29388922e-01 4.65532601e-01 -6.58504069e-02 -3.31034064e-01
-1.37897503e+00 7.48106420e-01 -9.42627713e-02 1.01634324e+00
-3.64419311e-01 -1.93267867e-01 5.02111018e-01 5.85574508e-01
6.43256366e-01 3.72458547e-01 -4.36681777e-01 -3.91644388e-01
-9.20158386e-01 8.62993449e-02 7.94088781e-01 2.24138692e-01
7.65689790e-01 -2.70912141e-01 1.68513834e-01 5.03467083e-01
-2.86532342e-01 3.19554687e-01 6.32699966e-01 -5.87624311e-01
-4.10505980e-02 1.14877748e+00 1.37224704e-01 -1.39601755e+00
-6.76888943e-01 -3.08930904e-01 -1.21405578e+00 8.36082101e-02
5.08635700e-01 1.74053237e-01 -4.86278802e-01 1.75099826e+00
6.42274618e-01 1.33744091e-01 -1.32388383e-01 2.96097815e-01
2.18448684e-01 7.27078855e-01 -2.71394730e-01 -6.41130447e-01
1.25545347e+00 -3.92041095e-02 -6.44874811e-01 4.84300286e-01
4.26615685e-01 -7.24725485e-01 5.98754227e-01 3.10684323e-01
-6.33139372e-01 -2.44205818e-01 -5.87009609e-01 6.58607721e-01
-7.24277020e-01 -1.84352189e-01 4.34473336e-01 4.40628350e-01
-1.10743165e+00 8.78580630e-01 -9.15915132e-01 -7.78107882e-01
-1.76685885e-01 4.71658796e-01 -3.59851241e-01 3.17800611e-01
-9.04827297e-01 5.90632915e-01 2.37557665e-01 1.03826068e-01
-2.37374187e-01 -3.77112925e-01 -3.94571126e-01 5.09652197e-02
4.73650336e-01 5.22820977e-04 6.16228640e-01 -1.18039155e+00
-8.46980393e-01 6.92314684e-01 -3.38090569e-01 -2.95305252e-01
5.47651649e-01 5.03394425e-01 -7.51648605e-01 7.02715889e-02
1.24902941e-01 -3.27149153e-01 8.79593074e-01 -9.90665674e-01
-5.36387324e-01 -3.77007365e-01 -3.65722120e-01 -2.40007684e-01
-6.01285458e-01 1.51085511e-01 -4.82838750e-02 -7.07627535e-01
1.67352811e-01 -1.13568318e+00 -3.40615153e-01 -4.57743227e-01
-3.47002804e-01 -4.61281866e-01 1.16840780e+00 -3.41957659e-01
1.48812020e+00 -2.04518199e+00 3.85710925e-01 6.53483748e-01
3.85902703e-01 1.51814133e-01 1.61115512e-01 1.15144050e+00
-7.72493929e-02 2.77814180e-01 -3.27883929e-01 -9.17694196e-02
-2.23583356e-01 3.22237611e-01 -3.68820995e-01 7.94800699e-01
1.40035555e-01 4.67540622e-01 -9.15904760e-01 -1.31107286e-01
2.67096996e-01 1.90725699e-01 -8.53834450e-02 -2.98970304e-02
1.10691853e-01 5.50879657e-01 -4.66747433e-01 4.54274833e-01
4.60813969e-01 -4.51237291e-01 4.42893147e-01 3.52154016e-01
-7.59876013e-01 -2.79875547e-01 -1.20815289e+00 6.80707276e-01
1.66985653e-02 7.49647975e-01 1.20965652e-01 -1.38416493e+00
9.33629453e-01 3.76125574e-01 1.21571124e+00 -4.82717395e-01
1.03461966e-01 1.97525397e-01 1.70967683e-01 -4.69225287e-01
5.19922674e-01 4.64355573e-02 -2.09668979e-01 7.46416807e-01
-2.76189715e-01 5.00129521e-01 4.83690470e-01 1.44373430e-02
1.45785010e+00 -6.85806453e-01 4.05672044e-01 -6.15030408e-01
6.51639402e-01 3.55855972e-02 6.63470328e-01 3.49437803e-01
8.30656365e-02 9.81492177e-02 6.94267929e-01 -5.67675352e-01
-9.84902740e-01 -8.75504136e-01 -1.87134936e-01 8.07420313e-01
1.18697375e-01 -4.03054386e-01 -1.73250828e-02 -2.10104182e-01
2.12556064e-01 -1.68849722e-01 -9.52281833e-01 -1.18398175e-01
-5.75486243e-01 -9.93874133e-01 5.48243104e-03 -5.98711446e-02
-1.09021038e-01 -1.06560302e+00 -6.18838787e-01 4.96024102e-01
-9.53264460e-02 -9.13071156e-01 -7.33589977e-02 8.63404013e-03
-8.80049169e-01 -1.29645801e+00 -6.11192524e-01 -4.84384596e-01
7.73101449e-01 4.29503858e-01 8.90928805e-01 7.90234208e-02
-5.75125456e-01 5.96147835e-01 -5.36673427e-01 -1.73492193e-01
-5.38399398e-01 1.82496190e-01 4.07848924e-01 6.25743449e-01
2.12552294e-01 -8.73229384e-01 -5.99259079e-01 8.42571080e-01
-1.02954829e+00 -2.72029459e-01 1.64872646e-01 5.25932670e-01
4.85621601e-01 5.29895425e-01 7.00510621e-01 -5.71672380e-01
7.69403458e-01 -1.03706181e+00 -6.09666884e-01 2.46120945e-01
-7.41022170e-01 -8.31352100e-02 8.38806450e-01 -7.74580002e-01
-5.24128139e-01 -8.50280449e-02 4.48436558e-01 -2.82649666e-01
-1.14878513e-01 7.48101413e-01 1.93225563e-01 6.74874261e-02
4.37875956e-01 3.46529901e-01 1.67810276e-01 -6.20712817e-01
4.98771369e-02 5.74262500e-01 3.31567377e-01 -5.09420395e-01
8.59033465e-01 8.67984056e-01 3.45339119e-01 -1.27237439e+00
-7.96210207e-03 -9.06894326e-01 -8.70038927e-01 -4.44192111e-01
3.55636895e-01 -3.44239414e-01 -8.29811573e-01 4.14224833e-01
-7.92060792e-01 -1.71582595e-01 -4.35105830e-01 2.04705968e-01
-1.68876439e-01 4.37553793e-01 -2.84435362e-01 -8.71119976e-01
-5.12146614e-02 -5.47598600e-01 4.97718245e-01 8.48973542e-02
-5.13144553e-01 -1.19023311e+00 4.29698795e-01 -3.85233313e-01
3.28418314e-01 8.99981499e-01 8.18283558e-01 -6.39664233e-01
-3.67816746e-01 -3.88496786e-01 1.66090369e-01 -2.49081522e-01
6.02984786e-01 5.60361564e-01 -4.75446731e-01 -6.82500482e-01
-1.12047419e-01 3.91581774e-01 6.22019649e-01 3.18716139e-01
7.12484241e-01 -5.65171778e-01 -6.78074300e-01 5.43904826e-02
1.07924998e+00 4.92414653e-01 1.68880984e-01 2.48362511e-01
4.01457995e-01 1.07522368e+00 4.82201844e-01 6.57407999e-01
1.53792784e-01 8.97734523e-01 2.99643219e-01 -8.00563022e-02
4.26500678e-01 3.17364305e-01 1.96979359e-01 9.77156460e-01
-3.73766452e-01 -1.78503245e-01 -1.16407251e+00 9.32120740e-01
-2.22772360e+00 -1.40107310e+00 -4.18268800e-01 2.44505405e+00
5.09792030e-01 3.79067771e-02 5.97929537e-01 5.14641404e-01
1.06761098e+00 7.28804395e-02 -6.18499517e-01 -3.56179565e-01
-2.13933438e-01 -1.28454287e-02 2.01063499e-01 3.52117978e-02
-9.39465761e-01 4.00953323e-01 6.41581392e+00 6.88546956e-01
-1.32913053e+00 -4.18626592e-02 3.88897508e-01 -2.64769681e-02
-9.17502642e-02 -9.52535197e-02 -4.11477178e-01 7.27676630e-01
1.09059870e+00 -7.00627565e-01 2.75971830e-01 3.74227613e-01
5.91735244e-01 -2.45726749e-01 -7.95740008e-01 7.35451579e-01
-4.46222216e-01 -1.27141356e+00 -3.59835744e-01 5.40426970e-01
7.45651066e-01 -1.09228246e-01 1.28171682e-01 -2.87231475e-01
2.87975550e-01 -9.01305079e-01 4.51023072e-01 4.60632861e-01
6.62148476e-01 -7.44896173e-01 4.25842553e-01 4.96818781e-01
-1.54943240e+00 -2.13902280e-01 -1.22262649e-01 -3.43901306e-01
1.61990777e-01 9.37625170e-01 -6.61926270e-01 4.37168390e-01
6.06049955e-01 1.08482277e+00 -3.92594427e-01 1.15086854e+00
4.23196107e-01 5.61351001e-01 -6.20243549e-01 -1.23965792e-01
2.01158017e-01 -4.50901151e-01 6.75531566e-01 1.09957421e+00
7.51729250e-01 -1.90842956e-01 1.65125594e-01 4.07423139e-01
4.21494275e-01 2.00626373e-01 -8.75980437e-01 -2.41125837e-01
5.49114287e-01 1.36644757e+00 -1.39929557e+00 -1.55540273e-01
-2.82421917e-01 5.73075831e-01 5.98503239e-02 1.75681457e-01
-3.48107278e-01 -2.60901242e-01 8.43762815e-01 5.77525854e-01
2.54029393e-01 -7.00787961e-01 2.25633323e-01 -1.22500396e+00
-4.61938493e-02 -5.39659202e-01 6.17262125e-01 1.40154034e-01
-1.33302820e+00 6.49596035e-01 9.78205651e-02 -1.43718708e+00
-4.91960496e-01 -2.34693363e-01 -8.45542610e-01 5.01747251e-01
-8.98388565e-01 -7.55668581e-01 -3.10316563e-01 9.13444281e-01
5.63937053e-02 4.76322882e-02 7.49281168e-01 1.97466180e-01
-5.41543186e-01 -2.15100590e-02 7.98842728e-01 -2.44775921e-01
2.85156697e-01 -1.13394868e+00 2.82799661e-01 8.58881533e-01
6.80314973e-02 4.90961850e-01 9.86975133e-01 -5.69255412e-01
-1.20527780e+00 -1.04749656e+00 1.01609111e+00 -3.21643382e-01
1.12627745e+00 -5.83888948e-01 -9.23501194e-01 3.60152721e-01
-1.13116823e-01 -1.28145173e-01 6.65520787e-01 2.89109319e-01
-1.29791185e-01 -1.82966292e-01 -9.33760524e-01 5.33187211e-01
9.35801148e-01 -3.14405680e-01 -2.74594963e-01 1.78943932e-01
2.29485616e-01 2.09529012e-01 -9.28025484e-01 3.96864176e-01
5.58144629e-01 -9.38448012e-01 8.81054282e-01 -4.11809415e-01
2.44245678e-02 -3.29976588e-01 1.67048723e-01 -1.09798062e+00
-4.33929771e-01 -1.10602331e+00 -2.16159672e-01 1.19352937e+00
1.31320432e-01 -9.56513941e-01 5.29373169e-01 2.66123325e-01
4.73298162e-01 -5.05900860e-01 -1.35138750e+00 -1.24746525e+00
-2.19793022e-01 -2.10205819e-02 5.79085946e-01 1.25898147e+00
3.61194342e-01 -2.35371236e-02 -3.80437940e-01 -2.05627307e-01
5.43749809e-01 7.34286368e-01 7.08574474e-01 -1.86322200e+00
2.82346793e-02 -7.48089910e-01 -7.63693154e-01 -3.63947153e-01
-1.95507199e-01 -6.41861141e-01 -3.46525967e-01 -1.04647076e+00
-4.92001064e-02 -6.09817803e-01 -3.35709512e-01 2.52850980e-01
2.41674662e-01 3.02997142e-01 -3.12527791e-02 7.49280512e-01
-4.28044170e-01 2.42852330e-01 7.23415434e-01 1.34321451e-01
-5.73567331e-01 3.17220747e-01 -3.67497057e-01 5.24796128e-01
8.18373978e-01 -4.37973410e-01 -2.19710171e-01 4.86755341e-01
4.93044525e-01 -1.19344860e-01 3.50372016e-01 -1.07301688e+00
2.34813929e-01 -4.08357292e-01 -8.15493427e-03 -6.37007058e-01
6.66764006e-02 -9.58360493e-01 8.93841207e-01 7.31253386e-01
8.63154158e-02 4.50828254e-01 9.19627845e-02 8.37738931e-01
-2.70891994e-01 9.06703845e-02 5.51712811e-01 4.06426787e-02
-6.07293606e-01 4.06823635e-01 -9.03348386e-01 -3.16570252e-01
1.47035205e+00 -2.79786438e-01 -1.97235122e-01 -3.38715017e-01
-1.02504230e+00 2.37848908e-01 6.50794148e-01 6.04354620e-01
2.72858500e-01 -1.27833295e+00 -4.74483520e-01 1.10202119e-01
1.39658645e-01 -3.63992363e-01 2.92447656e-01 1.05876338e+00
-1.91651821e-01 3.46164584e-01 -1.59717292e-01 -8.15706015e-01
-1.72450149e+00 7.06270397e-01 -2.88753230e-02 -5.21704018e-01
-8.63305628e-01 2.36093085e-02 -1.10733710e-01 8.29576552e-02
-7.92071074e-02 -3.15924883e-01 -5.41905642e-01 8.03586006e-01
5.56338310e-01 5.42441607e-01 -1.07197436e-02 -9.89838123e-01
-4.32470262e-01 7.70775676e-01 1.36704668e-01 1.02562666e-01
1.65105069e+00 -3.71877134e-01 -3.62463772e-01 9.85128999e-01
1.19830537e+00 6.17909990e-02 -1.03801000e+00 -3.02766979e-01
5.26285172e-01 -2.97850817e-01 -4.98364300e-01 -2.30657980e-01
-9.35360193e-01 4.93895531e-01 4.07339573e-01 1.34960127e+00
1.08128524e+00 3.24756950e-01 4.93247420e-01 8.41591135e-02
4.02441651e-01 -8.19054842e-01 1.15081124e-01 4.58262652e-01
7.73269653e-01 -8.37531269e-01 -1.28457025e-01 -2.74617016e-01
-7.18266144e-02 1.25945640e+00 -1.80573642e-01 -1.51516423e-01
9.86970425e-01 7.44279698e-02 -3.76605421e-01 -4.13424581e-01
-8.23038220e-01 -3.39149922e-01 1.62479401e-01 4.43863809e-01
-1.93403047e-02 2.26038098e-01 -7.53785491e-01 2.44532749e-01
-5.65378070e-02 -2.79863358e-01 6.45838082e-01 1.13671827e+00
-3.83197874e-01 -1.19873285e+00 -4.33125645e-01 5.77849686e-01
-5.54567456e-01 4.56475288e-01 -6.19426131e-01 7.16001809e-01
1.10996909e-01 9.90161717e-01 3.59525651e-01 -5.27011156e-01
2.16882005e-01 -1.25308316e-02 6.62002116e-02 -3.13266784e-01
-5.26064754e-01 2.66773775e-02 -1.24810725e-01 -5.80553889e-01
-5.82953870e-01 -1.14800191e+00 -7.72277832e-01 -7.37381518e-01
-1.91144049e-01 3.66783857e-01 6.20757580e-01 7.30648518e-01
6.17212117e-01 2.98213273e-01 1.13824165e+00 -8.35549235e-01
1.23879880e-01 -7.38268554e-01 -7.64249742e-01 6.48058116e-01
5.38740933e-01 -8.13116074e-01 -7.05054283e-01 5.50331064e-02] | [7.269399642944336, 3.4617550373077393] |
311e65bd-d97f-4a78-a5af-2ccb10b9a39c | visual-robot-task-planning | 1804.00062 | null | http://arxiv.org/abs/1804.00062v1 | http://arxiv.org/pdf/1804.00062v1.pdf | Visual Robot Task Planning | Prospection, the act of predicting the consequences of many possible futures,
is intrinsic to human planning and action, and may even be at the root of
consciousness. Surprisingly, this idea has been explored comparatively little
in robotics. In this work, we propose a neural network architecture and
associated planning algorithm that (1) learns a representation of the world
useful for generating prospective futures after the application of high-level
actions, (2) uses this generative model to simulate the result of sequences of
high-level actions in a variety of environments, and (3) uses this same
representation to evaluate these actions and perform tree search to find a
sequence of high-level actions in a new environment. Models are trained via
imitation learning on a variety of domains, including navigation,
pick-and-place, and a surgical robotics task. Our approach allows us to
visualize intermediate motion goals and learn to plan complex activity from
visual information. | ['Yotam Barnoy', 'Raman Arora', 'Chris Paxton', 'Kapil Katyal', 'Gregory D. Hager'] | 2018-03-30 | null | null | null | null | ['robot-task-planning'] | ['robots'] | [ 3.87128532e-01 4.54924911e-01 7.95759410e-02 -2.59795964e-01
-8.10441747e-03 -4.72665220e-01 9.48436081e-01 3.06934398e-02
-4.83297884e-01 8.50819647e-01 5.68476319e-01 -3.13407928e-01
-1.19377293e-01 -8.68503213e-01 -6.61265075e-01 -6.79580927e-01
-4.08321708e-01 5.54213345e-01 2.34999165e-01 -2.97514081e-01
6.13475621e-01 6.61272943e-01 -1.36333978e+00 2.61908829e-01
5.58374226e-01 4.45266545e-01 8.37986171e-01 5.44027686e-01
1.16701104e-01 1.01835048e+00 -2.28844553e-01 4.51041721e-02
1.74682498e-01 -9.87552822e-01 -1.04149783e+00 6.59647062e-02
-4.35431451e-01 -3.54682684e-01 -3.46386731e-01 8.48671436e-01
1.75770566e-01 3.00012738e-01 4.99820888e-01 -8.79102945e-01
-2.82537639e-01 5.63468039e-01 2.48468891e-01 -5.47782471e-03
6.69741511e-01 5.60987711e-01 5.32596350e-01 -4.09501612e-01
1.15714204e+00 1.12224913e+00 -6.14948124e-02 8.08511317e-01
-1.17570710e+00 -1.43265417e-02 -6.87657744e-02 1.75863892e-01
-6.78164423e-01 -2.36635894e-01 5.59569299e-01 -3.54732931e-01
9.74459589e-01 -3.34072948e-01 1.27358341e+00 1.11028922e+00
9.31591034e-01 5.73274970e-01 1.16196668e+00 -3.86103094e-01
6.27629638e-01 -6.17888749e-01 -5.51614761e-01 8.58454525e-01
-2.13327616e-01 6.11587465e-01 -7.27711618e-01 1.06357157e-01
1.07434714e+00 1.96288712e-02 -2.45474324e-01 -5.26289880e-01
-1.60448229e+00 4.85088736e-01 6.94140494e-01 4.83853847e-01
-8.01160216e-01 5.44698119e-01 -1.39178202e-01 1.88712955e-01
-4.28100914e-01 1.10976624e+00 -2.07882971e-01 -4.82067131e-02
-4.82117087e-01 3.53945017e-01 7.77152777e-01 3.46757114e-01
7.42222846e-01 7.49822930e-02 -5.36417663e-02 1.99692354e-01
2.64436334e-01 -3.24665988e-03 7.58725226e-01 -1.48635387e+00
1.23996045e-02 5.76133668e-01 1.75844863e-01 -5.41674197e-01
-7.27624118e-01 -1.03175528e-01 -2.67016977e-01 9.73071218e-01
3.47380191e-01 -2.01853976e-01 -1.14774096e+00 1.82230246e+00
1.43411249e-01 1.23676136e-02 1.99576765e-01 8.84797454e-01
3.01618069e-01 5.09473324e-01 2.23197088e-01 -1.55942962e-01
8.79285634e-01 -7.65821397e-01 -3.53308171e-01 -5.63023746e-01
6.61825359e-01 -5.75171225e-02 7.11966813e-01 4.12526309e-01
-1.07817924e+00 -2.73351997e-01 -9.00633931e-01 -1.61914062e-02
-2.75182158e-01 -3.00847769e-01 8.35219026e-01 -1.42791331e-01
-1.44476104e+00 1.18117356e+00 -1.03624940e+00 -6.30626798e-01
3.73315305e-01 3.54858905e-01 -4.94524300e-01 1.89000219e-01
-8.14977407e-01 1.41015089e+00 6.36952579e-01 -3.37455198e-02
-1.52658999e+00 -6.10026009e-02 -7.16940165e-01 1.64902732e-01
1.39491677e-01 -9.24035013e-01 1.21877134e+00 -7.54922092e-01
-1.75179935e+00 7.52959192e-01 6.00770256e-03 -5.78599870e-01
1.71996519e-01 -8.29933863e-03 8.86299908e-02 1.17755115e-01
1.20120421e-01 1.24465787e+00 4.84750003e-01 -9.45782304e-01
-2.73672849e-01 -5.28272569e-01 2.59211540e-01 6.59146249e-01
4.98631865e-01 -2.95309156e-01 -1.07430227e-01 -2.35647455e-01
5.25668561e-01 -9.28383112e-01 -9.59890306e-01 1.06682055e-01
-1.13741398e-01 -1.00899987e-01 6.96001053e-02 -3.83571923e-01
4.96925920e-01 -1.82337964e+00 6.58236563e-01 -3.98140587e-02
-2.79142596e-02 -1.42487600e-01 -2.81835943e-01 7.72114456e-01
-1.74959190e-02 -1.04103453e-01 -1.92478850e-01 4.27824855e-02
-2.38501638e-01 3.74407470e-01 -3.16259414e-01 1.55809432e-01
-8.06863084e-02 1.10282516e+00 -1.23884499e+00 -1.79755315e-01
3.48827600e-01 1.30500108e-01 -5.32914042e-01 1.99493974e-01
-6.33087158e-01 1.00092304e+00 -6.56596899e-01 3.69348854e-01
-2.06996545e-01 -1.07250884e-02 5.26334047e-01 3.82658452e-01
-3.56913537e-01 5.63518822e-01 -5.30193388e-01 2.08184481e+00
-2.56715447e-01 5.12283921e-01 -4.35210109e-01 -7.65075326e-01
8.52443695e-01 3.13260585e-01 2.34539524e-01 -8.18324208e-01
3.85204941e-01 2.11402044e-01 3.41336578e-01 -6.50710881e-01
8.06860030e-02 -3.65290672e-01 -1.02600411e-01 6.57025814e-01
1.68374926e-01 -5.08487165e-01 2.66954124e-01 -2.64816545e-02
1.33447051e+00 8.87105107e-01 8.16619694e-01 9.21333432e-02
1.91183135e-01 1.49412125e-01 3.83289665e-01 7.37490296e-01
-2.27141708e-01 3.38375926e-01 6.94781482e-01 -8.77095163e-01
-8.46068978e-01 -1.15348589e+00 3.84988397e-01 8.29165518e-01
3.76484871e-01 -3.03877667e-02 -5.66448748e-01 -2.80714214e-01
-5.44338405e-01 1.25471878e+00 -7.26135671e-01 -1.98403746e-01
-7.28053689e-01 -3.04061532e-01 -1.48888960e-01 4.35425222e-01
4.16453570e-01 -2.14869547e+00 -1.47122025e+00 4.22238231e-01
-3.13749872e-02 -6.38984263e-01 3.73395123e-02 4.15876001e-01
-1.10875797e+00 -1.12431538e+00 -3.26811463e-01 -8.65840852e-01
8.21939707e-01 -1.58692300e-01 8.70933056e-01 2.96348687e-02
-2.22792163e-01 1.86717033e-01 -1.61937013e-01 -2.04211354e-01
-6.12755477e-01 -4.75820452e-01 8.44649039e-03 -4.23155159e-01
-1.71027198e-01 -1.13920927e+00 -5.24172246e-01 -3.45889828e-03
-6.00982249e-01 6.03584707e-01 8.30606401e-01 7.83180714e-01
4.99353379e-01 -1.94570079e-01 3.55468750e-01 -3.33413124e-01
6.81978345e-01 -2.41928935e-01 -5.58572292e-01 1.31191775e-01
-3.36857706e-01 4.17010427e-01 5.40873885e-01 -3.24770331e-01
-1.03085864e+00 4.50420171e-01 -1.26633048e-01 -3.11781038e-02
-4.76074040e-01 5.18272340e-01 -1.93380546e-02 9.75590348e-02
7.94535935e-01 6.08614147e-01 -6.36432320e-02 -1.23186670e-01
6.00401998e-01 -1.42130181e-01 7.35623658e-01 -3.50636512e-01
3.56094539e-01 4.86708462e-01 2.98417985e-01 -4.36896771e-01
-4.41028416e-01 2.35857710e-01 -8.84484828e-01 -5.00525117e-01
1.02417064e+00 -3.62591624e-01 -8.43962491e-01 3.99601609e-01
-1.10386026e+00 -8.89081120e-01 -3.78049076e-01 6.57436848e-01
-1.32001233e+00 -8.93072877e-03 -2.06725597e-01 -5.76842308e-01
9.86721069e-02 -1.15967083e+00 5.95146835e-01 5.79319179e-01
-4.35932666e-01 -6.30523384e-01 3.68499994e-01 -3.25488538e-01
1.19950593e-01 4.85017955e-01 8.47669959e-01 -3.22971463e-01
-1.10833371e+00 1.20512940e-01 4.08403575e-01 -3.68958920e-01
-7.59887621e-02 -4.96662140e-01 -3.59315872e-01 1.02182895e-01
1.19357593e-01 -3.45176607e-01 7.23534524e-01 5.28259695e-01
1.10259414e+00 -2.97622919e-01 -6.24456525e-01 5.22483230e-01
1.17203915e+00 7.95869589e-01 1.07538366e+00 4.24218059e-01
-4.02321965e-02 8.65513921e-01 4.09040868e-01 2.10079819e-01
2.29917035e-01 6.01531565e-01 8.58119786e-01 3.86722744e-01
2.97047887e-02 -4.26752657e-01 3.83049011e-01 6.95430115e-02
-2.86641955e-01 1.24040231e-01 -8.70019972e-01 5.91590643e-01
-1.77199650e+00 -1.25068784e+00 5.33363044e-01 1.84569764e+00
5.78763366e-01 2.23692134e-01 -4.53946292e-02 -3.33690405e-01
3.69113684e-01 1.27090171e-01 -7.59271622e-01 -5.14774263e-01
3.34939986e-01 1.99210644e-01 5.27654327e-02 3.78513217e-01
-6.58166111e-01 1.20283127e+00 6.92995930e+00 1.02559356e-02
-1.05793512e+00 -2.38820538e-01 4.48826879e-01 4.91516814e-02
-2.02362970e-01 4.11346018e-01 -1.22376963e-01 9.23486575e-02
7.07464278e-01 -3.26634198e-01 9.64555144e-01 8.21031332e-01
2.37881035e-01 -5.03362417e-01 -1.28982127e+00 5.21115482e-01
-7.49901459e-02 -1.48597586e+00 -6.93714395e-02 4.54496853e-02
3.60667765e-01 1.37347922e-01 -6.64068982e-02 2.79897988e-01
7.54305303e-01 -1.21564746e+00 7.58383393e-01 7.09499180e-01
3.13488275e-01 -3.97708595e-01 1.42135710e-01 8.72157395e-01
-6.48360848e-01 -3.17364931e-01 -1.02201365e-01 -3.89284343e-01
4.51726586e-01 -1.93938255e-01 -1.12507641e+00 1.17958046e-01
1.94730476e-01 4.43243742e-01 -2.73020118e-01 1.16236782e+00
-9.94498730e-01 1.25818020e-02 -9.01892707e-02 -5.79733670e-01
2.58453876e-01 -2.78279066e-01 5.63552678e-01 5.49295187e-01
5.79703510e-01 2.95605183e-01 -1.49394482e-01 9.37183797e-01
1.63330302e-01 -2.66430736e-01 -1.02575564e+00 -6.19360171e-02
2.61227667e-01 1.09039855e+00 -8.96653712e-01 -7.86945596e-02
1.64203852e-01 9.91851568e-01 5.05058467e-01 4.39095795e-01
-4.51544732e-01 3.24698687e-02 3.98956865e-01 -1.05130188e-01
1.45941108e-01 -4.66946721e-01 -3.04537475e-01 -8.72073293e-01
-3.28462213e-01 -4.95056987e-01 -2.34203483e-03 -1.39122760e+00
-2.98384517e-01 5.62992275e-01 -9.63451639e-02 -1.14304852e+00
-8.38045180e-01 -4.87483978e-01 -9.18343604e-01 6.84568346e-01
-9.73090112e-01 -7.73722589e-01 -6.39464557e-02 3.04475754e-01
5.34086943e-01 -9.92877483e-02 1.07122469e+00 -7.16405630e-01
9.89344344e-02 -4.94702101e-01 -3.00232351e-01 -1.01095967e-01
-8.06962664e-04 -1.11592865e+00 5.19089997e-01 7.08325326e-01
3.56360227e-01 5.71373403e-01 8.06950688e-01 -7.37291694e-01
-9.78154123e-01 -6.26552641e-01 9.53092992e-01 -3.98602158e-01
3.01365286e-01 -3.22309881e-02 -5.31368434e-01 8.18894207e-01
2.14697748e-01 -4.69161421e-01 2.45274022e-01 -3.40895295e-01
1.30172566e-01 2.56955177e-01 -1.10438299e+00 1.28882933e+00
1.23576319e+00 -2.46755853e-01 -9.95458663e-01 1.45373389e-01
4.77750212e-01 -5.07032216e-01 -1.36514992e-01 1.33746564e-01
8.34631860e-01 -1.24091077e+00 7.75811732e-01 -7.94817805e-01
6.79940343e-01 -3.65975827e-01 8.22427464e-05 -1.55159509e+00
-3.78654450e-01 -6.88663304e-01 1.93735331e-01 1.92486659e-01
3.68901491e-01 -4.95902061e-01 6.46087110e-01 4.35937434e-01
-3.81839335e-01 -8.50235164e-01 -9.82463837e-01 -3.57089490e-01
-1.42750114e-01 -2.39464805e-01 5.16566634e-01 4.48757708e-01
4.11042660e-01 9.30610150e-02 -2.26055369e-01 -7.45577365e-02
3.64859343e-01 3.50510597e-01 6.48121357e-01 -9.15378213e-01
-2.44346127e-01 -4.20969427e-01 -3.90159160e-01 -9.26131368e-01
2.00430289e-01 -1.01242256e+00 4.26140994e-01 -2.10201240e+00
-1.02065004e-01 -1.47064235e-02 2.59269997e-02 6.76533401e-01
3.06488901e-01 -3.66855234e-01 4.16247070e-01 2.70973951e-01
-3.65160853e-01 6.26317620e-01 1.60551572e+00 3.83702815e-01
-4.70679313e-01 -9.64669660e-02 -5.55866003e-01 1.08679962e+00
8.69135678e-01 -4.61808741e-01 -2.71352202e-01 -2.38200381e-01
3.24309587e-01 6.32799804e-01 4.82779533e-01 -1.21635962e+00
3.17982286e-01 -5.20819306e-01 6.56439304e-01 -3.89162302e-01
3.90506953e-01 -5.65178752e-01 2.85351992e-01 1.12199390e+00
-5.25740206e-01 5.75228706e-02 -1.55748695e-01 5.47565758e-01
1.74393773e-01 -3.55403721e-01 6.27357960e-01 -7.20257580e-01
-1.17871618e+00 4.92778979e-02 -8.31562042e-01 -2.70895183e-01
1.11514711e+00 -2.78325588e-01 -1.69611081e-01 -4.68684167e-01
-1.26283038e+00 2.39563301e-01 7.04956412e-01 2.79039085e-01
7.92325318e-01 -1.18969440e+00 -3.25622976e-01 5.36327325e-02
-2.78988302e-01 -1.59033865e-01 -7.62737298e-04 5.38832068e-01
-8.17541718e-01 2.88266242e-01 -8.11992466e-01 -2.76433766e-01
-6.31536663e-01 5.16349792e-01 5.03774524e-01 -4.05279666e-01
-7.23685443e-01 5.13769209e-01 3.79242599e-01 -3.29624474e-01
-1.25095785e-01 -7.54075125e-02 -6.24427199e-01 -4.70356405e-01
4.15853590e-01 -2.13543288e-02 -5.69902122e-01 -4.25562322e-01
-2.31080890e-01 2.64577240e-01 4.52037722e-01 -6.79012358e-01
1.44468868e+00 3.87086831e-02 -1.73525587e-01 3.36763084e-01
4.75641280e-01 -5.58182001e-01 -1.47403646e+00 2.36443624e-01
1.18174508e-01 -2.00362220e-01 -4.95560855e-01 -9.85352695e-01
-4.14524376e-01 9.59671259e-01 1.50017902e-01 -1.84292480e-01
1.15745831e+00 1.99478894e-01 3.95177603e-01 7.15649426e-01
9.26308095e-01 -8.72300029e-01 5.06873488e-01 6.34827495e-01
1.12766635e+00 -7.43272007e-01 -2.23084360e-01 1.36864275e-01
-7.26405919e-01 1.20514524e+00 5.92883408e-01 -2.57185847e-01
2.70279437e-01 1.04481228e-01 -1.30283311e-01 -3.12138945e-01
-1.04385364e+00 -3.54875684e-01 -5.63455652e-03 8.44618082e-01
1.43817648e-01 1.70993507e-02 -4.12434131e-01 -9.11529586e-02
-2.74734825e-01 3.96962374e-01 7.66025603e-01 9.46838439e-01
-7.67005265e-01 -1.07041943e+00 -1.42790005e-02 3.34128767e-01
-1.23318369e-02 6.96133301e-02 -3.63529712e-01 5.16036510e-01
1.80392802e-01 5.11718750e-01 -7.50667825e-02 -2.45141923e-01
1.48783252e-02 2.93284386e-01 8.67131770e-01 -8.26625168e-01
-3.02660763e-01 -1.11503698e-01 -1.16153928e-02 -1.20580626e+00
-4.43819344e-01 -8.49457622e-01 -1.72532225e+00 3.36699784e-01
3.68806273e-01 -1.88304737e-01 7.72160411e-01 1.05927813e+00
2.92412937e-01 5.82640052e-01 2.32433096e-01 -1.17309475e+00
-2.67763764e-01 -6.86177969e-01 -4.07544017e-01 2.33373955e-01
1.98525012e-01 -6.75706267e-01 -1.22044295e-01 1.36576936e-01] | [4.391965389251709, 0.9728520512580872] |
93977b07-1e8d-4514-8cf0-a5e41044cf2c | feature-refinement-an-expression-specific | 2101.04838 | null | https://arxiv.org/abs/2101.04838v1 | https://arxiv.org/pdf/2101.04838v1.pdf | Feature refinement: An expression-specific feature learning and fusion method for micro-expression recognition | Micro-Expression Recognition has become challenging, as it is extremely difficult to extract the subtle facial changes of micro-expressions. Recently, several approaches proposed several expression-shared features algorithms for micro-expression recognition. However, they do not reveal the specific discriminative characteristics, which lead to sub-optimal performance. This paper proposes a novel Feature Refinement ({FR}) with expression-specific feature learning and fusion for micro-expression recognition. It aims to obtain salient and discriminative features for specific expressions and also predict expression by fusing the expression-specific features. FR consists of an expression proposal module with attention mechanism and a classification branch. First, an inception module is designed based on optical flow to obtain expression-shared features. Second, in order to extract salient and discriminative features for specific expression, expression-shared features are fed into an expression proposal module with attention factors and proposal loss. Last, in the classification branch, labels of categories are predicted by a fusion of the expression-specific features. Experiments on three publicly available databases validate the effectiveness of FR under different protocol. Results on public benchmarks demonstrate that our FR provides salient and discriminative information for micro-expression recognition. The results also show our FR achieves better or competitive performance with the existing state-of-the-art methods on micro-expression recognition. | ['Zhihong Zhang', 'Feifei Zhang', 'Xiaohua Huang', 'Qirong Mao', 'Ling Zhou'] | 2021-01-13 | null | null | null | null | ['micro-expression-recognition'] | ['computer-vision'] | [ 2.13183358e-01 -5.06500721e-01 -2.80715466e-01 -8.95297229e-01
-6.56320810e-01 -3.15344706e-02 2.11164728e-01 -2.90319234e-01
-2.49633178e-01 4.88605559e-01 2.60646850e-01 5.61236322e-01
6.35096729e-02 -4.75151211e-01 -2.59468496e-01 -1.21072245e+00
-7.54455999e-02 -2.61876941e-01 -2.86385626e-01 -3.37461203e-01
1.66976571e-01 7.90130377e-01 -1.83639050e+00 4.25914675e-01
7.19543576e-01 1.61018050e+00 -5.32545745e-02 4.40147698e-01
-4.44955856e-01 1.05794740e+00 -4.44409907e-01 -2.71175742e-01
-6.12119958e-02 -5.73983192e-01 -6.09926403e-01 2.24972129e-01
2.78245658e-01 -2.33687043e-01 1.37098446e-01 1.10479486e+00
5.45433939e-01 6.08478114e-03 6.51838183e-01 -1.49662602e+00
-4.70433027e-01 -2.10973680e-01 -1.01090884e+00 2.52320170e-01
6.36402309e-01 -1.05156444e-01 9.42750931e-01 -9.85244572e-01
5.69598079e-01 1.31581438e+00 4.74549085e-01 4.78791028e-01
-9.37807858e-01 -1.03332829e+00 2.46040002e-01 4.23145592e-01
-1.54279983e+00 -6.47256255e-01 1.08774030e+00 -4.23698425e-01
8.41850817e-01 2.71310061e-01 6.23952746e-01 7.31402814e-01
3.15411806e-01 1.13902628e+00 1.32641649e+00 -2.49704227e-01
-1.29116103e-01 2.91504711e-01 3.42798978e-01 1.03192019e+00
-5.68789363e-01 -1.42151058e-01 -5.76314270e-01 -2.85584658e-01
3.23285282e-01 1.95168152e-01 -3.92302245e-01 3.10661625e-02
-2.73771316e-01 6.16170824e-01 2.34196737e-01 5.12725294e-01
-5.43770254e-01 -1.96513966e-01 6.08111858e-01 4.47948784e-01
7.00759590e-01 -5.40989488e-02 -5.28326213e-01 -4.44037586e-01
-6.96262181e-01 1.47672921e-01 5.70494115e-01 6.44206524e-01
1.52291238e+00 -3.25042121e-02 -5.00778437e-01 1.14426875e+00
1.16282679e-01 3.94218117e-01 5.78981459e-01 -6.39012933e-01
4.07541022e-02 9.58827794e-01 -2.26630732e-01 -1.49201250e+00
-4.66863066e-01 -3.95746112e-01 -9.87625897e-01 8.80511627e-02
-8.28537047e-02 -3.05876851e-01 -5.24078310e-01 1.91889536e+00
3.41250569e-01 4.77343887e-01 4.32344005e-02 1.02592623e+00
7.74509430e-01 6.92856908e-01 3.66153896e-01 -6.77294493e-01
1.37227833e+00 -9.61637199e-01 -9.55082119e-01 2.78110087e-01
1.14617109e+00 -6.58936381e-01 8.93373787e-01 4.65081155e-01
-5.79565585e-01 -6.66989923e-01 -7.56058156e-01 1.84652239e-01
-2.52605915e-01 3.03780556e-01 9.62916732e-01 6.36580229e-01
-6.60058916e-01 4.63545710e-01 -3.84515136e-01 -1.98863834e-01
7.83941448e-01 6.45304620e-01 -7.85750687e-01 2.16674745e-01
-9.72481132e-01 6.29459679e-01 -1.49874896e-01 3.19543749e-01
-5.53355396e-01 -5.22452235e-01 -1.03090310e+00 4.97596860e-02
7.36460239e-02 -2.38227308e-01 9.28348243e-01 -2.10531425e+00
-1.88800299e+00 8.83852184e-01 -5.91644168e-01 1.29988730e-01
-6.40901625e-02 -1.74569994e-01 -6.94235325e-01 3.02534789e-01
-8.94194618e-02 3.22456747e-01 8.68486881e-01 -8.24808300e-01
-7.08971441e-01 -7.03004718e-01 -1.33422822e-01 1.60282105e-01
-5.31364143e-01 4.48100328e-01 -2.85644531e-01 -2.81749070e-01
-1.54369161e-01 -6.16681814e-01 -1.93135571e-02 8.54124129e-02
-8.35307464e-02 -4.87443298e-01 1.23243976e+00 -4.35028255e-01
1.24163139e+00 -2.37039566e+00 8.89370739e-02 2.67366618e-01
2.27741767e-02 1.92589939e-01 -2.83541173e-01 -1.43447384e-01
-1.79234579e-01 -3.71426903e-02 5.44721261e-02 -4.26793247e-01
-1.22988522e-01 1.40546620e-01 -2.04535156e-01 6.01528406e-01
5.17637670e-01 8.29670608e-01 -6.67231441e-01 -7.41609812e-01
2.35577151e-01 4.37905073e-01 -4.56298590e-01 4.95737761e-01
2.45130241e-01 4.92883146e-01 -9.39149499e-01 1.01641107e+00
9.24968779e-01 4.97249067e-02 -1.82378143e-01 -4.84992176e-01
-4.34600338e-02 -4.10493106e-01 -9.93665338e-01 1.44969738e+00
-6.71855807e-01 3.74181569e-01 3.52086574e-01 -1.21773350e+00
1.24476182e+00 1.19158529e-01 7.24549055e-01 -8.09175551e-01
4.63080883e-01 5.39626703e-02 -3.16222310e-01 -8.64830852e-01
2.54537255e-01 -3.85041267e-01 -1.16275355e-01 2.71897882e-01
3.46651196e-01 4.16404754e-01 -1.55720785e-01 -1.70772970e-01
8.52593958e-01 3.00499558e-01 3.33208412e-01 -2.17865601e-01
1.18965232e+00 -4.76634055e-01 1.10736322e+00 5.32406867e-02
-5.68239272e-01 1.88319653e-01 5.82199395e-01 -5.48688769e-01
-2.89999843e-01 -4.79872078e-01 -1.25591472e-01 1.46665347e+00
2.25562751e-01 -4.68976438e-01 -6.14764988e-01 -8.50432098e-01
-1.12014972e-01 1.77399635e-01 -1.01404595e+00 -3.51407558e-01
-2.38687009e-01 -8.36929262e-01 6.29636347e-01 5.14358461e-01
7.58213818e-01 -1.12539744e+00 -3.08437437e-01 1.20203353e-01
-5.75744323e-02 -9.03323352e-01 -4.06728208e-01 1.49036823e-02
-4.31740403e-01 -8.92856419e-01 -4.97395962e-01 -7.43998468e-01
7.39796698e-01 1.96331322e-01 5.66995919e-01 1.27608672e-01
-4.82019395e-01 2.58957595e-01 -4.96422201e-01 -3.49019855e-01
5.29860929e-02 -2.69244432e-01 -1.32784292e-01 1.00832653e+00
7.74656594e-01 -6.28109038e-01 -4.91501391e-01 2.50821799e-01
-6.07775211e-01 -2.94673771e-01 6.81144476e-01 8.68992269e-01
8.16055715e-01 -3.45577389e-01 4.50498521e-01 -6.40606821e-01
4.27556515e-01 -4.62942600e-01 -2.42952824e-01 2.14589521e-01
-2.79802114e-01 6.10108068e-03 7.60678709e-01 -2.94750154e-01
-1.39463937e+00 3.40869963e-01 -3.18118453e-01 -5.71531713e-01
-2.39205226e-01 3.47897679e-01 -4.61940646e-01 -3.74353409e-01
2.44302318e-01 5.02839327e-01 2.88134515e-01 -4.30614918e-01
-3.64240743e-02 8.78027916e-01 2.29009211e-01 -5.18109322e-01
3.17396790e-01 3.94373447e-01 1.16494335e-01 -1.00457764e+00
-9.26253080e-01 -6.31389916e-01 -5.03875434e-01 -2.24362567e-01
8.42727661e-01 -9.69479501e-01 -8.68171513e-01 7.18683183e-01
-9.32720602e-01 8.16270486e-02 1.36023343e-01 2.71171957e-01
-6.28861368e-01 5.01521647e-01 -5.66559732e-01 -9.32356119e-01
-5.52474797e-01 -1.22746956e+00 1.34089875e+00 6.67888224e-01
-1.96819156e-01 -7.63908684e-01 1.59811735e-01 2.51917303e-01
3.53380978e-01 4.13753390e-01 5.34390509e-01 -4.13572311e-01
-1.82852715e-01 -1.65749907e-01 -2.76346296e-01 5.14123023e-01
3.62791568e-01 4.33784395e-01 -1.20254576e+00 -5.68307266e-02
-1.36529088e-01 -6.85127974e-01 7.99311757e-01 7.85108507e-02
1.54761457e+00 -3.48105758e-01 -3.59381109e-01 1.03965271e+00
1.34204698e+00 1.45931631e-01 7.42754459e-01 3.21792439e-02
3.64156276e-01 7.25724697e-01 1.05265808e+00 8.52134705e-01
1.83261275e-01 8.81931901e-01 1.84471741e-01 -2.34377846e-01
3.80503982e-01 -7.91208521e-02 5.32891512e-01 5.19969523e-01
-2.07811907e-01 2.50226587e-01 -2.45246574e-01 2.71581024e-01
-1.83787644e+00 -1.13731194e+00 1.49543405e-01 1.58192348e+00
7.45063245e-01 -4.29457217e-01 1.27167091e-01 -1.06575787e-01
4.15325016e-01 3.33031684e-01 -3.96625727e-01 -8.28048825e-01
-4.48001534e-01 3.75155419e-01 -1.07413277e-01 2.82344848e-01
-1.20090508e+00 1.00701606e+00 5.33868408e+00 9.96437669e-01
-1.56064451e+00 2.02124357e-01 7.76517749e-01 8.76237527e-02
4.71676663e-02 -1.43481120e-01 -9.91751075e-01 3.19406211e-01
5.36319911e-01 -2.83543497e-01 -2.07390532e-01 1.15752709e+00
1.94944575e-01 -6.35925159e-02 -6.81113362e-01 1.34237063e+00
3.13157976e-01 -8.91055107e-01 1.43508911e-01 -2.31465876e-01
3.73995602e-01 -4.03596193e-01 -1.47481158e-01 4.97633219e-01
-1.95763707e-01 -1.08344781e+00 2.48181358e-01 8.64460886e-01
5.89775980e-01 -1.01313233e+00 9.97193873e-01 6.43311143e-02
-1.38614690e+00 -1.05420016e-01 -4.89112020e-01 -2.98435628e-01
-9.23819747e-03 3.53803039e-01 -2.55656511e-01 6.61147594e-01
6.06489599e-01 1.19446838e+00 -3.97561759e-01 5.59786141e-01
-1.71917990e-01 5.11302054e-01 -2.29682162e-01 -2.78289407e-01
2.60386676e-01 -3.75588208e-01 2.33816341e-01 1.61792207e+00
1.54786900e-01 3.59895438e-01 2.01227427e-01 7.13438451e-01
7.97617808e-02 8.86437416e-01 -4.55319136e-01 1.14603467e-01
-9.19832587e-02 1.82235396e+00 -1.54169253e-03 -2.86007732e-01
-5.03407419e-01 1.28803694e+00 5.44844806e-01 2.77825713e-01
-7.50972986e-01 -5.34140825e-01 1.19105601e+00 -2.67534941e-01
9.43588093e-02 2.72519201e-01 2.10933015e-01 -1.22254956e+00
-3.53166182e-03 -6.58517122e-01 4.68697250e-01 -7.03715324e-01
-1.25810075e+00 7.17817843e-01 -3.28850776e-01 -1.08894646e+00
-1.99627519e-01 -6.35398388e-01 -7.00481415e-01 9.19076383e-01
-1.64291453e+00 -1.34624505e+00 -7.83582687e-01 8.97334158e-01
3.87235671e-01 -2.79885173e-01 1.16521370e+00 3.58963877e-01
-9.23331022e-01 9.63685215e-01 -1.49659351e-01 2.71076113e-01
6.03305042e-01 -7.55649984e-01 -7.83114851e-01 4.55419600e-01
2.41543278e-02 4.46409166e-01 3.61469269e-01 -1.27952114e-01
-1.35504961e+00 -1.09317052e+00 7.24070311e-01 1.78574458e-01
3.58363181e-01 -2.44891182e-01 -9.57072496e-01 4.78600532e-01
-8.76625329e-02 5.33331275e-01 1.20024586e+00 1.34767711e-01
-3.80374640e-01 -7.38496482e-01 -1.19780123e+00 3.17987114e-01
8.93194258e-01 -5.94321370e-01 -2.96668917e-01 2.95402631e-02
1.64059728e-01 2.42093261e-02 -9.26561952e-01 5.34914315e-01
9.40286934e-01 -1.07955933e+00 3.94541055e-01 -6.46962821e-01
4.15675998e-01 -2.68475562e-01 -3.20227981e-01 -1.00814986e+00
-4.41646338e-01 -6.06116414e-01 1.66177139e-01 1.49355328e+00
2.78172474e-02 -6.54971361e-01 7.31984973e-01 4.39779311e-01
8.96694437e-02 -1.13838673e+00 -9.59853053e-01 -5.36170661e-01
-3.61190915e-01 -2.58747935e-01 6.94472849e-01 1.00759363e+00
1.99372530e-01 4.54384267e-01 -3.75347883e-01 -2.25925967e-01
2.68071443e-01 6.19355857e-01 9.73845065e-01 -9.37856913e-01
-1.17830686e-01 -5.97335041e-01 -9.47344363e-01 -1.08504725e+00
8.59524012e-01 -7.91034400e-01 -1.63415805e-01 -8.77388000e-01
5.87968111e-01 -2.78881311e-01 -7.95450032e-01 7.37428129e-01
-1.67764828e-01 2.67181218e-01 5.23135625e-02 -4.51648561e-03
-8.43076527e-01 1.07465482e+00 1.20428121e+00 -2.62854755e-01
-2.62166560e-01 -1.38553634e-01 -7.79758930e-01 7.95442104e-01
6.63521171e-01 -1.66082814e-01 -2.46845499e-01 1.15751877e-01
-3.20583969e-01 -2.35579852e-02 2.84853876e-01 -8.02604318e-01
-9.66442190e-03 -4.08011705e-01 4.93920147e-01 -4.56597120e-01
4.41820681e-01 -5.57980359e-01 -8.12221617e-02 3.08077663e-01
-1.40741214e-01 -2.61792570e-01 2.77893066e-01 3.24423850e-01
-6.89627111e-01 -9.43262596e-03 9.30274308e-01 1.40653461e-01
-1.25928378e+00 6.53607011e-01 -2.54472733e-01 -2.17662036e-01
1.24085748e+00 -4.48617905e-01 1.30569398e-01 -2.89033383e-01
-7.19517827e-01 1.99186027e-01 1.50970221e-01 5.67216873e-01
6.98783875e-01 -1.43849492e+00 -6.97579861e-01 3.83547872e-01
4.58862960e-01 -4.80001003e-01 4.88886714e-01 1.08893287e+00
-1.30652264e-01 1.60982125e-02 -5.74236631e-01 -6.12844229e-01
-1.89700985e+00 2.33853281e-01 6.60567701e-01 -3.30826610e-01
-1.21150941e-01 9.22338188e-01 5.92077553e-01 -1.86190680e-01
-7.47695491e-02 1.24173239e-01 -5.82507074e-01 2.06510127e-01
8.55346680e-01 1.07658252e-01 -2.75927395e-01 -1.30158138e+00
-5.15784323e-01 1.05075622e+00 -1.77543208e-01 4.44146156e-01
1.35442626e+00 -1.56299129e-01 -4.83255774e-01 1.97966799e-01
1.89887226e+00 -9.30800065e-02 -1.06780314e+00 -2.58603632e-01
-2.34307811e-01 -6.38030767e-01 4.53959666e-02 -5.52202344e-01
-1.37899411e+00 9.91662085e-01 8.01903427e-01 -6.24509513e-01
1.73925734e+00 -2.51383394e-01 5.56319535e-01 1.14703670e-01
3.72657865e-01 -1.01953864e+00 1.89219192e-01 3.98020446e-01
9.39765811e-01 -1.05406499e+00 -1.91083997e-01 -5.58143377e-01
-6.13894045e-01 1.24944234e+00 9.54286277e-01 -1.26255840e-01
9.59402204e-01 3.31158131e-01 1.69435009e-01 -1.82806998e-01
-7.66427517e-01 -9.14399251e-02 1.36251450e-01 3.83964032e-01
5.19999444e-01 -2.27255896e-02 -3.86681139e-01 1.12117386e+00
5.72393462e-02 3.84646207e-01 -8.64135697e-02 8.18506122e-01
-4.96468186e-01 -9.67328012e-01 -5.79211023e-03 4.41277057e-01
-7.65880346e-01 3.13573122e-01 -3.56984645e-01 4.87660110e-01
4.30643648e-01 7.95408607e-01 1.45910352e-01 -7.06717014e-01
3.52380902e-01 1.28945559e-01 4.81694549e-01 -2.95708895e-01
-5.85478663e-01 -7.30147399e-03 1.43012181e-01 -1.07581472e+00
-7.48094201e-01 -5.74337006e-01 -1.27937114e+00 -2.10965872e-02
-3.54601383e-01 2.94229835e-01 2.01753035e-01 8.89692843e-01
6.45624340e-01 6.67530000e-02 1.06800318e+00 -6.88922167e-01
-4.07278687e-01 -9.62790906e-01 -9.12102282e-01 9.82390285e-01
3.11054409e-01 -8.50799382e-01 -2.89756238e-01 -4.41962183e-02] | [13.648272514343262, 1.6817190647125244] |
7db470a3-6197-444c-bf73-a94b07359e35 | cfl-net-image-forgery-localization-using | 2210.02182 | null | https://arxiv.org/abs/2210.02182v1 | https://arxiv.org/pdf/2210.02182v1.pdf | CFL-Net: Image Forgery Localization Using Contrastive Learning | Conventional forgery localizing methods usually rely on different forgery footprints such as JPEG artifacts, edge inconsistency, camera noise, etc., with cross-entropy loss to locate manipulated regions. However, these methods have the disadvantage of over-fitting and focusing on only a few specific forgery footprints. On the other hand, real-life manipulated images are generated via a wide variety of forgery operations and thus, leave behind a wide variety of forgery footprints. Therefore, we need a more general approach for image forgery localization that can work well on a variety of forgery conditions. A key assumption in underlying forged region localization is that there remains a difference of feature distribution between untampered and manipulated regions in each forged image sample, irrespective of the forgery type. In this paper, we aim to leverage this difference of feature distribution to aid in image forgery localization. Specifically, we use contrastive loss to learn mapping into a feature space where the features between untampered and manipulated regions are well-separated for each image. Also, our method has the advantage of localizing manipulated region without requiring any prior knowledge or assumption about the forgery type. We demonstrate that our work outperforms several existing methods on three benchmark image manipulation datasets. Code is available at https://github.com/niloy193/CFLNet. | ['Simon S. Woo', 'Kishor Kumar Bhaumik', 'Fahim Faisal Niloy'] | 2022-10-04 | null | null | null | null | ['image-manipulation'] | ['computer-vision'] | [ 3.05275530e-01 -5.63012540e-01 -7.89650977e-02 1.36757176e-02
-7.63774514e-01 -6.93020701e-01 2.88133860e-01 -1.31871417e-01
5.01033440e-02 5.22787392e-01 -9.58224200e-03 4.39944193e-02
3.75734479e-03 -6.76743865e-01 -8.29467654e-01 -8.97318423e-01
1.43556744e-01 -5.80255806e-01 6.57650605e-02 -2.04808697e-01
4.49289709e-01 3.63231033e-01 -1.34154856e+00 3.18742484e-01
8.30939949e-01 1.08330047e+00 3.00549448e-01 5.39873064e-01
3.33441854e-01 5.72203577e-01 -7.65503466e-01 -3.40102494e-01
4.89630610e-01 -4.98487294e-01 -5.21883607e-01 3.76754254e-01
4.31549907e-01 -6.42238855e-01 -8.25744808e-01 1.47831702e+00
3.22202653e-01 1.62803844e-01 4.97576267e-01 -1.35165513e+00
-9.66121852e-01 1.81372434e-01 -8.06764722e-01 3.56178284e-01
3.82981062e-01 3.89831662e-01 5.42353153e-01 -1.19087696e+00
5.14744043e-01 1.02527976e+00 8.26307297e-01 2.54273981e-01
-8.43573868e-01 -7.61816561e-01 -1.95480451e-01 4.30941254e-01
-1.64382792e+00 -5.19366145e-01 9.80068088e-01 -1.51698127e-01
4.36845928e-01 4.78643954e-01 1.99717179e-01 1.04482698e+00
5.99483788e-01 6.64864242e-01 1.12650347e+00 -4.23448443e-01
2.41441908e-03 4.15084362e-02 -1.43927589e-01 6.79835677e-01
2.96071202e-01 2.48658165e-01 -4.95606124e-01 -2.77903050e-01
9.28969562e-01 3.39593291e-01 -1.00183618e+00 -3.36076379e-01
-1.13789260e+00 7.52820253e-01 3.85126501e-01 2.68109947e-01
-1.28295928e-01 2.69552618e-01 3.39462548e-01 4.63627219e-01
3.48866656e-02 2.83916324e-01 -2.15121612e-01 1.06085077e-01
-8.82806480e-01 2.79085577e-01 3.72520566e-01 9.58226681e-01
1.01533544e+00 -1.64768115e-01 -7.14592859e-02 7.13810682e-01
3.03492099e-02 2.06990674e-01 5.88599861e-01 -8.97393405e-01
5.68844616e-01 3.60944748e-01 2.46883258e-01 -1.61762857e+00
1.60218462e-01 -1.83849454e-01 -9.85967815e-01 3.62909824e-01
4.27686244e-01 8.31679180e-02 -8.87437642e-01 1.54529142e+00
8.37394968e-02 2.30600372e-01 -1.91138923e-01 1.01676202e+00
3.17082524e-01 3.74623477e-01 -4.03075904e-01 1.55673595e-02
1.33965778e+00 -8.79264057e-01 -9.05607700e-01 -2.68150389e-01
4.57486868e-01 -9.70600843e-01 1.23649585e+00 5.54521620e-01
-1.00550771e+00 -3.95076066e-01 -1.21201575e+00 -1.30441226e-02
-5.91227233e-01 8.02616775e-02 4.13150966e-01 8.65507185e-01
-9.34487641e-01 6.47338510e-01 -4.87788349e-01 -8.57724622e-02
4.17793930e-01 7.43317157e-02 -5.80999672e-01 -4.87882316e-01
-1.22838318e+00 8.16218436e-01 6.10257268e-01 2.55318135e-01
-7.99040377e-01 -6.17296100e-01 -8.42179835e-01 2.27669813e-03
4.91039246e-01 -4.15059566e-01 7.19413698e-01 -1.02622783e+00
-1.06062746e+00 7.49610305e-01 1.51780054e-01 -1.34711161e-01
7.86797583e-01 -9.18146372e-02 -6.00462079e-01 2.85052747e-01
-4.34841812e-02 1.60122633e-01 1.31465089e+00 -1.51295257e+00
-4.40388739e-01 -4.25043851e-01 -1.59199700e-01 1.21425055e-01
-2.85699308e-01 1.11990348e-01 -6.08426809e-01 -1.34049726e+00
1.57673836e-01 -5.90374708e-01 8.50622877e-02 3.69067073e-01
-3.11193496e-01 3.46842915e-01 1.40695953e+00 -9.26656604e-01
1.39738083e+00 -2.47303486e+00 -9.45903137e-02 3.35927382e-02
2.89804846e-01 1.30904108e-01 -5.95333874e-02 5.44943750e-01
-2.17449322e-01 3.70083541e-01 -5.46668291e-01 -2.02136829e-01
-9.99633148e-02 1.71463892e-01 -4.20996636e-01 1.00365794e+00
1.49641126e-01 7.55353987e-01 -1.00615418e+00 -3.34462881e-01
4.72103357e-01 4.98534381e-01 -2.60388821e-01 9.40838456e-03
2.35431582e-01 3.91380191e-01 -3.08997393e-01 7.75014400e-01
1.25483811e+00 5.73450029e-02 -2.92803913e-01 -3.66875857e-01
2.11505994e-01 -3.69378567e-01 -1.35450864e+00 1.80286920e+00
-3.89431864e-01 4.71227527e-01 3.46904337e-01 -8.92085314e-01
6.67084038e-01 2.08343774e-01 3.28403741e-01 -3.73938978e-01
2.78579980e-01 2.63742357e-01 -3.80013108e-01 -5.46770632e-01
6.84147000e-01 -5.09244353e-02 -2.52373189e-01 5.13626099e-01
-3.73240001e-03 2.67466009e-02 -1.65654287e-01 7.22615048e-02
1.22993779e+00 -1.43289477e-01 3.35088700e-01 -3.08425650e-02
3.25321674e-01 -4.47283298e-01 6.96358383e-01 7.60390162e-01
-5.53679526e-01 1.00011146e+00 9.08759311e-02 -1.62706003e-01
-8.49903703e-01 -1.09491408e+00 -2.33923107e-01 5.19161582e-01
7.64180005e-01 -3.88809264e-01 -6.47750974e-01 -8.70979786e-01
3.81140001e-02 6.07116044e-01 -4.68111545e-01 -5.76093376e-01
-6.60707653e-01 -6.61294758e-01 7.11114109e-01 1.95648924e-01
8.91112864e-01 -8.78282964e-01 -5.53007662e-01 -5.91940917e-02
-3.91835481e-01 -1.15731192e+00 -1.05665553e+00 -1.16586454e-01
-7.16684222e-01 -1.17733848e+00 -6.23944283e-01 -7.44126678e-01
8.45878959e-01 7.88407266e-01 8.66481662e-01 4.90263343e-01
-6.60701454e-01 3.75232607e-01 -5.31059861e-01 1.54624149e-01
-4.76847649e-01 -5.40506363e-01 -1.20483495e-01 1.68299377e-01
-4.46654931e-02 -3.47402453e-01 -8.95206392e-01 5.29294074e-01
-1.32790780e+00 -3.14825743e-01 4.90280420e-01 9.88988400e-01
3.28284621e-01 8.64083707e-01 4.30648297e-01 -4.31524634e-01
4.80219841e-01 -6.62499070e-01 -2.17633680e-01 4.04018968e-01
-3.96668136e-01 -1.11350484e-01 7.92594016e-01 -4.21898931e-01
-1.03656042e+00 -9.03105363e-02 2.47718066e-01 -8.38672996e-01
-2.09139735e-01 1.27357379e-01 -5.09105980e-01 -4.89502817e-01
4.93190080e-01 5.84951818e-01 1.80915929e-02 -4.24993873e-01
4.94742513e-01 6.70205474e-01 8.03333461e-01 -3.34756196e-01
9.73365426e-01 5.57076633e-01 -2.06379801e-01 -7.59823799e-01
-7.09599704e-02 -4.55104500e-01 -2.92971432e-01 -2.18673393e-01
3.73440653e-01 -6.35004044e-01 -3.93535376e-01 9.47013140e-01
-9.20331001e-01 -3.23509984e-02 -6.68086335e-02 1.85648412e-01
-5.12662649e-01 1.37148094e+00 -6.85330808e-01 -6.71397626e-01
-2.75427788e-01 -1.23136365e+00 1.04126048e+00 1.18372060e-01
2.05144912e-01 -9.61709678e-01 -5.43121994e-01 3.93627018e-01
3.72022271e-01 4.91222888e-01 6.90552831e-01 2.63308808e-02
-7.04616308e-01 -5.42804956e-01 -3.04249287e-01 6.11898184e-01
7.14553535e-01 -2.90465087e-01 -7.83212900e-01 -7.08422720e-01
5.13150454e-01 -2.73795307e-01 1.02552593e+00 1.52838618e-01
1.48353291e+00 -4.17145878e-01 -2.85631180e-01 7.64221013e-01
1.60688651e+00 3.43614183e-02 1.01119924e+00 2.26438358e-01
5.98119497e-01 3.22883248e-01 7.45171547e-01 5.34807265e-01
3.26918922e-02 7.07925022e-01 6.29532218e-01 3.38499919e-02
-2.09670603e-01 -1.63308993e-01 4.80881333e-01 3.48444313e-01
2.75172740e-01 -5.48067331e-01 -5.07522583e-01 5.81535935e-01
-1.64046216e+00 -9.64751005e-01 8.49417523e-02 2.59429264e+00
8.42468917e-01 -2.26865262e-01 -9.19184685e-02 3.29354018e-01
1.06385517e+00 3.04865181e-01 -4.55918938e-01 -1.15092797e-02
-3.00339103e-01 -3.36261354e-02 8.70241702e-01 3.39766949e-01
-1.30495453e+00 7.51308024e-01 5.70547247e+00 1.31371224e+00
-1.06837380e+00 -8.31387844e-03 4.61024672e-01 1.31402493e-01
-8.37549269e-02 1.21872880e-01 -4.50429022e-01 8.94595921e-01
3.75630319e-01 -9.55290347e-03 6.13713086e-01 4.29039925e-01
2.21579894e-01 -3.52980316e-01 -6.79048717e-01 1.31314635e+00
3.70211571e-01 -1.08391082e+00 -1.43439874e-01 -7.60511458e-02
4.81067747e-01 -6.08026743e-01 3.29116642e-01 -1.60093710e-01
3.45417820e-02 -1.01766050e+00 7.10563958e-01 3.47434014e-01
8.42899561e-01 -6.42981648e-01 5.81523180e-01 3.57735217e-01
-1.16005337e+00 -1.31301090e-01 -4.46196020e-01 2.09457815e-01
4.96888608e-02 7.11939156e-01 -5.48037946e-01 8.39871764e-01
7.60022104e-01 5.68443000e-01 -4.05327976e-01 9.85510170e-01
-1.18582994e-01 3.18301976e-01 -1.73416346e-01 6.94766939e-01
2.62100482e-03 -1.89863637e-01 5.30525923e-01 1.11229718e+00
5.64631104e-01 -1.45722911e-01 -3.16321738e-02 9.53525662e-01
-1.55598044e-01 -8.04533362e-02 -6.79406226e-01 2.47410282e-01
6.60546541e-01 1.03917027e+00 -7.52369642e-01 -6.83441609e-02
-3.89323980e-01 1.62047386e+00 -1.32774673e-02 4.04272407e-01
-1.03777313e+00 -5.85161686e-01 6.35739684e-01 -5.73424734e-02
4.95603710e-01 -1.19188622e-01 -2.15094447e-01 -1.44809616e+00
3.92749041e-01 -1.05132329e+00 4.25485700e-01 -6.25725508e-01
-1.45256066e+00 8.44989941e-02 -1.32551253e-01 -1.29429972e+00
2.53825963e-01 -4.38742578e-01 -6.47000372e-01 8.45784485e-01
-1.45824242e+00 -1.09424257e+00 -6.08237147e-01 7.93264389e-01
7.03675628e-01 1.32561371e-01 4.82724845e-01 3.89648914e-01
-6.62697971e-01 8.96865129e-01 2.19183728e-01 2.62313366e-01
9.40912783e-01 -9.85601425e-01 3.48134875e-01 1.10730982e+00
-2.16120452e-01 4.96468246e-01 6.73209071e-01 -7.29026794e-01
-1.58089185e+00 -1.31050003e+00 3.11615109e-01 -3.32904011e-01
4.44313586e-01 -3.06048661e-01 -1.10481191e+00 6.68726385e-01
2.26219386e-01 4.48345661e-01 1.75451517e-01 -7.36802161e-01
-4.31869388e-01 2.27196142e-01 -1.59206522e+00 4.97512281e-01
1.00720942e+00 -5.50613165e-01 -3.83926451e-01 1.81849346e-01
2.59661943e-01 -4.16065574e-01 -9.04223859e-01 3.16455722e-01
3.33392352e-01 -1.05257714e+00 1.24064374e+00 -6.66088685e-02
3.14173132e-01 -4.29945648e-01 -2.89536834e-01 -1.19650710e+00
-2.20830873e-01 -8.22775245e-01 -1.83168069e-01 1.14201856e+00
-2.04281047e-01 -7.92713463e-01 6.54420674e-01 3.26159149e-01
-9.39544961e-02 -6.74042165e-01 -8.84542942e-01 -1.08675635e+00
1.03634804e-01 -1.21370792e-01 5.57323694e-01 1.16704023e+00
-7.44089931e-02 -5.10696828e-01 -7.88313448e-01 3.49028826e-01
8.32316756e-01 -1.53633142e-02 5.69158733e-01 -3.30018550e-01
-2.91887671e-01 -2.48765469e-01 -7.20323861e-01 -9.86322105e-01
-1.04428299e-01 -6.53038502e-01 1.36677042e-01 -9.41106558e-01
2.17465267e-01 -5.29684663e-01 -2.49717191e-01 4.30887222e-01
-6.10054493e-01 4.76948917e-01 5.02439104e-02 3.07036191e-01
-1.56937718e-01 5.01849711e-01 1.28975999e+00 -8.95306990e-02
1.50596023e-01 -2.01970279e-01 -6.07335806e-01 5.83804727e-01
1.03185678e+00 -5.05307317e-01 -3.81379485e-01 -3.67557764e-01
-2.32203662e-01 1.52260944e-01 7.73687243e-01 -1.03448439e+00
1.07119679e-01 -1.44341305e-01 4.02295053e-01 -4.29115683e-01
2.89845496e-01 -8.18826795e-01 2.70340621e-01 3.98691922e-01
1.29477054e-01 5.10023572e-02 1.39380932e-01 9.36802506e-01
-3.14094990e-01 -4.01286125e-01 9.79983211e-01 -2.80968368e-01
-8.76295149e-01 1.87607825e-01 -1.91506341e-01 5.63407950e-02
1.26127970e+00 -7.62528181e-01 -3.56620789e-01 -3.56799155e-01
-3.33796710e-01 -1.91365659e-01 1.05540264e+00 4.29895282e-01
9.75808918e-01 -1.25291383e+00 -5.65497220e-01 4.71720427e-01
1.19356610e-01 -2.40529284e-01 5.96058905e-01 6.11418486e-01
-5.45757711e-01 -1.35836542e-01 -2.01970100e-01 -3.45086664e-01
-1.25672281e+00 9.22505140e-01 4.10703897e-01 -1.87716424e-01
-8.27161252e-01 7.21329093e-01 3.38390976e-01 -8.09340551e-02
-2.69686095e-02 -3.36331129e-01 4.17312801e-01 -2.93670148e-01
5.80107272e-01 5.48435152e-01 2.04656094e-01 -7.41175890e-01
-2.07200125e-01 7.32388496e-01 -1.35565281e-01 3.57397765e-01
7.24490464e-01 -5.39115012e-01 2.46727522e-02 -7.98566565e-02
1.45282578e+00 4.54318859e-02 -1.37179005e+00 -1.72504485e-01
-1.32995382e-01 -1.28188801e+00 3.45698357e-01 -5.28206229e-01
-1.46137452e+00 4.47109729e-01 7.14480639e-01 -3.66323255e-02
1.58025968e+00 -1.53589457e-01 1.10578203e+00 -1.29033640e-01
4.96207088e-01 -9.23345447e-01 2.25981325e-01 -5.31667843e-02
1.12556529e+00 -1.33249104e+00 2.06460252e-01 -6.31476641e-01
-4.95554686e-01 1.07182336e+00 3.08456719e-01 -2.85911113e-01
5.68477690e-01 3.25893968e-01 -4.41566594e-02 -1.57011703e-01
-2.06686556e-01 3.90331030e-01 -2.34075002e-02 5.95071375e-01
-1.55148990e-02 -2.36777402e-02 -1.62584484e-01 4.32275772e-01
-8.09214786e-02 -3.39146294e-02 5.84561348e-01 1.32447696e+00
-2.05448344e-01 -9.70448256e-01 -9.27807689e-01 3.79487008e-01
-5.48769534e-01 -6.49255216e-02 -2.11655796e-01 8.12650323e-01
3.00434381e-01 1.14083099e+00 -2.73612499e-01 -3.59878927e-01
1.20347440e-01 -2.84082115e-01 6.73200071e-01 -3.60388696e-01
-5.14688194e-01 -1.48622006e-01 -4.96961087e-01 -7.03124464e-01
2.34120004e-02 -6.18406475e-01 -9.25098121e-01 -5.18896997e-01
-7.50870764e-01 -2.94164866e-01 3.83642882e-01 6.47815883e-01
3.94383281e-01 1.62832662e-01 8.67624104e-01 -8.47406864e-01
-8.30543160e-01 -7.13670015e-01 -9.92633104e-01 7.67936528e-01
4.96617585e-01 -8.41454387e-01 -8.17710876e-01 2.15460718e-01] | [12.361136436462402, 0.9517622590065002] |
6c5ae8c0-d90a-4c97-9360-5304a11b6bf4 | data-driven-low-rank-neural-network | 2107.05787 | null | https://arxiv.org/abs/2107.05787v1 | https://arxiv.org/pdf/2107.05787v1.pdf | Data-Driven Low-Rank Neural Network Compression | Despite many modern applications of Deep Neural Networks (DNNs), the large number of parameters in the hidden layers makes them unattractive for deployment on devices with storage capacity constraints. In this paper we propose a Data-Driven Low-rank (DDLR) method to reduce the number of parameters of pretrained DNNs and expedite inference by imposing low-rank structure on the fully connected layers, while controlling for the overall accuracy and without requiring any retraining. We pose the problem as finding the lowest rank approximation of each fully connected layer with given performance guarantees and relax it to a tractable convex optimization problem. We show that it is possible to significantly reduce the number of parameters in common DNN architectures with only a small reduction in classification accuracy. We compare DDLR with Net-Trim, which is another data-driven DNN compression technique based on sparsity and show that DDLR consistently produces more compressed neural networks while maintaining higher accuracy. | ['Swayambhoo Jain', 'Dimitris Papadimitriou'] | 2021-07-13 | null | null | null | null | ['neural-network-compression', 'neural-network-compression'] | ['methodology', 'miscellaneous'] | [ 2.31352791e-01 2.35591367e-01 -2.54495382e-01 -6.64027452e-01
-4.51901734e-01 -5.05346417e-01 3.01053822e-01 -1.84769481e-01
-8.04598212e-01 7.51679122e-01 1.51581451e-01 -5.25925815e-01
-5.08043230e-01 -6.84080184e-01 -1.07962871e+00 -6.68899715e-01
7.45404214e-02 6.33008361e-01 4.89700399e-02 4.79086071e-01
-2.30015200e-02 8.95856380e-01 -1.75127745e+00 2.50251651e-01
5.42414606e-01 1.37408710e+00 4.35171425e-01 6.09363973e-01
8.57796818e-02 9.58674073e-01 -4.83234048e-01 -5.90417743e-01
6.09840333e-01 1.08426787e-01 -6.35896623e-01 -1.79625303e-01
1.26634419e+00 -8.99636626e-01 -9.74038720e-01 1.16412318e+00
4.71383095e-01 1.05030663e-01 2.69196063e-01 -1.20539439e+00
-4.28496033e-01 8.73573840e-01 -1.42093733e-01 3.68552208e-01
-6.02238655e-01 -2.99441606e-01 1.09900939e+00 -8.00861061e-01
5.33515692e-01 1.17835796e+00 7.25411355e-01 6.66972578e-01
-1.38342011e+00 -6.85415506e-01 1.86383665e-01 2.80960910e-02
-1.10126400e+00 -1.07936788e+00 4.28436935e-01 -8.29896554e-02
1.08784044e+00 1.56475335e-01 6.36938572e-01 7.52769411e-01
-4.56858188e-01 7.29378462e-01 3.82588446e-01 -1.13851205e-01
5.00641584e-01 -2.43995011e-01 3.01925629e-01 8.16461027e-01
9.26663220e-01 -3.03692609e-01 -6.66562319e-01 -9.94591266e-02
9.18552458e-01 3.15360695e-01 3.59334126e-02 -2.86634713e-01
-8.83991301e-01 7.57341087e-01 4.77674335e-01 1.12352274e-01
-2.81233490e-01 8.19689155e-01 5.14196277e-01 2.64770418e-01
4.28973466e-01 3.85895669e-02 -7.50961065e-01 -3.69064927e-01
-1.38346171e+00 3.29029739e-01 8.22818458e-01 1.17220271e+00
6.08974993e-01 3.49381953e-01 1.40080675e-01 1.04712462e+00
2.00358346e-01 3.86280864e-01 2.99588650e-01 -1.43806958e+00
7.23900974e-01 3.66989464e-01 -6.64494097e-01 -8.86899471e-01
-2.11064667e-01 -6.72665417e-01 -1.15495276e+00 -9.44358557e-02
2.26892695e-01 -2.03943029e-01 -9.30749834e-01 1.88181090e+00
-6.37126341e-02 1.49913281e-01 1.52416989e-01 7.35293746e-01
6.31397545e-01 6.92266703e-01 -1.71875760e-01 -1.14991650e-01
9.25459385e-01 -9.68975365e-01 -5.74438155e-01 -3.81542474e-01
9.84058499e-01 -2.54768640e-01 7.81813443e-01 6.20663106e-01
-1.43421662e+00 -1.25670493e-01 -1.11168218e+00 -7.29453504e-01
-6.88089505e-02 4.12164718e-01 9.26190317e-01 4.13577348e-01
-1.35269809e+00 8.31575751e-01 -9.67771649e-01 1.59408227e-01
9.05385733e-01 8.97279263e-01 -4.51457530e-01 -4.41302896e-01
-7.26751685e-01 6.77874088e-01 4.79486257e-01 2.23493114e-01
-9.73471522e-01 -1.08877540e+00 -5.55472791e-01 3.01413119e-01
2.34933961e-02 -4.59292531e-01 1.17249000e+00 -5.52746534e-01
-1.14398897e+00 6.51581466e-01 -2.03116566e-01 -9.79205310e-01
3.84072959e-01 -3.47556502e-01 -1.11073442e-01 2.31701374e-01
-3.78876299e-01 1.07765663e+00 6.25527143e-01 -7.43409514e-01
-6.40942454e-01 -5.06147444e-01 5.84097952e-02 -1.03364140e-01
-1.00009656e+00 -4.77021277e-01 -7.54929364e-01 -6.17339909e-01
4.80708957e-01 -9.26391423e-01 -2.23074123e-01 5.74156940e-01
-3.09369624e-01 -1.84848338e-01 1.05717313e+00 -6.41970694e-01
1.16376877e+00 -2.19356918e+00 1.32021889e-01 2.67620534e-01
6.49388850e-01 4.67518687e-01 -3.28193784e-01 -2.30519861e-01
1.65916726e-01 6.87469169e-02 -1.74110606e-01 -7.16429710e-01
-2.32510921e-02 8.87801826e-01 -3.69837314e-01 4.55648601e-01
-4.41402271e-02 6.34005308e-01 -4.88480091e-01 -4.16007221e-01
-2.19962478e-01 7.59085894e-01 -1.19356227e+00 -7.07792565e-02
-2.26409748e-01 -4.38022703e-01 -9.41480100e-02 3.59947294e-01
7.67572284e-01 -2.47706115e-01 2.36926228e-01 -5.79905331e-01
1.51938766e-01 7.31165409e-01 -9.23310935e-01 1.72284126e+00
-4.30157483e-01 1.12892461e+00 4.20526594e-01 -1.18569100e+00
8.19644868e-01 7.74872005e-02 3.79825801e-01 -6.51399553e-01
5.18997796e-02 3.40902001e-01 -7.27521628e-02 -1.03560276e-01
3.60321581e-01 1.69878602e-01 3.87597322e-01 1.99649855e-01
2.66523123e-01 4.03696567e-01 4.31421101e-01 2.82390386e-01
1.22486830e+00 -6.49179101e-01 -5.16727090e-01 -4.36394632e-01
-2.61140652e-02 -2.07223967e-01 6.18857503e-01 6.21402562e-01
1.62682027e-01 3.79147261e-01 5.83067060e-01 -5.68854809e-01
-1.48960233e+00 -9.69093740e-01 -3.76833826e-01 1.11145198e+00
-6.00049317e-01 -2.99366832e-01 -8.29882681e-01 -2.13226736e-01
2.03742579e-01 4.66102988e-01 -2.53422260e-01 -1.26011312e-01
-6.09306574e-01 -6.64208770e-01 8.91651511e-01 6.31485164e-01
6.43721759e-01 -3.67672175e-01 -4.59590524e-01 -4.35428917e-02
1.34176150e-01 -1.41603744e+00 -2.58322418e-01 5.59990108e-01
-1.67618728e+00 -5.27417064e-01 -6.60363793e-01 -9.43409085e-01
1.07595634e+00 2.15614095e-01 1.00954247e+00 1.24664448e-01
-2.09514901e-01 -2.14014754e-01 1.25669405e-01 -2.07704410e-01
2.51243841e-02 4.10057873e-01 3.74893188e-01 -4.21653181e-01
2.82696337e-01 -9.35152650e-01 -4.89609569e-01 -1.95465624e-01
-1.14764023e+00 1.14826530e-01 7.31923938e-01 5.34985900e-01
9.49314654e-01 2.56552935e-01 3.12933028e-01 -1.01483536e+00
3.90812695e-01 -2.51660913e-01 -9.63196337e-01 1.17326248e-02
-7.60706186e-01 5.37884355e-01 7.37060010e-01 -6.14407957e-01
-3.59008461e-01 2.02620491e-01 -1.14560492e-01 -9.11159933e-01
3.37443769e-01 5.09815872e-01 -1.75060689e-01 -1.26465812e-01
3.90007585e-01 1.86658159e-01 6.45907782e-03 -8.57866168e-01
3.46880943e-01 4.40456182e-01 7.98045516e-01 -5.25197208e-01
5.96682072e-01 3.67426693e-01 3.21949005e-01 -8.59204412e-01
-1.16944575e+00 -1.61694065e-02 -4.36061978e-01 1.99022859e-01
4.73280162e-01 -1.23012257e+00 -6.98404431e-01 6.95655644e-02
-1.19810116e+00 -4.69027221e-01 -4.52072144e-01 5.02808154e-01
-2.69320726e-01 6.71380088e-02 -7.93184280e-01 -5.22886693e-01
-5.71908951e-01 -8.72483432e-01 6.11219764e-01 -2.19761536e-01
2.01457962e-01 -7.75649488e-01 -1.81252554e-01 3.58587891e-01
5.32425165e-01 -1.41481400e-01 1.03742576e+00 -7.09558666e-01
-7.98691273e-01 -1.86835647e-01 -6.03991389e-01 6.90204680e-01
-3.91348302e-01 -2.23423228e-01 -9.87508535e-01 -5.26890755e-01
-1.00647368e-01 -4.30055082e-01 1.12781274e+00 4.94435281e-01
1.73281312e+00 -1.05557287e+00 -2.07017045e-02 1.14599347e+00
1.45198858e+00 -1.16920210e-01 5.27480900e-01 1.14632985e-02
8.76282275e-01 2.84407914e-01 -1.04402885e-01 4.85936940e-01
4.43954729e-02 2.91882813e-01 5.33173084e-01 8.19494650e-02
-1.49916321e-01 -3.36021781e-01 2.53770232e-01 1.19027305e+00
6.20769449e-02 -3.28977019e-01 -6.07186496e-01 6.03104234e-01
-1.74604738e+00 -7.94536531e-01 1.56184584e-01 2.09801984e+00
1.06327820e+00 2.03105807e-01 -1.24489129e-01 3.46328408e-01
4.66439426e-01 1.24132708e-01 -9.28954601e-01 -3.93138975e-01
-3.19886953e-01 1.30379871e-01 1.08877504e+00 3.62863243e-01
-7.59680986e-01 6.61103606e-01 6.89210224e+00 7.05466568e-01
-9.10651922e-01 7.42154643e-02 9.22438025e-01 -9.93672848e-01
-2.48759344e-01 -2.37136960e-01 -1.21736968e+00 2.88960099e-01
1.37994099e+00 -5.46750575e-02 7.67867744e-01 1.17190242e+00
-3.30972411e-02 4.32047099e-01 -1.46634829e+00 1.37663090e+00
-1.71117276e-01 -1.80774844e+00 4.85564291e-01 3.60981196e-01
9.05144274e-01 5.81792355e-01 1.23915434e-01 -8.14799145e-02
3.11272591e-01 -1.09195685e+00 7.11964667e-01 2.89665252e-01
1.05435264e+00 -1.01534140e+00 4.62262809e-01 2.82020450e-01
-7.29036033e-01 -3.04827452e-01 -1.12744987e+00 3.18338908e-02
-1.65528610e-01 1.18380272e+00 -3.49752575e-01 -4.98322248e-01
8.25982332e-01 6.80092514e-01 -2.78645426e-01 8.60840321e-01
3.20961565e-01 6.39271557e-01 -7.03926802e-01 9.40892994e-02
1.73325226e-01 -1.43486395e-01 2.17932209e-01 1.09186506e+00
4.39718872e-01 2.15199500e-01 -3.28722149e-01 7.31639802e-01
-8.56784046e-01 -3.41042489e-01 -5.93887329e-01 -2.96408892e-01
7.84938276e-01 9.21783864e-01 -4.21787947e-01 -3.88687015e-01
-1.45270437e-01 8.03009450e-01 7.44863331e-01 1.80424884e-01
-4.70904946e-01 -2.30407029e-01 9.35163617e-01 2.91590065e-01
5.67652583e-01 -4.82797712e-01 -4.87314761e-01 -9.98642445e-01
2.92982996e-01 -6.70936644e-01 1.58392683e-01 -4.05810297e-01
-1.03110015e+00 5.44359803e-01 -1.94714978e-01 -7.60005414e-01
-1.01206731e-02 -8.26117039e-01 4.14487086e-02 4.40028578e-01
-1.44934404e+00 -5.64730048e-01 -6.67347983e-02 5.46319842e-01
2.92196900e-01 -2.63523549e-01 5.58696747e-01 9.23879385e-01
-8.46807241e-01 1.06454611e+00 5.21779597e-01 1.91482887e-01
2.06897154e-01 -7.87203312e-01 1.70076132e-01 7.82242298e-01
8.52705389e-02 7.14708686e-01 5.67994535e-01 -3.04328024e-01
-1.75159395e+00 -1.36151993e+00 1.06594312e+00 2.31885493e-01
4.47833300e-01 -5.94061375e-01 -7.60327160e-01 7.01915860e-01
-4.17512178e-01 3.22572589e-01 5.70712328e-01 3.90831083e-02
-6.07822239e-01 -6.68962479e-01 -1.17202866e+00 5.25862515e-01
1.47330296e+00 -5.76743424e-01 -1.84529811e-01 6.38330460e-01
1.03260303e+00 -2.59538621e-01 -9.22910213e-01 2.28683189e-01
4.58533645e-01 -5.83343089e-01 1.09145641e+00 -6.33109987e-01
5.99700391e-01 8.09091777e-02 -6.09209657e-01 -6.84379041e-01
-2.38811955e-01 -5.04082680e-01 -8.07576060e-01 1.05865622e+00
5.15818775e-01 -4.52542037e-01 1.49978280e+00 1.11255348e+00
-1.08823575e-01 -8.21464956e-01 -1.14961338e+00 -8.86000454e-01
-1.10331371e-01 -4.42005813e-01 3.52498502e-01 7.04744101e-01
-3.40396196e-01 1.55806303e-01 -3.52291703e-01 1.29782870e-01
8.03287089e-01 -2.42854595e-01 3.40427518e-01 -1.47193313e+00
-2.66372412e-01 -3.53340030e-01 -4.86422181e-01 -1.52506018e+00
1.59276858e-01 -1.16744602e+00 -6.65595010e-02 -1.42179692e+00
2.34252185e-01 -7.50894070e-01 -2.70270556e-01 7.11649537e-01
6.77676857e-01 3.26426715e-01 1.01592965e-01 3.90606135e-01
-6.06717050e-01 4.30854708e-01 7.67071366e-01 -2.42066309e-02
1.67189147e-02 -2.93171465e-01 -7.03875422e-01 6.04591906e-01
6.24815822e-01 -6.80328667e-01 -7.91120589e-01 -1.44807065e+00
4.30984259e-01 -1.59840688e-01 2.33144537e-01 -1.32676148e+00
7.19927967e-01 3.66681784e-01 4.55130965e-01 -8.03559124e-01
4.51360434e-01 -9.99458551e-01 -5.32045402e-02 5.72950959e-01
-7.75318027e-01 2.84222573e-01 2.38960385e-01 4.58811760e-01
-7.48349726e-02 -2.67161787e-01 9.49187100e-01 1.96895421e-01
-2.55751252e-01 7.59032130e-01 -1.41925618e-01 1.23327784e-01
5.53346455e-01 -1.70804933e-01 -4.32892323e-01 -1.34460896e-01
-4.75978494e-01 1.68065354e-02 1.33173347e-01 5.88473082e-02
7.73032010e-01 -1.34529543e+00 -4.98527914e-01 2.71761239e-01
-5.11067510e-01 5.47297120e-01 1.40709519e-01 3.91768813e-01
-8.80131304e-01 7.95457542e-01 -1.48994088e-01 -3.64161998e-01
-1.23993945e+00 2.24348933e-01 1.85702637e-01 -1.63953573e-01
-6.54290617e-01 1.24634981e+00 -1.83143899e-01 -2.56089389e-01
9.84623075e-01 -4.64762747e-01 2.48303667e-01 -7.02649774e-03
6.42255902e-01 5.44199347e-01 3.69243592e-01 -2.20670342e-01
-2.09746912e-01 1.66448951e-01 -3.43308300e-01 2.13045999e-02
1.92075467e+00 1.85103580e-01 -2.98116773e-01 1.94480922e-02
1.80922902e+00 -5.55479646e-01 -1.59650803e+00 -4.83950078e-01
-5.58115356e-02 -2.12268785e-01 8.95794094e-01 -3.55515093e-01
-1.93388975e+00 9.10036266e-01 6.43587351e-01 -2.61769235e-01
1.21361184e+00 -1.47403210e-01 1.18298233e+00 1.21735847e+00
1.96216136e-01 -1.20308220e+00 -1.49201035e-01 5.60916722e-01
5.51712036e-01 -7.64019608e-01 2.81275272e-01 -1.20649941e-01
-1.99145116e-02 1.12197554e+00 4.76342529e-01 -2.92022645e-01
8.80422115e-01 7.30406523e-01 -4.82878178e-01 -1.47082210e-01
-1.21589029e+00 4.75746304e-01 1.84444375e-02 2.60862738e-01
1.47791564e-01 -2.62736738e-01 1.17295086e-01 3.23055625e-01
-5.19398808e-01 4.19902205e-02 1.44950405e-01 7.58757234e-01
-6.62037015e-01 -8.08446527e-01 1.81145772e-01 1.08380556e+00
-5.68554342e-01 -4.21194375e-01 -6.26978651e-02 3.44316274e-01
4.89031896e-02 5.49877703e-01 6.61745310e-01 -3.79567146e-01
1.11079179e-01 -1.01202220e-01 3.50651860e-01 -3.50435555e-01
-1.27237156e-01 -2.59837508e-01 9.62136537e-02 -6.25935972e-01
-8.36205557e-02 -4.82652038e-01 -1.20715272e+00 -9.38645065e-01
-1.51275858e-01 -1.16712630e-01 9.72403944e-01 7.40664840e-01
6.87676430e-01 3.28005552e-01 2.47767404e-01 -7.09751070e-01
-7.17074990e-01 -5.61475754e-01 -5.20212889e-01 3.14034447e-02
5.78218400e-01 -1.67454451e-01 -4.54983830e-01 2.04903185e-02] | [8.578060150146484, 3.1416361331939697] |
6320ea0b-d582-4435-9701-b10f982a6715 | wasserstein-image-local-analysis-histogram-of | 2205.05606 | null | https://arxiv.org/abs/2205.05606v1 | https://arxiv.org/pdf/2205.05606v1.pdf | Wasserstein Image Local Analysis: Histogram of Orientations, Smoothing and Edge Detection | The Histogram of Oriented Gradient is a widely used image feature, which describes local image directionality based on numerical differentiation. Due to its ill-posed nature, small noise may lead to large errors. Conventional HOG may fail to produce meaningful directionality results in the presence of noise, which is common in medical radiographic imaging. We approach the directionality problem from a novel perspective by the use of the optimal transport map of a local image patch to a uni-color patch of its mean. We decompose the transport map into sub-work costs in different directions. We evaluated the ability of the optimal transport to quantify tumor heterogeneity from brain MRI images of patients with glioblastoma multiforme from the TCIA. By considering the entropy difference of the extracted local directionality within tumor regions, we found that patients with higher entropy in their images, had statistically significant worse overall survival (p $=0.008$), which indicates that tumors exhibiting flows in many directions may be more malignant, perhaps reflecting high tumor histologic grade, a reflection of histologic disorganization. We also explored the possibility of solving classical image processing problems such as smoothing and edge detection via optimal transport. By looking for a 2-color patch with minimum transport distance to a local patch, we derive a nonlinear shock filter, which preserves edges. Moreover, we found that the color difference of the computed 2-color patch indicates whether there is a large change in color, i.e., an edge in the given patch. In summary, we expand the usefulness of optimal transport as an image local analysis tool, to extract robust measures of imaging tumor heterogeneity for outcomes prediction as well as image pre-processing. Because of its robust nature, we find it offers several advantages over the classical approaches. | ['Allen Tannenbaum', 'Joseph O. Deasy', 'Larry Norton', 'Harini Veeraraghavan', 'Jiening Zhu'] | 2022-05-11 | null | null | null | null | ['edge-detection'] | ['computer-vision'] | [ 1.34014040e-01 -8.59483033e-02 -6.57096207e-02 7.62160271e-02
-6.32835150e-01 -2.89074898e-01 2.48978317e-01 5.14231920e-01
-5.82336605e-01 6.40654743e-01 2.08402410e-01 -2.95536578e-01
-2.17383817e-01 -8.54634941e-01 -3.46578270e-01 -1.26133287e+00
-4.03192878e-01 -3.48658785e-02 2.71108866e-01 -7.77611732e-02
5.59520781e-01 7.80051887e-01 -9.98350024e-01 2.72557706e-01
8.02710295e-01 1.13970768e+00 4.40775096e-01 6.11756146e-01
6.46015629e-02 6.40944242e-01 -2.55038708e-01 1.33921862e-01
3.56593914e-02 -6.01215065e-01 -7.15606928e-01 4.29116040e-02
1.78862810e-01 -1.04500383e-01 -4.25971001e-02 1.36708713e+00
4.51451510e-01 -2.53332444e-02 9.69828725e-01 -9.46816385e-01
-3.24674666e-01 3.85732353e-02 -5.77909470e-01 4.95977998e-01
7.39603415e-02 3.50508332e-01 5.97620130e-01 -7.38898158e-01
1.00641012e+00 5.15829027e-01 7.15880036e-01 1.02458715e-01
-1.31509864e+00 -2.38847792e-01 -2.76971668e-01 3.68718445e-01
-1.35284615e+00 -9.86640155e-03 7.77192235e-01 -8.28947067e-01
6.53401554e-01 5.77105284e-01 9.67020512e-01 4.53827083e-01
1.05701816e+00 2.91613370e-01 1.62466383e+00 -4.25690532e-01
3.11410427e-01 -2.43522543e-02 7.14770257e-02 1.11695850e+00
2.74326473e-01 1.34373784e-01 -3.10416400e-01 -9.57048014e-02
7.36607075e-01 -8.91274661e-02 -6.63963854e-01 -3.10066611e-01
-1.42808127e+00 6.16497695e-01 6.64894223e-01 7.17843473e-01
-4.61864829e-01 1.36776432e-01 2.85207272e-01 6.09242450e-03
3.62444669e-01 3.75280589e-01 1.67805061e-01 3.63029093e-02
-7.30925024e-01 -1.09303132e-01 4.73583281e-01 1.27240792e-01
4.31989759e-01 -3.25792670e-01 -1.86989859e-01 3.73081416e-01
-1.90948881e-02 3.13577503e-01 7.95731366e-01 -9.10051763e-01
-5.14578968e-02 7.35785425e-01 -3.40798527e-01 -1.18037438e+00
-7.12981939e-01 -3.97848159e-01 -1.05787146e+00 6.79988563e-01
7.03995347e-01 1.63598806e-01 -6.78200364e-01 1.49724901e+00
2.36090392e-01 -9.17161331e-02 -6.90507516e-02 9.71682727e-01
4.61619079e-01 2.63628125e-01 3.06541938e-02 -4.33829129e-01
1.54331398e+00 -5.13330281e-01 -6.61368728e-01 1.64090514e-01
9.29663420e-01 -6.84351742e-01 9.61323321e-01 1.92600191e-01
-1.07870328e+00 8.10643137e-02 -8.61668706e-01 1.93456382e-01
-5.44901907e-01 -1.70227185e-01 4.32286739e-01 4.65409011e-01
-1.29789221e+00 6.99350297e-01 -9.04039979e-01 -4.64322329e-01
2.00417221e-01 6.59271926e-02 -5.70535004e-01 -2.85715330e-02
-7.30696917e-01 9.69208121e-01 5.75521402e-02 -4.02030200e-02
-1.98415011e-01 -8.18561316e-01 -6.76536798e-01 -5.13793565e-02
-1.18259713e-01 -6.30558729e-01 5.04548669e-01 -8.63810956e-01
-1.09857583e+00 8.58277142e-01 -2.50420392e-01 -1.05383150e-01
7.82001436e-01 8.31713140e-01 -2.07926154e-01 5.67058206e-01
9.75713432e-02 3.16049159e-01 6.74923241e-01 -1.16110563e+00
-5.40637553e-01 -6.61931574e-01 -4.45458293e-01 2.67247051e-01
-2.07910568e-01 -1.25067383e-01 7.78246298e-02 -7.01120615e-01
4.89789903e-01 -9.09707367e-01 -3.69533181e-01 2.24083692e-01
-3.32947582e-01 2.36609757e-01 5.68961024e-01 -8.31317127e-01
1.17445827e+00 -2.25081849e+00 -9.45889279e-02 6.20092750e-01
5.19006670e-01 -4.50548053e-01 2.80099630e-01 -6.75282553e-02
-1.79031953e-01 2.78293729e-01 -6.27220392e-01 4.14224982e-01
-3.55416328e-01 -1.89815626e-01 4.78763521e-01 9.81924772e-01
-1.08565316e-01 7.77185321e-01 -1.04467905e+00 -7.51232922e-01
-3.12664313e-03 4.67870146e-01 -3.46568465e-01 -1.83103547e-01
6.20944917e-01 3.28287393e-01 -1.67625561e-01 5.62333763e-01
7.09008098e-01 -1.15816429e-01 -1.96985062e-02 -3.89182925e-01
-5.76063812e-01 -3.17918390e-01 -9.59551454e-01 1.12539279e+00
-1.63602710e-01 1.08194005e+00 7.21288100e-02 -9.69259024e-01
6.51069045e-01 2.51563430e-01 1.04112029e+00 -8.51251423e-01
6.31199181e-02 4.95678961e-01 2.69515604e-01 -6.75828516e-01
3.01993787e-01 -4.58255559e-01 2.37490714e-01 2.76310802e-01
-5.42177618e-01 -2.20779195e-01 6.04977347e-02 -3.01212538e-02
1.32822359e+00 -6.44210577e-01 3.97918016e-01 -8.63853157e-01
5.45183480e-01 1.54783413e-01 2.07640603e-01 6.13315165e-01
-5.74263394e-01 8.26696455e-01 8.62630010e-01 -2.26899385e-01
-8.98191452e-01 -9.25346315e-01 -6.28314018e-01 2.73819745e-01
2.68470764e-01 1.38030499e-01 -7.34570503e-01 -4.94882643e-01
-3.96641456e-02 2.92120457e-01 -7.66532481e-01 -2.53692865e-01
-5.24668097e-01 -1.27328861e+00 1.53583810e-01 2.86092371e-01
4.71832097e-01 -8.98087621e-01 -1.11052966e+00 2.98045158e-01
-3.93197894e-01 -7.16580093e-01 -5.35428405e-01 2.91798174e-01
-9.53312576e-01 -1.06194282e+00 -1.13263786e+00 -8.02513599e-01
1.16584468e+00 2.73949623e-01 7.79349625e-01 3.26688766e-01
-8.18078935e-01 3.51548582e-01 -1.63421288e-01 9.42240357e-02
-3.71199220e-01 -2.90031970e-01 -2.29767889e-01 -2.25167815e-02
-5.55730313e-02 -3.98274511e-02 -1.05829906e+00 3.40937972e-01
-1.02681315e+00 -4.91426187e-03 5.35670400e-01 1.01052928e+00
7.19312370e-01 3.74190450e-01 -4.51061577e-02 -4.13849920e-01
7.81409979e-01 -3.94851387e-01 -3.15918326e-01 1.38623640e-01
-8.27041626e-01 1.23051114e-01 4.37287211e-01 -1.54100090e-01
-9.07676399e-01 -1.65437669e-01 2.59216517e-01 2.28826419e-01
-1.11580148e-01 6.28981054e-01 4.55687553e-01 -5.45867324e-01
6.50582552e-01 1.75531924e-01 3.88287693e-01 2.38846809e-01
-1.83521539e-01 3.58282089e-01 3.20896327e-01 -1.83607921e-01
3.64210367e-01 1.03739381e+00 5.39670289e-01 -9.25084412e-01
4.60604876e-02 -7.73294508e-01 -4.44043010e-01 -6.21383905e-01
1.14864194e+00 -6.40408844e-02 -9.62805629e-01 5.31746805e-01
-8.34588468e-01 -4.69339132e-01 -2.75092632e-01 8.36403489e-01
-6.76927745e-01 4.08446312e-01 -7.08274782e-01 -4.31849658e-01
-2.51870334e-01 -1.55749726e+00 6.97274506e-01 2.27562457e-01
-8.44557807e-02 -1.33690202e+00 5.35455272e-02 -9.06418934e-02
5.71524799e-01 4.43624705e-01 1.38048112e+00 -5.88843860e-02
-4.43135738e-01 -1.99689552e-01 -3.08005750e-01 5.95846660e-02
1.29433855e-01 2.10405856e-01 -6.02210462e-01 -1.88162163e-01
9.79672670e-02 2.70348817e-01 7.87001967e-01 9.57920074e-01
1.01130080e+00 -6.05559535e-02 -4.95159298e-01 7.65141606e-01
1.80206347e+00 5.02521753e-01 7.10200191e-01 7.00384855e-01
3.93906415e-01 7.53577590e-01 3.44498605e-01 3.31084818e-01
4.35770452e-02 5.36334991e-01 2.88026482e-01 -6.39815509e-01
-3.14668030e-01 2.75903225e-01 1.47420377e-01 6.56016827e-01
-5.74027896e-01 2.20218644e-01 -1.03478312e+00 4.45519358e-01
-1.39264941e+00 -8.53338599e-01 -4.18129534e-01 2.35772157e+00
5.50360978e-01 -1.52406050e-02 -1.22502811e-01 1.13005042e-01
5.96312344e-01 -1.39801830e-01 -1.89560592e-01 -3.27093184e-01
-1.14741214e-01 -2.00077578e-01 1.03847897e+00 6.88239396e-01
-6.08626187e-01 2.48842120e-01 6.49634504e+00 7.18239307e-01
-1.52091730e+00 -8.01348388e-02 9.83734846e-01 2.29249924e-01
-4.05754358e-01 -8.31006318e-02 -2.23893002e-01 5.73266625e-01
4.88937587e-01 -8.82005915e-02 1.44102275e-01 3.37076843e-01
5.34874201e-01 -7.93671012e-01 -6.11439586e-01 8.04664552e-01
-3.40811908e-02 -1.29120505e+00 -2.32808679e-01 6.14270568e-01
4.64422345e-01 -2.75433749e-01 2.27126122e-01 -4.46517080e-01
-1.57215148e-01 -8.34157169e-01 6.48574054e-01 6.68367743e-01
6.87838376e-01 -4.66525823e-01 9.62675154e-01 1.16919493e-02
-1.15540648e+00 1.39177382e-01 -2.30359957e-01 3.42046857e-01
8.37457553e-02 7.19294846e-01 -1.10532546e+00 2.96012878e-01
6.91137910e-01 3.50768834e-01 -5.19338727e-01 1.31780601e+00
3.30970168e-01 1.21257044e-01 -2.51041740e-01 -2.50869989e-02
2.30671927e-01 -5.74431658e-01 6.79299474e-01 1.33903313e+00
6.24795556e-01 9.21903327e-02 -1.69472039e-01 7.32831180e-01
3.96907449e-01 4.70724910e-01 -6.04045033e-01 2.91425645e-01
3.82967107e-02 1.36605799e+00 -1.52070343e+00 -4.04645085e-01
-4.57539499e-01 8.54172885e-01 1.91915389e-02 5.67205667e-01
-4.65851009e-01 -3.96078587e-01 3.81133437e-01 3.42965603e-01
-2.69532144e-01 -1.26782626e-01 -7.21622944e-01 -8.28145146e-01
-8.65644217e-03 -4.31793332e-01 4.83840376e-01 -7.43252814e-01
-8.87521684e-01 6.51264429e-01 -1.01101391e-01 -1.25650048e+00
3.54160145e-02 -8.27149808e-01 -6.25978887e-01 8.13677132e-01
-1.50998008e+00 -5.39116800e-01 -5.84593832e-01 6.39811218e-01
8.82347673e-02 3.83479655e-01 5.27705789e-01 -8.59671533e-02
-3.46973211e-01 2.83018708e-01 1.60509616e-01 -1.17342636e-01
4.82195526e-01 -1.47771752e+00 -5.13650060e-01 8.56675327e-01
-6.13717914e-01 2.48374224e-01 6.61284089e-01 -6.63742006e-01
-1.08247900e+00 -4.76308167e-01 7.58243263e-01 4.84760627e-02
9.35702682e-01 1.90364823e-01 -8.69218111e-01 2.04170004e-01
1.60062373e-01 1.90592393e-01 4.68795180e-01 -6.70378506e-01
2.92690128e-01 5.95098212e-02 -1.37253404e+00 8.15465808e-01
6.88007057e-01 -3.23253840e-01 -5.27844131e-02 2.71578640e-01
-1.03853203e-01 -7.42174089e-02 -1.29041612e+00 4.58531857e-01
5.44588804e-01 -1.29944062e+00 8.85118246e-01 -4.29153591e-02
2.93689400e-01 -7.12506697e-02 8.94816071e-02 -1.54674006e+00
-6.15716636e-01 -1.73505098e-01 8.35619390e-01 3.98462474e-01
4.23651248e-01 -9.22576547e-01 7.20922291e-01 6.71081841e-01
-2.59899914e-01 -8.71162117e-01 -1.12960768e+00 -7.34234333e-01
1.19945966e-01 -8.04587007e-02 -8.77036992e-03 8.93196166e-01
6.56238079e-01 -5.62316895e-01 5.51844776e-01 -1.30967628e-02
6.74307466e-01 -1.37238875e-01 -2.31961850e-02 -8.79599929e-01
7.09232688e-02 -1.13886631e+00 -8.55118394e-01 -2.40888819e-01
-2.02054709e-01 -1.13052738e+00 1.07824892e-01 -1.64016879e+00
3.15818280e-01 -5.59904695e-01 -3.69311184e-01 1.13914073e-01
-8.29465389e-02 3.22534472e-01 -1.70272458e-02 4.22894061e-01
2.81055242e-01 -1.80941224e-02 1.72117937e+00 -2.62930065e-01
-2.09257632e-01 -4.75987732e-01 -2.41610676e-01 7.90932178e-01
6.95224881e-01 -3.69165570e-01 -1.09070372e-02 1.54168338e-01
-8.07116032e-02 1.94487110e-01 4.61547375e-01 -1.04522860e+00
2.80439198e-01 -2.20799953e-01 3.29610735e-01 -1.30643174e-01
5.97434230e-02 -9.01250720e-01 9.59777310e-02 1.08024561e+00
-1.42431468e-01 2.27445945e-01 -2.18697026e-01 1.90515980e-01
-2.57744372e-01 -4.70720351e-01 9.77561891e-01 -3.59524220e-01
-7.31919408e-01 1.04137070e-01 -7.90629685e-01 -3.02079618e-01
1.28378308e+00 -5.45085967e-01 -4.34052110e-01 -1.19822733e-01
-7.47956097e-01 -4.77863520e-01 7.52149642e-01 -4.15370226e-01
7.23252535e-01 -1.12503314e+00 -5.57005644e-01 2.90003002e-01
1.08562164e-01 -4.41122353e-01 3.61923426e-01 1.86537147e+00
-1.14525163e+00 2.07816899e-01 -4.06444341e-01 -7.46807516e-01
-1.23731077e+00 3.92212033e-01 7.55249202e-01 -2.55053192e-01
-9.45612252e-01 6.25428259e-01 3.08489591e-01 3.92815948e-01
-1.22257113e-01 -7.26312399e-01 -2.26278901e-01 1.96404338e-01
3.59022796e-01 5.73065639e-01 4.20490384e-01 -7.62032807e-01
-4.40993667e-01 8.59511256e-01 1.95627525e-01 -3.35765630e-01
8.36421013e-01 -9.26320180e-02 -4.07841533e-01 2.81673968e-01
1.58392477e+00 8.11356977e-02 -1.15593839e+00 2.63806760e-01
9.87895671e-03 -6.06410503e-01 5.11576176e-01 -5.38879454e-01
-1.26250947e+00 4.54640985e-01 1.04486167e+00 5.28346300e-01
1.24000192e+00 5.48562184e-02 4.74016726e-01 -1.73597232e-01
1.99398518e-01 -1.06072342e+00 -4.93848324e-02 4.75998260e-02
7.17532396e-01 -1.15857601e+00 -1.98372453e-01 -4.70204681e-01
-4.44742054e-01 1.28234875e+00 1.45858556e-01 2.68152189e-02
7.93275833e-01 5.52024722e-01 2.53752232e-01 -4.20832098e-01
-1.79445460e-01 -1.91158026e-01 2.09377065e-01 4.98578280e-01
4.11403120e-01 3.87831062e-01 -7.97163069e-01 -1.50619566e-01
-1.26112849e-01 -1.62637569e-02 8.90286267e-01 1.02818751e+00
-6.19775593e-01 -4.89423841e-01 -6.47164941e-01 5.29611230e-01
-3.47757339e-01 -1.41511485e-01 8.31090882e-02 8.88070762e-01
-2.02019997e-02 5.22504151e-01 2.47119963e-01 -1.10641785e-01
3.82346183e-01 -1.14068061e-01 4.35806960e-01 1.23553351e-01
-2.81749457e-01 1.87578514e-01 -2.60118902e-01 -5.11716306e-01
-3.67130607e-01 -7.31562853e-01 -1.53360939e+00 -2.63439834e-01
1.44852325e-02 -1.04407705e-02 8.81793916e-01 6.78483903e-01
-1.10775582e-01 6.02771938e-01 4.93904322e-01 -5.92622697e-01
-5.72988465e-02 -5.15994668e-01 -1.00793600e+00 6.55825257e-01
4.70492870e-01 -6.72752261e-01 -7.20506132e-01 1.25224322e-01] | [14.353963851928711, -2.5819859504699707] |
ec369e0e-31ea-4f1f-b352-4ca632c7d064 | speech-augmentation-based-unsupervised | 2205.14329 | null | https://arxiv.org/abs/2205.14329v1 | https://arxiv.org/pdf/2205.14329v1.pdf | Speech Augmentation Based Unsupervised Learning for Keyword Spotting | In this paper, we investigated a speech augmentation based unsupervised learning approach for keyword spotting (KWS) task. KWS is a useful speech application, yet also heavily depends on the labeled data. We designed a CNN-Attention architecture to conduct the KWS task. CNN layers focus on the local acoustic features, and attention layers model the long-time dependency. To improve the robustness of KWS model, we also proposed an unsupervised learning method. The unsupervised loss is based on the similarity between the original and augmented speech features, as well as the audio reconstructing information. Two speech augmentation methods are explored in the unsupervised learning: speed and intensity. The experiments on Google Speech Commands V2 Dataset demonstrated that our CNN-Attention model has competitive results. Moreover, the augmentation based unsupervised learning could further improve the classification accuracy of KWS task. In our experiments, with augmentation based unsupervised learning, our KWS model achieves better performance than other unsupervised methods, such as CPC, APC, and MPC. | ['Jing Xiao', 'Haobin Tang', 'Ning Cheng', 'Jianzong Wang', 'Jian Luo'] | 2022-05-28 | null | null | null | null | ['keyword-spotting'] | ['speech'] | [-5.30068055e-02 8.83622691e-02 -1.45017833e-01 -5.20950377e-01
-7.82519519e-01 2.55211473e-01 3.12032223e-01 -1.01378299e-01
-5.00845075e-01 2.35438585e-01 6.65568590e-01 -4.87682611e-01
2.80227184e-01 -4.68433976e-01 -5.41409194e-01 -7.29245842e-01
3.67231995e-01 5.10356463e-02 2.54149586e-01 -1.25130057e-01
2.34721061e-02 1.71809420e-01 -1.35218537e+00 2.65468627e-01
8.66208971e-01 1.16099429e+00 5.24467111e-01 5.62118709e-01
-4.64407921e-01 9.38253701e-01 -6.98943198e-01 1.19380496e-01
-1.15639791e-01 -1.78130239e-01 -6.09045863e-01 -1.36244535e-01
-3.46310794e-01 -2.92287946e-01 -7.82012045e-01 8.15806031e-01
8.52034032e-01 2.86931753e-01 5.34732997e-01 -1.36991203e+00
-6.65086508e-01 1.02056992e+00 -1.54824957e-01 3.23841572e-01
4.47757542e-02 2.33773753e-01 1.04027879e+00 -1.14661694e+00
-1.79642558e-01 1.07410192e+00 6.58737957e-01 3.95694911e-01
-7.14188457e-01 -1.00715971e+00 4.06781107e-01 6.14962280e-01
-1.43604398e+00 -7.95851469e-01 9.62459445e-01 -2.38303691e-01
1.16800678e+00 2.38351017e-01 4.62534577e-01 1.13394511e+00
-3.20570320e-01 1.36793709e+00 6.95826888e-01 -5.62235177e-01
8.89111534e-02 2.54705876e-01 1.32896751e-01 2.68021673e-01
-6.74567640e-01 1.48360059e-01 -6.43339038e-01 1.10866390e-02
3.98273051e-01 -1.16713410e-02 -3.74333680e-01 3.00287783e-01
-1.24075305e+00 7.46368647e-01 3.55654269e-01 3.88839394e-01
-3.98012996e-02 1.73661243e-02 5.55693507e-01 2.09581286e-01
8.13920736e-01 3.05072069e-01 -9.36594665e-01 -3.83635581e-01
-7.08122194e-01 -2.96280235e-01 3.41167331e-01 1.05170584e+00
5.92911303e-01 4.70407009e-01 -3.61343116e-01 1.15138757e+00
4.65392083e-01 4.68714207e-01 1.04559433e+00 -2.70426482e-01
7.45040894e-01 4.48435426e-01 -5.29436469e-01 -8.28403592e-01
-3.23073924e-01 -4.11243439e-01 -1.07322383e+00 -4.08121556e-01
-4.19199884e-01 -1.83955565e-01 -1.12287462e+00 1.74715769e+00
-6.22193329e-02 4.32306737e-01 1.08432904e-01 8.19168031e-01
1.17857516e+00 1.06381238e+00 2.30533898e-01 -4.55081642e-01
8.96985471e-01 -1.37420595e+00 -1.35036349e+00 -9.15059168e-03
7.40357220e-01 -8.55580449e-01 1.56669569e+00 4.51671332e-02
-6.79415643e-01 -7.17727065e-01 -9.33512151e-01 2.27908380e-02
-5.47035098e-01 5.72248399e-01 5.13424575e-01 5.30492663e-01
-9.19424653e-01 2.52892435e-01 -7.09620833e-01 -1.89183339e-01
3.14475685e-01 3.87750864e-01 -9.28844437e-02 4.74779874e-01
-1.57037354e+00 4.21580702e-01 6.20630324e-01 1.50963143e-01
-8.25166285e-01 -6.20243371e-01 -8.44547510e-01 4.70446587e-01
4.04358774e-01 2.01840028e-02 1.44255412e+00 -9.08752799e-01
-2.01514053e+00 1.91088974e-01 -9.75884572e-02 -6.12811863e-01
-1.08181529e-01 -1.69876993e-01 -7.54231572e-01 -4.94534895e-02
-8.57329816e-02 5.34493387e-01 1.03448045e+00 -7.62844563e-01
-4.53518093e-01 -8.23574811e-02 -2.15528354e-01 3.57426852e-01
-1.01668334e+00 -6.36437684e-02 -5.54495573e-01 -1.17392600e+00
3.57411839e-02 -6.18563294e-01 -7.72509351e-02 -5.42968750e-01
-6.47027791e-01 -4.96951312e-01 1.11692762e+00 -8.64700854e-01
1.44950151e+00 -2.66663694e+00 -6.83261305e-02 8.96000415e-02
-1.07412927e-01 6.60943985e-01 -1.89336225e-01 4.53413039e-01
-2.94109792e-01 3.61163080e-01 -3.34768176e-01 -5.87544799e-01
-7.86598474e-02 1.94250166e-01 -5.60738027e-01 3.94084938e-02
4.17771786e-01 9.15717542e-01 -5.76782167e-01 -4.06008959e-01
1.64112985e-01 1.99650541e-01 -4.39198107e-01 7.20500350e-01
-1.65272400e-01 4.89155114e-01 -1.93645597e-01 4.09701645e-01
3.96899730e-01 8.48644823e-02 -3.90019268e-01 -1.48599759e-01
-1.13546595e-01 6.75503969e-01 -7.38707304e-01 1.72412050e+00
-7.27350295e-01 7.10083485e-01 -2.45308161e-01 -1.30249667e+00
1.02014494e+00 5.80947340e-01 2.64902204e-01 -8.63439441e-01
4.03276272e-03 6.81091323e-02 -1.70670953e-02 -5.85087776e-01
3.08500558e-01 1.83164433e-01 2.34509513e-01 3.06772679e-01
3.90319526e-01 -1.61942244e-02 -4.74691480e-01 3.26580793e-01
7.27939487e-01 -3.50547969e-01 -2.31230304e-01 -1.75400719e-01
6.77290142e-01 -4.13576514e-01 5.52840352e-01 4.97166097e-01
4.33180779e-02 6.50309205e-01 3.14953804e-01 -1.51935667e-01
-8.29486191e-01 -5.55618227e-01 -5.36523852e-03 1.22735572e+00
7.23132044e-02 -6.58946395e-01 -7.54819214e-01 -7.82878757e-01
-3.24452400e-01 6.41291678e-01 -3.31876069e-01 -8.16022336e-01
-1.65131807e-01 -5.30581415e-01 7.64704108e-01 8.31763923e-01
5.72813869e-01 -1.32860923e+00 3.53889793e-01 2.18065664e-01
-3.42178941e-01 -1.31526768e+00 -9.34933424e-01 3.89752328e-01
-4.56246316e-01 -7.10663319e-01 -8.47141504e-01 -1.09416914e+00
5.77714205e-01 2.60906458e-01 4.17915374e-01 1.85063586e-01
7.00152665e-02 3.65036428e-01 -8.95020366e-01 -6.86166286e-01
-2.07553446e-01 4.89434421e-01 3.45867693e-01 4.51213032e-01
2.49161914e-01 -6.33279026e-01 -4.16065395e-01 1.93531513e-01
-6.73056602e-01 1.46893248e-01 6.58746481e-01 1.01325643e+00
3.85347039e-01 1.47932559e-01 8.41483831e-01 -5.21308303e-01
8.55193019e-01 -2.40662336e-01 -3.04148108e-01 -4.80326079e-03
-7.58087158e-01 -1.83774512e-02 6.95624173e-01 -7.10768580e-01
-1.16379535e+00 5.54436930e-02 -6.48146510e-01 -6.59129381e-01
-9.84366760e-02 7.69562900e-01 -5.70592821e-01 9.58305970e-02
4.78752479e-02 5.97709179e-01 -1.71501711e-02 -8.34220946e-01
1.54097080e-01 1.42962527e+00 4.08347636e-01 -2.16399729e-01
6.48426473e-01 -1.64401561e-01 -8.82155001e-01 -1.19241631e+00
-6.76772475e-01 -6.86723173e-01 -2.61218905e-01 7.96664134e-02
7.85386264e-01 -1.03975046e+00 -5.44873893e-01 7.40050912e-01
-1.40339220e+00 -4.42098171e-01 -4.56297129e-01 8.85793388e-01
-3.03473532e-01 4.77060974e-01 -6.00241244e-01 -1.15344954e+00
-4.76161510e-01 -1.27234674e+00 1.06354618e+00 1.22001268e-01
2.69468427e-01 -7.40127563e-01 -6.34870455e-02 3.05820048e-01
5.96489906e-01 -7.08435357e-01 9.55282688e-01 -1.29745090e+00
-1.67088822e-01 -1.17036231e-01 -8.82091597e-02 8.21035504e-01
1.72591478e-01 -1.00907728e-01 -1.31473756e+00 -1.21826813e-01
7.20666498e-02 -3.75096411e-01 9.43809569e-01 1.32388413e-01
1.95821512e+00 -5.81414819e-01 1.29204188e-02 6.11135602e-01
6.91248953e-01 6.52466893e-01 8.40383649e-01 1.09848067e-01
8.32440972e-01 3.69234383e-01 5.89873910e-01 2.62820333e-01
1.41668409e-01 6.66425645e-01 2.81373203e-01 -4.23614502e-01
-1.34670600e-01 -4.85791147e-01 3.93334568e-01 1.98440480e+00
2.27122545e-01 -2.43088186e-01 -7.93936729e-01 5.41168630e-01
-1.88178062e+00 -6.30091727e-01 7.21419603e-02 2.07242203e+00
9.49977756e-01 2.52583385e-01 -9.75456238e-02 4.75730211e-01
7.68484056e-01 2.05691352e-01 -4.39009041e-01 -2.57371575e-01
-9.81108919e-02 3.38501453e-01 1.17876783e-01 3.68694574e-01
-1.32309067e+00 1.16080415e+00 5.77951670e+00 1.34174752e+00
-1.12661731e+00 2.28281632e-01 8.31574559e-01 3.24076504e-01
-3.27433378e-01 -2.13559628e-01 -6.41306221e-01 8.74365807e-01
1.01435745e+00 4.82133543e-03 2.78728396e-01 8.66388023e-01
3.73975843e-01 4.46104586e-01 -7.10923731e-01 1.09000492e+00
-1.06260870e-02 -1.07289505e+00 2.40357131e-01 -1.42644018e-01
4.31962729e-01 -8.22984334e-03 1.71532616e-01 8.05739880e-01
-9.64580029e-02 -1.00730908e+00 4.08130825e-01 2.04866350e-01
9.27393079e-01 -9.05313790e-01 1.10713351e+00 4.91860420e-01
-1.29722691e+00 -9.93428007e-03 -2.83671081e-01 3.95917408e-02
-1.46279240e-03 4.00494516e-01 -8.92991185e-01 5.16213655e-01
7.67859280e-01 6.75140440e-01 -3.81171346e-01 9.13429499e-01
-3.93645495e-01 1.13594806e+00 -3.25667590e-01 -1.93971768e-01
3.60531330e-01 -6.47678152e-02 3.38398695e-01 1.46804249e+00
2.59453487e-02 2.55475253e-01 1.57255739e-01 4.81135756e-01
-1.90847412e-01 5.42212009e-01 -3.48324686e-01 -5.41752756e-01
5.65593243e-01 9.82329488e-01 -2.97140151e-01 -3.82731378e-01
-4.43400502e-01 8.15426648e-01 2.60639489e-01 5.60844302e-01
-8.61977935e-01 -7.43641913e-01 4.90448892e-01 1.31696388e-02
2.53225148e-01 -1.77049354e-01 -2.44625583e-02 -1.23186302e+00
9.06186700e-02 -7.41068065e-01 -2.73034982e-02 -7.99383461e-01
-1.07147455e+00 6.24179423e-01 -3.13866407e-01 -1.18915236e+00
5.40988855e-02 -3.65192682e-01 -1.02617431e+00 8.79494250e-01
-1.64417624e+00 -1.09860873e+00 -2.56867677e-01 7.56914675e-01
1.04477191e+00 -7.93065727e-01 8.43678415e-01 4.84382659e-01
-7.49719918e-01 9.10958409e-01 7.40136206e-03 4.08975065e-01
6.23698294e-01 -1.03818560e+00 5.41229427e-01 7.63279080e-01
4.77261066e-01 3.63481134e-01 -1.30441990e-02 -4.12571222e-01
-1.04240751e+00 -1.16692221e+00 6.59861565e-01 3.41643989e-01
6.83686972e-01 -3.93356234e-01 -1.10484695e+00 4.68352407e-01
2.66970843e-01 -4.18251716e-02 6.96167052e-01 7.21605867e-02
-2.73535818e-01 -2.22528055e-01 -4.71509725e-01 4.75907832e-01
7.54658937e-01 -1.01488531e+00 -5.23829520e-01 5.50951421e-01
1.69569349e+00 -2.56890446e-01 -4.99053538e-01 6.36148453e-01
1.62881732e-01 -4.40892726e-01 8.55921209e-01 -8.79903495e-01
2.81053662e-01 -9.59180668e-02 -8.28505233e-02 -1.44117665e+00
-1.65473491e-01 -5.53066373e-01 -2.22012281e-01 1.60709882e+00
6.96966767e-01 -5.66864908e-01 5.16156912e-01 1.10655643e-01
-6.13175333e-01 -9.02198851e-01 -9.18633044e-01 -7.88628578e-01
-2.27813452e-01 -8.80282700e-01 6.89149439e-01 9.74243104e-01
6.49740323e-02 5.25952339e-01 -4.66466457e-01 3.92209053e-01
1.81658119e-02 -3.73441905e-01 6.81148291e-01 -8.85136127e-01
-1.53160110e-01 -3.17223042e-01 -5.18235981e-01 -1.21495771e+00
3.87878507e-01 -8.96660388e-01 2.05204532e-01 -1.20652366e+00
1.75443571e-02 -4.17328566e-01 -5.84665000e-01 6.83791578e-01
-3.65810961e-01 -2.34286681e-01 3.79070528e-02 3.18464220e-01
-5.09825051e-01 1.25050616e+00 9.16512430e-01 -6.11150444e-01
-3.80676985e-01 1.28839508e-01 -3.78364801e-01 6.56214774e-01
1.08609843e+00 -3.11555535e-01 -5.51798403e-01 -4.82981831e-01
-4.29247856e-01 -2.20775053e-01 -4.26706448e-02 -8.62745166e-01
6.63030803e-01 8.28889683e-02 2.71539073e-02 -7.64656663e-01
4.28488135e-01 -7.82819390e-01 -5.12115717e-01 2.48619154e-01
-6.11446142e-01 -4.19304579e-01 2.82786667e-01 7.24309862e-01
-6.16796076e-01 -6.02338351e-02 6.43305719e-01 1.51600376e-01
-6.83531046e-01 6.46910191e-01 -5.07457972e-01 -7.69957341e-03
5.49775958e-01 3.17290992e-01 1.06807366e-01 -7.63832569e-01
-8.47664535e-01 3.33681643e-01 -4.00095165e-01 7.41203129e-01
9.54285264e-01 -1.62798798e+00 -5.15872359e-01 5.59982240e-01
3.47445279e-01 1.50332808e-01 7.67591596e-02 7.73915231e-01
-1.07646376e-01 4.19074297e-01 4.59855080e-01 -6.24368846e-01
-9.80099201e-01 6.26079977e-01 9.98209789e-02 4.23699468e-02
-4.82594371e-01 8.87965798e-01 4.44406271e-01 -7.89245546e-01
9.53745425e-01 -4.32379097e-01 -4.73024666e-01 -1.86585978e-01
5.56125522e-01 -3.74791166e-03 3.32767963e-01 -4.43226546e-01
-3.22455615e-01 2.70123243e-01 -3.34141940e-01 -1.16353780e-01
1.16236472e+00 -1.15223683e-01 1.13585159e-01 5.84359825e-01
1.34298885e+00 -4.79464307e-02 -8.36104155e-01 -4.28941041e-01
1.89330913e-02 -1.56549402e-02 3.73882413e-01 -6.09114707e-01
-1.37052560e+00 1.07518041e+00 4.79479194e-01 2.23045915e-01
1.19858432e+00 -5.02644442e-02 1.00657177e+00 4.54443067e-01
-3.64795662e-02 -1.38429785e+00 2.52912223e-01 7.67306030e-01
9.18553591e-01 -1.36905527e+00 -4.64135200e-01 -4.78999227e-01
-6.60640895e-01 9.79369044e-01 8.37520778e-01 2.55003631e-01
8.75403285e-01 3.97247553e-01 4.29380685e-02 1.20367110e-01
-5.70623219e-01 -3.76771599e-01 2.70115554e-01 5.36269546e-01
2.27310136e-01 -1.52159274e-01 -9.40486044e-02 1.20482516e+00
-1.66932166e-01 -4.33139473e-01 1.08102731e-01 6.01346314e-01
-2.49874637e-01 -8.57723117e-01 -1.66165680e-02 3.96905690e-01
-2.26744384e-01 -5.82139134e-01 -5.71344197e-01 3.91635209e-01
-1.00759886e-01 1.10318315e+00 1.84040353e-01 -9.43972290e-01
5.91701627e-01 1.61351427e-01 -3.41041595e-01 -7.16552734e-01
-4.17889059e-01 2.47046173e-01 -1.41125247e-01 -4.00903195e-01
-3.23476881e-01 4.07715514e-02 -1.11832309e+00 1.14981249e-01
-9.40290987e-01 6.48934722e-01 7.53015935e-01 1.17258549e+00
3.86962295e-01 8.90307009e-01 1.28379869e+00 -5.94979286e-01
-5.17329216e-01 -1.40334785e+00 -5.50714076e-01 2.48875231e-01
2.68330097e-01 -4.99310642e-01 -6.25458837e-01 1.22317776e-01] | [14.546890258789062, 6.464897155761719] |
3ba32c2c-b51b-4aaf-aeb7-98bd0261dd6c | r-transformer-recurrent-neural-network | 1907.05572 | null | https://arxiv.org/abs/1907.05572v1 | https://arxiv.org/pdf/1907.05572v1.pdf | R-Transformer: Recurrent Neural Network Enhanced Transformer | Recurrent Neural Networks have long been the dominating choice for sequence modeling. However, it severely suffers from two issues: impotent in capturing very long-term dependencies and unable to parallelize the sequential computation procedure. Therefore, many non-recurrent sequence models that are built on convolution and attention operations have been proposed recently. Notably, models with multi-head attention such as Transformer have demonstrated extreme effectiveness in capturing long-term dependencies in a variety of sequence modeling tasks. Despite their success, however, these models lack necessary components to model local structures in sequences and heavily rely on position embeddings that have limited effects and require a considerable amount of design efforts. In this paper, we propose the R-Transformer which enjoys the advantages of both RNNs and the multi-head attention mechanism while avoids their respective drawbacks. The proposed model can effectively capture both local structures and global long-term dependencies in sequences without any use of position embeddings. We evaluate R-Transformer through extensive experiments with data from a wide range of domains and the empirical results show that R-Transformer outperforms the state-of-the-art methods by a large margin in most of the tasks. We have made the code publicly available at \url{https://github.com/DSE-MSU/R-transformer}. | ['Zitao Liu', 'Zhiwei Wang', 'Jiliang Tang', 'Yao Ma'] | 2019-07-12 | r-transformer-recurrent-neural-network-1 | https://openreview.net/forum?id=HJx4PAEYDH | https://openreview.net/pdf?id=HJx4PAEYDH | iclr-2020-1 | ['sequential-image-classification', 'music-modeling'] | ['computer-vision', 'music'] | [-3.42797749e-02 -3.52488369e-01 -5.57549745e-02 -2.36177623e-01
-4.44450289e-01 -4.17675287e-01 6.33333623e-01 -9.70981941e-02
-5.01759112e-01 5.80901742e-01 3.21662694e-01 -5.83610594e-01
2.11499110e-01 -5.13520896e-01 -5.92816532e-01 -7.16482937e-01
-5.92733696e-02 2.51890391e-01 1.88891485e-01 -4.32207942e-01
1.48460329e-01 2.60547638e-01 -1.24144435e+00 1.43855646e-01
7.65804648e-01 8.20549548e-01 7.18626261e-01 5.46643496e-01
-2.68041521e-01 1.04653800e+00 -2.96247602e-01 -4.17245150e-01
-9.32469293e-02 -4.46691811e-01 -6.37173831e-01 -3.26789588e-01
-2.79407412e-01 -2.23334417e-01 -6.06132150e-01 8.76299083e-01
7.76533961e-01 7.21685141e-02 3.50921571e-01 -9.30060685e-01
-9.35158014e-01 8.44917417e-01 -5.70798159e-01 5.31545758e-01
1.76880151e-01 1.12425841e-01 1.23981810e+00 -1.13898218e+00
2.07895637e-01 9.83624339e-01 9.67685640e-01 3.83480102e-01
-9.55020547e-01 -6.17107570e-01 2.97867119e-01 4.43696737e-01
-1.40061021e+00 -4.57454264e-01 7.41834402e-01 -4.10358548e-01
1.58805919e+00 2.73257166e-01 2.95991510e-01 1.32052255e+00
1.77518383e-01 9.18658078e-01 6.98334455e-01 -4.26346809e-01
-2.32795969e-01 -2.03429103e-01 4.93694395e-01 6.48489237e-01
-1.26567200e-01 -1.78472131e-01 -3.87518197e-01 -1.08717091e-01
8.08341622e-01 3.49195927e-01 -2.83339858e-01 2.07364365e-01
-1.12949777e+00 7.42433786e-01 2.89918065e-01 6.55701518e-01
-5.40195942e-01 1.78567290e-01 6.30417407e-01 2.44154543e-01
4.82703656e-01 3.71244103e-02 -6.04113996e-01 -5.68160236e-01
-5.59079945e-01 -7.03186765e-02 4.53955263e-01 9.84044313e-01
4.29591984e-01 1.22159325e-01 -8.59630704e-02 1.14167619e+00
-2.46601440e-02 -2.65378077e-02 9.03040707e-01 -2.47585163e-01
5.79259932e-01 5.04319370e-01 -1.54671982e-01 -9.25337434e-01
-4.42585230e-01 -7.97854483e-01 -1.21235108e+00 -3.79037023e-01
4.07909751e-02 -9.13887322e-02 -8.91253650e-01 1.90940678e+00
-1.06338210e-01 4.40417379e-01 -1.03128150e-01 7.19224095e-01
7.00357139e-01 9.57867265e-01 -5.18756453e-03 -1.45353138e-01
1.24492407e+00 -1.28828371e+00 -8.65957320e-01 -3.69543970e-01
7.11794198e-01 -8.21375251e-01 1.20963657e+00 7.36942664e-02
-1.14401472e+00 -7.79263079e-01 -9.53911126e-01 -2.58230180e-01
-1.83686301e-01 1.26170397e-01 5.79749405e-01 4.68094498e-01
-1.06172609e+00 5.03668249e-01 -9.56908941e-01 -3.20406586e-01
1.84160694e-01 1.97367206e-01 -1.40227750e-01 1.67502742e-02
-1.50789940e+00 9.81017828e-01 4.24899608e-01 4.15845096e-01
-7.96717584e-01 -5.14182746e-01 -7.80369699e-01 3.41587633e-01
1.98054209e-01 -4.97000873e-01 1.22278643e+00 -7.23333418e-01
-1.42371762e+00 4.50013995e-01 -4.42911536e-01 -5.13547182e-01
3.37779462e-01 -5.45962274e-01 -3.87765795e-01 -2.04017237e-01
-2.90379196e-01 2.10014388e-01 4.28373218e-01 -5.58068454e-01
-4.89937097e-01 -3.05262357e-01 7.04558333e-03 2.02041164e-01
-6.34519637e-01 2.57801116e-01 -6.08358741e-01 -1.00143790e+00
-1.66805372e-01 -1.02523673e+00 -2.80338347e-01 -5.91398120e-01
-2.06571534e-01 -3.21631312e-01 8.18948269e-01 -7.34847724e-01
1.72080684e+00 -2.00168014e+00 1.31232753e-01 -3.51464570e-01
5.85159054e-03 7.77415454e-01 -2.64985800e-01 8.90974164e-01
-2.68045694e-01 3.20735693e-01 -2.87957847e-01 -4.25988734e-01
-1.89848747e-02 2.33583599e-01 -3.32760721e-01 3.37209970e-01
1.95316851e-01 1.20001233e+00 -6.25957608e-01 -2.01764315e-01
1.76445961e-01 7.76139081e-01 -3.57151151e-01 3.80672783e-01
-2.21937358e-01 3.07676256e-01 -2.97383457e-01 3.53228956e-01
4.97583777e-01 -5.76787591e-01 3.26658726e-01 2.14548379e-01
-1.46752581e-01 6.55485094e-01 -6.69163346e-01 1.71796203e+00
-4.74676430e-01 4.92084682e-01 -1.48681328e-01 -1.16364729e+00
9.19399798e-01 6.01656139e-01 2.86540747e-01 -6.71519458e-01
1.48061946e-01 2.15808049e-01 2.42473021e-01 -5.62073767e-01
4.36463922e-01 -2.02827543e-01 9.58040953e-02 5.57332814e-01
-2.28763912e-02 7.02012479e-01 1.43844023e-01 1.94604516e-01
1.24133360e+00 1.23634249e-01 3.00872505e-01 -9.95352715e-02
6.32590890e-01 -5.36176920e-01 8.47841799e-01 5.75230181e-01
-1.06571123e-01 6.54727399e-01 3.63541573e-01 -5.05129874e-01
-1.18713188e+00 -6.09105647e-01 1.21822402e-01 1.40003800e+00
-1.52764902e-01 -5.34572065e-01 -4.67421860e-01 -3.75359178e-01
-3.17855537e-01 5.21003962e-01 -6.46061063e-01 -5.12002707e-02
-9.02866960e-01 -9.01919186e-01 8.16689968e-01 8.93161118e-01
3.61178368e-01 -1.27406478e+00 -4.02147055e-01 4.72965866e-01
-5.29122889e-01 -1.07542455e+00 -7.45673358e-01 2.71315306e-01
-8.27157259e-01 -6.30808771e-01 -8.76169205e-01 -8.93498003e-01
3.05157512e-01 3.69428217e-01 1.04066455e+00 2.62045592e-01
-1.61486313e-01 -3.27388406e-01 -6.67665541e-01 -2.62176514e-01
-7.19517097e-02 4.52463925e-01 -1.66540995e-01 -3.20671545e-03
5.28481483e-01 -9.37702358e-01 -5.05499184e-01 3.71661663e-01
-9.31974471e-01 1.34441823e-01 7.63907552e-01 1.01084495e+00
2.42542684e-01 -1.77694365e-01 9.24827814e-01 -8.31036031e-01
7.24185228e-01 -7.15959549e-01 -3.63546878e-01 2.69324601e-01
-3.47866207e-01 1.01675987e-01 8.77718389e-01 -4.28315997e-01
-1.04750180e+00 -3.36958259e-01 -6.07051849e-01 -5.55483997e-01
1.31012812e-01 9.38587368e-01 -1.55985266e-01 5.15386105e-01
2.10930228e-01 7.00690687e-01 1.89932752e-02 -7.45010853e-01
1.07330441e-01 6.88870788e-01 8.79338756e-02 -3.48455846e-01
3.34383994e-01 1.19402356e-01 -4.63891804e-01 -9.58102763e-01
-6.35118723e-01 -4.55985665e-01 -4.99379307e-01 2.51204491e-01
6.47439361e-01 -9.13680196e-01 -5.24263322e-01 7.71000743e-01
-1.34155786e+00 -2.63924003e-01 1.66003808e-01 4.37078357e-01
-4.21830565e-01 6.16199911e-01 -1.12763453e+00 -8.27439964e-01
-5.50435483e-01 -1.21984839e+00 5.23481965e-01 1.91752352e-02
-2.36555263e-01 -1.03426111e+00 7.52895474e-02 1.13303050e-01
7.50022888e-01 -1.17536843e-01 1.11354160e+00 -7.84839571e-01
-5.59070766e-01 -5.71777485e-02 -1.31914526e-01 3.51621091e-01
1.95382223e-01 -1.63498655e-01 -1.03250408e+00 -3.09511483e-01
1.74220540e-02 -2.72788227e-01 9.81538117e-01 3.67987633e-01
1.18504441e+00 -1.04108773e-01 -2.35518634e-01 6.90733731e-01
1.31300318e+00 4.18006808e-01 7.90948033e-01 2.24512756e-01
8.11626792e-01 3.56304437e-01 3.62643570e-01 5.59057295e-01
3.28044266e-01 5.95927298e-01 3.17939878e-01 3.24992836e-02
6.21449575e-02 -3.50944728e-01 3.98683161e-01 1.54443133e+00
-1.67474404e-01 -5.43277860e-01 -9.24239337e-01 8.03553700e-01
-2.12642694e+00 -1.07238328e+00 -7.85077885e-02 1.80831587e+00
8.44968259e-01 2.23741949e-01 1.05566934e-01 8.69107246e-02
7.90146887e-01 4.43286270e-01 -4.40342635e-01 -6.26154900e-01
-1.07669100e-01 1.76198691e-01 1.79679587e-01 3.33081156e-01
-8.93297076e-01 9.06292081e-01 6.11758661e+00 9.19170916e-01
-1.04381812e+00 2.41764799e-01 5.03915787e-01 2.27857288e-02
-3.33485186e-01 -1.33407280e-01 -9.87401247e-01 6.49348617e-01
1.22403932e+00 -4.82314453e-02 2.09218785e-01 6.70676112e-01
1.26762882e-01 4.65048760e-01 -9.48449612e-01 8.71688783e-01
-1.29096583e-01 -1.36255693e+00 3.15947272e-02 9.53188818e-03
4.10438985e-01 2.99728870e-01 4.78290282e-02 4.44536030e-01
1.82618797e-01 -1.24112642e+00 6.44116044e-01 3.28112662e-01
7.18270063e-01 -8.26216221e-01 8.65969419e-01 6.24064982e-01
-1.37769818e+00 -1.70568138e-01 -4.50203836e-01 -4.74031240e-01
3.62809092e-01 6.46915019e-01 -5.00165403e-01 6.45932734e-01
5.64174056e-01 8.81241202e-01 -1.95763588e-01 6.96890056e-01
-1.94282368e-01 8.34312201e-01 -1.22109927e-01 -1.22643188e-01
5.47369540e-01 -2.97122329e-01 3.09616834e-01 1.60122693e+00
3.58078182e-01 7.47799277e-02 -2.16254387e-02 6.40350580e-01
-4.35957275e-02 1.14699587e-01 -5.93005359e-01 -3.38464886e-01
5.41095912e-01 8.50250840e-01 -4.52816337e-01 -3.16308588e-01
-7.18017101e-01 7.87172675e-01 7.01271892e-01 4.28857207e-01
-1.15050459e+00 -5.10800660e-01 9.42781687e-01 4.31160145e-02
7.87403166e-01 -4.86912668e-01 -2.48874336e-01 -1.16414821e+00
2.46270061e-01 -9.81438100e-01 4.16127652e-01 -5.72473347e-01
-1.18859255e+00 1.06148016e+00 -4.27216947e-01 -9.58814740e-01
-3.55849892e-01 -4.68663156e-01 -5.90490639e-01 1.20736980e+00
-1.58596206e+00 -1.18626857e+00 2.31602695e-02 5.46484828e-01
8.49319100e-01 -5.49192838e-02 8.95671666e-01 6.21802747e-01
-9.06827569e-01 7.35455453e-01 2.88699478e-01 2.99762785e-01
3.13364655e-01 -8.74437034e-01 8.86598766e-01 9.52452898e-01
6.81593735e-03 1.09613121e+00 4.69346493e-01 -4.76221919e-01
-1.34650838e+00 -9.54741359e-01 1.18394840e+00 -2.54064113e-01
7.36052513e-01 -5.66722274e-01 -1.20556116e+00 1.13067007e+00
4.80452597e-01 -1.02522612e-01 7.43842542e-01 2.57919908e-01
-4.22221839e-01 1.53377429e-01 -3.52176905e-01 6.74224973e-01
1.26676381e+00 -6.56434953e-01 -6.33822322e-01 -5.57801351e-02
8.27400684e-01 -3.21714543e-02 -6.59498274e-01 4.71492767e-01
6.63072109e-01 -1.06809556e+00 8.23008001e-01 -6.14328802e-01
5.32419920e-01 -1.19062297e-01 -1.49713680e-01 -1.18363571e+00
-7.30517745e-01 -7.14685738e-01 -1.29587173e-01 1.26110446e+00
5.32885253e-01 -9.12921965e-01 4.99431372e-01 2.01094449e-01
-4.37998950e-01 -1.00879025e+00 -6.29277825e-01 -9.03037786e-01
3.68866511e-02 -4.26000059e-01 5.27823150e-01 1.04768789e+00
9.13644806e-02 5.78290284e-01 -7.74862766e-01 4.08358350e-02
2.21621811e-01 2.59264231e-01 5.62896669e-01 -7.63539374e-01
-4.62669492e-01 -4.70273674e-01 -2.48809844e-01 -1.49912846e+00
1.41005397e-01 -8.85085821e-01 -4.05866876e-02 -1.46822357e+00
3.06031466e-01 -4.15666461e-01 -6.84658289e-01 4.60384250e-01
-3.89888972e-01 -2.46308148e-02 7.26653412e-02 3.40216428e-01
-3.43315035e-01 8.16371620e-01 8.70895684e-01 1.37399569e-01
2.68177185e-02 -1.40145242e-01 -6.75587416e-01 6.25721991e-01
1.04835105e+00 -3.97415489e-01 -5.47923684e-01 -7.18118668e-01
1.49872035e-01 9.89969075e-02 -1.42716914e-01 -7.33231068e-01
2.00645342e-01 -1.51189873e-02 1.63535848e-01 -7.33651996e-01
5.15991628e-01 -5.47428370e-01 3.50451887e-01 4.71209049e-01
-4.56231236e-01 5.60172915e-01 1.21007904e-01 4.88754153e-01
-3.80761832e-01 -1.56373814e-01 6.73439026e-01 -2.43177295e-01
-7.21474230e-01 3.80143374e-01 -4.90314335e-01 3.99747007e-02
8.60762596e-01 -1.26556531e-01 -2.11425409e-01 -3.88970137e-01
-5.81736982e-01 1.67376503e-01 3.23028088e-01 6.28641367e-01
6.14595950e-01 -1.19691610e+00 -7.81935632e-01 3.64811361e-01
-1.01282177e-02 -2.71102697e-01 3.69395643e-01 9.30841506e-01
-4.38223451e-01 8.17871869e-01 -1.33052304e-01 -4.14594054e-01
-1.25612652e+00 7.39345372e-01 7.81054646e-02 -6.34848058e-01
-7.59771526e-01 1.00222564e+00 4.97867078e-01 -4.70423013e-01
2.99513102e-01 -2.81836420e-01 -1.06422469e-01 -2.35125460e-02
7.19963014e-01 2.26342604e-01 7.12929741e-02 -6.82680428e-01
-4.54480410e-01 4.62920278e-01 -4.43996310e-01 1.93633497e-01
1.57026136e+00 -2.31294945e-01 -5.12950532e-02 5.91284156e-01
1.33697283e+00 -3.26528311e-01 -1.18133605e+00 -4.33745325e-01
7.10972995e-02 -3.61754864e-01 -1.73473373e-01 -5.70570886e-01
-1.09953606e+00 1.30182123e+00 1.10359572e-01 2.51172870e-01
1.11059093e+00 -3.28179032e-01 1.13604343e+00 6.18310049e-02
2.90175050e-01 -7.45902836e-01 9.90161523e-02 1.08535957e+00
7.87984133e-01 -8.94526541e-01 -3.85273218e-01 -2.62333870e-01
-7.06571460e-01 9.37347472e-01 4.58594024e-01 -7.20246211e-02
6.58471167e-01 5.02884448e-01 1.56547397e-01 3.45942527e-02
-1.16191745e+00 -2.56291538e-01 -1.95825383e-01 3.33353639e-01
9.43134904e-01 -7.90479109e-02 -4.85812515e-01 5.64636528e-01
1.10119209e-01 9.72766057e-02 4.02246714e-01 1.06632411e+00
-3.53937417e-01 -1.31385648e+00 -3.24098654e-02 1.65861949e-01
-8.90051663e-01 -4.83190507e-01 -1.18703142e-01 5.85292697e-01
-2.75711805e-01 7.87208140e-01 -1.05410807e-01 -5.03845751e-01
1.38604909e-01 3.35094452e-01 1.90093845e-01 -3.91263813e-01
-7.02862382e-01 2.42985591e-01 1.23976238e-01 -4.00367230e-01
-2.81820208e-01 -4.47973669e-01 -1.10481358e+00 -4.65662301e-01
-2.15473339e-01 1.84676513e-01 3.81759882e-01 7.70920813e-01
7.48695076e-01 6.61999404e-01 3.47544014e-01 -6.86108172e-01
-7.46068597e-01 -1.36548936e+00 -5.57769656e-01 3.22176695e-01
3.12518030e-01 -4.02708977e-01 -7.50012994e-02 -6.57168478e-02] | [10.877964973449707, 6.643746852874756] |
8bc5e0a2-b4fe-4c97-85f4-4ba402019a6e | learning-to-see-through-obstructions | 2004.01180 | null | https://arxiv.org/abs/2004.01180v1 | https://arxiv.org/pdf/2004.01180v1.pdf | Learning to See Through Obstructions | We present a learning-based approach for removing unwanted obstructions, such as window reflections, fence occlusions or raindrops, from a short sequence of images captured by a moving camera. Our method leverages the motion differences between the background and the obstructing elements to recover both layers. Specifically, we alternate between estimating dense optical flow fields of the two layers and reconstructing each layer from the flow-warped images via a deep convolutional neural network. The learning-based layer reconstruction allows us to accommodate potential errors in the flow estimation and brittle assumptions such as brightness consistency. We show that training on synthetically generated data transfers well to real images. Our results on numerous challenging scenarios of reflection and fence removal demonstrate the effectiveness of the proposed method. | ['Jia-Bin Huang', 'Yung-Yu Chuang', 'Ming-Hsuan Yang', 'Wei-Sheng Lai', 'Yu-Lun Liu'] | 2020-04-02 | learning-to-see-through-obstructions-1 | http://openaccess.thecvf.com/content_CVPR_2020/html/Liu_Learning_to_See_Through_Obstructions_CVPR_2020_paper.html | http://openaccess.thecvf.com/content_CVPR_2020/papers/Liu_Learning_to_See_Through_Obstructions_CVPR_2020_paper.pdf | cvpr-2020-6 | ['reflection-removal'] | ['computer-vision'] | [ 3.17179352e-01 -2.45720699e-01 3.98504436e-01 -2.06171423e-01
-3.27162743e-01 -7.45855570e-01 4.93413150e-01 -4.89899874e-01
-2.83328325e-01 7.16629803e-01 3.26659530e-01 -2.47411758e-01
3.11821640e-01 -5.89194715e-01 -9.49175835e-01 -4.97808784e-01
-3.48142385e-01 -1.84590429e-01 4.00859833e-01 9.50602591e-02
2.98649579e-01 5.67523301e-01 -1.50204027e+00 1.24394305e-01
8.27996731e-01 7.29881287e-01 8.88600200e-02 1.12160146e+00
1.36220530e-01 1.21368980e+00 -4.34601933e-01 -4.08571661e-02
7.91332662e-01 -1.48263380e-01 -6.85146153e-01 4.08218592e-01
1.23157167e+00 -1.27263331e+00 -7.92210639e-01 7.26605058e-01
3.80160026e-02 4.49923784e-01 3.24183106e-01 -8.80815506e-01
-1.04675852e-01 -2.19010144e-01 -6.85029685e-01 4.09668922e-01
2.54223853e-01 5.68356276e-01 6.40067637e-01 -8.24292243e-01
8.03913355e-01 1.19265854e+00 6.81038916e-01 4.20504868e-01
-1.19190800e+00 -5.25120795e-01 3.00473511e-01 9.01219249e-02
-1.05517757e+00 -7.97488809e-01 6.76020503e-01 -7.41733968e-01
5.81887782e-01 7.26340786e-02 6.38243377e-01 1.01393461e+00
7.22647682e-02 6.15357459e-01 8.63125265e-01 -2.24993914e-01
3.79172787e-02 -2.63502687e-01 -1.08833663e-01 9.19797599e-01
4.63456601e-01 2.48526886e-01 -3.73275101e-01 3.16087268e-02
1.11288154e+00 -3.36137451e-02 -6.25684261e-01 -3.28567088e-01
-1.03265882e+00 1.61524177e-01 3.79145682e-01 -3.77055585e-01
-1.97723269e-01 3.71780008e-01 -4.65154089e-03 -3.62935141e-02
5.46356201e-01 2.47928068e-01 -3.00678164e-01 7.31144473e-02
-1.18952203e+00 2.83686459e-01 7.31038988e-01 7.88405597e-01
9.55740273e-01 4.25115108e-01 2.31889948e-01 5.30312896e-01
2.75481492e-01 7.61678874e-01 -1.85467213e-01 -1.63903415e+00
5.76338112e-01 8.14628229e-02 6.79155529e-01 -1.11333537e+00
-1.23376682e-01 -1.37521565e-01 -7.47239172e-01 6.26379311e-01
7.75285304e-01 -5.03018677e-01 -1.03249550e+00 1.65632343e+00
4.61607248e-01 1.02183521e+00 -9.77116898e-02 1.13925922e+00
4.27352816e-01 5.89629769e-01 -3.21565241e-01 -5.05503379e-02
7.65696466e-01 -1.20768535e+00 -5.89518905e-01 -5.47767758e-01
2.71548986e-01 -8.03672910e-01 6.96384490e-01 3.15306693e-01
-1.10261834e+00 -3.87781858e-01 -9.44345772e-01 -1.85989454e-01
2.12976336e-01 -2.70152807e-01 4.41368163e-01 2.46085227e-01
-9.08225417e-01 6.68230712e-01 -1.07129443e+00 -5.80579042e-02
5.97824693e-01 1.65907234e-01 -4.28258419e-01 -5.45197070e-01
-8.15224886e-01 6.29343569e-01 -2.43839085e-01 5.33849418e-01
-1.13637507e+00 -8.70967329e-01 -9.86342788e-01 3.37613672e-02
1.43222690e-01 -8.90600204e-01 1.02020860e+00 -1.17915821e+00
-1.44292045e+00 5.36064327e-01 -4.57435101e-01 -2.77939409e-01
7.45459437e-01 -6.31229699e-01 9.24708974e-03 4.28438544e-01
-1.28459379e-01 2.82255232e-01 1.01127565e+00 -1.51389170e+00
-6.24187171e-01 1.17527753e-01 2.24490970e-01 1.59671158e-01
5.30027971e-02 -2.78982192e-01 -3.68521392e-01 -4.12200093e-01
-1.21835910e-01 -9.69237566e-01 -2.41272643e-01 4.70503569e-01
-5.26262045e-01 1.04027355e+00 9.65092540e-01 -9.32698607e-01
9.23901439e-01 -1.83142233e+00 -6.25400916e-02 5.72069772e-02
2.98702508e-01 4.93728548e-01 -4.19034779e-01 6.42999932e-02
9.14878845e-02 -7.66079873e-02 -4.41431016e-01 -5.88161349e-01
-5.66943705e-01 4.84419793e-01 -5.90173423e-01 7.53349721e-01
2.98970610e-01 6.57934129e-01 -1.03809917e+00 -2.61058450e-01
6.99603975e-01 8.18739772e-01 -7.42762327e-01 5.51601052e-01
4.04842719e-02 8.71899903e-01 -1.71047464e-01 4.97965723e-01
1.08305943e+00 -9.10362601e-02 1.53234512e-01 -8.83373395e-02
-2.56038159e-01 4.05467540e-01 -1.27326167e+00 1.36853230e+00
-6.38392568e-01 1.06709218e+00 6.16881132e-01 -4.73070621e-01
3.33066165e-01 7.48910159e-02 5.40036798e-01 -5.82982600e-01
-8.46436396e-02 4.99481931e-02 -2.81213969e-02 -8.20715487e-01
3.41007978e-01 -2.34674066e-02 5.41533411e-01 4.26075220e-01
-1.85089365e-01 -3.37353423e-02 -4.09920104e-02 1.60588846e-01
1.28384256e+00 4.52066779e-01 -1.64600596e-01 -7.06635118e-02
4.92348820e-01 -1.62047908e-01 7.74161875e-01 7.89071679e-01
-3.04163694e-01 1.13878870e+00 2.69391894e-01 -9.94944096e-01
-1.18765700e+00 -1.22504807e+00 1.58442244e-01 5.17553926e-01
5.40650189e-01 -2.27990434e-01 -6.80165946e-01 -5.49277544e-01
-7.99869597e-02 1.87495217e-01 -5.19622087e-01 1.85739845e-02
-1.04002631e+00 -5.20389140e-01 4.58039582e-01 2.34886572e-01
6.75685644e-01 -6.40490234e-01 -8.95886481e-01 1.44725755e-01
-7.09000111e-01 -1.73569119e+00 -5.49893081e-01 -3.57661217e-01
-8.58349979e-01 -1.45591164e+00 -4.28644836e-01 -5.11267066e-01
8.39311779e-01 8.52704704e-01 1.09138906e+00 4.29697603e-01
-4.34627384e-01 3.51231724e-01 7.06699118e-02 1.05381250e-01
-2.36691073e-01 -4.35949147e-01 -1.38559863e-01 4.15056527e-01
-5.13795435e-01 -6.93811297e-01 -1.14573586e+00 2.23715842e-01
-9.71062720e-01 1.87700361e-01 6.03266582e-02 6.40750229e-01
1.51578084e-01 -9.70218703e-02 -3.64260077e-01 -8.15146923e-01
1.74302561e-03 -2.92857736e-01 -8.60773087e-01 -1.92173660e-01
-2.54132523e-04 -2.79643759e-02 6.89375639e-01 -3.33334029e-01
-1.29006660e+00 2.08024755e-01 1.83426932e-01 -7.23177433e-01
-1.80089846e-01 -2.76643276e-01 4.51237671e-02 -2.29635611e-01
3.92290920e-01 4.10336163e-03 -1.17653832e-01 -3.31667453e-01
2.59636462e-01 1.73064888e-01 9.42329526e-01 -5.29348731e-01
1.18831444e+00 1.30785298e+00 1.47506312e-01 -1.06933284e+00
-9.93931949e-01 -2.94863373e-01 -7.09969103e-01 -3.77807766e-01
7.97830045e-01 -8.97703290e-01 -6.78298116e-01 7.86397159e-01
-1.53625405e+00 -7.03060985e-01 -9.27351490e-02 4.83281910e-01
-2.66064078e-01 5.04690826e-01 -7.43492603e-01 -8.28286529e-01
-1.65218994e-01 -1.10696244e+00 1.00965285e+00 4.56179619e-01
2.15124384e-01 -1.26379609e+00 1.82241708e-01 3.58890891e-01
4.55807775e-01 5.65083146e-01 4.76286560e-01 2.71396607e-01
-1.11739814e+00 9.52142701e-02 -2.25401402e-01 6.29233599e-01
3.41133058e-01 4.83356118e-01 -1.30035126e+00 -2.60001391e-01
-1.20707750e-01 -1.46346748e-01 1.06736469e+00 4.74195689e-01
7.77367771e-01 -5.11223614e-01 -1.05772428e-01 1.35047829e+00
1.49234903e+00 -1.62133679e-01 9.76830304e-01 3.62229377e-01
1.13704073e+00 7.06746221e-01 3.01923037e-01 4.15840536e-01
3.52247208e-01 4.43881631e-01 7.99545586e-01 -5.75744331e-01
-3.89049590e-01 -1.13263242e-01 3.58865440e-01 2.63339400e-01
-2.55989701e-01 -3.14559221e-01 -8.29080582e-01 5.82618415e-01
-1.72170305e+00 -1.07160008e+00 -3.31081122e-01 2.37457991e+00
3.68044436e-01 9.91714653e-03 -3.89577985e-01 -2.59854019e-01
6.03080988e-01 5.30593932e-01 -5.08089960e-01 -1.04884349e-01
-2.59737283e-01 1.33511722e-01 6.30743921e-01 1.34035897e+00
-1.20275640e+00 1.05842590e+00 6.84502077e+00 -1.77985251e-01
-1.34830165e+00 -9.74934697e-02 5.19525230e-01 -4.01887268e-01
-3.15864474e-01 2.62148172e-01 -3.98257613e-01 3.64957780e-01
5.02109468e-01 4.04469639e-01 6.59666598e-01 8.76680315e-02
7.75106311e-01 -3.01334590e-01 -7.25188792e-01 7.15681911e-01
1.16924733e-01 -1.42658782e+00 -6.02439977e-02 3.75480764e-02
8.66272151e-01 2.83750594e-01 -5.52031286e-02 -3.70683908e-01
5.94800472e-01 -8.99177253e-01 7.38937199e-01 6.63533092e-01
5.86640120e-01 -2.06898421e-01 3.88236552e-01 1.32748142e-01
-9.84336078e-01 -1.74453426e-02 -1.15233235e-01 -4.58656520e-01
4.61103648e-01 6.32590055e-01 -6.06257141e-01 2.46811464e-01
5.84393382e-01 9.69122529e-01 -3.63288283e-01 1.11033094e+00
-3.61882895e-01 5.81571400e-01 -5.61024249e-01 1.01248610e+00
1.98742658e-01 -5.35878241e-01 8.89010072e-01 1.27610219e+00
1.93319172e-02 -5.24878949e-02 -3.87897529e-02 8.97594333e-01
4.36660089e-02 -5.79064131e-01 -8.02306712e-01 5.54290950e-01
2.34037325e-01 1.34827828e+00 -4.83333200e-01 -2.38866776e-01
-6.34892523e-01 1.06460989e+00 2.54627168e-01 9.88820076e-01
-9.33106661e-01 -1.43453419e-01 1.36827135e+00 5.88577747e-01
4.40951705e-01 -5.77669442e-01 -6.47083595e-02 -1.76044381e+00
1.73045814e-01 -5.00009239e-01 -1.10016530e-02 -8.66213143e-01
-8.40544462e-01 3.76596391e-01 -1.82897851e-01 -1.24667776e+00
9.51211751e-02 -4.87889290e-01 -9.52568769e-01 8.57048333e-01
-2.21810508e+00 -9.63571846e-01 -7.54663169e-01 5.39557278e-01
3.46345395e-01 4.51394171e-01 2.91221321e-01 4.93483812e-01
-7.31524885e-01 -8.01115260e-02 1.34372577e-01 4.00787801e-01
8.79160404e-01 -9.98011589e-01 5.86738288e-01 1.61291587e+00
-8.80052522e-02 5.50162673e-01 6.35768652e-01 -3.63835782e-01
-1.27473545e+00 -1.15321970e+00 4.34343696e-01 -4.52897370e-01
4.72435802e-01 -5.91986895e-01 -1.06001961e+00 9.21810865e-01
1.37931645e-01 7.91643858e-01 8.34756047e-02 -5.62377512e-01
-5.79593539e-01 -1.92586347e-01 -7.48060346e-01 5.52984238e-01
1.01198554e+00 -3.37173253e-01 -1.11754328e-01 1.55190691e-01
3.74697715e-01 -6.63452864e-01 -3.26697797e-01 2.62880743e-01
7.78164685e-01 -1.26233089e+00 1.27471173e+00 -6.76873267e-01
5.74015498e-01 -4.53956604e-01 -4.76986170e-02 -1.22488570e+00
4.62123007e-03 -1.01481271e+00 -2.95433193e-01 7.83160090e-01
-1.01630343e-02 -4.77135897e-01 7.02221334e-01 8.49391699e-01
-3.47314216e-02 -1.94354698e-01 -7.06352055e-01 -3.23135436e-01
-6.08159490e-02 -3.59837294e-01 1.13047414e-01 8.61983061e-01
-6.34867907e-01 1.39969662e-01 -9.09936368e-01 5.39875746e-01
9.10236120e-01 4.25150655e-02 1.12459624e+00 -1.03579104e+00
-2.78270543e-01 1.34821096e-02 -1.87948152e-01 -1.36517012e+00
2.96498030e-01 -2.89332509e-01 3.03481460e-01 -1.46908557e+00
-1.63847432e-01 -2.86494255e-01 1.06193170e-01 3.08596075e-01
-4.13210183e-01 4.32099879e-01 2.39348799e-01 2.36989856e-01
-4.99954224e-01 3.25703621e-01 1.49706316e+00 9.36681870e-03
-3.31058592e-01 -1.92295775e-01 -3.70828420e-01 1.14032674e+00
5.79819441e-01 -5.00369251e-01 -3.80237103e-01 -1.17962003e+00
4.42816243e-02 3.63244265e-02 6.92911744e-01 -9.47144628e-01
1.01768866e-01 -3.80140424e-01 5.55545390e-01 -1.41085247e-02
2.76359260e-01 -8.86789620e-01 -3.36615928e-02 4.93819207e-01
-6.63860962e-02 -1.84353217e-01 4.07014221e-01 7.34731674e-01
3.16439420e-02 1.99225023e-01 1.09792149e+00 -2.14626566e-01
-5.15302658e-01 3.60895544e-01 -3.48517835e-01 4.64128762e-01
6.23842299e-01 -2.39204064e-01 -5.94028592e-01 -5.96785426e-01
-4.16962951e-01 2.19361991e-01 6.54660344e-01 3.49442452e-01
7.05480516e-01 -7.63135910e-01 -7.22387910e-01 5.93087375e-01
-2.68187821e-01 3.05848271e-01 2.21788451e-01 8.77883315e-01
-1.24894738e+00 -1.40970692e-01 -1.98133022e-01 -5.53598285e-01
-1.10299182e+00 3.98058519e-02 8.65861952e-01 -7.05574825e-02
-1.07456660e+00 6.06435895e-01 6.11298025e-01 -1.32238716e-01
8.30817744e-02 -4.39648479e-01 1.69500977e-01 -4.48035508e-01
8.00523400e-01 5.50504625e-01 -1.50448978e-01 -6.49389267e-01
-3.15540403e-01 8.43497336e-01 8.44327509e-02 -2.40679801e-01
1.33570540e+00 -6.52204931e-01 7.70170912e-02 1.42021790e-01
1.23148727e+00 2.67667860e-01 -2.26079535e+00 -5.81129119e-02
-4.82330412e-01 -9.13420558e-01 1.06983013e-01 -3.35943252e-01
-1.40115261e+00 9.17510986e-01 8.64830539e-02 -2.87358254e-01
9.65273499e-01 -6.74586535e-01 1.13014901e+00 2.17684850e-01
1.67948976e-02 -6.95824265e-01 -2.08853986e-02 5.67680418e-01
3.69924068e-01 -1.23972714e+00 1.26964822e-01 -4.30160314e-01
-2.58662939e-01 1.19532681e+00 5.96491516e-01 -4.58319128e-01
5.25285602e-01 5.99893868e-01 4.99161452e-01 2.33055905e-01
-8.35710824e-01 -8.33774656e-02 1.67974010e-02 7.73854911e-01
-1.72881354e-02 -4.54272985e-01 3.99503350e-01 -3.75320882e-01
2.25610167e-01 -7.89704397e-02 1.17677784e+00 9.96228695e-01
-2.83318460e-01 -6.44699872e-01 -4.74582821e-01 8.48195404e-02
-2.99621373e-01 -1.99188232e-01 -1.75217181e-01 7.77011037e-01
1.13549180e-01 9.17571545e-01 3.31854820e-01 -1.67677924e-01
2.98215508e-01 -3.26639354e-01 5.62499881e-01 -3.32251102e-01
-3.50373328e-01 1.14846639e-01 6.09986931e-02 -1.03704917e+00
-6.29324853e-01 -4.85530555e-01 -1.20130742e+00 -3.98456931e-01
-6.73111603e-02 -3.32493037e-01 4.46812481e-01 8.28804195e-01
3.58841389e-01 4.64317083e-01 6.67912245e-01 -1.30792046e+00
1.30055830e-01 -4.95381057e-01 -4.94359344e-01 7.75613666e-01
1.27812636e+00 -6.01561189e-01 -9.71917808e-01 5.57901025e-01] | [10.53453254699707, -1.597476840019226] |
873a3617-7c59-464a-aac0-7f39d6116f3e | grammar-based-grounded-lexicon-learning-1 | 2202.08806 | null | https://arxiv.org/abs/2202.08806v1 | https://arxiv.org/pdf/2202.08806v1.pdf | Grammar-Based Grounded Lexicon Learning | We present Grammar-Based Grounded Lexicon Learning (G2L2), a lexicalist approach toward learning a compositional and grounded meaning representation of language from grounded data, such as paired images and texts. At the core of G2L2 is a collection of lexicon entries, which map each word to a tuple of a syntactic type and a neuro-symbolic semantic program. For example, the word shiny has a syntactic type of adjective; its neuro-symbolic semantic program has the symbolic form {\lambda}x. filter(x, SHINY), where the concept SHINY is associated with a neural network embedding, which will be used to classify shiny objects. Given an input sentence, G2L2 first looks up the lexicon entries associated with each token. It then derives the meaning of the sentence as an executable neuro-symbolic program by composing lexical meanings based on syntax. The recovered meaning programs can be executed on grounded inputs. To facilitate learning in an exponentially-growing compositional space, we introduce a joint parsing and expected execution algorithm, which does local marginalization over derivations to reduce the training time. We evaluate G2L2 on two domains: visual reasoning and language-driven navigation. Results show that G2L2 can generalize from small amounts of data to novel compositions of words. | ['Joshua B. Tenenbaum', 'Roger P. Levy', 'Jiajun Wu', 'Haoyue Shi', 'Jiayuan Mao'] | 2022-02-17 | grammar-based-grounded-lexicon-learning | http://proceedings.neurips.cc/paper/2021/hash/4158f6d19559955bae372bb00f6204e4-Abstract.html | http://proceedings.neurips.cc/paper/2021/file/4158f6d19559955bae372bb00f6204e4-Paper.pdf | neurips-2021-12 | ['network-embedding'] | ['methodology'] | [ 4.40261215e-01 4.47654337e-01 -2.21789852e-01 -4.98206079e-01
-7.93154180e-01 -8.18135977e-01 4.73640531e-01 2.83189416e-01
-2.37857655e-01 4.69374150e-01 2.50240088e-01 -5.57818830e-01
1.22757345e-01 -1.37064528e+00 -1.12594926e+00 -6.25580847e-01
-9.67727527e-02 6.58741415e-01 6.10534325e-02 -2.87554413e-01
2.05667287e-01 1.11923553e-01 -1.66740811e+00 7.72076309e-01
5.55029213e-01 9.79538798e-01 6.27398014e-01 5.39426148e-01
-8.75179648e-01 1.26236606e+00 -4.82098162e-01 -3.97265524e-01
-8.66596550e-02 -8.53631139e-01 -8.50673020e-01 -2.34660596e-01
2.37235382e-01 -1.10918514e-01 1.39801428e-01 1.31122088e+00
-3.06856055e-02 2.25949615e-01 6.62777543e-01 -1.36340559e+00
-9.28691328e-01 1.22538280e+00 1.84096113e-01 -2.61286914e-01
5.55454254e-01 9.39560458e-02 1.30389500e+00 -1.07856238e+00
1.00422132e+00 1.78477144e+00 5.61028659e-01 8.26507449e-01
-1.38130128e+00 -6.32422924e-01 2.22961903e-01 1.53921887e-01
-1.27459002e+00 -1.06267668e-01 5.26168406e-01 -4.08352077e-01
1.42871463e+00 -8.66516382e-02 9.39683855e-01 1.06163418e+00
3.49906653e-01 8.92902553e-01 9.42289412e-01 -7.82626212e-01
7.52707303e-01 7.64230117e-02 2.94724733e-01 1.35461569e+00
2.57118214e-02 8.80247578e-02 -7.80721426e-01 -3.58329117e-02
5.95215023e-01 -2.00061291e-01 2.73236543e-01 -3.21891010e-01
-1.24317133e+00 9.49084163e-01 5.14276564e-01 1.28384426e-01
-2.06107780e-01 5.35957336e-01 4.33114856e-01 4.24267620e-01
8.54057297e-02 6.25588357e-01 -1.24257691e-01 2.64097959e-01
-6.07365549e-01 4.07242000e-01 8.34737062e-01 1.20379412e+00
1.11024368e+00 1.16029851e-01 1.92228198e-01 3.34994733e-01
4.26606655e-01 8.85675192e-01 5.97924232e-01 -1.04969788e+00
2.48913750e-01 9.23435628e-01 -5.36934078e-01 -7.77269959e-01
-2.96692133e-01 1.86303899e-01 -3.14303637e-01 2.59379506e-01
1.32964641e-01 1.51665792e-01 -7.59254515e-01 2.12370467e+00
5.96051861e-04 1.68284208e-01 5.00013292e-01 6.40790462e-01
1.09983766e+00 8.27308774e-01 4.79857981e-01 7.59879127e-02
1.45033669e+00 -7.95188785e-01 -4.38925564e-01 -5.15300333e-01
1.00089526e+00 2.45680213e-02 1.67508912e+00 3.28379631e-01
-1.17735612e+00 -4.78748053e-01 -1.22322583e+00 -3.01940531e-01
-7.51546443e-01 -2.80341268e-01 7.37269461e-01 2.95538574e-01
-1.21655619e+00 3.93031448e-01 -8.34359527e-01 -4.78798032e-01
3.54369432e-01 1.39486253e-01 -3.27491671e-01 -2.65416056e-02
-1.20182371e+00 8.19666266e-01 8.23052347e-01 -3.69982898e-01
-1.23544037e+00 -4.52754438e-01 -1.41606402e+00 6.89726789e-03
5.31978667e-01 -8.87635529e-01 1.15879786e+00 -1.18284023e+00
-1.19972801e+00 1.39605832e+00 -2.03739807e-01 -7.22624362e-01
-3.41326803e-01 2.19742373e-01 -3.47866297e-01 3.21567208e-01
3.99130195e-01 9.67895985e-01 8.23978186e-01 -1.18052697e+00
-5.61902106e-01 -5.05263746e-01 4.08395052e-01 2.28045672e-01
-1.27565965e-01 -1.32694334e-01 -2.87673384e-01 -5.95348299e-01
3.33082944e-01 -9.10454750e-01 1.09615006e-01 -5.20074144e-02
-4.77245539e-01 -2.50575632e-01 2.18816891e-01 -4.50284570e-01
1.10998023e+00 -2.16297913e+00 5.32332122e-01 2.57450879e-01
3.85056287e-01 -4.41069692e-01 -2.53567576e-01 3.22589099e-01
-4.18297797e-02 1.74241871e-01 -4.62820381e-01 -1.15082711e-01
1.56034961e-01 4.70759273e-01 -7.26314306e-01 2.89158486e-02
2.19234228e-01 1.26039994e+00 -1.21385694e+00 -6.57646418e-01
-4.79669397e-04 -7.33826384e-02 -7.80004859e-01 2.32196689e-01
-8.18523824e-01 -5.37713394e-02 -4.64125037e-01 6.08459055e-01
-7.64031187e-02 -2.03577608e-01 4.08015370e-01 -9.33265910e-02
2.19624206e-01 4.32587594e-01 -9.14814055e-01 2.08805680e+00
-9.08191025e-01 5.18313766e-01 -3.28795969e-01 -1.00028133e+00
9.49599504e-01 7.32136294e-02 -2.53838569e-01 -7.12163329e-01
1.48921952e-01 2.36983344e-01 -1.89419582e-01 -6.19762540e-01
3.31820011e-01 -6.58933699e-01 -6.45531476e-01 6.41378641e-01
3.99338573e-01 -5.06291807e-01 1.41879246e-01 4.00768280e-01
1.02248216e+00 5.16929150e-01 5.60299635e-01 -4.57593113e-01
5.87894082e-01 3.88268054e-01 1.49494126e-01 7.02400506e-01
3.32348585e-01 2.07597986e-02 6.28355563e-01 -4.85182405e-01
-8.31366599e-01 -1.40966415e+00 4.21063185e-01 1.58446348e+00
2.00363815e-01 -5.09607553e-01 -1.03565037e+00 -4.48092073e-01
-2.28566527e-01 1.48525214e+00 -6.05033517e-01 -5.29374421e-01
-6.29133701e-01 -3.09691727e-02 6.97812140e-01 6.95214808e-01
2.32564211e-01 -1.84144866e+00 -1.13704765e+00 2.32059285e-01
-5.56636900e-02 -9.02167082e-01 -2.29961485e-01 5.81086934e-01
-9.30536926e-01 -9.45669830e-01 2.54261345e-01 -1.30819249e+00
8.82783711e-01 -3.51128906e-01 1.45483518e+00 9.65898857e-02
-1.55265346e-01 3.44879210e-01 -2.32641488e-01 -3.10033292e-01
-8.75307322e-01 -5.90694509e-02 -8.10130462e-02 -3.11693162e-01
5.32175779e-01 -4.55342084e-01 2.11087689e-01 -3.31316769e-01
-9.93981004e-01 4.49786514e-01 1.77997991e-01 9.22453761e-01
9.19778824e-01 -4.67734933e-02 3.09778750e-01 -1.09481072e+00
5.79270959e-01 -4.11380768e-01 -6.43027961e-01 4.46631402e-01
-2.21230790e-01 6.23309791e-01 9.39582646e-01 -4.68249232e-01
-8.67162943e-01 -6.38399348e-02 8.96940678e-02 -3.28465074e-01
-1.44332856e-01 8.52954924e-01 -5.11341572e-01 5.96807003e-01
9.40369964e-01 3.86222392e-01 5.92331812e-02 -1.37246512e-02
9.12804008e-01 2.75419056e-01 1.02037454e+00 -1.25211918e+00
4.24735665e-01 4.06614929e-01 3.39385904e-02 -7.16753483e-01
-8.91657174e-01 2.49953941e-01 -4.33399379e-01 5.82007505e-02
1.21699190e+00 -7.32588768e-01 -6.22826219e-01 6.15352839e-02
-1.25250661e+00 -7.74639189e-01 -7.63946712e-01 2.23802567e-01
-1.18590343e+00 -2.67403901e-01 -5.25318623e-01 -7.06727803e-01
-2.54656732e-01 -1.18308103e+00 1.14536822e+00 -2.23003939e-01
-6.23199224e-01 -1.22253954e+00 -1.48310527e-01 -2.34071717e-01
-1.91019371e-01 3.17475945e-01 2.01225042e+00 -8.88163149e-01
-5.85948825e-01 4.75222096e-02 1.44966662e-01 2.08026558e-01
-2.17304036e-01 -4.32119459e-01 -7.59216189e-01 3.41248214e-02
2.14338690e-01 -7.89835989e-01 5.26526213e-01 9.01802182e-02
1.01390243e+00 -4.03883785e-01 -2.45440647e-01 7.68184483e-01
1.56846762e+00 4.60518420e-01 4.14854974e-01 4.16092426e-01
6.59083068e-01 8.01468968e-01 4.50274676e-01 -6.39212951e-02
4.82842416e-01 2.55225062e-01 3.85331452e-01 3.36288780e-01
-3.13158095e-01 -7.89351881e-01 7.00676560e-01 7.53010631e-01
3.14528316e-01 9.06175897e-02 -1.14246500e+00 4.42190051e-01
-1.62034440e+00 -9.12616432e-01 4.08796310e-01 1.97853422e+00
9.87824380e-01 1.50260016e-01 -2.26216957e-01 -1.15961686e-01
5.93159676e-01 -9.93190408e-02 -8.18298578e-01 -5.01689494e-01
-3.22110653e-01 5.64197838e-01 -2.58046426e-02 7.58644521e-01
-5.41648567e-01 1.51796103e+00 6.00896215e+00 7.40510702e-01
-9.59802091e-01 1.31459430e-01 2.83125043e-01 7.46763200e-02
-7.81755209e-01 6.04275614e-02 -7.00689316e-01 -1.30151799e-02
1.16685200e+00 -5.29704571e-01 9.09968734e-01 9.59541798e-01
-3.49204153e-01 4.02375050e-02 -1.70695412e+00 9.98481810e-01
3.42623323e-01 -1.51525164e+00 5.34061790e-01 -3.82232606e-01
3.83171111e-01 -1.77015275e-01 -7.08442852e-02 6.43210590e-01
4.33136582e-01 -1.15964031e+00 1.45863378e+00 4.25700754e-01
1.02481186e+00 -7.81963527e-01 1.67179003e-01 4.29335952e-01
-1.29225957e+00 -1.73126236e-01 -2.96637118e-01 -6.93502873e-02
-5.30474596e-02 -4.14830782e-02 -4.96713430e-01 4.86463010e-02
4.48495537e-01 6.79970980e-01 -4.43899781e-01 -9.71959382e-02
-9.45111632e-01 4.20974106e-01 -9.39584337e-03 -3.38216871e-01
3.01492721e-01 -1.56265691e-01 3.76072675e-01 1.16986012e+00
5.06957650e-01 2.29745492e-01 2.54601181e-01 1.45587289e+00
-1.30963475e-01 2.29872495e-01 -1.18290341e+00 -2.41342694e-01
6.84232235e-01 1.01240373e+00 -9.86902773e-01 -8.49484861e-01
-3.65170628e-01 8.08230102e-01 3.66844028e-01 3.83952081e-01
-7.23744392e-01 -3.31243902e-01 4.48352158e-01 -1.72285195e-02
9.24841017e-02 -8.21903348e-02 -4.31160748e-01 -1.02656543e+00
-2.42008507e-01 -1.04035032e+00 3.24429542e-01 -1.23505652e+00
-8.50912809e-01 8.71026814e-01 3.84709209e-01 -7.95104682e-01
-6.99302852e-01 -9.85226035e-01 -3.48539591e-01 9.05060053e-01
-7.36821890e-01 -1.19701397e+00 -7.16200992e-02 7.17143178e-01
6.38503730e-01 -7.26984918e-01 1.34262145e+00 -4.47691858e-01
-5.16487397e-02 2.93851048e-01 -6.15416288e-01 8.38357303e-03
-2.49912683e-02 -1.25012636e+00 6.42908990e-01 6.85206950e-01
4.22472447e-01 1.12856770e+00 5.74499547e-01 -6.99273229e-01
-1.60989511e+00 -1.21292567e+00 1.02733088e+00 -5.71715713e-01
9.44201529e-01 -7.06237674e-01 -8.65324616e-01 1.06213999e+00
1.16032988e-01 -1.67969912e-01 6.81510985e-01 -1.61479607e-01
-7.88189590e-01 3.04509014e-01 -1.05094397e+00 1.00607562e+00
1.36138558e+00 -1.11358333e+00 -1.22116649e+00 2.53341883e-01
1.16168833e+00 -3.84847105e-01 -3.90815079e-01 -1.84832394e-01
3.87383342e-01 -7.53086865e-01 8.38113129e-01 -8.35812211e-01
5.78783572e-01 -2.65192330e-01 -7.62053549e-01 -1.18749547e+00
-3.25411148e-02 -2.92540461e-01 -1.89755499e-01 7.86356807e-01
4.67949957e-01 -4.98352855e-01 6.74912989e-01 3.15641135e-01
-2.68899053e-01 -7.76577711e-01 -6.93303406e-01 -8.05521369e-01
1.94739312e-01 -1.06704605e+00 7.87285984e-01 5.13359249e-01
2.48595327e-01 6.66839659e-01 3.04937810e-01 -7.36489072e-02
6.20675027e-01 4.87186521e-01 3.00972641e-01 -9.57963169e-01
-4.58656222e-01 -4.32966024e-01 -3.98003489e-01 -7.50947773e-01
1.02841496e+00 -1.86875093e+00 4.00055438e-01 -1.42888522e+00
1.51943728e-01 -3.37038152e-02 2.33196113e-02 8.92469585e-01
2.14685366e-01 4.73135673e-02 2.34121367e-01 8.76848120e-03
-4.46234137e-01 2.65183121e-01 9.40628886e-01 -4.44653958e-01
-1.56608596e-01 -6.82597041e-01 -7.72569597e-01 1.12751544e+00
6.13600254e-01 -5.40064812e-01 -8.95504236e-01 -4.67286468e-01
7.65852928e-01 1.59821019e-01 5.33190429e-01 -7.87195981e-01
2.51931995e-01 -1.51105300e-01 -1.30867034e-01 -3.64098161e-01
7.40186572e-02 -6.68457627e-01 1.19311139e-01 4.98694658e-01
-7.55711138e-01 2.09619343e-01 2.20728055e-01 1.94971979e-01
-2.31665447e-01 -5.20767033e-01 4.71840590e-01 -4.99964535e-01
-1.13371456e+00 -1.96900278e-01 -3.83901238e-01 3.63493145e-01
6.96822464e-01 -1.94251120e-01 -1.15740612e-01 -1.99461281e-01
-9.82498586e-01 -3.66885886e-02 5.20217597e-01 3.73252392e-01
9.80733395e-01 -1.48795378e+00 -1.96809664e-01 4.86934781e-01
4.26543027e-01 -4.94730240e-03 -8.18010122e-02 2.85479099e-01
-5.39371252e-01 2.12655202e-01 -2.79542744e-01 -4.35232937e-01
-8.52407098e-01 9.61847246e-01 3.08919817e-01 8.38555843e-02
-7.80334592e-01 8.98830891e-01 6.05088949e-01 -6.79253340e-01
1.78475842e-01 -8.77311945e-01 -4.90117585e-03 4.14691940e-02
6.63316250e-01 -9.13903564e-02 -1.49916485e-01 -5.83044231e-01
-3.78247589e-01 6.56405807e-01 3.06317121e-01 -5.32664835e-01
1.11198425e+00 1.91352800e-01 -6.23833776e-01 9.11423564e-01
1.27925014e+00 -3.91830206e-01 -9.11452413e-01 -2.19365984e-01
3.66260767e-01 2.84007370e-01 -3.00976902e-01 -6.28833234e-01
-7.60077000e-01 8.73421073e-01 1.66023314e-01 -8.01958516e-02
1.06210256e+00 5.73086560e-01 5.58514357e-01 9.22314703e-01
7.97570825e-01 -7.07036197e-01 2.55598247e-01 9.03285325e-01
9.73892272e-01 -5.64871967e-01 -5.09059370e-01 -1.56281441e-01
-6.87300563e-01 1.20352757e+00 3.82759422e-01 -2.42982283e-01
3.57981741e-01 5.19373596e-01 1.14549370e-02 -4.02213871e-01
-8.46791983e-01 -2.14711517e-01 7.44804787e-03 8.36588204e-01
1.93366349e-01 9.01769996e-02 9.95318592e-02 9.73925769e-01
-6.48528814e-01 -7.90577456e-02 2.02171549e-01 9.76442754e-01
-5.56156099e-01 -8.74300301e-01 -2.28228360e-01 1.46422356e-01
1.70346856e-01 -5.67083597e-01 -2.22465336e-01 7.93168008e-01
4.53786016e-01 4.77200925e-01 2.57952780e-01 -3.27062398e-01
3.40018153e-01 6.40830100e-01 8.05348873e-01 -1.26990330e+00
-4.09989685e-01 -1.28660262e-01 -6.66654110e-02 -8.86578321e-01
-3.57774675e-01 -4.89949942e-01 -2.44863462e+00 4.27041855e-03
4.54161793e-01 -5.15866019e-02 5.68921566e-01 1.06245768e+00
-8.54156464e-02 6.14308178e-01 1.51631073e-03 -5.84745228e-01
-3.96672070e-01 -4.09159780e-01 -6.58573210e-01 5.05214632e-01
9.47017297e-02 -4.57805991e-01 -1.63606077e-01 5.90767682e-01] | [9.60303783416748, 7.254092216491699] |
566911e6-6dc7-42a6-9735-5ca5bb57f628 | multilingual-abusiveness-identification-on | 2204.01848 | null | https://arxiv.org/abs/2204.01848v1 | https://arxiv.org/pdf/2204.01848v1.pdf | Multilingual Abusiveness Identification on Code-Mixed Social Media Text | Social Media platforms have been seeing adoption and growth in their usage over time. This growth has been further accelerated with the lockdown in the past year when people's interaction, conversation, and expression were limited physically. It is becoming increasingly important to keep the platform safe from abusive content for better user experience. Much work has been done on English social media content but text analysis on non-English social media is relatively underexplored. Non-English social media content have the additional challenges of code-mixing, transliteration and using different scripture in same sentence. In this work, we propose an approach for abusiveness identification on the multilingual Moj dataset which comprises of Indic languages. Our approach tackles the common challenges of non-English social media content and can be extended to other languages as well. | ['Naman Poddar', 'Ekagra Ranjan'] | 2022-03-01 | null | null | null | null | ['transliteration'] | ['natural-language-processing'] | [-4.34122294e-01 -1.73508570e-01 -1.19800933e-01 -8.80949274e-02
-5.59163094e-01 -7.95071900e-01 4.47926998e-01 4.47082430e-01
-4.69899267e-01 5.76480389e-01 3.36426616e-01 -4.02507097e-01
3.48462552e-01 -3.12164754e-01 2.54349224e-02 -8.46563578e-02
2.03211725e-01 -1.05495453e-01 3.54814261e-01 -6.15441978e-01
3.14494401e-01 5.69751486e-03 -1.17761171e+00 3.72629553e-01
1.01387346e+00 2.07177192e-01 1.47058874e-01 7.22471535e-01
-5.95204890e-01 1.22955465e+00 -6.02739513e-01 -8.99455786e-01
-9.51930210e-02 -3.71430963e-01 -9.22507346e-01 8.61547068e-02
3.59960020e-01 -4.86559421e-02 -1.93992704e-01 1.45391977e+00
4.68361557e-01 -2.15975702e-01 2.50135034e-01 -1.08608711e+00
-5.87686002e-01 8.96172047e-01 -8.47713351e-01 5.04898190e-01
7.73204327e-01 -2.43139103e-01 7.84749985e-01 -5.25988162e-01
9.45679247e-01 1.01774621e+00 8.09654117e-01 3.26588035e-01
-9.47107434e-01 -7.01447010e-01 -1.11032501e-01 2.93490607e-02
-1.61442661e+00 -2.99751282e-01 4.95087624e-01 -9.15253401e-01
1.06035602e+00 2.71637380e-01 5.27458012e-01 1.29754114e+00
2.36960888e-01 5.77128232e-01 1.10109043e+00 -6.81432784e-01
-2.69096136e-01 6.94586158e-01 1.84551880e-01 6.21585429e-01
8.37005973e-02 -8.41845155e-01 -4.43895787e-01 -2.84907758e-01
-5.73163144e-02 -9.27526206e-02 9.64485481e-03 3.26449603e-01
-6.56647086e-01 1.11341691e+00 -3.10989231e-01 8.31031322e-01
1.05983149e-02 -3.56466323e-01 8.95866692e-01 5.69058776e-01
7.03599334e-01 3.07470411e-01 -1.46105498e-01 -8.18495810e-01
-1.05239332e+00 2.31090501e-01 1.02355301e+00 1.11500442e+00
3.98481250e-01 -2.00378582e-01 4.46813941e-01 1.42180848e+00
3.69511098e-01 4.23773348e-01 7.63517201e-01 -6.44458950e-01
7.34635115e-01 7.12128937e-01 -1.58780068e-01 -1.36551368e+00
-3.36713970e-01 -1.74979672e-01 -3.58406216e-01 -1.43572435e-01
6.89299405e-01 -3.55960190e-01 -2.08012104e-01 1.39814663e+00
2.38692299e-01 -4.83892560e-01 -2.22157463e-01 3.01841289e-01
5.23349345e-01 5.28283656e-01 1.28717750e-01 -2.39071608e-01
1.33326781e+00 -7.49907255e-01 -1.00521111e+00 -2.73300484e-02
9.76732731e-01 -1.54781091e+00 1.26807046e+00 4.73472774e-01
-8.49879980e-01 1.39274094e-02 -9.94816780e-01 -2.64253885e-01
-6.17238402e-01 -4.38188910e-01 2.45763734e-01 1.27140319e+00
-6.22472346e-01 4.48008925e-01 -5.62167168e-01 -8.88694286e-01
3.19132596e-01 1.51948288e-01 -4.12094563e-01 2.34152511e-01
-1.11227000e+00 9.69625354e-01 2.43797563e-02 -4.60176408e-01
-2.32578456e-01 -2.60105968e-01 -7.02675939e-01 -5.82805514e-01
4.14701253e-01 7.10522607e-02 1.25303173e+00 -1.37261760e+00
-1.68737423e+00 1.23072374e+00 1.64659426e-01 1.86902415e-02
6.85298026e-01 -5.22008300e-01 -1.01338017e+00 -2.68249601e-01
2.06044853e-01 -9.21399519e-02 7.27357984e-01 -6.87322140e-01
-4.56680208e-01 -2.97595412e-01 1.20483615e-01 -9.53578800e-02
-7.95911849e-01 1.10328376e+00 -2.22060934e-01 -7.51122832e-01
-1.56646609e-01 -1.23084688e+00 2.46990845e-01 -3.33579659e-01
-6.27816439e-01 1.11394837e-01 1.03476536e+00 -1.12826335e+00
1.84009540e+00 -2.37640738e+00 1.11087963e-01 5.25626801e-02
2.58681983e-01 2.39709228e-01 3.87072176e-01 9.62138653e-01
1.71385139e-01 6.74021721e-01 -1.87017694e-01 -5.53907275e-01
-1.40825734e-01 1.17945015e-01 -1.15508117e-01 6.41169190e-01
-2.45024621e-01 3.13065767e-01 -8.92090321e-01 -8.40068758e-01
-2.44718462e-01 4.99258339e-01 -5.85809410e-01 -5.20704091e-02
2.73113191e-01 3.92851084e-01 -7.37868026e-02 5.88343501e-01
5.47331035e-01 1.16619065e-01 1.06569692e-01 5.87174356e-01
-6.62196696e-01 3.04610699e-01 -1.03014958e+00 1.66844523e+00
-5.80517948e-01 1.03716016e+00 2.50353843e-01 -9.58008617e-02
5.59707165e-01 3.19766670e-01 2.89319515e-01 -2.91484684e-01
4.20898288e-01 3.49202007e-01 3.02155703e-01 -1.05608618e+00
8.79680336e-01 -4.78310026e-02 -3.30434442e-01 5.39329588e-01
-5.04249707e-02 -6.38689324e-02 3.67154926e-01 5.51355183e-01
7.27656424e-01 5.76358177e-02 6.97295964e-01 -4.73780930e-01
6.25026464e-01 -1.66701972e-01 2.20707029e-01 4.25772190e-01
-5.32189727e-01 4.15888876e-01 6.29071712e-01 9.74683762e-02
-1.07565570e+00 -6.63404942e-01 -3.22961599e-01 1.28582597e+00
-2.92979956e-01 -9.90006626e-01 -9.80730891e-01 -6.56728327e-01
-1.66665688e-01 5.91058314e-01 -1.88072830e-01 1.95783734e-01
-5.30462503e-01 -6.04047418e-01 9.21156526e-01 -1.29994214e-01
3.84114295e-01 -6.65879071e-01 -2.21443132e-01 2.86204278e-01
-4.09929991e-01 -1.25631654e+00 -5.51375151e-01 -3.58978391e-01
-5.71900606e-01 -7.96139061e-01 -3.40901375e-01 -5.59328258e-01
2.84027159e-01 1.26395345e-01 7.07356930e-01 2.45650053e-01
-2.49289110e-01 2.22988561e-01 -6.76528335e-01 -6.10074818e-01
-9.17368710e-01 5.18231571e-01 8.14828724e-02 5.86400591e-02
6.10571921e-01 -5.93752682e-01 -9.40973461e-02 1.71047688e-01
-9.61019874e-01 -1.54487893e-01 -2.50397742e-01 2.65606523e-01
-4.00584251e-01 -3.07951093e-01 4.67305005e-01 -1.39200890e+00
7.13869870e-01 -1.06161380e+00 -1.12145111e-01 -2.67858505e-01
-3.56020927e-01 -4.53717321e-01 6.82842314e-01 -4.58612382e-01
-1.07050335e+00 -4.70660567e-01 -4.57404524e-01 4.20685232e-01
-9.10570547e-02 4.43562061e-01 7.83320069e-02 -3.49640027e-02
6.15458310e-01 -3.09563696e-01 -1.36205060e-02 -6.53727353e-01
6.83884174e-02 1.57530153e+00 -2.60528140e-02 -2.14309692e-01
7.49629200e-01 4.71358091e-01 -7.68895626e-01 -1.56672442e+00
-6.57653987e-01 -6.98724329e-01 -5.77594876e-01 -4.00702477e-01
1.01952100e+00 -8.32298815e-01 -4.93389189e-01 8.25339258e-01
-1.14102304e+00 -2.42554713e-02 3.42890501e-01 2.10802287e-01
5.44385202e-02 8.94079208e-01 -1.12128341e+00 -7.22823262e-01
-7.95318931e-02 -9.53356266e-01 3.67304981e-01 -3.38980667e-02
-9.63850260e-01 -1.19501173e+00 4.11195695e-01 5.52376270e-01
5.23337245e-01 1.56072155e-01 6.43173099e-01 -6.79977298e-01
4.97714803e-02 -3.21232498e-01 8.18275958e-02 3.94256264e-01
1.86803728e-01 6.39210343e-01 -8.62897217e-01 -2.47496113e-01
2.02224046e-01 -3.79701763e-01 8.52544904e-02 -3.34667087e-01
5.48844755e-01 -4.10611570e-01 -4.12434246e-03 2.67412663e-01
1.27979624e+00 6.94216564e-02 5.68266451e-01 6.73320651e-01
7.06918538e-01 6.27348661e-01 3.19918126e-01 5.56054115e-01
3.47138405e-01 5.76650500e-01 1.96518064e-01 4.01545435e-01
8.40133727e-02 -1.65834054e-01 5.96401095e-01 1.62535703e+00
-4.83303405e-02 -1.87795147e-01 -1.22613239e+00 3.89762223e-01
-1.60197675e+00 -1.07470131e+00 -5.49365222e-01 1.92496145e+00
8.85888159e-01 1.41088367e-01 3.72641951e-01 1.24958038e-01
6.88718617e-01 1.89600676e-01 1.26367107e-01 -9.95620370e-01
-1.09972239e-01 -1.26748100e-01 3.36574048e-01 7.37218201e-01
-1.02600777e+00 7.65973687e-01 6.09156895e+00 7.41262019e-01
-1.30448008e+00 8.15919280e-01 2.84205645e-01 -1.27085567e-01
-4.53097783e-02 4.41438593e-02 -9.88376141e-01 8.74885678e-01
8.88157845e-01 -5.63068502e-02 4.32473570e-01 8.69263709e-01
1.46958932e-01 -3.85686636e-01 -4.78863686e-01 1.03706586e+00
3.91025871e-01 -7.35014737e-01 -6.55449927e-01 7.79727474e-02
7.54280984e-01 3.45938563e-01 -7.70473108e-02 4.74874467e-01
1.12119995e-01 -7.75495887e-01 9.57466424e-01 3.04935183e-02
4.62184459e-01 -6.88110709e-01 5.89416027e-01 6.27957642e-01
-7.16039240e-01 -9.97973159e-02 -2.01292709e-01 -4.44342911e-01
2.55690545e-01 4.68762875e-01 -4.68114913e-01 1.98695213e-01
7.58667827e-01 7.16513634e-01 -8.96970153e-01 7.07276762e-01
-5.35686463e-02 7.91858613e-01 -3.55931222e-01 -1.95302203e-01
1.39945045e-01 -3.60830456e-01 8.50125790e-01 1.72257388e+00
1.68327048e-01 -4.25256878e-01 7.29987621e-02 3.16647440e-01
7.60511830e-02 8.37730289e-01 -7.86812305e-01 -4.09589320e-01
3.09757918e-01 1.35012913e+00 -8.46348763e-01 -1.26997158e-01
-8.68692398e-01 1.25227833e+00 6.61363065e-01 5.73558994e-02
-9.12120402e-01 -3.67396355e-01 6.79870129e-01 4.84604210e-01
-2.22111300e-01 -6.22395277e-01 -2.60503143e-02 -1.37746572e+00
1.44837135e-02 -1.19997859e+00 5.12433708e-01 -2.70241022e-01
-1.32831836e+00 7.37354875e-01 -1.58224314e-01 -1.15349925e+00
-1.29508883e-01 -3.74556541e-01 -2.48700038e-01 5.20249426e-01
-9.97857332e-01 -9.47265089e-01 3.06286756e-02 3.88432384e-01
5.70943713e-01 -1.53218448e-01 7.39609897e-01 9.33626831e-01
-8.80205750e-01 6.26097381e-01 2.02766389e-01 1.52422071e-01
9.85103011e-01 -9.72588241e-01 1.84462983e-02 9.93554294e-01
-5.41121364e-02 9.44330871e-01 9.10636723e-01 -8.02986622e-01
-1.00227487e+00 -6.08479977e-01 1.34998143e+00 -7.62388587e-01
1.32924068e+00 -7.41311848e-01 -7.39095449e-01 8.28792870e-01
7.21767306e-01 -5.53506851e-01 1.21617508e+00 2.41184875e-01
-5.62507987e-01 2.89187431e-01 -1.04999888e+00 8.95876706e-01
9.58441138e-01 -9.40339029e-01 -3.59984070e-01 3.42172831e-01
3.36629122e-01 -3.19949359e-01 -8.29805374e-01 -4.26557422e-01
4.77282912e-01 -1.11702228e+00 2.34609887e-01 -5.19397140e-01
7.27678478e-01 -1.12766817e-01 -1.29461408e-01 -1.14347947e+00
9.24822409e-03 -1.19079816e+00 2.66722888e-01 1.73237586e+00
4.20698971e-01 -6.03451014e-01 2.47324109e-01 5.41814625e-01
-2.89813057e-02 -1.45223513e-01 -7.89473832e-01 -5.78768730e-01
3.02239358e-01 -5.23758292e-01 3.86409834e-02 1.57455850e+00
6.85979784e-01 4.21181649e-01 -6.27616405e-01 -1.73041016e-01
9.50620919e-02 -4.27895606e-01 8.21769416e-01 -9.35870051e-01
-4.03765947e-01 -4.56232637e-01 -5.49328566e-01 -5.51236093e-01
-3.31758894e-02 -1.03274083e+00 -4.04203385e-01 -8.33001912e-01
4.47326660e-01 2.12033708e-02 5.38578272e-01 1.13910185e-02
-2.92607453e-02 5.85727870e-01 4.58162278e-01 2.90803671e-01
-5.91606438e-01 -1.18617252e-01 7.96365321e-01 2.40971789e-01
-2.48501897e-01 -3.73050421e-01 -5.08588850e-01 9.74124968e-01
8.13740194e-01 -7.70362020e-01 -1.27942577e-01 -1.40031621e-01
1.16004777e+00 -3.87172282e-01 -2.96464354e-01 -1.05636370e+00
-6.75024092e-02 2.76729744e-02 -2.58230329e-01 -5.20138443e-02
-4.70211394e-02 -8.29818189e-01 2.13255480e-01 2.99442887e-01
-5.20395022e-03 2.97544479e-01 9.37343240e-02 2.03422263e-01
-1.73005804e-01 -7.16113865e-01 9.63946342e-01 -1.42560959e-01
-2.48757303e-01 -1.30426884e-01 -1.25249267e+00 3.76910836e-01
1.21097112e+00 -3.20697010e-01 -1.49733573e-01 -4.47590113e-01
-8.30503881e-01 -1.70326754e-01 8.46920133e-01 6.41804755e-01
-6.68170080e-02 -8.18376303e-01 -6.38423502e-01 -2.25299504e-02
2.37425476e-01 -7.02167034e-01 2.20167711e-01 1.07484365e+00
-1.10342526e+00 -8.98920447e-02 -2.22931981e-01 -3.37093771e-01
-1.61852384e+00 3.89080912e-01 1.42856121e-01 -1.05340295e-01
-3.94862533e-01 5.65590978e-01 -5.77439725e-01 -4.26176518e-01
-1.33078143e-01 1.71778008e-01 -4.14292037e-01 4.56966728e-01
6.79456770e-01 7.23376751e-01 3.27018090e-02 -1.18738055e+00
-1.77706227e-01 3.55262309e-01 -4.90744591e-01 -2.03899711e-01
1.08882129e+00 -5.69396615e-01 -7.07512617e-01 1.11979175e+00
1.45714426e+00 1.08518922e+00 -4.47426081e-01 1.26866698e-01
4.27329242e-01 -6.80023849e-01 -3.05991381e-01 -3.18179220e-01
-6.44475460e-01 5.78328967e-01 3.46052408e-01 7.15333581e-01
4.76443082e-01 -4.44721617e-02 9.55517292e-01 2.22918298e-02
2.80391335e-01 -1.54400635e+00 9.90003124e-02 9.32739258e-01
6.99895680e-01 -1.19621611e+00 6.55790642e-02 -7.07436740e-01
-6.17505252e-01 1.06591451e+00 4.09076095e-01 1.12660825e-01
1.02070689e+00 5.53521216e-01 4.42062050e-01 -2.68824488e-01
-8.08869973e-02 5.56076020e-02 -2.87975907e-01 1.66453555e-01
1.23962784e+00 5.07429130e-02 -8.71971130e-01 5.18801451e-01
-3.88867110e-01 -3.06032807e-01 1.09361613e+00 1.20282090e+00
-1.94519684e-01 -1.32977533e+00 -4.01417732e-01 3.22842181e-01
-1.40689242e+00 -1.97001144e-01 -6.00626647e-01 9.04628336e-01
2.31795251e-01 1.15507925e+00 -1.60051316e-01 -3.17385882e-01
1.97885949e-02 4.65929329e-01 2.05871284e-01 -9.28410172e-01
-1.08184469e+00 -5.32774925e-02 5.48457205e-01 -3.14617902e-01
-3.07059079e-01 -1.01309645e+00 -1.04058743e+00 -8.66977096e-01
-3.50097418e-01 -5.19555807e-02 8.40516269e-01 8.30800235e-01
2.03467906e-01 -5.45406435e-03 3.75455946e-01 -2.86240011e-01
-3.37108523e-01 -1.28346217e+00 -5.74066281e-01 7.66537726e-01
8.26731622e-02 -3.55132192e-01 -4.58862871e-01 7.23531023e-02] | [8.962770462036133, 10.507708549499512] |
93928c17-8f34-4afe-a32c-e43f9aa5cf42 | s4r-self-supervised-semantic-scene | 2302.03640 | null | https://arxiv.org/abs/2302.03640v3 | https://arxiv.org/pdf/2302.03640v3.pdf | SSR-2D: Semantic 3D Scene Reconstruction from 2D Images | Most deep learning approaches to comprehensive semantic modeling of 3D indoor spaces require costly dense annotations in the 3D domain. In this work, we explore a central 3D scene modeling task, namely, semantic scene reconstruction without using any 3D annotations. The key idea of our approach is to design a trainable model that employs both incomplete 3D reconstructions and their corresponding source RGB-D images, fusing cross-domain features into volumetric embeddings to predict complete 3D geometry, color, and semantics with only 2D labeling which can be either manual or machine-generated. Our key technical innovation is to leverage differentiable rendering of color and semantics to bridge 2D observations and unknown 3D space, using the observed RGB images and 2D semantics as supervision, respectively. We additionally develop a learning pipeline and corresponding method to enable learning from imperfect predicted 2D labels, which could be additionally acquired by synthesizing in an augmented set of virtual training views complementing the original real captures, enabling more efficient self-supervision loop for semantics. In this work, we propose an end-to-end trainable solution jointly addressing geometry completion, colorization, and semantic mapping from limited RGB-D images, without relying on any 3D ground-truth information. Our method achieves state-of-the-art performance of semantic scene reconstruction on two large-scale benchmark datasets MatterPort3D and ScanNet, surpasses baselines even with costly 3D annotations. To our knowledge, our method is also the first 2D-driven method addressing completion and semantic segmentation of real-world 3D scans. | ['Matthias Nießner', 'Alexey Artemov', 'Kai Xu', 'Shuaifeng Zhi', 'Yujin Chen', 'Junwen Huang'] | 2023-02-07 | null | null | null | null | ['colorization', '3d-scene-reconstruction'] | ['computer-vision', 'computer-vision'] | [ 3.21364075e-01 5.08684814e-01 9.79443640e-02 -5.56027949e-01
-9.16726828e-01 -8.15840840e-01 4.41305995e-01 6.94166962e-03
-2.25792810e-01 2.59776622e-01 -1.40487120e-01 -3.75275701e-01
1.71387121e-01 -8.69945347e-01 -1.25484169e+00 -1.96828142e-01
2.36273423e-01 8.36560667e-01 1.79684058e-01 -3.45578380e-02
-1.06431078e-02 8.12839210e-01 -1.60916853e+00 -1.00501604e-01
7.95950830e-01 1.30567539e+00 3.92777175e-01 5.71922779e-01
-6.31494224e-01 4.27652001e-01 9.17098448e-02 -2.01954037e-01
6.53416991e-01 2.16326304e-02 -7.54061580e-01 6.23116314e-01
6.56895339e-01 -5.61553836e-01 -3.11201006e-01 7.71864593e-01
3.39841217e-01 4.14158497e-03 4.45017308e-01 -1.13561618e+00
-5.20103812e-01 -6.41932338e-02 -4.11610663e-01 -6.15422130e-01
5.08206189e-01 1.84771359e-01 6.33241534e-01 -9.83699381e-01
6.69220090e-01 1.22665095e+00 7.24863291e-01 5.22309422e-01
-1.30930507e+00 -4.05909956e-01 3.69953036e-01 -2.96701640e-01
-1.36821258e+00 -1.00377314e-01 1.20185471e+00 -4.89891320e-01
8.99581790e-01 7.31065422e-02 9.50866342e-01 1.03990078e+00
-4.87389296e-01 6.45221949e-01 1.30197620e+00 -2.51861036e-01
5.22319198e-01 5.55519946e-02 -2.93710589e-01 1.06059945e+00
-1.18808724e-01 -4.88371029e-03 -3.93251866e-01 1.06511943e-01
1.17543530e+00 3.35966468e-01 -1.57293499e-01 -1.05598378e+00
-1.13952386e+00 6.19929075e-01 6.67982638e-01 -3.42496634e-01
-3.32640976e-01 3.25039059e-01 -1.59132302e-01 1.25269797e-02
8.21996987e-01 1.86755478e-01 -9.35125172e-01 -2.11115070e-02
-6.82573080e-01 1.34172291e-01 5.53250372e-01 1.29429185e+00
1.30393231e+00 -9.63238403e-02 3.24448407e-01 5.20840943e-01
4.14210200e-01 9.26822841e-01 -5.11564314e-02 -1.37790287e+00
6.38517559e-01 9.26716685e-01 1.54309109e-01 -4.21747088e-01
-5.38423181e-01 -3.04617882e-01 -4.37992781e-01 3.28946024e-01
3.51777077e-01 3.32812190e-01 -1.41743898e+00 1.51983333e+00
7.11709797e-01 3.80041569e-01 -1.07293375e-01 9.63675261e-01
8.14264596e-01 2.26482093e-01 -2.86607653e-01 4.37767446e-01
9.54372764e-01 -8.13818395e-01 -3.30812223e-02 -2.45069742e-01
4.97906774e-01 -4.42311734e-01 1.39527917e+00 1.83147177e-01
-1.02110624e+00 -4.47720796e-01 -9.51447964e-01 -6.00709200e-01
-4.65162426e-01 -7.77842104e-02 8.16481054e-01 6.14336073e-01
-1.06637549e+00 5.24048924e-01 -1.15577030e+00 -3.86036217e-01
7.39276767e-01 1.98353976e-01 -6.57336950e-01 -5.62609732e-01
-6.47694528e-01 6.32857203e-01 1.93476275e-01 -5.95935248e-02
-1.24294245e+00 -1.09472215e+00 -1.19044518e+00 -3.52693707e-01
6.39151931e-01 -1.03473067e+00 1.07311308e+00 -3.98403555e-01
-1.40378153e+00 1.29838026e+00 -4.31195460e-02 -1.17104091e-01
7.76336730e-01 -2.79366255e-01 2.29064196e-01 2.16608480e-01
2.02447832e-01 7.27017939e-01 6.48313642e-01 -1.79156077e+00
-1.79614097e-01 -8.25895011e-01 2.88518459e-01 2.95982301e-01
1.03694066e-01 -9.48390365e-01 -7.86043048e-01 -2.44332716e-01
8.48832250e-01 -8.52253675e-01 -3.99305135e-01 5.80772460e-01
-6.29540145e-01 2.66874552e-01 6.38329983e-01 -6.28964782e-01
7.81230405e-02 -1.95824909e+00 2.69988537e-01 2.57288069e-01
2.14828759e-01 -2.21755147e-01 -2.46719271e-01 4.79464345e-02
1.50332525e-01 1.62279397e-01 -6.45744026e-01 -9.55099165e-01
2.20893711e-01 5.59426665e-01 -3.89789313e-01 5.55847347e-01
3.59563500e-01 1.08360088e+00 -1.06475198e+00 -2.92528272e-01
8.42719495e-01 7.82365024e-01 -7.25322187e-01 4.95619923e-01
-5.83232939e-01 9.17074919e-01 -6.55046821e-01 9.01194215e-01
1.07738626e+00 -3.66039634e-01 -6.41941950e-02 -3.15709144e-01
1.41783757e-03 3.40968639e-01 -1.14044762e+00 2.86565757e+00
-8.99566531e-01 8.19384828e-02 7.34310448e-02 -9.72644567e-01
9.39762414e-01 -3.28034237e-02 7.24920273e-01 -7.27252066e-01
5.09974035e-03 2.84736067e-01 -1.01144516e+00 -3.31688106e-01
2.72705793e-01 -1.26005739e-01 -2.07111731e-01 4.48149145e-01
2.88459569e-01 -1.28476989e+00 -5.66360056e-01 2.13037327e-01
1.02680147e+00 8.39394331e-01 -1.42868310e-01 3.27134460e-01
2.26208895e-01 1.81904659e-01 3.05819184e-01 5.99554479e-01
3.56947988e-01 1.01130760e+00 3.32203388e-01 -3.30462337e-01
-1.30818236e+00 -1.59007180e+00 -3.27441543e-02 4.64866370e-01
4.81257796e-01 -1.21687070e-01 -5.72272301e-01 -9.24171090e-01
3.34347874e-01 6.07225776e-01 -6.57319188e-01 9.33102816e-02
-4.69407618e-01 -2.18921244e-01 2.26833731e-01 4.84097481e-01
4.93330508e-01 -4.58951384e-01 -7.29972899e-01 9.50525142e-03
-1.36131719e-02 -1.60828459e+00 -9.36457738e-02 4.48970169e-01
-1.07997215e+00 -1.26591933e+00 -5.13898015e-01 -3.40501875e-01
7.82328486e-01 4.78562683e-01 1.17003977e+00 -1.90416537e-02
-3.73407483e-01 8.22561622e-01 -3.57064515e-01 -1.53703094e-01
-1.75194114e-01 -4.43391278e-02 -2.87857175e-01 -2.01472983e-01
-1.40307620e-01 -9.37741816e-01 -6.90487206e-01 1.05306000e-01
-8.21259379e-01 6.23311698e-01 5.43883622e-01 4.79141563e-01
1.19536996e+00 -2.78335989e-01 -1.40677288e-01 -8.82434130e-01
-3.95032108e-01 -3.10873628e-01 -7.09966600e-01 2.73173675e-02
-4.46793377e-01 2.22053483e-01 4.54555452e-01 -9.78858443e-04
-1.03892350e+00 6.11945927e-01 -3.18270415e-01 -9.59486425e-01
-5.36152124e-01 5.88234002e-03 -4.01976496e-01 -1.07710883e-01
3.53546679e-01 2.01666370e-01 -7.49271810e-02 -8.93652558e-01
8.65265369e-01 2.63178140e-01 6.06125236e-01 -7.68801808e-01
1.20189071e+00 9.96123314e-01 2.56567061e-01 -5.40020287e-01
-1.09817302e+00 -4.87886161e-01 -1.02413893e+00 1.95422824e-02
1.07060337e+00 -1.27113330e+00 -4.56218749e-01 3.94134820e-01
-1.17138934e+00 -7.54042327e-01 -5.90958476e-01 3.79136354e-01
-9.27003860e-01 3.08016747e-01 -3.12788099e-01 -6.88320220e-01
8.18732455e-02 -1.13804746e+00 1.93205810e+00 -1.50465250e-01
3.39903623e-01 -8.19174826e-01 -2.39092067e-01 6.81151032e-01
2.63458653e-03 7.30372787e-01 7.84967542e-01 -6.28209561e-02
-1.13885593e+00 2.52677985e-02 -4.56609458e-01 3.68320942e-01
1.56705216e-01 -5.42829931e-01 -1.30579662e+00 7.33411908e-02
-1.11331783e-01 -5.64933598e-01 7.21526623e-01 1.23814225e-01
1.51612997e+00 2.52961844e-01 -1.76683262e-01 1.32163823e+00
1.66090357e+00 -3.98699075e-01 2.34143391e-01 1.62712678e-01
1.23777676e+00 6.50768399e-01 5.64529359e-01 5.01586080e-01
9.57103014e-01 5.61069965e-01 1.05729723e+00 -4.49208647e-01
-3.89858007e-01 -8.57351482e-01 -2.30637386e-01 5.40394247e-01
-1.41360033e-02 -5.73973032e-03 -9.01776612e-01 3.48086119e-01
-1.62161076e+00 -2.20580727e-01 -1.67033762e-01 2.15622473e+00
6.35681689e-01 1.47917848e-02 -2.81193703e-01 -4.27053683e-03
1.82274401e-01 1.32249951e-01 -9.86110389e-01 2.62654573e-01
-5.53534813e-02 5.58801949e-01 8.67979884e-01 6.49950743e-01
-9.77237523e-01 1.06704617e+00 4.76037741e+00 3.84859383e-01
-1.01359022e+00 3.04609448e-01 5.64914107e-01 -2.01018229e-01
-9.85994220e-01 1.54270306e-01 -3.32047671e-01 8.40038806e-02
3.70091408e-01 5.47597528e-01 6.00743830e-01 9.68304276e-01
1.79547474e-01 -4.55799028e-02 -1.26244509e+00 1.26553226e+00
-7.16479961e-04 -1.38393760e+00 -1.17933238e-02 2.63712287e-01
8.50454211e-01 3.02210897e-01 -1.28243729e-01 4.75623459e-02
5.33028007e-01 -8.70986402e-01 1.14163268e+00 5.21072388e-01
1.21924770e+00 -4.03437227e-01 2.86444396e-01 4.17799145e-01
-1.21540940e+00 1.57881752e-01 -1.62475452e-01 1.07938483e-01
4.88865614e-01 6.94215894e-01 -7.10473895e-01 8.98091018e-01
7.49536574e-01 9.09522772e-01 -4.73932952e-01 6.05589747e-01
-4.41288143e-01 1.28291875e-01 -6.25243008e-01 5.20813942e-01
1.49282694e-01 -1.62397861e-01 2.46022671e-01 6.10927463e-01
4.39008951e-01 9.29461271e-02 3.75643969e-01 1.26190186e+00
-2.18398467e-01 -2.79914111e-01 -5.84933221e-01 2.11954385e-01
4.86258715e-01 1.06886888e+00 -9.12095010e-01 -2.29563311e-01
-3.93581599e-01 1.36243570e+00 3.58463377e-01 4.42179412e-01
-8.10329139e-01 7.56634474e-02 6.71030521e-01 3.59948367e-01
3.35107774e-01 -5.91864705e-01 -6.88656509e-01 -1.34329677e+00
2.66909450e-01 -6.88088592e-03 -3.10997088e-02 -9.99513745e-01
-1.21377599e+00 2.67927527e-01 3.09888460e-02 -1.09202933e+00
-2.73405276e-02 -6.49989665e-01 -8.25074911e-02 9.42809701e-01
-2.08710146e+00 -1.69049239e+00 -8.69162142e-01 8.38959336e-01
4.16842192e-01 3.54915172e-01 6.93767667e-01 2.05211967e-01
3.21584358e-03 1.23121567e-01 -3.28593045e-01 -1.68038651e-01
2.40586117e-01 -1.47644031e+00 6.57199502e-01 5.78011274e-01
1.49778038e-01 4.36844453e-02 2.93909043e-01 -4.37413037e-01
-1.75169981e+00 -1.31090164e+00 2.78909802e-01 -7.77790606e-01
2.75574505e-01 -8.58225405e-01 -6.18914366e-01 6.26313567e-01
-7.57839203e-01 4.33960825e-01 2.62424827e-01 -1.96956515e-01
-5.98470628e-01 8.71563144e-03 -1.29173899e+00 2.35791251e-01
1.81946528e+00 -8.04999948e-01 -3.31387222e-01 4.23882753e-01
1.33009899e+00 -9.61822629e-01 -8.84533346e-01 4.87744093e-01
2.78222382e-01 -9.99987185e-01 1.47322834e+00 -1.69916257e-01
4.52827901e-01 -5.56985915e-01 -6.78620696e-01 -1.06314039e+00
4.50729996e-01 -1.75931543e-01 -4.53421101e-02 9.79710281e-01
1.08287252e-01 -4.60040450e-01 1.15212393e+00 7.80771554e-01
-6.95148587e-01 -5.74576855e-01 -9.86178100e-01 -4.91521239e-01
-7.96547085e-02 -1.09829330e+00 1.00788403e+00 9.23213184e-01
-9.39787567e-01 -1.25823513e-01 -1.02649748e-01 4.17845249e-01
9.37301040e-01 4.84612435e-01 1.23025095e+00 -1.24910545e+00
-3.31515700e-01 -6.38720915e-02 -4.70069230e-01 -1.51206470e+00
3.76823306e-01 -1.04408062e+00 -2.67907437e-02 -1.80431402e+00
-2.03347683e-01 -9.22047019e-01 5.08310311e-02 6.12140357e-01
2.10474074e-01 4.52455819e-01 8.63744766e-02 -6.79922774e-02
-4.50529367e-01 1.02917194e+00 1.48037541e+00 -1.66575462e-01
-2.65176415e-01 -3.17793667e-01 -5.39870143e-01 6.29442215e-01
4.75469172e-01 -2.99722642e-01 -6.59809828e-01 -8.30026925e-01
9.71990973e-02 6.79538175e-02 8.55935812e-01 -7.20766306e-01
-1.27297953e-01 -2.43620291e-01 5.49416244e-01 -8.28472674e-01
8.88762593e-01 -1.10346496e+00 1.47151291e-01 -4.27713878e-02
2.67254170e-02 -3.70994300e-01 1.02934845e-01 6.91577494e-01
2.09531292e-01 3.20946634e-01 4.45964187e-01 -5.21956801e-01
-8.95777404e-01 7.38900423e-01 5.02171695e-01 -4.74919751e-02
8.91591132e-01 -3.26650262e-01 9.55575481e-02 -1.05197534e-01
-7.88502276e-01 1.70743898e-01 1.15391958e+00 3.75797510e-01
9.27961469e-01 -1.27609718e+00 -4.03714329e-01 5.56318879e-01
2.85524517e-01 1.01952744e+00 5.41879892e-01 4.12645906e-01
-7.01250732e-01 2.46171102e-01 -3.39213962e-04 -1.04840839e+00
-6.33175910e-01 4.27703053e-01 4.11908865e-01 7.49209598e-02
-9.47506905e-01 8.21480751e-01 3.69690418e-01 -1.27857089e+00
1.58681646e-01 -6.49722517e-01 5.91638863e-01 -4.76831138e-01
-5.31154498e-02 -4.71574580e-03 3.11967582e-01 -5.45708835e-01
-3.44795823e-01 9.97837186e-01 5.00223756e-01 -1.56143591e-01
1.50524783e+00 -3.39078456e-01 7.00369850e-02 5.15284419e-01
1.43575239e+00 -2.37546533e-01 -1.80425739e+00 -2.99313962e-01
-3.87091100e-01 -8.34326088e-01 2.46654063e-01 -7.51953244e-01
-1.18153811e+00 1.02410042e+00 5.67074180e-01 -3.57857227e-01
9.58934069e-01 5.03227770e-01 8.61506283e-01 1.73898578e-01
8.96164298e-01 -8.39905739e-01 2.05230400e-01 2.54444480e-01
6.70998275e-01 -1.54336882e+00 -7.65329599e-02 -6.76360190e-01
-3.28374118e-01 9.40780282e-01 6.06667995e-01 -8.12395364e-02
7.26510763e-01 -8.94026458e-02 6.01010537e-03 -4.56594676e-01
-2.41893619e-01 -3.75585467e-01 2.49352768e-01 9.14163709e-01
-1.46983601e-02 1.56763434e-01 6.63416862e-01 3.36572856e-01
-3.04999232e-01 -1.18234828e-01 2.07615897e-01 8.65098596e-01
-8.23741406e-02 -1.11938787e+00 -3.54623169e-01 1.41928971e-01
2.67779708e-01 2.61069424e-02 -2.07177669e-01 7.45065987e-01
3.86692166e-01 6.44596756e-01 2.40588993e-01 -4.21573400e-01
5.85807979e-01 -8.10378492e-02 7.81888008e-01 -7.90216029e-01
2.63228536e-01 -3.48617248e-02 -1.45490155e-01 -1.12970698e+00
-4.58166808e-01 -4.89539742e-01 -1.47768784e+00 -1.75700262e-01
1.18368790e-01 -3.00226122e-01 1.24320066e+00 8.64225090e-01
3.53295594e-01 4.53937680e-01 7.09019482e-01 -1.38817310e+00
-1.27396628e-01 -5.13382077e-01 -6.27648413e-01 5.64068854e-01
5.75397074e-01 -8.85243773e-01 -3.20658177e-01 2.38043275e-02] | [8.444332122802734, -2.9431982040405273] |
ac27ffa6-a4cb-43f9-9f13-b7ae764951a4 | adapting-a-framenet-semantic-parser-for | 1910.02734 | null | https://arxiv.org/abs/1910.02734v1 | https://arxiv.org/pdf/1910.02734v1.pdf | Adapting a FrameNet Semantic Parser for Spoken Language Understanding Using Adversarial Learning | This paper presents a new semantic frame parsing model, based on Berkeley FrameNet, adapted to process spoken documents in order to perform information extraction from broadcast contents. Building upon previous work that had shown the effectiveness of adversarial learning for domain generalization in the context of semantic parsing of encyclopedic written documents, we propose to extend this approach to elocutionary style generalization. The underlying question throughout this study is whether adversarial learning can be used to combine data from different sources and train models on a higher level of abstraction in order to increase their robustness to lexical and stylistic variations as well as automatic speech recognition errors. The proposed strategy is evaluated on a French corpus of encyclopedic written documents and a smaller corpus of radio podcast transcriptions, both annotated with a FrameNet paradigm. We show that adversarial learning increases all models generalization capabilities both on manual and automatic speech transcription as well as on encyclopedic data. | ['Frédéric Béchet', 'Geraldine Damnati', 'Gabriel Marzinotto'] | 2019-10-07 | null | null | null | null | ['style-generalization'] | ['computer-vision'] | [ 3.51403385e-01 5.78389585e-01 2.05713511e-01 -3.80825698e-01
-1.05525434e+00 -1.04314244e+00 9.35990930e-01 1.45508513e-01
-5.54618537e-01 1.07774711e+00 5.80946207e-01 -1.52485251e-01
-6.02709092e-02 -8.16899359e-01 -6.34168208e-01 -3.32951576e-01
2.00064868e-01 6.22595966e-01 5.24346888e-01 -8.15222502e-01
1.96939427e-02 4.95174199e-01 -1.43453312e+00 6.95529997e-01
3.59647155e-01 7.99097955e-01 2.62132555e-01 1.05385900e+00
-3.02929699e-01 1.00998652e+00 -1.22608149e+00 -4.99427885e-01
-1.76001377e-02 -3.42521399e-01 -1.33931422e+00 1.41619086e-01
3.10524583e-01 5.15618399e-02 -4.01937991e-01 9.88198757e-01
5.83983541e-01 2.74607003e-01 5.88668585e-01 -9.57086146e-01
-4.16155577e-01 1.05106366e+00 5.52619040e-01 2.81009793e-01
7.10714936e-01 -1.85111314e-01 8.36235106e-01 -4.42587763e-01
8.49155724e-01 1.62974858e+00 6.08132601e-01 8.38152528e-01
-1.08465254e+00 -5.67651749e-01 1.55106857e-01 4.64147739e-02
-1.12405288e+00 -4.07831758e-01 7.00535834e-01 -2.10772514e-01
9.50976670e-01 3.19774240e-01 1.88277557e-01 1.55704105e+00
-9.98298004e-02 7.55777836e-01 1.04124165e+00 -8.70318890e-01
5.04635572e-01 1.07385702e-01 2.47848064e-01 3.81188273e-01
-2.03005314e-01 -9.25436690e-02 -6.56863153e-01 -7.95227438e-02
1.65125504e-01 -8.96160603e-01 -3.63985568e-01 4.93576117e-02
-8.60966325e-01 9.44125891e-01 -1.15553297e-01 7.53050685e-01
6.63107727e-03 6.36021420e-02 9.24831092e-01 4.97937202e-01
6.38425291e-01 6.42602980e-01 -5.46685874e-01 -2.89658129e-01
-7.58890033e-01 6.80123806e-01 1.15610433e+00 9.88352299e-01
4.77975845e-01 3.14180076e-01 -3.10966279e-02 9.45191741e-01
8.28245878e-02 6.58122718e-01 8.69696856e-01 -7.82521546e-01
6.98505640e-01 2.08591059e-01 -5.66896610e-02 -8.15156102e-01
-5.16256690e-01 -6.99886978e-02 -4.04423475e-01 4.85351570e-02
4.61990803e-01 -3.58056128e-01 -1.06046319e+00 1.86327457e+00
8.94870460e-02 4.19999994e-02 7.49624670e-01 4.03262347e-01
1.05077362e+00 7.59803891e-01 3.36713105e-01 -6.54427633e-02
1.47452390e+00 -4.97287154e-01 -7.99415231e-01 -1.17409945e-01
6.65792286e-01 -8.34257245e-01 1.13733721e+00 5.86001992e-01
-1.10226274e+00 -6.18659377e-01 -1.05696440e+00 -9.58167315e-02
-8.35690916e-01 -2.47439072e-01 8.31484497e-02 1.11327362e+00
-8.93978238e-01 4.72066522e-01 -7.27760732e-01 -6.07312977e-01
3.12086623e-02 2.83798546e-01 -3.24116170e-01 4.44617718e-02
-1.67605543e+00 9.24322367e-01 9.22977209e-01 -4.35956448e-01
-8.14961076e-01 -4.73407030e-01 -9.09355223e-01 9.94688869e-02
4.54790294e-01 -2.82982230e-01 1.60548198e+00 -1.08720100e+00
-1.72955418e+00 7.05662727e-01 1.32523969e-01 -1.04390442e+00
5.56170583e-01 -2.30626956e-01 -8.12828839e-01 3.35554540e-01
-1.36850819e-01 4.84579742e-01 9.05314028e-01 -1.17507339e+00
-8.43613744e-01 -3.11022282e-01 5.22275090e-01 1.02768570e-01
-4.64470446e-01 1.32927150e-01 -2.34844126e-02 -1.18968964e+00
-1.15978301e-01 -1.06257915e+00 2.02826142e-01 -8.23892415e-01
-1.59758672e-01 -1.33706614e-01 1.15070641e+00 -8.52681935e-01
1.10842454e+00 -1.94622409e+00 4.57120299e-01 2.05230221e-01
-3.85729343e-01 3.74941051e-01 1.47635102e-01 4.23300177e-01
-3.73313576e-02 2.51083970e-01 -3.47060323e-01 -2.07530335e-01
1.00593962e-01 6.68008566e-01 -5.33902287e-01 1.26655564e-01
6.83386102e-02 3.50254089e-01 -1.01371098e+00 -2.44138077e-01
1.74761340e-01 3.28153670e-01 -5.86678624e-01 -2.76364777e-02
-6.07514203e-01 3.97203237e-01 -5.02662599e-01 1.81733370e-01
3.16317469e-01 5.81086278e-01 1.93281591e-01 1.41143903e-01
1.00821272e-01 4.44434196e-01 -1.21107650e+00 1.88843000e+00
-8.68147254e-01 6.09211743e-01 -2.24809349e-02 -1.13986552e+00
9.84451830e-01 6.84399307e-01 1.67684659e-01 -4.40591961e-01
2.40245312e-01 1.39256328e-01 -3.17613095e-01 -3.32672775e-01
8.59983504e-01 -3.26936841e-01 -7.18898654e-01 2.94901460e-01
6.22660697e-01 -6.81242108e-01 2.74805337e-01 9.49748009e-02
1.11928785e+00 1.46374390e-01 2.95625925e-01 -4.43036973e-01
1.11160409e+00 1.53116196e-01 2.24909872e-01 7.75465846e-01
6.30901158e-02 6.79948449e-01 5.29068410e-01 -4.51484889e-01
-9.10297394e-01 -8.23344052e-01 4.90109548e-02 1.41083753e+00
-1.87562585e-01 -5.60383141e-01 -1.26227522e+00 -9.60921168e-01
-4.32440341e-01 1.20298767e+00 -4.68047321e-01 -1.97406024e-01
-7.70727634e-01 -5.34193635e-01 1.13675416e+00 3.61709356e-01
5.59600532e-01 -1.17660570e+00 -3.60043615e-01 5.52403748e-01
-2.77645767e-01 -1.47684407e+00 1.15485862e-01 3.11834663e-01
-3.07632357e-01 -9.08145666e-01 -4.62594986e-01 -7.71152854e-01
-1.57810047e-01 -5.59839427e-01 1.19179261e+00 -2.91873306e-01
1.18437715e-01 8.96148205e-01 -1.04850507e+00 -9.06071842e-01
-1.58026338e+00 5.48101187e-01 1.46623656e-01 -3.92683707e-02
2.78719425e-01 -3.21088016e-01 3.62383813e-01 1.31573737e-01
-1.37232614e+00 -5.19563735e-01 -2.36007616e-01 9.19538736e-01
1.14960253e-01 8.39245021e-02 7.08013117e-01 -1.07176447e+00
6.80999994e-01 -2.15189472e-01 -5.37084281e-01 2.90516496e-01
1.20420987e-02 2.37272397e-01 8.61252904e-01 -1.72981724e-01
-1.45533872e+00 -3.11957393e-02 -7.09546983e-01 1.92741659e-02
-4.82111365e-01 2.16384649e-01 -4.16135609e-01 2.72207875e-02
1.04156363e+00 1.08669989e-01 -2.95118839e-01 -3.39922071e-01
4.98616725e-01 7.62631714e-01 8.45588326e-01 -7.80455887e-01
8.21956277e-01 2.36087784e-01 -1.29560798e-01 -1.34328568e+00
-5.91651976e-01 -2.11764485e-01 -7.15955794e-01 -8.64816457e-02
1.09526455e+00 -7.52154231e-01 -1.70561254e-01 5.10157764e-01
-1.40625417e+00 -1.74496308e-01 -3.23451161e-01 3.81859988e-01
-9.64981019e-01 4.12658572e-01 -3.66465330e-01 -4.93375838e-01
-7.58800283e-02 -9.40552413e-01 1.01512623e+00 -1.08895540e-01
-3.91814202e-01 -1.27883077e+00 2.12047815e-01 1.41277328e-01
1.35311708e-01 2.16981158e-01 9.83899951e-01 -1.30965960e+00
-1.54526591e-01 -2.23890096e-01 6.03146970e-01 7.44166791e-01
4.10696901e-02 -1.85775146e-01 -1.36003280e+00 -1.20689996e-01
2.31960118e-01 -5.38890600e-01 5.44548810e-01 -1.38252348e-01
7.93767214e-01 -3.42608422e-01 3.73144113e-02 3.10248196e-01
1.10713708e+00 3.19472104e-01 8.66029263e-01 6.47764325e-01
3.72129172e-01 4.66918290e-01 5.11755407e-01 2.13255659e-01
4.80924249e-02 7.34276891e-01 2.12955639e-01 3.31632614e-01
-2.27092177e-01 -2.35528588e-01 4.55180526e-01 6.70714319e-01
4.01617214e-02 -7.36030161e-01 -1.18076217e+00 4.46169168e-01
-1.43152726e+00 -9.55832183e-01 3.11216801e-01 1.83126426e+00
5.42444408e-01 4.71164733e-01 7.13869333e-02 5.08132994e-01
8.21810305e-01 2.42027491e-01 2.65209466e-01 -8.05555880e-01
-3.16799879e-01 5.70472956e-01 4.34417367e-01 6.89409792e-01
-1.32377470e+00 1.37425077e+00 6.00699091e+00 9.37661171e-01
-1.03809774e+00 3.26423734e-01 2.89089829e-01 2.14238182e-01
-2.17449486e-01 -1.74151421e-01 -8.54910374e-01 3.49206239e-01
1.27917290e+00 -1.09115541e-01 2.83877343e-01 7.58995473e-01
-1.87700570e-01 2.86514431e-01 -1.00794721e+00 5.51432908e-01
2.01746762e-01 -1.47615659e+00 2.72823244e-01 -6.21197402e-01
4.71181393e-01 -2.04278789e-02 -1.94863215e-01 5.48861325e-01
4.57819551e-01 -6.86518431e-01 9.07577097e-01 1.36131898e-01
4.88125592e-01 -8.76795530e-01 8.42165589e-01 3.05371344e-01
-8.76664758e-01 -8.66355821e-02 -1.60773814e-01 4.03785938e-03
2.04171062e-01 -2.48751000e-01 -1.16649652e+00 7.88424671e-01
6.86991215e-01 2.63089895e-01 -5.80619752e-01 2.40951985e-01
-2.52462596e-01 6.65711462e-01 -4.08222139e-01 -6.22506104e-02
5.57879984e-01 3.91947269e-01 8.65367115e-01 1.56037211e+00
2.38539353e-01 5.13779186e-02 9.37134475e-02 2.04431161e-01
8.84439573e-02 2.85994768e-01 -6.80557966e-01 7.96703435e-03
5.01759350e-01 7.56908596e-01 -9.57560301e-01 -4.64849025e-01
-4.10547078e-01 8.02321017e-01 -1.64420322e-01 3.33461642e-01
-7.14038014e-01 -3.68351281e-01 3.48924339e-01 -2.69414410e-02
3.63662004e-01 -1.48334116e-01 1.06666489e-02 -1.00249279e+00
-2.79008031e-01 -1.28484702e+00 4.86099660e-01 -7.33928859e-01
-8.46321464e-01 1.11654437e+00 3.26042742e-01 -1.00284731e+00
-8.84583712e-01 -6.86714113e-01 -2.11004615e-01 7.42462635e-01
-1.03092980e+00 -1.40592217e+00 8.64568725e-03 7.06877768e-01
9.31501865e-01 -8.43563557e-01 1.14994383e+00 3.12892604e-03
2.26809084e-02 5.35487533e-01 1.22484088e-01 2.36277953e-01
4.84685749e-01 -1.29318023e+00 4.81900871e-01 9.48335111e-01
4.35092568e-01 5.45738451e-02 1.13702989e+00 -4.10048455e-01
-7.99632907e-01 -1.09928894e+00 8.49615216e-01 -6.03859067e-01
7.79448628e-01 -6.67793453e-01 -1.01510882e+00 8.41907024e-01
2.84787565e-01 -2.12516516e-01 4.66279864e-01 -1.03738852e-01
-1.84834942e-01 7.50348419e-02 -1.31377840e+00 6.21034920e-01
8.68822038e-01 -7.57249713e-01 -1.33906078e+00 4.73737270e-01
9.69373584e-01 -6.22578859e-01 -8.38964164e-01 3.54259640e-01
1.83644757e-01 -8.00116003e-01 8.18218470e-01 -6.98040843e-01
2.54271626e-01 -4.61724102e-02 -7.04290807e-01 -1.38835025e+00
1.98096558e-01 -9.08531785e-01 2.33026177e-01 1.67811418e+00
2.05974132e-01 -4.35623318e-01 6.46401167e-01 2.84362912e-01
-4.50975597e-01 2.14175716e-01 -1.35638213e+00 -7.94078052e-01
4.08878833e-01 -8.41328859e-01 5.88554084e-01 6.61912143e-01
-1.86173066e-01 5.32522917e-01 -1.92730680e-01 4.13955599e-01
6.81296885e-02 -5.84520757e-01 8.89011621e-01 -1.12082875e+00
-3.10529381e-01 -1.38245955e-01 -8.74007642e-01 -4.41001654e-01
9.05795932e-01 -9.23232079e-01 5.20443879e-02 -8.82090330e-01
-8.66817117e-01 -1.53869316e-01 -5.86993098e-02 2.83666015e-01
2.21528351e-01 2.38471299e-01 5.22548199e-01 -2.00273097e-01
-3.05835128e-01 4.25268173e-01 5.64665675e-01 -2.74480790e-01
-1.42183453e-01 1.86146811e-01 -3.68910611e-01 1.07005334e+00
5.87460876e-01 -5.86647093e-01 -5.51500916e-01 -2.39370108e-01
-1.60978317e-01 3.91259119e-02 2.72563070e-01 -1.27773285e+00
-9.48764458e-02 2.50878274e-01 1.64282992e-01 -2.69843936e-01
2.73855120e-01 -9.00336504e-01 -1.14176214e-01 3.76291662e-01
-5.00082850e-01 -1.90212168e-02 6.30725563e-01 4.89740968e-01
-5.38272262e-01 -7.54323661e-01 7.77688980e-01 -3.83929312e-01
-1.14237189e+00 -2.59991050e-01 -8.56858134e-01 3.35521102e-01
9.20196116e-01 2.96040587e-02 -2.56576002e-01 -6.65633678e-01
-1.15554309e+00 -1.99427202e-01 3.20434242e-01 7.57514596e-01
1.40276939e-01 -1.07010055e+00 -5.52897215e-01 2.93363780e-01
1.07005022e-01 -2.16195852e-01 1.15526430e-01 -1.00808561e-01
-8.20473909e-01 5.51295877e-01 -2.41050646e-01 -5.29570460e-01
-1.31655872e+00 5.48139393e-01 2.92791158e-01 -1.97647840e-01
-4.80704099e-01 7.10658491e-01 3.61581822e-03 -5.04313648e-01
3.48687798e-01 -6.24110639e-01 -3.68725032e-01 1.58460826e-01
4.93105173e-01 2.83776343e-01 5.17553747e-01 -8.41668367e-01
-2.65370935e-01 3.40335190e-01 -4.84162495e-02 -5.54561496e-01
1.09696019e+00 -2.45277733e-01 3.07307601e-01 4.35124725e-01
1.02860856e+00 2.58083314e-01 -9.33966756e-01 -2.22033903e-01
4.70536649e-01 -4.09023836e-02 2.15827348e-03 -7.40543127e-01
-4.83659774e-01 6.14219844e-01 6.39616132e-01 6.72811627e-01
1.01267242e+00 4.37014503e-03 5.11466622e-01 8.38675678e-01
4.22710270e-01 -1.46362674e+00 -1.18113481e-01 1.05750704e+00
9.60819602e-01 -8.84912729e-01 -2.09651783e-01 -3.45932990e-01
-8.01493108e-01 1.25445974e+00 1.97924867e-01 -7.87087604e-02
5.57547629e-01 3.71032238e-01 1.95154548e-01 1.01687051e-01
-4.06663448e-01 -1.13210164e-01 4.62865969e-03 9.67250049e-01
2.27350086e-01 -2.14819610e-01 -2.59674758e-01 7.10310519e-01
-5.48149824e-01 -4.21530120e-02 8.99877071e-01 1.05128872e+00
-4.46994036e-01 -1.40569580e+00 -7.04690695e-01 -2.06810683e-01
-8.82923126e-01 -4.53513563e-02 -5.43655097e-01 1.32086730e+00
1.43588170e-01 9.18474615e-01 1.29892141e-01 -2.24430948e-01
5.18674195e-01 6.38527453e-01 4.29847032e-01 -8.76809478e-01
-7.82746017e-01 -1.48197249e-01 4.90665227e-01 -2.12709457e-01
-7.82294452e-01 -6.39332950e-01 -1.15488231e+00 1.34610444e-01
8.35338160e-02 2.61638552e-01 6.59437478e-01 1.27176785e+00
-2.02393427e-01 7.28890061e-01 3.67768914e-01 -7.92819858e-01
-4.91885751e-01 -1.12746680e+00 -4.97115105e-01 6.17096364e-01
3.06001276e-01 -6.17371440e-01 -1.86102673e-01 3.98761123e-01] | [14.369271278381348, 6.734138011932373] |
cba8b554-b8e9-415b-8742-05713b94f7cf | towards-better-graph-representation-learning | 2305.06102 | null | https://arxiv.org/abs/2305.06102v1 | https://arxiv.org/pdf/2305.06102v1.pdf | Towards Better Graph Representation Learning with Parameterized Decomposition & Filtering | Proposing an effective and flexible matrix to represent a graph is a fundamental challenge that has been explored from multiple perspectives, e.g., filtering in Graph Fourier Transforms. In this work, we develop a novel and general framework which unifies many existing GNN models from the view of parameterized decomposition and filtering, and show how it helps to enhance the flexibility of GNNs while alleviating the smoothness and amplification issues of existing models. Essentially, we show that the extensively studied spectral graph convolutions with learnable polynomial filters are constrained variants of this formulation, and releasing these constraints enables our model to express the desired decomposition and filtering simultaneously. Based on this generalized framework, we develop models that are simple in implementation but achieve significant improvements and computational efficiency on a variety of graph learning tasks. Code is available at https://github.com/qslim/PDF. | ['Bryan Hooi', 'Yanming Shen', 'Wenjie Feng', 'Mingqi Yang'] | 2023-05-10 | null | null | null | null | ['graph-regression'] | ['graphs'] | [ 1.51559591e-01 7.13466853e-02 -3.46016772e-02 9.90043879e-02
-2.09962890e-01 -7.37832785e-01 4.67322439e-01 -2.80181944e-01
3.84061821e-02 3.83409500e-01 1.62587777e-01 -3.03354949e-01
-4.72902566e-01 -8.62450659e-01 -7.07073450e-01 -6.66842520e-01
-8.32482278e-02 -2.75763217e-02 -1.00616328e-02 -2.79669583e-01
-1.22305520e-01 6.39381945e-01 -1.13161409e+00 8.22224095e-02
9.69547868e-01 6.42145514e-01 2.23908558e-01 6.21135771e-01
4.01940756e-02 5.24732769e-01 4.88716364e-02 -4.36270982e-01
5.24456859e-01 -1.73358828e-01 -7.10659742e-01 3.60681117e-01
7.22234964e-01 -2.76903193e-02 -8.57625902e-01 1.21591496e+00
1.89330444e-01 3.09620380e-01 2.63496667e-01 -1.30044353e+00
-9.92009163e-01 7.15290785e-01 -2.99983919e-01 -8.62610117e-02
1.08210690e-01 -1.05051205e-01 1.23295915e+00 -7.79114366e-01
5.02248645e-01 1.29931259e+00 9.06109571e-01 5.27215779e-01
-1.79681849e+00 -4.49553400e-01 4.85513002e-01 1.42379776e-01
-1.41978502e+00 -4.81513590e-01 8.82036567e-01 -3.90992194e-01
6.57708406e-01 4.03066516e-01 8.15123975e-01 9.47454274e-01
-3.34618725e-02 7.72483468e-01 8.99889171e-01 -4.05937880e-01
-1.88198105e-01 -2.60390520e-01 3.00314784e-01 1.06984675e+00
3.74179751e-01 -1.52601942e-01 -5.33630311e-01 -2.59882927e-01
1.13216388e+00 1.96189255e-01 -6.41076624e-01 -8.36021841e-01
-9.11402941e-01 9.42783535e-01 4.85138386e-01 2.17199326e-01
-1.18095249e-01 5.28125465e-01 -9.07463953e-03 4.62005883e-01
5.97409844e-01 2.27777183e-01 -2.29278609e-01 2.94142962e-01
-8.40452373e-01 2.30968967e-01 8.83557916e-01 8.74044061e-01
9.82267737e-01 2.43759111e-01 7.27676228e-02 5.85724771e-01
2.91293323e-01 3.47587615e-01 -3.16634268e-01 -1.00438583e+00
1.09183975e-01 5.09186387e-01 -1.64089039e-01 -1.16209114e+00
-5.80827892e-01 -9.54316854e-01 -9.93333757e-01 -1.35526180e-01
4.23334390e-01 -2.49987524e-02 -8.47703397e-01 2.06313872e+00
1.60132185e-01 5.05525053e-01 -2.37084284e-01 6.27206266e-01
6.69936121e-01 5.33392489e-01 -1.86082035e-01 -1.18352562e-01
9.82242584e-01 -1.06558561e+00 -6.71360195e-01 -1.16035849e-01
4.65724260e-01 -7.65251517e-01 1.05834544e+00 4.52626914e-01
-1.07217300e+00 -3.66637886e-01 -8.13968837e-01 -2.34793141e-01
-1.25429690e-01 2.41341069e-01 1.21680260e+00 7.38348186e-01
-1.67648709e+00 9.48389947e-01 -1.04258919e+00 -4.86893147e-01
2.73104399e-01 4.07942593e-01 -1.72208548e-01 -1.38275787e-01
-1.02566648e+00 4.84049708e-01 1.86918695e-02 1.93368971e-01
-5.29843748e-01 -8.27790201e-01 -8.44041646e-01 3.51630777e-01
6.51700377e-01 -1.22105968e+00 1.12806284e+00 -8.05012643e-01
-1.43097270e+00 4.35691386e-01 -5.81446066e-02 -6.24720931e-01
4.57572430e-01 -1.98004380e-01 -2.90646046e-01 1.75506458e-01
-2.64715940e-01 2.62795448e-01 1.03321552e+00 -8.98649931e-01
-2.17215065e-02 -1.77680224e-01 4.35745835e-01 7.53591731e-02
-5.70079744e-01 -1.78176180e-01 -6.33112371e-01 -8.02141190e-01
1.95486411e-01 -1.06500709e+00 -5.95277548e-01 1.40119419e-01
-3.20827305e-01 -2.12018117e-02 6.01070821e-01 -4.67349142e-01
1.35914254e+00 -2.06592083e+00 6.44054651e-01 4.02335197e-01
7.66701102e-01 1.87415794e-01 -3.28186542e-01 9.25142467e-01
-1.57667100e-01 1.11465476e-01 -1.61159053e-01 -4.18803811e-01
2.19710693e-01 2.55339175e-01 -3.38915914e-01 5.54481983e-01
1.77711338e-01 9.52660561e-01 -7.24349618e-01 1.06551044e-01
2.86538720e-01 8.78921866e-01 -7.11166799e-01 -2.26428598e-01
-1.79783836e-01 4.50765580e-01 -4.39952672e-01 3.77506614e-01
8.74693334e-01 -6.93071604e-01 4.33791339e-01 -5.09655356e-01
-1.56409387e-02 2.48258933e-01 -1.52528310e+00 1.69863749e+00
-2.22716108e-01 4.53117728e-01 5.55806220e-01 -1.33189809e+00
4.91037637e-01 8.67393240e-02 5.52117944e-01 -2.46540532e-01
-1.65514797e-01 3.93153466e-02 -7.99586102e-02 -3.47807519e-02
3.19095373e-01 1.58018887e-01 2.45162621e-01 3.04988027e-01
3.40819329e-01 1.15928650e-01 4.07009810e-01 4.84053701e-01
1.24908936e+00 7.46755302e-02 1.64124250e-01 -4.76392269e-01
4.31683809e-01 -2.43522584e-01 2.59357363e-01 8.54840696e-01
1.73090398e-01 2.20201716e-01 5.30120790e-01 -2.82724530e-01
-7.95190454e-01 -1.17833853e+00 3.65166143e-02 1.10341597e+00
-4.30644937e-02 -9.24479902e-01 -6.72370434e-01 -3.21011186e-01
8.20367187e-02 2.19982147e-01 -4.48437244e-01 -1.82577707e-02
-4.95350540e-01 -6.86334193e-01 4.26965028e-01 3.86158079e-01
1.87253281e-01 -3.97557914e-01 1.35232881e-01 5.84598146e-02
-7.68431649e-02 -1.05717921e+00 -6.17988527e-01 -8.25233106e-03
-9.92442548e-01 -1.00255811e+00 -4.81581509e-01 -6.05733633e-01
6.99580967e-01 8.46782744e-01 1.17557764e+00 3.39057148e-01
-1.48949564e-01 8.60279441e-01 -2.04550698e-01 6.39922693e-02
-1.47062451e-01 1.05785087e-01 8.22742507e-02 1.69063821e-01
-1.17523067e-01 -9.77876127e-01 -4.89877045e-01 -1.89442467e-02
-1.11692786e+00 3.89442414e-01 3.96724612e-01 7.88289964e-01
5.64468861e-01 5.68322167e-02 2.34061241e-01 -1.01349056e+00
7.50364542e-01 -3.20962727e-01 -8.53787363e-01 2.55709946e-01
-6.59364402e-01 2.42447466e-01 8.84596884e-01 -3.85631651e-01
-5.65387189e-01 1.17758296e-01 -4.80063222e-02 -5.35272956e-01
3.57925028e-01 7.75625050e-01 -2.30006198e-03 -8.94741952e-01
4.60275292e-01 2.69582927e-01 1.27765983e-01 -6.33235753e-01
7.72680819e-01 -1.55986547e-01 4.06633586e-01 -6.73474073e-01
1.33584404e+00 5.52472413e-01 2.62199759e-01 -9.62907970e-01
-7.02811003e-01 -4.80726063e-01 -4.38868523e-01 -1.86281398e-01
2.67315745e-01 -9.61602032e-01 -7.61759758e-01 3.69768620e-01
-8.25604022e-01 -3.86880517e-01 -3.08047310e-02 4.07517880e-01
-6.07910037e-01 7.10060894e-01 -8.31979394e-01 -5.87733686e-01
-2.73224354e-01 -9.46497321e-01 7.00450242e-01 1.47793576e-01
1.68245152e-01 -1.24902999e+00 -9.80417281e-02 -4.16394509e-02
5.62473714e-01 1.49009809e-01 8.68891478e-01 -2.11752847e-01
-1.07917833e+00 1.05101772e-01 -3.33222181e-01 3.38161558e-01
2.64334958e-03 9.59581062e-02 -6.42499089e-01 -6.63962305e-01
-7.50214560e-03 3.84282917e-02 1.24990153e+00 5.87039888e-01
1.13135219e+00 -3.46602648e-01 -3.20799679e-01 1.18916428e+00
1.73234284e+00 -4.37443346e-01 3.85845095e-01 -2.69640595e-01
9.26416576e-01 2.12800786e-01 -1.28854170e-01 3.40269893e-01
2.28874341e-01 5.71287394e-01 3.45429927e-01 -1.93936065e-01
-2.40413100e-01 -1.95086226e-01 2.86275268e-01 1.09892285e+00
-4.25232291e-01 -2.83114791e-01 -6.36351645e-01 3.43395352e-01
-2.27715921e+00 -9.96075034e-01 -2.45766267e-01 1.99244738e+00
4.59530234e-01 -1.58042341e-01 2.45453104e-01 -5.98174222e-02
7.15495408e-01 3.88841450e-01 -4.18086320e-01 -1.40220061e-01
-1.23786874e-01 6.51207328e-01 7.05866218e-01 7.27990866e-01
-1.20207477e+00 9.12046611e-01 6.72936726e+00 8.43192875e-01
-9.59906995e-01 -2.02957280e-02 -3.63355912e-02 9.68705639e-02
-5.75956941e-01 1.37700096e-01 -5.55651724e-01 -2.66154706e-02
7.21767783e-01 -3.98770601e-01 1.19098830e+00 6.85680091e-01
1.48718134e-01 5.08163869e-01 -1.02705562e+00 9.19890821e-01
-1.50523186e-01 -1.71494532e+00 1.80110410e-01 1.94624558e-01
6.10051095e-01 9.91704389e-02 3.10906768e-01 7.86464363e-02
4.75636840e-01 -1.02975273e+00 5.20674288e-01 5.78309774e-01
5.14785588e-01 -6.48738146e-01 5.25096199e-03 2.07308725e-01
-1.57967937e+00 -1.34745864e-02 -4.26168948e-01 -4.54372197e-01
1.31431386e-01 6.54455960e-01 -3.63804728e-01 9.54139709e-01
3.77982110e-01 9.44529772e-01 -4.95345980e-01 1.12155068e+00
-3.12019944e-01 6.82479203e-01 -4.70424473e-01 1.67851090e-01
2.37752587e-01 -6.42018259e-01 6.91885531e-01 1.09884310e+00
3.98714244e-01 -5.26231304e-02 4.86652285e-01 1.20777774e+00
-1.93161830e-01 9.06836912e-02 -6.04319990e-01 -2.83467680e-01
2.19493344e-01 1.47777700e+00 -6.81898057e-01 4.71321829e-02
-8.13745916e-01 8.10218751e-01 5.89870095e-01 7.90633142e-01
-9.21410322e-01 -2.25442275e-01 7.63740182e-01 2.19643246e-02
4.99084979e-01 -7.38563716e-01 6.09939061e-02 -1.59947705e+00
1.60006359e-02 -7.76812732e-01 3.32852066e-01 -3.95601958e-01
-1.32081223e+00 3.74970257e-01 -2.63154387e-01 -9.86671627e-01
1.11161634e-01 -9.13581610e-01 -4.34461772e-01 8.88662338e-01
-1.38733840e+00 -1.39382339e+00 -2.87641644e-01 8.55108440e-01
3.64052281e-02 2.99809337e-01 7.93016791e-01 4.91435498e-01
-6.49039268e-01 5.53668439e-01 2.12225020e-01 3.99804041e-02
4.16182607e-01 -1.30261338e+00 7.13795066e-01 1.28574741e+00
6.12327993e-01 1.02320361e+00 6.75614893e-01 -4.69022661e-01
-1.87009251e+00 -1.13445818e+00 4.69776243e-01 -2.17833087e-01
1.07640004e+00 -5.98447144e-01 -9.26545203e-01 9.62385118e-01
4.66756672e-02 8.59617814e-03 6.05038464e-01 3.10076714e-01
-4.79885131e-01 5.74997514e-02 -7.01747775e-01 7.86543608e-01
1.59670377e+00 -6.49313152e-01 1.82196107e-02 5.83282232e-01
8.14980567e-01 -4.19707477e-01 -8.49992037e-01 2.62032837e-01
4.11520094e-01 -8.27936530e-01 1.25050175e+00 -6.51702881e-01
-4.45618331e-02 -4.19713110e-01 -1.44305870e-01 -1.30461073e+00
-9.08531010e-01 -1.20724714e+00 -4.49261487e-01 8.50571573e-01
1.96039289e-01 -1.04523134e+00 6.85149252e-01 3.01691562e-01
-3.64121228e-01 -7.44218647e-01 -6.24118268e-01 -8.78371000e-01
-9.41587985e-02 -4.02596444e-01 3.90219867e-01 9.47099447e-01
-1.05243556e-01 2.26511702e-01 -5.15615106e-01 4.44876164e-01
7.92035162e-01 2.50866801e-01 6.95043504e-01 -1.28763473e+00
-6.69256210e-01 -6.13883436e-01 -3.79706562e-01 -1.40183723e+00
1.52351707e-01 -1.33177745e+00 -6.24902785e-01 -1.59408951e+00
7.84788802e-02 -1.40711486e-01 -2.84986287e-01 4.25907314e-01
6.47318140e-02 2.23847046e-01 4.01065916e-01 9.91014466e-02
-5.02441943e-01 3.33701968e-01 1.32005405e+00 6.05294481e-02
-3.19893956e-02 2.63234049e-01 -9.88959551e-01 6.94880128e-01
7.65562594e-01 -1.04232311e-01 -6.50520742e-01 -5.18472910e-01
4.57700551e-01 -2.39544213e-02 6.67727411e-01 -8.61657202e-01
2.93593049e-01 -1.61712930e-01 1.43616945e-01 -3.03971946e-01
3.49189609e-01 -6.15305543e-01 5.42019606e-01 5.39282203e-01
-6.81659719e-03 5.04868068e-02 1.93970501e-01 8.32489610e-01
8.86937901e-02 4.06738855e-02 6.49247587e-01 -1.75512403e-01
-5.78649521e-01 7.47191489e-01 -8.02939013e-02 -7.92717189e-02
4.99176234e-01 -5.48777096e-02 -4.55265641e-01 -6.17146611e-01
-1.02166641e+00 1.27243415e-01 6.26971364e-01 3.23308446e-02
3.95140141e-01 -1.39534950e+00 -5.71921527e-01 1.75749496e-01
-2.31612250e-01 -3.48424405e-01 3.64657849e-01 1.15338957e+00
-4.07311946e-01 4.11321133e-01 -4.25110497e-02 -4.30559993e-01
-1.06098711e+00 6.25451505e-01 2.96475381e-01 -4.02342707e-01
-7.28980124e-01 9.52642500e-01 3.70335072e-01 -3.55572492e-01
9.99422520e-02 -5.48996687e-01 1.74555108e-01 -2.90721625e-01
2.60306865e-01 3.94245625e-01 -3.08530144e-02 -4.01219845e-01
-2.37725794e-01 6.90627754e-01 8.75282586e-02 4.15172130e-01
1.40917110e+00 -1.66620165e-01 -4.19274420e-01 4.37114201e-02
9.13273096e-01 3.89026821e-01 -1.27990663e+00 -5.43597817e-01
-1.69986948e-01 -3.27165335e-01 -8.76606703e-02 -2.57350862e-01
-1.24318922e+00 6.58488095e-01 9.75060910e-02 7.24574625e-01
1.25810504e+00 1.34867216e-02 6.65251970e-01 3.96535367e-01
2.96762019e-01 -6.43415451e-01 -1.72096357e-01 6.13176882e-01
7.87414372e-01 -6.00455344e-01 8.56410787e-02 -1.08478498e+00
7.18751624e-02 1.28156769e+00 2.00512946e-01 -5.23977458e-01
7.48004556e-01 2.76053250e-01 -4.21460658e-01 -2.77622640e-01
-7.41914213e-01 -4.85331088e-01 5.20194292e-01 6.31361723e-01
4.62549865e-01 2.54649907e-01 -4.89396602e-01 5.36015570e-01
-1.80394188e-01 -1.48453653e-01 5.65735281e-01 4.27134335e-01
-2.98144817e-01 -1.33541715e+00 -6.54335395e-02 4.01659966e-01
-3.66987497e-01 -3.30200702e-01 -3.05982262e-01 7.24577606e-01
-2.80950129e-01 7.35763192e-01 -3.36467713e-01 -4.93377000e-01
3.62505428e-02 5.51753631e-03 7.97419965e-01 -6.92133904e-01
-2.11130366e-01 2.84732938e-01 1.56726182e-01 -6.59376383e-01
-5.54791391e-01 -4.80932683e-01 -9.35152590e-01 -5.65957844e-01
-3.79635304e-01 -5.05332612e-02 2.53188133e-01 6.55976236e-01
5.31139791e-01 5.66707432e-01 3.83609593e-01 -7.86387503e-01
-6.94766223e-01 -5.46048701e-01 -6.59091651e-01 2.10131660e-01
4.14772511e-01 -7.63760567e-01 -4.30424869e-01 -7.99696371e-02] | [6.858516216278076, 6.023688316345215] |
17c33857-efd8-4fd6-a7fd-97efb23ff6b0 | cross-domain-generalization-for-amr-parsing | 2210.12445 | null | https://arxiv.org/abs/2210.12445v1 | https://arxiv.org/pdf/2210.12445v1.pdf | Cross-domain Generalization for AMR Parsing | Abstract Meaning Representation (AMR) parsing aims to predict an AMR graph from textual input. Recently, there has been notable growth in AMR parsing performance. However, most existing work focuses on improving the performance in the specific domain, ignoring the potential domain dependence of AMR parsing systems. To address this, we extensively evaluate five representative AMR parsers on five domains and analyze challenges to cross-domain AMR parsing. We observe that challenges to cross-domain AMR parsing mainly arise from the distribution shift of words and AMR concepts. Based on our observation, we investigate two approaches to reduce the domain distribution divergence of text and AMR features, respectively. Experimental results on two out-of-domain test sets show the superiority of our method. | ['Yue Zhang', 'Linfeng Song', 'Leyang Cui', 'Seng Yang', 'Xuefeng Bai'] | 2022-10-22 | null | null | null | null | ['amr-parsing'] | ['natural-language-processing'] | [ 4.33003694e-01 2.60530502e-01 -2.54155338e-01 -5.47971308e-01
-1.07666790e+00 -8.82729709e-01 4.58600014e-01 2.37621725e-01
-1.00398865e-02 4.93947208e-01 4.34576035e-01 -5.85790277e-01
1.06041178e-01 -8.43103468e-01 -4.69746619e-01 -4.69482914e-02
4.10290301e-01 5.77225208e-01 1.63239792e-01 -4.64322805e-01
2.32144773e-01 3.01709414e-01 -8.14278126e-01 6.98666930e-01
8.13810885e-01 4.96044070e-01 1.39325157e-01 6.97190046e-01
-9.33289766e-01 9.64551628e-01 -1.03807211e+00 -8.09808969e-01
1.35021940e-01 -5.01985431e-01 -1.20248508e+00 -9.91282240e-02
2.92632818e-01 1.71746105e-01 -4.41418678e-01 1.19120085e+00
2.05912128e-01 8.14508006e-04 6.04727030e-01 -9.18010056e-01
-1.03156579e+00 1.29071701e+00 -7.91404843e-01 4.05099571e-01
6.20017767e-01 -2.94223726e-01 1.41833651e+00 -5.34495115e-01
7.94844091e-01 1.61741197e+00 4.60572392e-01 9.25660789e-01
-1.23374915e+00 -6.62621558e-01 5.45190811e-01 -3.59916180e-01
-1.09601831e+00 -2.24147085e-02 1.05459058e+00 -1.02834530e-01
1.27331185e+00 -5.72820380e-02 8.29087272e-02 1.38475144e+00
1.60232529e-01 9.67590451e-01 1.03970790e+00 -7.06000745e-01
2.44171798e-01 -1.60383105e-01 7.84522057e-01 3.83998990e-01
4.00502503e-01 -5.67786813e-01 -4.38531309e-01 -3.23103189e-01
5.82701445e-01 -4.60518539e-01 1.60881519e-01 6.68214634e-02
-7.03081310e-01 1.16102111e+00 -2.29241818e-01 3.20421278e-01
-1.00026708e-02 -8.01647678e-02 5.35552382e-01 5.32750726e-01
8.30795944e-01 3.83262664e-01 -9.02537107e-01 -2.16715425e-01
-2.83717960e-01 1.92278132e-01 1.05151248e+00 1.19811571e+00
5.44388413e-01 -6.40840083e-02 5.43172248e-02 1.25214016e+00
2.57788152e-01 5.13790190e-01 6.67268574e-01 -5.57103395e-01
1.21559465e+00 7.89339781e-01 -3.88279557e-01 -8.10209215e-01
-4.30190533e-01 -1.11798503e-01 -5.47733009e-01 -3.97495389e-01
5.99197328e-01 -2.41100580e-01 -9.70733523e-01 1.77812541e+00
-6.85334206e-04 -2.99181432e-01 6.57507658e-01 4.67626929e-01
1.02306163e+00 7.11544275e-01 4.99381363e-01 1.66778758e-01
1.63186145e+00 -7.24701166e-01 -6.81043446e-01 -9.35132086e-01
1.13137186e+00 -7.59838045e-01 1.15787756e+00 7.94549286e-02
-8.36157978e-01 -4.89488780e-01 -7.96025932e-01 -2.13973001e-01
-4.31942612e-01 3.35446075e-02 6.20125175e-01 8.38773847e-01
-6.00026667e-01 3.69632483e-01 -6.70514882e-01 -4.28731292e-01
2.40331516e-01 1.25438735e-01 -4.41579044e-01 -5.12571871e-01
-1.49509692e+00 7.11714506e-01 5.97238898e-01 -2.69437343e-01
-1.56293854e-01 -7.20213532e-01 -1.07522619e+00 -1.64120927e-01
1.47649154e-01 -6.08718395e-01 1.40717053e+00 -1.09329736e+00
-1.43023789e+00 1.10094333e+00 -2.42930964e-01 -5.46960056e-01
2.54615963e-01 -6.07608557e-01 -7.71726251e-01 -1.06801189e-01
2.13664860e-01 4.84020263e-01 5.70366740e-01 -1.05102730e+00
-6.75767601e-01 -5.59194207e-01 2.27529295e-02 2.21551239e-01
-7.76828080e-02 3.01606953e-01 -3.30999970e-01 -7.80571759e-01
2.28284925e-01 -7.68282652e-01 -3.10352147e-01 -1.18995023e+00
-2.93782264e-01 -5.67453325e-01 5.39396107e-01 -8.19310784e-01
1.29359186e+00 -2.30677700e+00 -3.15738320e-02 -1.49603233e-01
9.75928679e-02 4.13032025e-02 -5.50224245e-01 4.10153449e-01
-4.28415924e-01 4.73279029e-01 -3.00107181e-01 -2.40574956e-01
-1.47033885e-01 5.55969238e-01 -8.05813611e-01 7.46377036e-02
6.58853650e-01 9.31324899e-01 -7.38715768e-01 -4.43178952e-01
-8.72592181e-02 1.31221280e-01 -3.62364918e-01 3.07648659e-01
-5.29410779e-01 4.50035632e-01 -7.65680552e-01 6.88559651e-01
7.88017809e-01 -2.02173427e-01 6.46399796e-01 3.16303760e-01
3.81712317e-01 8.50408792e-01 -8.24711621e-01 1.86851668e+00
-5.13601661e-01 4.67524797e-01 -2.98790127e-01 -1.11240304e+00
1.34329617e+00 -5.65269589e-03 1.86236814e-01 -8.03973615e-01
-4.98097427e-02 1.83275044e-01 1.77630365e-01 1.83141753e-02
6.85557723e-01 -2.11823434e-01 -6.88946903e-01 2.42189571e-01
5.91326877e-02 -6.07811175e-02 6.05658740e-02 1.91620126e-01
1.53349018e+00 1.74719259e-01 6.99148774e-01 -2.20248803e-01
3.66993636e-01 4.42918569e-01 4.98055637e-01 8.00010026e-01
-2.22823899e-02 6.04370594e-01 7.80938566e-01 -4.65005219e-01
-6.64781868e-01 -1.17461777e+00 -1.56740785e-01 1.33871675e+00
1.20382883e-01 -5.77799559e-01 -6.88325763e-01 -1.31473935e+00
-1.31317601e-01 1.14722431e+00 -5.41848004e-01 7.45019317e-02
-1.14494491e+00 -9.50650096e-01 7.47341633e-01 8.52837145e-01
2.64418095e-01 -1.15525913e+00 -1.72161072e-01 4.00022656e-01
-3.73742551e-01 -1.67788422e+00 1.67209476e-01 4.00529116e-01
-1.10884857e+00 -7.35614955e-01 -3.97888809e-01 -8.57762575e-01
4.01842862e-01 2.29205891e-01 1.83838212e+00 -2.62262940e-01
-1.16424769e-01 4.71426219e-01 -8.28871548e-01 -3.66339684e-01
-1.03524876e+00 5.88117421e-01 -5.26587248e-01 -6.46684051e-01
1.09394705e+00 -3.39070499e-01 -1.41235456e-01 1.55805573e-01
-8.01249862e-01 -6.93889782e-02 5.89470685e-01 4.77295727e-01
5.31006932e-01 1.91789381e-02 7.58807003e-01 -1.55585659e+00
8.88914049e-01 -7.11557627e-01 -4.64687228e-01 2.87889212e-01
-2.53498256e-01 2.74741679e-01 7.98217893e-01 -3.56607646e-01
-1.38806701e+00 -1.35475183e-02 -3.38185489e-01 1.68140784e-01
-4.45010424e-01 3.22539300e-01 -4.32443798e-01 6.50130451e-01
5.16140640e-01 -1.40822418e-02 -5.05773187e-01 -5.64224720e-01
6.81532919e-01 6.53897345e-01 2.60440201e-01 -1.02681756e+00
6.33722305e-01 1.10178232e-01 -2.19803631e-01 -9.65513349e-01
-1.14893198e+00 -5.52733183e-01 -4.89218324e-01 4.73985910e-01
1.13806033e+00 -1.11803842e+00 7.09231272e-02 -2.51108687e-03
-1.46452439e+00 -2.43648037e-01 -3.19533676e-01 2.16068640e-01
-3.81653488e-01 4.37273800e-01 -7.75801897e-01 -6.75078154e-01
-3.74467373e-01 -8.04423034e-01 1.19405413e+00 7.76754618e-02
-6.38431966e-01 -1.22515261e+00 2.74204493e-01 3.28706414e-01
6.81005418e-03 1.61120470e-03 1.44383073e+00 -1.38309216e+00
-2.25002512e-01 -5.85199632e-02 -3.26961577e-01 2.01926455e-01
2.66144156e-01 -4.21558470e-01 -9.90301609e-01 5.37495837e-02
-8.93703029e-02 -1.93789095e-01 8.64957333e-01 2.94848531e-01
1.04099751e+00 1.34776577e-01 -3.63894731e-01 2.19420493e-01
1.32823205e+00 2.39685282e-01 6.88049495e-01 4.30632859e-01
8.06029379e-01 7.94277012e-01 9.11100507e-01 1.45159557e-01
4.34846669e-01 3.09730798e-01 1.67690128e-01 -3.61692645e-02
-2.88586617e-01 -5.57563484e-01 5.51424384e-01 9.20507371e-01
2.93102950e-01 -6.11809909e-01 -1.00366318e+00 4.72249538e-01
-1.67853022e+00 -3.67877185e-01 -2.47272000e-01 1.82981086e+00
6.71754837e-01 3.12599927e-01 -3.20566744e-02 -3.67232472e-01
8.19068313e-01 3.54855210e-01 -3.79819423e-01 -8.95707488e-01
-2.65694976e-01 5.62756181e-01 4.56842363e-01 1.67225942e-01
-1.09663498e+00 1.55622828e+00 6.91240120e+00 6.66952908e-01
-5.31446934e-01 1.01751434e-02 6.16301715e-01 5.99764228e-01
-5.38480878e-01 1.09684438e-01 -1.33666813e+00 9.56235006e-02
1.10897124e+00 1.49969524e-02 1.13628790e-01 1.27990627e+00
-4.34624761e-01 3.56645644e-01 -1.23315275e+00 7.91182578e-01
-8.04989412e-02 -8.59844148e-01 3.47025782e-01 -1.22198366e-01
3.05991411e-01 1.42223388e-01 -1.90511018e-01 6.77460313e-01
9.53511536e-01 -1.05744851e+00 3.21527332e-01 -3.79699349e-01
5.65286994e-01 -8.80374074e-01 7.45056808e-01 1.30398363e-01
-1.44419909e+00 1.06463701e-01 -7.61127532e-01 -1.76131517e-01
-1.96900256e-02 4.13449168e-01 -7.34300077e-01 7.80643463e-01
5.43197036e-01 7.48996675e-01 -8.27590704e-01 -9.02776122e-02
-5.11200070e-01 8.49796891e-01 7.74281248e-02 8.62882007e-04
7.51688778e-02 -2.85067081e-01 5.14100492e-01 1.69667876e+00
5.18584885e-02 7.96623826e-02 2.24567816e-01 7.62326121e-01
-2.80371159e-01 3.04055452e-01 -8.77572477e-01 -3.22154611e-01
3.89887512e-01 1.08236551e+00 -9.12114978e-01 -2.40854055e-01
-9.58647668e-01 1.07839513e+00 7.01935589e-01 2.19638869e-01
-7.66261756e-01 -1.51666477e-01 1.01154196e+00 -4.52123098e-02
2.84602195e-01 -2.83084542e-01 -5.45153975e-01 -1.28625679e+00
1.04561627e-01 -1.05611897e+00 9.81300175e-01 -3.90732318e-01
-1.89704132e+00 6.91515744e-01 1.65250540e-01 -8.80931377e-01
-2.37684056e-01 -9.79668915e-01 -3.15010667e-01 9.36947703e-01
-1.60462630e+00 -1.08588445e+00 2.39856988e-01 2.55024761e-01
1.08874595e+00 -3.20630729e-01 1.02314496e+00 -7.99656659e-02
-4.94928300e-01 5.63491285e-01 -2.60689527e-01 5.84640324e-01
5.97266078e-01 -1.51993203e+00 1.34087873e+00 1.03354442e+00
3.89140904e-01 7.80009568e-01 5.69991946e-01 -7.86999047e-01
-1.37633574e+00 -1.14714837e+00 6.92097187e-01 -9.39385116e-01
9.46866989e-01 -5.94491422e-01 -1.16490698e+00 1.08702540e+00
1.81688547e-01 -1.67831615e-01 9.59708393e-01 5.57454109e-01
-7.87791908e-01 3.52283001e-01 -9.32791889e-01 6.23260915e-01
1.34980726e+00 -4.19390917e-01 -1.23352015e+00 8.69021565e-02
1.19905603e+00 -4.06179994e-01 -9.97809708e-01 4.32191849e-01
1.75898328e-01 -4.83606488e-01 8.65327597e-01 -1.12056208e+00
8.05527389e-01 2.55807699e-03 -4.86888111e-01 -1.30519152e+00
-2.97475815e-01 -3.63089502e-01 6.31229132e-02 1.65993261e+00
6.15647435e-01 -7.64617145e-01 7.83901393e-01 6.81387603e-01
3.18640247e-02 -5.30486517e-02 -6.84857011e-01 -8.47780108e-01
6.61413252e-01 -7.28761256e-01 3.64100814e-01 9.48353589e-01
-7.13859349e-02 1.09422600e+00 -3.17130759e-02 4.01966870e-01
3.92496258e-01 4.14637804e-01 7.39762068e-01 -1.29661286e+00
-5.03645897e-01 -8.94407481e-02 -2.56797373e-01 -1.28858244e+00
6.75661385e-01 -1.01306510e+00 -1.03415884e-01 -1.64807153e+00
2.43675634e-01 -2.64114767e-01 -1.68330491e-01 4.47993815e-01
-3.60375375e-01 -2.75372088e-01 6.52285144e-02 -3.09278332e-02
-6.03358924e-01 6.76036701e-02 8.78926694e-01 3.65672074e-02
-2.20817745e-01 -1.67037487e-01 -9.84743714e-01 9.02395010e-01
1.12312853e+00 -8.26933920e-01 -5.16902030e-01 -6.92203403e-01
4.23261315e-01 -8.66743503e-04 -2.99718857e-01 -6.14135265e-01
-4.34525132e-01 -1.24225445e-01 4.17303711e-01 -6.16812050e-01
-1.26440421e-01 -6.62032127e-01 -3.50489140e-01 1.74987674e-01
-4.40661550e-01 4.36689377e-01 4.08378303e-01 7.00269699e-01
-3.93336713e-01 -4.27443653e-01 7.56895363e-01 -3.03609371e-01
-9.90539551e-01 -1.47163048e-01 -5.20280004e-01 8.81104529e-01
5.43492436e-01 2.20946036e-02 -6.37946844e-01 -3.16643901e-02
-4.16124374e-01 -5.07711284e-02 1.85576454e-01 8.80925596e-01
4.57915276e-01 -9.03120637e-01 -6.84860945e-01 4.80105318e-02
4.50211674e-01 1.88586682e-01 -1.66135672e-02 -1.33402953e-02
-3.31084847e-01 3.06710005e-01 9.26196799e-02 -4.11405563e-01
-1.39886785e+00 6.45224452e-01 -4.62641269e-02 -9.18423533e-01
-9.33519661e-01 5.90384007e-01 6.39858246e-01 -7.53288388e-01
-1.60982788e-01 -4.22343642e-01 -3.61903369e-01 -2.69692034e-01
5.27043879e-01 2.05489844e-01 6.26184931e-03 -3.89190346e-01
-5.19374669e-01 5.36606014e-01 -5.05511940e-01 4.68758028e-03
1.24966884e+00 -1.61451876e-01 4.81230626e-03 3.63035977e-01
1.06208217e+00 2.61579812e-01 -7.19687879e-01 -2.36181214e-01
6.20185018e-01 -2.28250712e-01 -4.60735291e-01 -7.38284349e-01
-7.19415128e-01 8.42091084e-01 2.44902015e-01 5.60733020e-01
1.09825444e+00 4.48108137e-01 9.50593233e-01 2.70148426e-01
2.50160426e-01 -1.08345568e+00 1.32641375e-01 9.43354011e-01
5.28406918e-01 -1.08524704e+00 -3.21377665e-01 -8.86876345e-01
-1.01287949e+00 1.30153537e+00 7.31772661e-01 -2.05365285e-01
4.45513546e-01 4.38145757e-01 2.45883971e-01 -1.39969289e-01
-5.46381414e-01 -2.11199015e-01 1.88332680e-03 8.35304618e-01
8.96316409e-01 1.60481349e-01 -4.44440067e-01 1.03052115e+00
-3.59886765e-01 -4.33936864e-01 4.75506365e-01 9.15930450e-01
-2.88969487e-01 -1.61625195e+00 -8.42376426e-02 1.34781450e-01
-9.96174157e-01 -3.91551197e-01 -8.47551227e-01 1.14457273e+00
-5.35659730e-01 1.06403100e+00 1.24880761e-01 -1.39458194e-01
5.27291179e-01 3.87683481e-01 5.14080763e-01 -9.96008992e-01
-4.25625980e-01 -1.42716095e-01 6.19875848e-01 -3.52731705e-01
-1.70685157e-01 -3.82475883e-01 -1.53040683e+00 4.09720317e-02
1.35145877e-02 6.97613358e-02 5.33745944e-01 9.12593782e-01
3.64879847e-01 6.27743125e-01 2.34499037e-01 2.04595357e-01
-6.12628698e-01 -9.73255038e-01 -5.95984995e-01 6.86992824e-01
-3.06627125e-01 -2.30100438e-01 -2.83731043e-01 -9.11733955e-02] | [10.457590103149414, 9.378049850463867] |
57c52e0d-a49d-4be8-9d29-aeaa4fab94db | evaluation-of-unsupervised-information | null | null | https://aclanthology.org/L12-1313 | https://aclanthology.org/L12-1313.pdf | Evaluation of Unsupervised Information Extraction | Unsupervised methods gain more and more attention nowadays in information extraction area, which allows to design more open extraction systems. In the domain of unsupervised information extraction, clustering methods are of particular importance. However, evaluating the results of clustering remains difficult at a large scale, especially in the absence of reliable reference. On the basis of our experiments on unsupervised relation extraction, we first discuss in this article how to evaluate clustering quality without a reference by relying on internal measures. Then we propose a method, supported by a dedicated annotation tool, for building a set of reference clusters of relations from a corpus. Moreover, we apply it to our experimental framework and illustrate in this way how to build a significant reference for unsupervised relation extraction, more precisely made of 80 clusters gathering more than 4,000 relation instances, in a short time. Finally, we present how such reference is exploited for the evaluation of clustering with external measures and analyze the results of the application of these measures to the clusters of relations produced by our unsupervised relation extraction system. | ['Romaric Besan{\\c{c}}on', 'Olivier Ferret', 'Brigitte Grau', 'Wei Wang'] | 2012-05-01 | null | null | null | lrec-2012-5 | ['open-information-extraction'] | ['natural-language-processing'] | [ 8.18622932e-02 7.04239428e-01 1.60122126e-01 -2.10028589e-01
-5.13512015e-01 -4.74392533e-01 9.27351236e-01 1.10803616e+00
-6.78590894e-01 8.51138115e-01 4.31802198e-02 -1.88005920e-02
-6.87924743e-01 -1.05468524e+00 -1.65684611e-01 -4.99946713e-01
-7.54793137e-02 9.85150099e-01 3.51337224e-01 -8.45881850e-02
-7.68541591e-03 7.55259991e-01 -1.64917243e+00 6.19119667e-02
8.87022853e-01 5.81357658e-01 9.20787528e-02 8.22932497e-02
-3.01074266e-01 6.21995270e-01 -8.44028473e-01 -5.26421010e-01
-2.00651661e-01 -5.65224886e-01 -1.26303792e+00 2.69606829e-01
-4.40320015e-01 4.22668338e-01 3.76439780e-01 8.66115451e-01
7.75380507e-02 1.88187018e-01 1.03720403e+00 -8.20213854e-01
3.27265710e-01 1.12031698e+00 4.00650054e-02 -1.54304236e-01
3.83861661e-01 -4.05188411e-01 9.95510995e-01 -4.84287441e-01
1.25835037e+00 6.75607741e-01 1.08362779e-01 5.04134521e-02
-1.36745930e+00 -1.00680850e-01 -1.82388261e-01 3.36437047e-01
-1.40268040e+00 -4.53797877e-01 6.25846386e-01 -7.15915918e-01
7.05139339e-01 3.32207471e-01 5.57108760e-01 6.55333042e-01
-5.85450232e-01 3.02472293e-01 1.09249151e+00 -7.79870987e-01
3.35088879e-01 5.73872447e-01 5.63870013e-01 1.91869467e-01
6.57495856e-01 -4.09557164e-01 -6.39812127e-02 3.12921889e-02
1.62367865e-01 -5.08098960e-01 -2.35940367e-01 -4.81078207e-01
-8.49916339e-01 5.30375183e-01 2.65268773e-01 1.46130061e+00
-3.02918166e-01 -6.22838199e-01 3.75679284e-01 -3.27413529e-02
5.66097200e-01 6.90995097e-01 -3.51390481e-01 9.20769293e-03
-8.84453833e-01 -3.31564173e-02 1.20135117e+00 9.80048358e-01
8.46936405e-01 -5.64388156e-01 1.36866361e-01 5.38463473e-01
-3.86427082e-02 8.95428285e-02 2.73659289e-01 -3.42649937e-01
4.12003160e-01 1.22097564e+00 -1.38020277e-01 -1.12978935e+00
-6.76193058e-01 -4.89716440e-01 -8.95942926e-01 1.08852863e-01
4.27468300e-01 5.32422401e-03 -1.77198425e-02 1.32914841e+00
4.08570468e-01 -2.61909783e-01 3.86193603e-01 3.88377488e-01
7.22797096e-01 -8.66490602e-03 5.00996113e-02 -9.25887108e-01
1.29918408e+00 -4.18744594e-01 -9.61076736e-01 6.96509659e-01
1.06508875e+00 -6.56617045e-01 4.27798331e-01 5.42378485e-01
-6.81777716e-01 -3.68773162e-01 -8.50283146e-01 2.41770536e-01
-7.34353364e-01 4.92392540e-01 4.29423302e-01 3.91754240e-01
-5.99046886e-01 9.51481104e-01 -6.98099971e-01 -7.20658183e-01
8.23905021e-02 4.49928284e-01 -6.76367700e-01 5.25258422e-01
-8.96499157e-01 1.19940519e+00 8.41482460e-01 1.57809258e-01
1.19477369e-01 4.95143235e-02 -5.50239384e-01 1.99235678e-01
7.70518124e-01 -2.04733342e-01 7.04466939e-01 -7.81811476e-01
-1.18409145e+00 1.05265212e+00 2.38747131e-02 -6.38828456e-01
4.24789280e-01 -5.87432506e-03 -6.27949834e-01 4.03776079e-01
1.16698846e-01 3.67871001e-02 1.93969652e-01 -1.39253509e+00
-5.23737192e-01 -4.63594139e-01 -1.53985918e-01 -1.90828338e-01
-5.08260131e-01 6.00769073e-02 -4.27067131e-01 -1.10848740e-01
1.78217946e-03 -5.92164516e-01 -2.26899624e-01 -1.04170072e+00
-7.24326730e-01 -5.98014116e-01 4.22791243e-01 -4.74044949e-01
1.51463866e+00 -1.89496017e+00 5.58741748e-01 7.49459207e-01
5.45182824e-01 3.85286897e-01 4.53824639e-01 6.23694837e-01
-2.35668257e-01 2.20707104e-01 -5.04811883e-01 -2.80180126e-01
3.12440470e-02 2.10183397e-01 1.41728401e-01 4.02378216e-02
2.09457397e-01 2.86577940e-01 -8.96925151e-01 -1.14503920e+00
3.31010967e-01 1.81588039e-01 -2.28024244e-01 2.37011090e-01
-1.51084378e-01 4.32057947e-01 -6.48991227e-01 7.65688419e-02
2.73306310e-01 5.94762713e-02 8.13244879e-01 -3.48493665e-01
-3.79694253e-01 7.26196766e-02 -1.16393936e+00 1.29619372e+00
-2.25297391e-01 3.40919644e-01 -2.93331027e-01 -1.23196197e+00
1.37247467e+00 5.96086025e-01 7.83138156e-01 -2.43891925e-01
5.35556197e-01 4.24283117e-01 1.38482705e-01 -4.63782996e-01
5.04408658e-01 2.89913774e-01 1.62724137e-01 3.42066258e-01
4.06509310e-01 -6.84523433e-02 1.15932727e+00 1.62193775e-01
1.09003770e+00 1.50747970e-01 6.07256293e-01 -2.63084024e-01
9.51071143e-01 1.97270159e-02 9.13886055e-02 1.03213795e-01
3.67212981e-01 1.77791014e-01 9.52579975e-01 -3.07092339e-01
-9.15211499e-01 -3.89942497e-01 -4.18144077e-01 3.03391814e-01
-2.59808332e-01 -8.44221115e-01 -9.15260673e-01 -9.59988296e-01
-3.72684866e-01 3.40012789e-01 -4.61922020e-01 2.56752640e-01
-2.70393848e-01 -8.60901237e-01 2.24740312e-01 -2.36821279e-01
3.75732869e-01 -1.29618132e+00 -5.55518568e-01 2.10790858e-01
-3.07573289e-01 -1.20847249e+00 7.43497491e-01 3.70046258e-01
-7.92924106e-01 -1.44508553e+00 -2.55800694e-01 -4.92358059e-01
4.60425764e-01 -4.85316426e-01 1.36658061e+00 4.65385437e-01
-6.49757087e-02 -6.35686610e-03 -9.09334838e-01 -1.68095693e-01
-8.63398671e-01 8.44706953e-01 -6.66727424e-02 -1.78472474e-02
5.46973109e-01 -7.33368456e-01 4.14223932e-02 2.44319290e-01
-1.03572702e+00 -2.45757684e-01 5.03325164e-01 1.17038898e-01
2.56309479e-01 2.11850092e-01 4.68101203e-01 -1.31188142e+00
6.45499229e-01 -2.56499797e-01 -6.28382444e-01 3.33159655e-01
-7.01274693e-01 3.29247415e-01 5.11942685e-01 1.08174175e-01
-8.56263399e-01 1.76793098e-01 -3.10291171e-01 3.01414400e-01
-6.36780739e-01 5.30919135e-01 -3.69739562e-01 2.19510049e-01
9.78421569e-01 -3.71000022e-01 -2.11421877e-01 -7.57617056e-01
4.62143242e-01 6.97336435e-01 1.88379183e-01 -6.51545465e-01
7.73642421e-01 6.90274611e-02 3.04476261e-01 -9.44135368e-01
-4.41425353e-01 -6.37779355e-01 -1.20859122e+00 -1.78176641e-01
8.45578969e-01 -2.07431883e-01 -7.58855641e-01 -1.08382672e-01
-1.18835008e+00 -2.30856910e-01 -5.65433145e-01 6.15215302e-01
-4.29139286e-01 4.68460619e-01 -2.77476043e-01 -8.38590562e-01
-1.04638107e-01 -9.42349732e-01 5.64143121e-01 -2.30713505e-02
-5.30591249e-01 -9.22058702e-01 3.34482282e-01 9.11050588e-02
-2.67255008e-01 3.52011740e-01 8.46635699e-01 -1.02755630e+00
-4.30263042e-01 -2.12110922e-01 1.68482424e-03 1.53000712e-01
3.21754873e-01 2.36309767e-01 -6.93700790e-01 2.93601722e-01
-1.81830198e-01 2.82560229e-01 7.45632052e-01 -3.50676447e-01
6.28399730e-01 3.52482274e-02 -5.71233213e-01 1.31040439e-01
1.42336392e+00 1.39046922e-01 7.09157944e-01 4.81351167e-01
4.89536047e-01 1.02942491e+00 8.83706808e-01 1.72671691e-01
5.67738991e-03 1.07208657e+00 7.46927857e-02 -3.54870819e-02
7.30330274e-02 3.23268503e-01 -2.53487825e-01 1.01379097e+00
-7.58117378e-01 -1.94685049e-02 -9.47169065e-01 5.14171362e-01
-1.75501025e+00 -7.82262564e-01 -7.14775801e-01 2.25761890e+00
9.40471172e-01 4.18983608e-01 4.09217119e-01 8.47751081e-01
6.03606820e-01 -2.66717613e-01 4.57107633e-01 -3.25353652e-01
-6.31413087e-02 4.78984654e-01 -3.76379793e-03 6.05577469e-01
-9.99456227e-01 8.18978071e-01 4.66224337e+00 7.94012487e-01
-5.20013154e-01 -7.20528066e-02 3.00272197e-01 5.03596663e-01
-5.46670221e-02 1.80495083e-01 -7.08832145e-01 2.82164186e-01
9.28494334e-01 1.71873227e-01 -1.09246835e-01 6.27759457e-01
4.06260826e-02 -4.97286201e-01 -1.01041114e+00 5.99626482e-01
-1.21738583e-01 -1.03377903e+00 -8.26980546e-02 2.25078270e-01
4.53115165e-01 -4.67794687e-01 -6.89091921e-01 -4.73133102e-02
4.36917841e-02 -6.66019678e-01 3.12353849e-01 6.69776678e-01
4.44571555e-01 -7.38303423e-01 1.06972599e+00 2.85339892e-01
-1.15459895e+00 3.30425918e-01 -1.24652781e-01 4.90263663e-02
1.38565153e-01 9.99180257e-01 -9.34824824e-01 1.18801749e+00
1.65415764e-01 4.50147569e-01 -6.75580382e-01 1.01083481e+00
-4.26758796e-01 4.36953545e-01 -3.50114822e-01 -3.80355686e-01
-1.48297846e-01 -4.54653889e-01 4.14312959e-01 1.41368783e+00
1.42326936e-01 5.27119748e-02 -2.45095223e-01 7.38396168e-01
1.00212999e-01 9.03527796e-01 -6.27726316e-01 -2.10518762e-02
3.10495973e-01 1.59500539e+00 -1.29303980e+00 -5.25323629e-01
1.33607566e-01 3.90768528e-01 4.58601654e-01 8.83292332e-02
-2.86286175e-01 -6.44005537e-01 2.36292426e-02 7.22971633e-02
3.68463099e-02 -3.60584855e-01 -8.18253309e-03 -1.01351058e+00
-2.47156210e-02 -5.96220136e-01 1.82293415e-01 -3.18640679e-01
-7.21216440e-01 1.15584433e+00 2.97505826e-01 -1.06790686e+00
-7.05122828e-01 -4.46445018e-01 -7.23289251e-02 5.44902325e-01
-1.00628531e+00 -7.96024621e-01 -2.91014791e-01 4.06694859e-01
3.71165834e-02 -5.96340448e-02 1.18236041e+00 4.62790877e-01
-6.32859707e-01 -1.18846847e-02 -1.59506515e-01 3.39880437e-01
5.75114667e-01 -1.47596693e+00 -5.32711372e-02 6.74319685e-01
5.78304291e-01 5.20364940e-01 7.13655531e-01 -5.98981738e-01
-4.00030941e-01 -4.83542919e-01 1.56353605e+00 -4.83095467e-01
6.77080452e-01 -2.29816884e-01 -8.66674244e-01 2.96686679e-01
2.79464543e-01 -3.31208885e-01 6.03513241e-01 5.80037594e-01
-2.27432139e-02 -2.64302701e-01 -8.69020343e-01 2.42121473e-01
8.92126977e-01 -3.41220379e-01 -7.09453881e-01 3.87480199e-01
4.52535987e-01 9.90210921e-02 -1.43388438e+00 4.32200760e-01
7.46811926e-02 -1.14915740e+00 4.91506130e-01 -1.66575208e-01
2.74299502e-01 -3.35604727e-01 3.97108704e-01 -1.00106061e+00
8.72330144e-02 -5.72067499e-01 1.66245013e-01 1.72850120e+00
5.87972462e-01 -4.79155153e-01 6.02899611e-01 9.20748115e-02
3.11071545e-01 -5.08007348e-01 -5.66388071e-01 -4.47867811e-01
-2.21899569e-01 -3.00769717e-01 5.52299201e-01 1.08531117e+00
5.37741363e-01 6.24684751e-01 2.23295614e-01 -1.38884902e-01
5.15884459e-01 2.27767751e-01 8.61205280e-01 -1.74102855e+00
-2.64250964e-01 -5.04495800e-01 -7.93464780e-01 -6.07455522e-02
1.02109753e-01 -7.83538282e-01 -1.27359778e-01 -1.37156129e+00
-5.70825860e-02 -5.30661345e-01 -1.69236436e-01 2.36496836e-01
4.09399495e-02 2.62466967e-01 2.62116522e-01 3.61684263e-01
-6.65124178e-01 8.90393183e-02 6.50010407e-01 1.38510272e-01
-2.27202520e-01 1.57826766e-01 -3.41103554e-01 7.60053217e-01
7.22208798e-01 -5.72968602e-01 -1.82084069e-01 2.59006858e-01
4.96093571e-01 -2.39123717e-01 -6.55855313e-02 -1.20354581e+00
1.23956211e-01 2.58814722e-01 1.28228739e-01 -4.09966826e-01
-4.81406897e-02 -1.11424243e+00 3.06725532e-01 2.85706550e-01
-1.62200138e-01 -4.08049256e-01 -2.27276146e-01 -5.89830754e-03
-4.42323834e-01 -6.88753486e-01 3.62096637e-01 -2.29311269e-02
-3.74743223e-01 -1.87613994e-01 -2.67639130e-01 -1.48441583e-01
8.80469739e-01 6.31996021e-02 -2.37534312e-03 2.17782799e-02
-1.42603922e+00 -4.83906269e-02 2.95089394e-01 -2.10117474e-01
8.96104500e-02 -9.91615951e-01 -5.01086354e-01 -3.13294441e-01
2.39313245e-01 1.36371618e-02 -1.35803610e-01 9.62412477e-01
-4.77577657e-01 3.43157321e-01 -2.33590305e-02 -3.43852818e-01
-1.55194330e+00 8.84803653e-01 -9.46065411e-02 -8.16667199e-01
-3.77356946e-01 8.50325152e-02 -3.99076253e-01 -6.01704419e-02
1.13070726e-01 -5.37507057e-01 -1.04238892e+00 6.08498096e-01
1.01752080e-01 2.02160835e-01 4.15986389e-01 -7.70873785e-01
-2.08144516e-01 6.37581825e-01 4.94022161e-01 -2.08953127e-01
1.37963271e+00 6.12560567e-03 -5.57432950e-01 5.33281863e-01
8.79098833e-01 4.31208313e-01 -4.66783673e-01 8.39746930e-03
8.90370965e-01 9.04879794e-02 -5.80773473e-01 -4.02645022e-01
-6.52617812e-01 5.61572731e-01 1.31575510e-01 1.13404238e+00
1.06756294e+00 2.83776730e-01 4.26617861e-02 6.13325775e-01
4.70592260e-01 -8.84255826e-01 -4.69638854e-01 2.04521924e-01
5.24472594e-01 -9.17381704e-01 2.15201423e-01 -9.99004781e-01
-1.98527858e-01 1.39288366e+00 -2.17103399e-02 2.00447608e-02
7.27442980e-01 3.75801712e-01 -2.76941389e-01 -3.51934373e-01
-2.88862020e-01 -1.02205741e+00 2.43599683e-01 5.70250511e-01
6.17274106e-01 2.29682654e-01 -1.24960077e+00 4.79668975e-01
-2.82574445e-01 9.36063007e-02 2.90231526e-01 5.44114828e-01
-5.12147784e-01 -1.87144971e+00 -1.78799555e-01 2.00247616e-01
-3.55368227e-01 2.65523419e-02 -8.20297778e-01 1.15253592e+00
6.22971117e-01 1.10951710e+00 -1.82967275e-01 -4.86876249e-01
4.06693637e-01 8.45520571e-02 5.88351965e-01 -7.28058815e-01
-7.51330733e-01 6.47350820e-03 5.58659613e-01 -1.28687143e-01
-1.12577963e+00 -5.33498526e-01 -1.09823072e+00 8.55764970e-02
-6.42786264e-01 9.28796947e-01 7.66586244e-01 1.24836278e+00
-8.84644240e-02 6.24370992e-01 4.54454124e-01 -4.98107046e-01
1.49632081e-01 -1.29171014e+00 -7.24851429e-01 6.02301061e-01
-2.88995832e-01 -5.98767996e-01 -2.20058158e-01 1.61657065e-01] | [9.398929595947266, 8.766794204711914] |
a51fe8ea-05d0-4ca9-8fa3-93a1930c9ed2 | multinomial-adversarial-networks-for-multi | 1802.05694 | null | http://arxiv.org/abs/1802.05694v1 | http://arxiv.org/pdf/1802.05694v1.pdf | Multinomial Adversarial Networks for Multi-Domain Text Classification | Many text classification tasks are known to be highly domain-dependent.
Unfortunately, the availability of training data can vary drastically across
domains. Worse still, for some domains there may not be any annotated data at
all. In this work, we propose a multinomial adversarial network (MAN) to tackle
the text classification problem in this real-world multidomain setting (MDTC).
We provide theoretical justifications for the MAN framework, proving that
different instances of MANs are essentially minimizers of various f-divergence
metrics (Ali and Silvey, 1966) among multiple probability distributions. MANs
are thus a theoretically sound generalization of traditional adversarial
networks that discriminate over two distributions. More specifically, for the
MDTC task, MAN learns features that are invariant across multiple domains by
resorting to its ability to reduce the divergence among the feature
distributions of each domain. We present experimental results showing that MANs
significantly outperform the prior art on the MDTC task. We also show that MANs
achieve state-of-the-art performance for domains with no labeled data. | ['Xilun Chen', 'Claire Cardie'] | 2018-02-15 | multinomial-adversarial-networks-for-multi-1 | https://aclanthology.org/N18-1111 | https://aclanthology.org/N18-1111.pdf | naacl-2018-6 | ['cross-domain-text-classification'] | ['natural-language-processing'] | [ 3.00484687e-01 -9.83091667e-02 -4.85507809e-02 -4.68222082e-01
-9.24488664e-01 -1.01350296e+00 6.67566359e-01 1.40341803e-01
-4.23484623e-01 1.07949245e+00 -1.79906398e-01 -2.79022157e-01
-2.41983280e-01 -7.79340625e-01 -6.74035907e-01 -7.36154497e-01
3.62232663e-02 8.93053532e-01 8.36411025e-03 -3.29293549e-01
-5.23749813e-02 2.82590061e-01 -1.18553591e+00 2.12025717e-01
9.34996188e-01 8.27292144e-01 -2.95729965e-01 8.23937595e-01
9.70109478e-02 7.23385632e-01 -9.19884145e-01 -7.62044847e-01
3.42685938e-01 -2.72209436e-01 -1.03043044e+00 -3.50703187e-02
6.14167511e-01 -1.96232408e-01 -4.54139143e-01 1.22226012e+00
3.29733491e-01 2.30029061e-01 1.26391041e+00 -1.79724991e+00
-1.05151916e+00 6.49305999e-01 -4.02401447e-01 6.31922558e-02
1.92296863e-01 -2.66573995e-01 1.25483453e+00 -5.81107140e-01
4.03794497e-01 1.20978069e+00 7.24956214e-01 6.86309338e-01
-1.26267064e+00 -5.46994746e-01 1.19383045e-01 4.49226145e-03
-1.17580509e+00 -1.31277561e-01 5.41390896e-01 -3.75813991e-01
5.54162323e-01 2.61334419e-01 -1.45154551e-01 1.43460202e+00
3.49322557e-01 1.05862665e+00 9.84243989e-01 -4.45679367e-01
2.86899805e-01 2.51375973e-01 1.12910770e-01 3.10348868e-01
3.67918521e-01 -1.24645606e-01 -4.00955498e-01 -4.26913768e-01
4.53070611e-01 -9.51891169e-02 -2.33538002e-01 -5.39637685e-01
-9.94894505e-01 1.12880099e+00 8.34122598e-02 3.59344244e-01
-9.33571011e-02 -2.16660842e-01 6.21052802e-01 8.32626700e-01
7.04456210e-01 5.36511004e-01 -7.71050334e-01 1.93222731e-01
-6.75702691e-01 5.34534454e-01 1.07015860e+00 1.08129454e+00
2.93729186e-01 -3.12668756e-02 3.30277346e-02 8.44443560e-01
-7.08872173e-03 5.59745014e-01 6.73423290e-01 -6.66514158e-01
6.68877602e-01 3.45642745e-01 5.24425805e-02 -7.36996770e-01
-3.44453722e-01 -2.17901975e-01 -1.21541882e+00 1.21457025e-01
9.40550327e-01 -3.42147470e-01 -5.56341290e-01 2.03473616e+00
9.25209820e-02 -1.07542463e-01 4.08306330e-01 5.40247202e-01
3.52108985e-01 4.31989759e-01 1.75423279e-01 -2.69965846e-02
8.98601532e-01 -6.68775856e-01 -6.07221127e-01 -3.64802867e-01
6.23982191e-01 -4.79445547e-01 1.07117331e+00 5.35101771e-01
-8.32057416e-01 -3.38633835e-01 -1.06914186e+00 7.63102770e-02
-7.20705628e-01 -2.81830072e-01 5.34976065e-01 9.62288797e-01
-8.06270599e-01 7.17751324e-01 -5.34773946e-01 -3.12521130e-01
5.07190049e-01 4.07415152e-01 -6.36986017e-01 -2.02203035e-01
-1.62979221e+00 8.40919614e-01 6.12606943e-01 -4.01810527e-01
-7.09268153e-01 -7.14323938e-01 -8.08535039e-01 -1.08446598e-01
1.33855551e-01 -6.21602714e-01 1.42594850e+00 -1.31955004e+00
-1.24151838e+00 9.94322300e-01 2.58672893e-01 -6.46503031e-01
9.29814637e-01 -1.63049668e-01 -5.08069754e-01 1.06710389e-01
3.90318811e-01 2.46034414e-01 9.86415923e-01 -1.01899660e+00
-6.54240787e-01 -4.27394271e-01 2.15960681e-01 1.67071283e-01
-7.43824899e-01 -5.01474850e-02 1.82401508e-01 -8.43333304e-01
-4.82391655e-01 -8.83915603e-01 -7.73418471e-02 1.71647489e-01
-4.81678545e-01 -3.63906413e-01 8.02336216e-01 -4.56499577e-01
8.27124834e-01 -2.12769318e+00 3.38204920e-01 3.44457477e-02
1.78592905e-01 2.00644284e-01 -2.32697234e-01 2.82012641e-01
-2.02433288e-01 2.46249169e-01 -5.40626585e-01 -2.76681840e-01
3.82114440e-01 3.78860652e-01 -3.82374853e-01 7.78350830e-01
1.99109152e-01 6.14990652e-01 -8.67175281e-01 -3.87090951e-01
1.45790149e-02 1.78993672e-01 -3.80757779e-01 2.22276807e-01
-3.86130035e-01 2.89267868e-01 -5.16804218e-01 3.55968505e-01
8.19748998e-01 -3.89686078e-01 4.20828700e-01 3.95928174e-01
5.25356114e-01 -6.82802200e-02 -1.18193293e+00 1.42897904e+00
-5.74531794e-01 7.86357582e-01 7.06806965e-03 -1.45979226e+00
6.29545689e-01 3.81007075e-01 5.93528807e-01 -2.20859766e-01
1.10763639e-01 2.60214865e-01 -3.71538326e-02 -6.85763583e-02
4.05885756e-01 -3.87947470e-01 -5.61476886e-01 5.42076170e-01
3.37996751e-01 -2.14647666e-01 1.22651361e-01 1.60053268e-01
1.21261871e+00 -2.72237360e-01 4.68553215e-01 -4.06329155e-01
5.29566944e-01 -7.45805800e-02 2.79509008e-01 1.11819506e+00
-4.92310792e-01 5.68275690e-01 6.80432498e-01 -2.55754113e-01
-1.24671340e+00 -1.21358836e+00 -4.30487871e-01 1.40049839e+00
1.02589600e-01 1.00364514e-01 -7.09826112e-01 -1.11598778e+00
2.44085923e-01 7.53616810e-01 -8.13408673e-01 -2.39201695e-01
-2.97333330e-01 -8.74314725e-01 9.92731094e-01 4.52536076e-01
6.07969463e-01 -8.33467007e-01 -8.11102092e-02 1.54198766e-01
-5.46728075e-02 -1.12779391e+00 -5.25633156e-01 5.07810533e-01
-5.38774550e-01 -1.21089339e+00 -8.49421203e-01 -9.65382159e-01
4.18841839e-01 -8.05702358e-02 1.28336239e+00 -3.98541629e-01
2.32124478e-02 4.36347336e-01 -4.17350799e-01 -4.89723593e-01
-8.88990879e-01 3.02291214e-01 2.87524760e-01 -1.36286184e-01
6.13053143e-01 -5.41797578e-01 -5.38892075e-02 4.52969462e-01
-1.27800882e+00 -4.49447125e-01 2.98648924e-01 1.18826926e+00
2.86927968e-01 5.06459832e-01 8.88304353e-01 -1.31536603e+00
7.94354677e-01 -8.40752244e-01 -3.71412188e-01 4.23628628e-01
-3.39766115e-01 1.15417451e-01 1.37807500e+00 -7.72145450e-01
-7.79419720e-01 -1.40558511e-01 5.67651875e-02 -1.87585369e-01
-5.18729866e-01 2.93362647e-01 -3.90567482e-01 4.78839427e-02
9.03951645e-01 1.33644059e-01 -1.71692342e-01 -1.42163828e-01
2.78772175e-01 8.88423085e-01 3.83290499e-01 -8.16125691e-01
9.86547530e-01 3.82438689e-01 2.96270140e-02 -7.02835977e-01
-1.14680123e+00 -2.92381823e-01 -8.69493723e-01 1.98599696e-01
6.77957833e-01 -7.13329434e-01 -4.03823614e-01 5.97556233e-01
-9.54968333e-01 -4.36529338e-01 -1.90381259e-01 1.34560376e-01
-7.30765045e-01 4.16860521e-01 -4.88979220e-01 -5.71370184e-01
-2.89598256e-01 -7.34460771e-01 7.64347315e-01 8.84993467e-03
-1.14020750e-01 -1.55074513e+00 2.32980195e-02 2.94799712e-02
2.28954330e-01 3.02994877e-01 1.05213761e+00 -1.56331444e+00
1.93545952e-01 -3.14533174e-01 2.27844678e-02 6.33775055e-01
2.19997644e-01 -3.34926009e-01 -1.08952081e+00 -4.27139461e-01
7.39825442e-02 -6.69864655e-01 6.16858900e-01 2.80906826e-01
1.35306931e+00 -3.48325819e-01 -1.27543688e-01 4.78948504e-01
1.33841598e+00 1.12075126e-02 3.03378135e-01 5.51478922e-01
5.71723640e-01 4.91144329e-01 5.47437370e-01 4.20023710e-01
9.51276422e-02 5.20739436e-01 2.44181111e-01 8.67801681e-02
3.51163387e-01 -5.31554669e-02 3.80621374e-01 4.07162130e-01
3.28723550e-01 -8.09660435e-01 -8.62481892e-01 6.06555402e-01
-1.70647979e+00 -9.90086854e-01 -6.63423240e-02 2.07629251e+00
1.03584206e+00 3.25660318e-01 2.86359459e-01 1.82555735e-01
9.85976934e-01 9.58790258e-02 -8.54461312e-01 -4.17109162e-01
-4.05697852e-01 4.41019922e-01 7.00198770e-01 2.48952478e-01
-1.70703614e+00 6.14314914e-01 6.71624136e+00 9.62617397e-01
-5.81278503e-01 2.21253540e-02 4.38730597e-01 2.92643070e-01
-7.50331506e-02 -4.82523799e-01 -5.77532709e-01 5.10170758e-01
1.05075741e+00 -5.75484335e-01 2.90022641e-01 1.01702809e+00
-5.41841984e-01 3.31031114e-01 -1.40508163e+00 6.83228433e-01
4.75568324e-02 -7.72373378e-01 1.20710813e-01 5.92381787e-03
8.10183287e-01 -1.69130191e-01 4.57463801e-01 4.73570704e-01
9.29612696e-01 -1.13073874e+00 4.45581168e-01 -4.89615351e-02
1.00826240e+00 -9.95062351e-01 8.06733072e-01 6.23261869e-01
-7.33578086e-01 -5.47114089e-02 -6.30002201e-01 1.94352195e-01
-3.22158575e-01 4.64851022e-01 -6.78214073e-01 7.51128852e-01
2.72838324e-01 7.44833171e-01 -4.33511674e-01 4.93751466e-01
-4.72601838e-02 4.93830740e-01 -2.70788044e-01 -1.66756406e-01
1.99656412e-01 1.83358878e-01 5.00782967e-01 1.19447660e+00
-5.74165732e-02 -1.40363067e-01 3.90696406e-01 5.91024816e-01
-4.64788795e-01 -2.89652888e-02 -8.52706432e-01 -5.39650209e-02
3.93270642e-01 8.35179210e-01 -3.80999744e-01 -1.39009118e-01
-5.90161741e-01 1.28149152e+00 3.41864556e-01 2.67532468e-01
-7.68192410e-01 -7.72660434e-01 8.91353488e-01 -2.56082565e-01
2.64396429e-01 5.86921312e-02 -3.41371655e-01 -1.37577665e+00
-2.14426275e-02 -1.11253822e+00 6.81782365e-01 -2.59995371e-01
-2.29510570e+00 5.26502192e-01 -5.01613803e-02 -1.32328582e+00
-3.50789160e-01 -1.11261451e+00 -2.05120504e-01 8.69296134e-01
-1.46901786e+00 -1.09871340e+00 1.32196724e-01 9.64800656e-01
6.29760742e-01 -5.42000771e-01 1.21486545e+00 2.03227624e-01
-2.90103614e-01 1.02902782e+00 8.83514166e-01 5.76017499e-01
9.63595569e-01 -1.80235744e+00 5.42210937e-01 6.75503433e-01
1.05893463e-01 3.33777428e-01 7.08684981e-01 -4.13846165e-01
-1.08882761e+00 -1.21639621e+00 6.11194193e-01 -6.81341648e-01
1.09385240e+00 -3.69664848e-01 -1.01062095e+00 9.88496542e-01
1.77719258e-02 3.38617265e-02 9.63702261e-01 2.35019661e-02
-7.02802300e-01 2.49966025e-01 -1.53182185e+00 4.04822975e-01
7.49420464e-01 -6.43337011e-01 -7.51537085e-01 6.90797329e-01
4.34122384e-01 -3.54090095e-01 -9.30774152e-01 -2.39721849e-03
4.56211656e-01 -7.85386622e-01 9.08373058e-01 -1.06137311e+00
6.94689989e-01 1.69051215e-01 -3.76644105e-01 -1.64620483e+00
-1.15559004e-01 -3.57979357e-01 6.21588454e-02 1.06964529e+00
3.33396226e-01 -8.26405048e-01 6.17770076e-01 5.92027068e-01
1.56003729e-01 -1.25567868e-01 -1.08101594e+00 -1.15915036e+00
1.12375903e+00 -3.51327121e-01 4.17146862e-01 1.38579154e+00
-4.60404679e-02 3.19605678e-01 -3.42813969e-01 3.13637674e-01
7.93051898e-01 -7.13417679e-02 5.89352489e-01 -1.58863342e+00
-3.98967087e-01 -3.34355295e-01 -4.86234397e-01 -9.58645046e-01
8.74894798e-01 -1.04151690e+00 6.15887567e-02 -1.04351926e+00
6.20954596e-02 -4.13839042e-01 -2.55620122e-01 3.61201704e-01
-2.39771545e-01 7.94244185e-03 1.06441766e-01 4.67210971e-02
-5.84261000e-01 2.87863880e-01 1.09029937e+00 -3.55581433e-01
1.01059251e-01 2.79569358e-01 -8.17494452e-01 7.22900510e-01
9.98114645e-01 -6.85407102e-01 -3.46954733e-01 -3.33160073e-01
-5.39863296e-02 -7.75354505e-02 3.17676008e-01 -8.99392128e-01
1.34359375e-01 -3.44577581e-01 5.73837578e-01 -1.03949286e-01
5.32887802e-02 -8.74405682e-01 -3.10890079e-01 1.91211626e-01
-6.90750539e-01 -2.34019309e-01 2.16838315e-01 7.73221135e-01
-1.65067300e-01 -4.58763480e-01 1.12198532e+00 -3.11215185e-02
-3.87076199e-01 3.67313683e-01 -3.30005407e-01 6.82892799e-01
1.03949928e+00 1.62018105e-01 -4.28167552e-01 -4.35041010e-01
-5.98835707e-01 1.81767434e-01 3.42704713e-01 2.37715855e-01
1.98473856e-01 -1.17501807e+00 -9.08726037e-01 2.08902836e-01
5.43228798e-02 6.93160295e-02 8.31687525e-02 1.54320002e-01
-2.63073176e-01 1.68386132e-01 -2.97514766e-01 -3.62534583e-01
-1.06102157e+00 6.32162571e-01 5.35013914e-01 -6.11635625e-01
-3.09917569e-01 8.00481439e-01 2.50831246e-01 -7.82156467e-01
1.38185725e-01 1.16755299e-01 -5.81222065e-02 6.21925713e-03
4.28470254e-01 2.03999355e-01 6.11376166e-02 -3.90813470e-01
-2.93815583e-01 9.79679525e-02 -3.89502972e-01 -1.97195202e-01
1.28036976e+00 1.38205230e-01 1.73357993e-01 4.66761917e-01
1.39870858e+00 -2.10215777e-01 -1.08840203e+00 -3.77768427e-01
3.50728966e-02 -4.16955262e-01 -3.45085442e-01 -8.44199419e-01
-6.92319393e-01 7.32494891e-01 3.04470539e-01 7.67560065e-01
1.11056447e+00 -6.98080212e-02 5.76142490e-01 5.53523242e-01
2.52382010e-01 -1.08471823e+00 1.27875969e-01 7.84480214e-01
5.35239935e-01 -1.29340398e+00 -1.67840302e-01 -1.11361533e-01
-8.74970376e-01 1.29467499e+00 5.58664799e-01 -3.02251518e-01
6.71022952e-01 2.65110999e-01 -7.29964003e-02 2.58104920e-01
-6.31870270e-01 9.49853752e-03 1.45166948e-01 1.11560094e+00
3.36408168e-01 1.42522871e-01 6.69564530e-02 6.55186057e-01
-2.91398019e-01 -3.72735828e-01 5.70783913e-01 8.75156343e-01
-3.06633353e-01 -1.22255611e+00 -3.83695424e-01 5.61917245e-01
-8.42042387e-01 1.27248699e-02 -4.67221558e-01 9.98373866e-01
-1.69212952e-01 1.02588320e+00 -2.24754866e-02 -2.00364932e-01
3.10430437e-01 2.92049855e-01 4.66992557e-01 -5.97763300e-01
-3.91629666e-01 -4.99036282e-01 -1.04487665e-01 1.52736291e-01
-4.53964382e-01 -8.85547996e-01 -7.99105465e-01 -6.79524064e-01
-1.53090417e-01 2.10149571e-01 3.20880234e-01 1.09366965e+00
-1.19597271e-01 4.15531754e-01 8.27841818e-01 -3.74051273e-01
-1.28063238e+00 -1.05724406e+00 -1.08664405e+00 6.88497663e-01
6.23196185e-01 -6.26574457e-01 -6.93750799e-01 2.68742710e-01] | [10.290606498718262, 3.204988718032837] |
16eb2de2-1afe-4a9b-9685-487136e337c9 | gsim-a-graph-neural-network-based-relevance | 2208.06144 | null | https://arxiv.org/abs/2208.06144v2 | https://arxiv.org/pdf/2208.06144v2.pdf | GSim: A Graph Neural Network based Relevance Measure for Heterogeneous Graphs | Heterogeneous graphs, which contain nodes and edges of multiple types, are prevalent in various domains, including bibliographic networks, social media, and knowledge graphs. As a fundamental task in analyzing heterogeneous graphs, relevance measure aims to calculate the relevance between two objects of different types, which has been used in many applications such as web search, recommendation, and community detection. Most of existing relevance measures focus on homogeneous networks where objects are of the same type, and a few measures are developed for heterogeneous graphs, but they often need the pre-defined meta-path. Defining meaningful meta-paths requires much domain knowledge, which largely limits their applications, especially on schema-rich heterogeneous graphs like knowledge graphs. Recently, the Graph Neural Network (GNN) has been widely applied in many graph mining tasks, but it has not been applied for measuring relevance yet. To address the aforementioned problems, we propose a novel GNN-based relevance measure, namely GSim. Specifically, we first theoretically analyze and show that GNN is effective for measuring the relevance of nodes in the graph. We then propose a context path-based graph neural network (CP-GNN) to automatically leverage the semantics in heterogeneous graphs. Moreover, we exploit CP-GNN to support relevance measures between two objects of any type. Extensive experiments demonstrate that GSim outperforms existing measures. | ['Wenjie Zhang', 'Xiaofeng Zhang', 'Xin Cao', 'Moli Lu', 'Yixiang Fang', 'Linhao Luo'] | 2022-08-12 | null | null | null | null | ['graph-mining'] | ['graphs'] | [ 4.68381234e-02 9.66316275e-03 -5.10012507e-01 7.86174759e-02
-4.74718306e-03 -2.35167846e-01 4.53807443e-01 8.64653051e-01
-1.00853063e-01 6.44845963e-01 1.21183448e-01 -3.75297576e-01
-6.85414672e-01 -1.63113761e+00 -2.73928702e-01 -3.82042080e-01
-1.71754181e-01 1.98362559e-01 6.22423947e-01 -4.43084270e-01
1.59323037e-01 1.78102225e-01 -1.11208594e+00 -2.26186424e-01
9.00772154e-01 8.67404044e-01 6.97755814e-02 5.72490692e-03
-3.90528858e-01 6.28435194e-01 -5.57128906e-01 -5.96273661e-01
-1.21993693e-02 -3.87786955e-01 -9.10573840e-01 -3.07251453e-01
-4.20344546e-02 2.67021537e-01 -5.57048857e-01 1.43735075e+00
4.26761627e-01 1.09069996e-01 5.38917124e-01 -1.33526361e+00
-8.03665102e-01 9.49748695e-01 -6.90785468e-01 3.49596322e-01
3.67458493e-01 -3.62136692e-01 1.42240250e+00 -4.73659873e-01
6.34905040e-01 1.21411276e+00 5.92283368e-01 -4.26344685e-02
-6.82421744e-01 -6.41778231e-01 2.81357676e-01 6.04823232e-01
-1.30725229e+00 2.51475930e-01 9.76882398e-01 -1.73978195e-01
5.79748988e-01 3.40358138e-01 8.04925501e-01 7.59027064e-01
-1.19628139e-01 5.08859575e-01 7.61627495e-01 -4.00265634e-01
1.96085330e-02 -1.89332768e-01 3.53654772e-01 4.84726131e-01
8.49584281e-01 -3.20663720e-01 -2.25910708e-01 -8.32921267e-02
6.44647479e-01 1.15122400e-01 -4.73302722e-01 -2.14787960e-01
-1.12789106e+00 7.16378570e-01 8.39113176e-01 7.80096471e-01
-3.49896699e-01 -1.49037465e-01 6.05756581e-01 4.61569250e-01
3.54541183e-01 4.30870712e-01 -1.30280897e-01 4.05740172e-01
-2.87005514e-01 -1.52066424e-01 9.07704294e-01 1.02256215e+00
8.35271478e-01 -3.77066225e-01 -3.96892726e-01 7.77688682e-01
2.61604339e-01 9.22662690e-02 5.27787924e-01 -4.34490204e-01
6.27015591e-01 1.39873028e+00 -3.49731147e-01 -1.90710223e+00
-4.88387495e-01 -5.74082613e-01 -1.29479980e+00 -5.17977953e-01
8.20838884e-02 2.35072136e-01 -3.61519039e-01 1.71619713e+00
6.37682378e-01 2.16338962e-01 -4.20853607e-02 7.73510814e-01
1.31748962e+00 3.76320153e-01 7.03576878e-02 -3.65881324e-01
1.15735698e+00 -7.91172326e-01 -4.57797199e-01 -8.75677739e-04
6.36760592e-01 -5.14018416e-01 8.48650277e-01 -1.32007420e-01
-8.07920694e-01 -2.69400746e-01 -9.54727352e-01 8.84304792e-02
-5.01853526e-01 -6.14868224e-01 7.08845258e-01 4.03892636e-01
-1.07994926e+00 6.76238775e-01 4.38731536e-02 -6.03470743e-01
2.65802503e-01 2.13781297e-01 -3.49706501e-01 -3.24222118e-01
-1.69572735e+00 5.86182654e-01 9.48201060e-01 5.49333282e-02
-1.72255382e-01 -2.13977993e-01 -8.44273806e-01 3.18939656e-01
8.74680877e-01 -8.92497361e-01 6.45186305e-01 -9.26344395e-01
-8.31165075e-01 6.22297347e-01 8.81318375e-02 -1.59119219e-02
2.23669529e-01 5.63058853e-01 -1.08252847e+00 2.07110807e-01
2.98260421e-01 -1.64395720e-01 3.05732608e-01 -9.48408902e-01
-6.90203369e-01 -4.67738092e-01 5.15050709e-01 3.27495486e-01
-8.76792789e-01 -1.44110680e-01 -9.50417995e-01 -5.16350508e-01
2.74030596e-01 -5.18157065e-01 -1.24766938e-01 -4.03407842e-01
-6.85767591e-01 -5.88729799e-01 6.81527615e-01 -4.04671967e-01
1.72837484e+00 -1.64409316e+00 1.65384393e-02 7.48669028e-01
8.75341237e-01 2.51044691e-01 -2.59335279e-01 3.63418609e-01
5.50440326e-02 3.45573097e-01 -1.40516907e-01 4.65259671e-01
-1.74326733e-01 -3.64608206e-02 2.91487664e-01 1.30154461e-01
-4.18222286e-02 1.01544082e+00 -1.16383874e+00 -7.82145858e-01
-1.56384885e-01 1.63669288e-01 -1.76813200e-01 3.59529108e-02
-1.86249629e-01 4.84006666e-02 -7.26139843e-01 6.04633629e-01
5.11956632e-01 -8.59711468e-01 6.75245523e-01 -2.74560869e-01
3.03417951e-01 7.85868913e-02 -1.06303775e+00 1.31345892e+00
-2.68896013e-01 3.27540964e-01 -2.65293062e-01 -1.31929827e+00
1.02482569e+00 1.07625641e-01 4.40284431e-01 -1.01961410e+00
1.40946537e-01 3.84068429e-01 2.29019597e-01 -6.02945030e-01
5.73643565e-01 1.12728707e-01 9.16099921e-02 3.70151371e-01
-3.04996282e-01 4.80302095e-01 6.82562828e-01 5.30899346e-01
1.53410101e+00 -3.59060138e-01 5.54248512e-01 -1.80378795e-01
9.13965702e-01 -2.34309226e-01 5.30251443e-01 5.61265290e-01
-2.18466390e-02 2.93453693e-01 8.01459670e-01 -2.38733679e-01
-6.32075846e-01 -6.72761738e-01 2.21922442e-01 8.36257577e-01
8.31870854e-01 -5.95116854e-01 -4.04561520e-01 -8.84304106e-01
3.19357142e-02 1.00299932e-01 -4.90880579e-01 -4.63334620e-01
-5.87929428e-01 -6.50678754e-01 1.98452026e-01 4.14495885e-01
4.44488138e-01 -1.23017836e+00 2.50957727e-01 2.83707917e-01
-3.51008207e-01 -9.78626370e-01 -5.18577933e-01 -3.80996197e-01
-8.06309879e-01 -1.55306423e+00 -6.13302827e-01 -9.03309166e-01
6.27762675e-01 7.11515248e-01 1.42203379e+00 6.63145065e-01
1.24160737e-01 2.93382257e-01 -4.75348830e-01 3.02629080e-02
-1.82546899e-01 5.24641514e-01 -1.35538429e-01 -1.21325158e-01
4.47484553e-01 -8.57131004e-01 -4.89333808e-01 5.05925715e-01
-1.10072434e+00 -5.96475005e-02 7.60101378e-01 6.56846344e-01
5.56077421e-01 2.79603720e-01 1.00308049e+00 -1.13575304e+00
1.12592566e+00 -9.46203589e-01 -3.20727825e-01 5.98892748e-01
-9.72053707e-01 -7.47394189e-03 7.91161060e-01 -2.20218658e-01
-6.12804234e-01 -9.60140169e-01 -5.27931238e-03 -1.99962229e-01
2.93310344e-01 1.33261311e+00 -5.32779932e-01 -1.61335796e-01
5.02597630e-01 -7.37473220e-02 -3.74260783e-01 -1.58188373e-01
1.00649104e-01 4.63112921e-01 2.93567657e-01 -5.49912512e-01
7.09468246e-01 1.71028450e-01 4.53117937e-01 -5.84093869e-01
-7.58139729e-01 -6.93207622e-01 -2.59036601e-01 -1.23238616e-01
2.43333727e-01 -4.46045220e-01 -8.94008875e-01 6.55121058e-02
-1.12159181e+00 2.20064983e-01 6.33225217e-02 3.49024832e-01
1.63171103e-03 8.20026278e-01 -4.69886631e-01 -2.80074030e-01
-6.87125206e-01 -7.83108413e-01 4.51480687e-01 5.66458941e-01
7.17126504e-02 -1.21974766e+00 1.13797881e-01 5.10895103e-02
3.78168970e-01 3.23180854e-01 1.30258298e+00 -7.93793499e-01
-6.58494830e-01 -2.25382730e-01 -7.63574779e-01 -1.13921940e-01
3.85631680e-01 -2.92298168e-01 -3.11607540e-01 -1.80595234e-01
-6.29863977e-01 5.11817411e-02 8.76621425e-01 1.30941972e-01
1.22675598e+00 -4.76766378e-01 -6.79360211e-01 4.20071125e-01
1.40813410e+00 4.42118980e-02 4.53220636e-01 6.01070344e-01
9.93495643e-01 6.78547323e-01 4.56975460e-01 1.76718369e-01
5.78027487e-01 5.28485060e-01 6.15060449e-01 1.37681723e-01
2.18186393e-01 -2.68456995e-01 -1.81891426e-01 1.31883705e+00
-3.65204990e-01 -5.47136068e-01 -7.36010253e-01 5.44753850e-01
-2.16250014e+00 -8.33911180e-01 -3.93492788e-01 2.27158093e+00
7.04075813e-01 1.87672526e-01 1.21456511e-01 1.19109794e-01
1.30545163e+00 2.03899279e-01 -5.89776158e-01 1.28901660e-01
-1.21394135e-01 4.30761985e-02 8.79889131e-02 7.99964964e-02
-7.86686778e-01 6.36486411e-01 4.58752680e+00 9.50191975e-01
-7.55663335e-01 8.47498626e-02 3.91502827e-01 4.07425642e-01
-8.02623272e-01 1.33148551e-01 -3.91893536e-01 4.83772248e-01
5.10212839e-01 -8.55232954e-01 1.24858499e-01 8.18474770e-01
-1.43108994e-01 2.97425061e-01 -7.57195890e-01 9.28834379e-01
-8.06377083e-02 -1.16030169e+00 1.16407834e-01 9.03686360e-02
7.77971685e-01 -2.43701234e-01 -2.25409403e-01 4.68877077e-01
2.08357692e-01 -7.74132073e-01 5.68350777e-02 3.52531582e-01
6.95202410e-01 -9.71539259e-01 8.97499561e-01 1.30218744e-01
-1.71766353e+00 1.54235721e-01 -6.67769611e-01 2.47313514e-01
-5.15474286e-03 1.10848594e+00 -6.53479218e-01 1.25665390e+00
6.06733263e-01 8.74737501e-01 -6.51000679e-01 1.27661800e+00
-3.10129285e-01 3.61546725e-01 -1.08107440e-01 -2.83795416e-01
5.81210256e-02 -3.19966614e-01 6.03431880e-01 8.93916309e-01
5.94636381e-01 -1.27367899e-01 3.05787325e-01 7.22819388e-01
-6.36246443e-01 5.80868423e-01 -6.29392087e-01 -2.33617812e-01
6.35654569e-01 1.55228651e+00 -8.11618745e-01 -2.48215914e-01
-5.69041967e-01 6.05706632e-01 5.96119106e-01 2.69465417e-01
-5.48491418e-01 -7.17931092e-01 3.64843369e-01 1.99925721e-01
-2.21120358e-01 6.46003112e-02 1.43784508e-01 -1.15207350e+00
3.01427722e-01 -6.13408983e-01 8.48635614e-01 -4.02447432e-01
-1.66028321e+00 6.39725149e-01 -4.97026108e-02 -1.27293718e+00
-1.10563740e-01 -2.93771297e-01 -8.09784412e-01 8.62543702e-01
-1.71661019e+00 -1.01783276e+00 -9.21101034e-01 7.48976707e-01
-2.20560730e-01 -9.58062038e-02 4.12623763e-01 4.15248662e-01
-7.00545847e-01 5.25292814e-01 -6.25981614e-02 3.15616190e-01
5.40943623e-01 -1.02160132e+00 3.77560586e-01 7.69497693e-01
-1.52912121e-02 8.33235621e-01 2.83223987e-01 -9.78093803e-01
-1.24497688e+00 -1.28125823e+00 7.10986018e-01 1.44218534e-01
7.82496035e-01 4.33661073e-01 -1.19677305e+00 4.17935133e-01
2.20992472e-02 2.56271005e-01 5.57976961e-01 4.54757392e-01
-3.09569806e-01 -1.79943860e-01 -7.98506737e-01 7.35326886e-01
1.46038270e+00 -3.28767627e-01 -1.33374721e-01 3.73639643e-01
1.03550887e+00 -7.13020340e-02 -1.07236135e+00 7.34321237e-01
4.59084660e-01 -1.06390667e+00 9.59738135e-01 -5.43343008e-01
4.53434378e-01 -3.44321787e-01 1.64056972e-01 -1.43592870e+00
-5.91540396e-01 -2.43710548e-01 -4.74675715e-01 1.21308458e+00
2.45982096e-01 -9.67963576e-01 7.44692326e-01 -7.50803761e-03
5.84075553e-03 -6.83043778e-01 -6.95521593e-01 -8.31069112e-01
-3.74847144e-01 -2.07414374e-01 1.00006711e+00 1.28228557e+00
1.59069106e-01 6.19153082e-01 3.80139165e-02 8.05533230e-02
5.76950431e-01 3.35768998e-01 6.19711816e-01 -1.84399533e+00
-2.40827680e-01 -8.63131046e-01 -7.83961892e-01 -5.33536971e-01
1.75173715e-01 -1.22207892e+00 -3.50830615e-01 -2.06855035e+00
4.40234333e-01 -7.59864688e-01 -6.73264802e-01 2.69801050e-01
-6.03261471e-01 -2.43449137e-02 -1.41608283e-01 3.97062778e-01
-7.91274607e-01 4.26085830e-01 1.41958082e+00 -5.12248814e-01
-2.75914401e-01 2.39922907e-02 -9.46027875e-01 6.28209233e-01
7.48511076e-01 -3.10981154e-01 -5.68333089e-01 -9.47758555e-02
8.58276248e-01 5.59777766e-03 1.46084353e-01 -8.93297493e-01
3.36139441e-01 -3.59115273e-01 -4.95599173e-02 -2.31963471e-01
-2.75685936e-01 -7.52936661e-01 2.97353178e-01 3.34505320e-01
-1.71878070e-01 1.51197493e-01 -2.61791229e-01 9.09955323e-01
-4.05218184e-01 -3.55223805e-01 4.56477702e-01 -1.44670427e-01
-9.14918363e-01 7.57972002e-01 5.13505399e-01 3.17613691e-01
9.40812230e-01 -8.64281654e-02 -7.13371992e-01 -3.57354671e-01
-3.65630716e-01 3.25913906e-01 4.02925044e-01 5.17760634e-01
6.66270673e-01 -1.59106863e+00 -6.14150286e-01 -2.62532562e-01
5.66124141e-01 2.00850785e-01 2.62114525e-01 9.44635212e-01
-3.34776461e-01 1.87422663e-01 7.40778819e-02 -4.20612723e-01
-1.18258893e+00 7.29262888e-01 3.97815965e-02 -7.35324919e-01
-5.76241314e-01 4.00170207e-01 2.74066776e-01 -3.13047349e-01
-8.18737745e-02 2.37405561e-02 -8.08287680e-01 1.13036469e-01
4.05881822e-01 4.64661926e-01 1.86162338e-01 -5.98357260e-01
-2.06454977e-01 4.13683921e-01 -1.13965884e-01 4.09033358e-01
1.02571666e+00 -8.52176026e-02 -6.20992780e-01 1.71274975e-01
1.07638836e+00 1.42095760e-02 -1.17133752e-01 -7.29899585e-01
2.54947424e-01 -2.89441466e-01 -1.77142292e-01 -2.98995465e-01
-1.33757460e+00 5.86770773e-01 2.73707826e-02 7.47628927e-01
1.24478734e+00 9.30904225e-02 9.49622512e-01 5.34396589e-01
5.95153570e-01 -8.89559269e-01 8.71916786e-02 4.88343865e-01
4.80523139e-01 -1.08119881e+00 5.37492372e-02 -6.16753221e-01
-2.27893457e-01 1.07360506e+00 8.43278825e-01 3.96689951e-01
6.81983590e-01 -2.86774069e-01 -5.77188671e-01 -4.65904832e-01
-2.36606956e-01 -5.80495059e-01 5.26573777e-01 5.30294180e-01
2.85624295e-01 1.11474670e-01 -6.31000519e-01 5.88633180e-01
2.39093855e-01 -6.64031971e-03 3.45514566e-01 6.89733863e-01
-5.12933016e-01 -1.07469785e+00 -5.53850494e-02 8.57785642e-01
-2.09659651e-01 -2.16637552e-01 -4.43645835e-01 7.30098605e-01
-6.07145242e-02 8.74477923e-01 -3.16685736e-01 -6.66772962e-01
3.39937955e-01 -4.66535181e-01 2.40811035e-01 -4.96285647e-01
-4.89459813e-01 -3.75929326e-01 2.90273000e-02 -2.63133556e-01
-6.72186255e-01 2.27344222e-02 -1.18292952e+00 -6.01477385e-01
-6.13203645e-01 2.75208950e-01 1.13991134e-01 9.64177132e-01
3.82512003e-01 6.09555602e-01 6.06538355e-01 -2.69339234e-01
4.14956175e-02 -9.77615952e-01 -8.00293684e-01 5.25272191e-01
-2.10144296e-01 -6.97385669e-01 -1.85866237e-01 -6.96665823e-01] | [7.527002811431885, 6.543099880218506] |
2add8501-81e4-4caa-b9ef-9d27cd43bc37 | a-meta-learning-approach-to-reservoir | 2110.03722 | null | https://arxiv.org/abs/2110.03722v1 | https://arxiv.org/pdf/2110.03722v1.pdf | A Meta-learning Approach to Reservoir Computing: Time Series Prediction with Limited Data | Recent research has established the effectiveness of machine learning for data-driven prediction of the future evolution of unknown dynamical systems, including chaotic systems. However, these approaches require large amounts of measured time series data from the process to be predicted. When only limited data is available, forecasters are forced to impose significant model structure that may or may not accurately represent the process of interest. In this work, we present a Meta-learning Approach to Reservoir Computing (MARC), a data-driven approach to automatically extract an appropriate model structure from experimentally observed "related" processes that can be used to vastly reduce the amount of data required to successfully train a predictive model. We demonstrate our approach on a simple benchmark problem, where it beats the state of the art meta-learning techniques, as well as a challenging chaotic problem. | ['Michelle Girvan', 'Andrew Pomerance', 'Daniel Canaday'] | 2021-10-07 | null | null | null | null | ['time-series-prediction'] | ['time-series'] | [ 1.75676093e-01 -3.98316503e-01 1.60813987e-01 1.55642167e-01
-4.94884312e-01 -5.63481033e-01 1.07603586e+00 4.30422336e-01
-7.02675879e-02 7.03945339e-01 -2.17204392e-01 -4.31232452e-01
-1.70064732e-01 -6.21337593e-01 -4.90522236e-01 -8.52171242e-01
-4.39689547e-01 7.73039341e-01 8.60853642e-02 -4.50932980e-01
7.80824423e-01 5.30932665e-01 -1.41956544e+00 1.22116111e-01
7.73552656e-01 9.06885386e-01 2.11355865e-01 7.31021881e-01
-6.14892915e-02 1.00064373e+00 -5.23243964e-01 3.14395219e-01
1.46883518e-01 -7.01196134e-01 -5.46148717e-01 -1.99256524e-01
-2.12657094e-01 1.10493205e-01 -4.50919360e-01 7.36280978e-01
2.74681777e-01 6.01450279e-02 8.02325904e-01 -9.31105733e-01
-2.44617924e-01 4.85316724e-01 -9.95248258e-02 6.55497193e-01
-1.48724109e-01 6.05977893e-01 4.99693006e-01 -6.34903491e-01
5.01468301e-01 7.76832044e-01 7.09140420e-01 5.46222389e-01
-1.58553314e+00 -4.91414517e-01 -9.33932662e-02 1.43889248e-01
-1.06974792e+00 -4.75956619e-01 1.03007078e+00 -6.64872944e-01
1.13311899e+00 -1.27571717e-01 1.12423587e+00 9.63853955e-01
8.21606159e-01 3.73101294e-01 1.47047842e+00 -2.84371644e-01
7.80609608e-01 8.93688574e-02 -1.95836097e-01 4.60478902e-01
7.59395212e-02 7.72622943e-01 -5.50241053e-01 -5.29020369e-01
5.63009858e-01 -8.42031762e-02 -6.01016879e-02 -6.41517863e-02
-1.13484764e+00 6.03944123e-01 7.31162280e-02 3.97706628e-01
-5.66996038e-01 2.09128916e-01 2.81733125e-01 5.83760381e-01
6.92130566e-01 1.30189061e+00 -5.47741890e-01 -4.74621713e-01
-1.16114342e+00 2.99823284e-01 1.14719748e+00 5.27015150e-01
6.82602882e-01 5.02275705e-01 2.49659359e-01 3.17457438e-01
-1.01174645e-01 3.88407469e-01 8.52002978e-01 -7.99508631e-01
1.17420368e-01 5.78620255e-01 4.71639574e-01 -6.57633781e-01
-1.95291147e-01 -2.17999667e-01 -9.26982164e-01 1.18329972e-02
2.86728263e-01 -2.73325235e-01 -8.06277037e-01 1.22932947e+00
-8.92828405e-02 9.59902525e-01 3.68522227e-01 3.69350493e-01
9.93099660e-02 1.13292289e+00 -4.14649606e-01 -4.63439107e-01
6.23847842e-01 -9.89337087e-01 -3.30731004e-01 -3.12119424e-01
4.13391441e-01 -3.32442224e-01 6.83892250e-01 3.63111943e-01
-1.13512397e+00 -4.03899312e-01 -1.20695066e+00 7.58531272e-01
-3.11712772e-01 -2.82251596e-01 4.67715055e-01 1.92928597e-01
-9.11198497e-01 1.25231898e+00 -1.34756362e+00 -1.58050582e-02
2.49396209e-02 3.93322736e-01 1.34817645e-01 2.60137290e-01
-9.01781023e-01 1.23303497e+00 5.65311551e-01 9.81735811e-02
-1.52101076e+00 -9.51363742e-01 -4.20193821e-01 1.32254258e-01
1.96430281e-01 -3.89330626e-01 1.20933294e+00 -5.57321250e-01
-1.57786787e+00 1.84541911e-01 -2.58441359e-01 -8.15338790e-01
4.08416837e-01 2.45368287e-01 -4.35318351e-01 8.98185968e-02
-6.11983776e-01 -3.17960829e-02 1.13407731e+00 -1.30385435e+00
-3.46998811e-01 -1.03765920e-01 -5.02114534e-01 -2.68984526e-01
-2.21024811e-01 -2.42883399e-01 2.11065337e-01 -3.58891279e-01
1.57975703e-01 -1.14128315e+00 -6.69386446e-01 -5.73907137e-01
-9.11337063e-02 -7.27447793e-02 9.77796793e-01 -4.81775790e-01
1.14321506e+00 -1.70534718e+00 3.51903588e-01 1.82304040e-01
1.66125327e-01 3.48533034e-01 -2.03747451e-02 1.05775690e+00
1.54366627e-01 2.80679613e-01 -5.92397571e-01 -4.44123894e-01
-4.14397269e-01 3.72989833e-01 -8.11664104e-01 2.60156780e-01
6.02883995e-01 7.46548355e-01 -9.37344313e-01 -1.93235412e-01
3.13167572e-01 1.96709961e-01 -8.38599820e-03 6.51933908e-01
-5.71857274e-01 1.00439095e+00 -4.60146189e-01 4.42128062e-01
2.13941574e-01 -4.77147251e-01 2.33674303e-01 3.91570479e-01
-2.51927197e-01 1.45943180e-01 -7.74991095e-01 9.06751692e-01
-7.67111182e-01 6.77432954e-01 -4.17351991e-01 -1.38462877e+00
1.30823803e+00 2.85600990e-01 8.25246096e-01 -5.75663805e-01
7.19860867e-02 3.38968515e-01 2.98679411e-01 -4.27631050e-01
4.38881189e-01 -1.12415873e-01 8.03491399e-02 8.89433801e-01
-8.99609998e-02 -5.29111803e-01 1.13430813e-01 -2.91625887e-01
1.17492485e+00 -1.15886694e-02 4.06257778e-01 -3.81024510e-01
6.39343619e-01 4.37150389e-01 4.93202150e-01 7.93906450e-01
1.43734306e-01 1.39369160e-01 3.90796006e-01 -8.21371675e-01
-1.59580469e+00 -6.57396793e-01 -8.08329508e-03 3.87713701e-01
9.88485515e-02 -4.51091200e-01 -4.45785463e-01 8.91841128e-02
3.52927893e-02 5.91317236e-01 -8.38248849e-01 -4.50323284e-01
-9.12422359e-01 -1.15560412e+00 2.75655597e-01 3.68615925e-01
3.90781760e-02 -1.29129660e+00 -8.37329209e-01 7.51195788e-01
3.94868702e-01 -7.72736251e-01 1.22617185e-01 2.60266930e-01
-1.21440196e+00 -9.80508208e-01 -3.00543457e-01 -5.46859086e-01
6.61951900e-01 -6.65266141e-02 1.06216371e+00 1.62413254e-01
-4.20560569e-01 6.13915175e-02 -1.64068714e-01 -4.88923967e-01
-1.09353209e+00 2.97516119e-02 2.50455379e-01 -6.85414672e-02
6.85610548e-02 -8.79901230e-01 -3.98877829e-01 1.42378345e-01
-6.06690645e-01 -5.74483955e-03 3.77045691e-01 1.17348242e+00
5.52742660e-01 3.21673334e-01 7.43463218e-01 -4.60152149e-01
8.68374705e-01 -7.69110620e-01 -9.48278487e-01 5.24667144e-01
-1.05161476e+00 4.12815660e-01 1.22466898e+00 -9.21637297e-01
-7.42825031e-01 -8.61805826e-02 5.08099198e-01 -7.26532042e-01
2.32524171e-01 9.42120075e-01 6.00665033e-01 -1.29621834e-01
5.39602578e-01 1.03843057e+00 2.91021287e-01 -4.12881345e-01
-1.74446374e-01 5.70188105e-01 4.56858516e-01 -7.59265423e-01
6.48032367e-01 4.02271092e-01 3.74254346e-01 -8.01319003e-01
-4.47726935e-01 -2.74925977e-01 -7.18718529e-01 -1.17141955e-01
7.90188089e-02 -6.14803851e-01 -6.52034461e-01 5.75964391e-01
-8.25805306e-01 -6.15632236e-01 -3.57425749e-01 3.34467858e-01
-7.81373024e-01 -1.07928209e-01 -5.42917967e-01 -1.12730515e+00
-4.46122229e-01 -9.57230568e-01 6.07029498e-01 2.16504306e-01
-7.64739886e-02 -1.28348672e+00 5.94373107e-01 -4.42817956e-01
7.32404470e-01 4.49176967e-01 1.05633104e+00 -7.82222092e-01
-5.42784929e-01 -2.34627932e-01 5.09481251e-01 1.99756861e-01
2.62386143e-01 4.66052592e-01 -6.98448896e-01 -5.70317805e-01
4.07337755e-01 -1.37455761e-01 7.95741558e-01 1.23006225e-01
8.82853627e-01 -3.33263248e-01 -4.52220559e-01 3.63472432e-01
1.39826274e+00 5.42239130e-01 3.49799871e-01 3.14050764e-01
2.79205889e-01 5.98648071e-01 3.53630871e-01 7.50231087e-01
1.68297201e-01 3.07135373e-01 1.64987534e-01 5.08374691e-01
4.59374219e-01 -3.84311318e-01 3.58224005e-01 1.23888111e+00
-7.18848854e-02 -7.61226565e-02 -1.43785727e+00 6.73819482e-01
-1.91315699e+00 -1.16220307e+00 4.65193652e-02 2.09999514e+00
7.17247009e-01 9.07556340e-02 -1.00377444e-02 3.01732663e-02
5.72468579e-01 9.36691910e-02 -1.15502131e+00 -3.91582727e-01
3.77444364e-02 1.92329377e-01 1.27019852e-01 2.18440264e-01
-9.53646600e-01 9.50265110e-01 7.28539467e+00 2.71778464e-01
-1.65917909e+00 -9.64897871e-02 4.72423702e-01 -7.87699036e-03
-7.13880174e-03 2.58089572e-01 -6.37447834e-01 7.55285680e-01
1.73502517e+00 -9.73691404e-01 7.71103442e-01 6.18109465e-01
4.13330555e-01 7.34819379e-03 -1.31057513e+00 8.16651762e-01
-2.60054350e-01 -1.83372164e+00 -1.97358340e-01 2.34126911e-01
1.09823275e+00 3.04856002e-01 5.42218499e-02 2.15098873e-01
3.95998657e-01 -1.13878548e+00 4.57230359e-01 1.03358757e+00
2.40477532e-01 -6.55048490e-01 6.04850233e-01 8.29413056e-01
-1.09607756e+00 -5.09898484e-01 -4.33895350e-01 -3.67201775e-01
5.69709800e-02 3.66599768e-01 -1.09801650e+00 2.61788011e-01
4.35061663e-01 9.53182876e-01 -5.66687405e-01 1.02728379e+00
1.13099031e-01 9.91051078e-01 -3.90730381e-01 -3.19043010e-01
3.22877437e-01 -2.22767428e-01 6.16657138e-01 7.54999220e-01
6.39515281e-01 -5.23319878e-02 2.47144446e-01 9.45993662e-01
2.60936469e-01 -2.33496264e-01 -7.65707135e-01 -5.20529509e-01
6.91090941e-01 9.58175361e-01 -5.43722451e-01 -4.15512979e-01
6.99702203e-02 4.16300416e-01 2.80168980e-01 2.20457539e-02
-4.35984731e-01 1.71699077e-01 3.10566276e-01 -2.91712973e-02
3.76602948e-01 -7.27411330e-01 -2.15834215e-01 -1.28120983e+00
-5.20880520e-02 -8.91277134e-01 -2.70800292e-02 -5.34477592e-01
-1.52940595e+00 9.23068345e-01 -1.36233196e-01 -1.51933217e+00
-1.07781744e+00 -3.78609687e-01 -1.15323138e+00 8.90382826e-01
-1.46797276e+00 -8.06846738e-01 -1.27677888e-01 2.63777107e-01
4.30283427e-01 -6.77923322e-01 9.79233086e-01 -4.91958886e-01
-4.56637532e-01 -1.81962892e-01 9.76389170e-01 -3.03566784e-01
2.69870251e-01 -1.31127214e+00 6.97968900e-01 6.57118142e-01
1.66952319e-03 5.57645857e-01 1.00376701e+00 -7.58678317e-01
-1.87556398e+00 -1.22251403e+00 5.43886781e-01 -5.33088326e-01
1.11201692e+00 -1.45831227e-01 -1.16926539e+00 4.45851773e-01
-4.99089211e-02 2.49254689e-01 4.80472654e-01 -1.93432167e-01
-4.41638678e-02 -1.19407348e-01 -1.12879920e+00 4.53933716e-01
4.49947596e-01 -5.49658537e-01 -6.88059151e-01 2.86247253e-01
2.40935996e-01 -1.02985755e-01 -1.13398552e+00 2.27054670e-01
4.02386785e-01 -4.24519986e-01 5.64832628e-01 -9.58598435e-01
6.78395569e-01 -2.65522599e-01 1.35397255e-01 -1.79054749e+00
-1.99948605e-02 -9.12512422e-01 -9.48107004e-01 8.81233215e-01
4.42941189e-01 -6.79723203e-01 6.58319414e-01 5.57596266e-01
1.45858213e-01 -8.64647269e-01 -9.17676032e-01 -1.02804160e+00
4.89696175e-01 -1.21217519e-01 8.87658298e-01 1.00473261e+00
1.84719801e-01 -8.23094696e-02 -4.12857234e-01 3.78752083e-01
5.98990023e-01 5.44710577e-01 6.86647534e-01 -1.36592758e+00
-4.06830192e-01 -3.88746589e-01 -2.35198423e-01 -4.00862545e-01
2.46676147e-01 -6.27338409e-01 9.48414952e-02 -1.04015493e+00
-1.17418885e-01 -7.11507082e-01 -4.98180658e-01 7.90723711e-02
2.07026470e-02 -1.47041589e-01 1.93677038e-01 8.48150373e-01
-6.96402714e-02 7.80621111e-01 1.07117140e+00 -3.22648108e-01
-4.73133743e-01 2.26840943e-01 -3.25384021e-01 4.46941644e-01
1.14128304e+00 -6.28508687e-01 -4.25024837e-01 -3.89217883e-02
5.90312667e-02 5.79370320e-01 4.08148617e-01 -1.16443169e+00
5.83810449e-01 -5.18041193e-01 2.98390925e-01 -5.41078985e-01
4.24077392e-01 -5.58081210e-01 3.47341806e-01 9.10127699e-01
-4.38246936e-01 4.52591211e-01 2.63963699e-01 6.55889630e-01
-3.92575890e-01 -2.94842362e-01 6.05062962e-01 -4.21979666e-01
-7.19326377e-01 4.39159453e-01 -6.53768837e-01 -1.13436908e-01
1.06977594e+00 1.70212030e-01 -3.57911795e-01 -3.39451253e-01
-7.86371350e-01 3.19567651e-01 5.06318927e-01 4.28287446e-01
5.75476229e-01 -7.39240050e-01 -7.80784607e-01 2.25798577e-01
1.46780862e-02 -2.62440413e-01 2.67533259e-03 6.12380028e-01
-3.97403628e-01 3.96126628e-01 -3.29994261e-01 -6.41417921e-01
-6.21091843e-01 8.14529955e-01 6.95968747e-01 -4.76442933e-01
-7.44292974e-01 1.08851306e-01 -6.88095927e-01 -3.17202836e-01
-4.92034078e-01 -1.92758694e-01 6.41271174e-02 -2.30833799e-01
6.07168794e-01 2.78078139e-01 -1.63917206e-02 -3.18585813e-01
-8.86479095e-02 2.63841629e-01 -4.32141498e-02 -2.49918729e-01
1.81540740e+00 1.78769991e-01 -2.81729102e-01 8.94854486e-01
8.30599368e-01 -6.01819158e-01 -1.73974216e+00 -3.46554428e-01
4.36179221e-01 -2.14359105e-01 4.90791351e-02 -7.08087206e-01
-6.27671957e-01 1.10620606e+00 5.32169282e-01 6.39376223e-01
8.97875726e-01 -1.85646877e-01 6.44137502e-01 8.98909330e-01
6.74847543e-01 -1.12874079e+00 1.27102986e-01 6.85452044e-01
7.96444952e-01 -1.17938232e+00 8.21990371e-02 3.17599833e-01
-4.72356617e-01 1.41036749e+00 5.03787279e-01 -5.61735332e-01
9.12223458e-01 5.81271768e-01 -3.61985803e-01 -5.89651205e-02
-1.41196847e+00 2.61868298e-01 5.98028712e-02 5.03913343e-01
-7.90378377e-02 -1.12505451e-01 1.43618241e-01 2.95400709e-01
-2.70269990e-01 5.17335758e-02 8.36130083e-01 1.23143184e+00
-4.17708784e-01 -1.14982569e+00 -2.61678815e-01 4.85195756e-01
-5.51495887e-02 -1.39399499e-01 -3.88796508e-01 6.09094203e-01
-2.76505411e-01 8.14749002e-01 3.35376412e-01 -3.94638717e-01
-8.57738554e-02 1.15673974e-01 3.75749648e-01 -4.25347030e-01
-6.73252285e-01 -1.78649083e-01 -2.73095280e-01 -1.29051819e-01
-3.04883540e-01 -8.97076011e-01 -9.91451144e-01 -4.47001755e-01
-1.93505094e-03 3.92684281e-01 7.15971410e-01 1.17105436e+00
6.38800263e-01 3.85706276e-01 1.21126688e+00 -1.15979195e+00
-8.46944392e-01 -1.05698693e+00 -3.08949679e-01 -1.41875325e-02
5.12630224e-01 -5.60814917e-01 -4.02174354e-01 1.63562894e-01] | [6.5900373458862305, 3.351055383682251] |
742e4edd-611b-4ad3-8dea-f76211090bd8 | magnetic-properties-of-poly-trimethylene | 2012.05819 | null | https://arxiv.org/abs/2012.05819v1 | https://arxiv.org/pdf/2012.05819v1.pdf | Magnetic properties of poly(trimethylene terephthalate-block-poly(tetramethylene oxide) copolymer nanocomposites reinforced by graphene oxide-Fe3O4 hybrid nanoparticles | Thermoplastic elastomeric nanocomposites based on poly(trimethylene terephthalate-block-poly(tetramethylene oxide) copolymer (PTT-PTMO) and graphene oxide-Fe3O4 nanoparticle hybrid were prepared by in situ polymerization. Superparamagnetic GO-Fe3O4 hybrid nanoparticles before introducing to elastomeric matrix were characterized by X-ray photoelectron spectroscopy (XPS), X-ray diffraction (XRD), thermogravimetric analysis (TGA) and scanning electron microscopy (SEM). The effect of loading (0.3 and 0.5 wt%) of GO-Fe3O4 nanoparticle hybrid on the phase structure, tensile and magnetic properties of synthesized nanocomposites was investigated. The phase structure of nanocomposites was evaluated by differential scanning calorimetry (DSC) and dynamic mechanical thermal analysis (DMTA). Dispersion of GO-Fe3O4 nanoparticles in elastomeric matrix was evaluated by transmission electron microscopy (TEM). Magnetic properties of GO-Fe3O4 nanoparticle hybrid and nanocomposites with their content were characterized by using two different techniques: dc SQUID magnetization measurements as a function of temperature (from 2 to 300 K) and external magnetic field and ferromagnetic resonance (FMR) at microwave frequency. | ['Nikos Guskos', 'Grzegorz Żołnierkiewicz', 'Izabela Janowska', 'Zdenko Špitalský', 'Janusz Typek', 'Sandra Paszkiewicz', 'Anna Szymczyk'] | 2020-12-10 | null | null | null | null | ['x-ray-diffraction'] | ['miscellaneous'] | [ 5.38766384e-01 1.11294739e-01 -1.39653891e-01 3.89557421e-01
-5.15468307e-02 -3.47239345e-01 4.67412800e-01 1.12714037e-01
-4.84606922e-01 1.13185728e+00 4.50984120e-01 -2.03494132e-01
-4.33546789e-02 -1.25885403e+00 -5.60356259e-01 -1.05192423e+00
-1.77052870e-01 6.45277321e-01 7.53394246e-01 -4.42324758e-01
2.09255174e-01 6.88194633e-01 -1.31262398e+00 3.03730518e-01
9.07655001e-01 9.29827631e-01 4.97309953e-01 7.02536047e-01
1.27333120e-01 8.79739583e-01 5.36476150e-02 -1.65595874e-01
-1.34428190e-02 1.45886555e-01 -8.46771121e-01 2.16160461e-01
-2.75426656e-01 -6.21141791e-01 3.97118963e-02 1.10999942e+00
-1.49019152e-01 3.40653419e-01 5.21118164e-01 -4.93464142e-01
-1.16515279e+00 6.97012305e-01 -6.76259875e-01 4.58910391e-02
5.41273475e-01 2.17873320e-01 5.47078192e-01 -5.39165199e-01
8.27922106e-01 1.39707267e+00 3.91910493e-01 6.50204957e-01
-1.13885677e+00 -1.69694871e-01 -3.64581138e-01 -3.89785622e-03
-3.69929284e-01 7.37522468e-02 6.61615133e-01 -6.17959440e-01
1.10865521e+00 8.22011352e-01 6.81532443e-01 8.60018373e-01
1.04143858e+00 1.25751866e-03 1.72116756e+00 -4.94665474e-01
2.08670050e-01 1.10139422e-01 1.50493801e-01 3.17166150e-01
7.24170744e-01 2.16065928e-01 7.19055712e-01 -3.09135228e-01
7.52442002e-01 8.16929564e-02 2.78312355e-01 5.35909534e-01
-9.61513579e-01 4.20801431e-01 -2.10107446e-01 7.84207821e-01
-6.85658514e-01 -2.46682949e-02 7.22790241e-01 4.12442267e-01
4.27816540e-01 2.11738691e-01 -2.49519330e-02 -1.14594899e-01
-3.63119751e-01 -1.54480468e-02 1.02688503e+00 4.95734811e-01
7.50921816e-02 2.24068373e-01 4.60423678e-01 9.98484910e-01
7.18278110e-01 6.97100163e-01 5.77763557e-01 -9.04909849e-01
1.65556520e-01 3.00375462e-01 5.65951943e-01 -7.34228134e-01
4.95449230e-02 8.42219234e-01 -4.37213063e-01 3.91550601e-01
1.39693022e-01 -3.81739408e-01 -3.05241108e-01 8.91380847e-01
5.74082136e-01 -6.01144969e-01 1.41996711e-01 1.09063566e+00
9.18170333e-01 6.93505883e-01 4.40275103e-01 -3.80067706e-01
1.72904563e+00 -9.02407527e-01 -6.55378342e-01 3.52977484e-01
-1.82567999e-01 -6.49014950e-01 7.71908820e-01 2.46951692e-02
-1.19675541e+00 -2.02511266e-01 -9.67015743e-01 3.91131550e-01
-1.09316856e-01 -2.22800612e-01 3.64810973e-01 9.99072731e-01
-3.51618767e-01 1.17073071e+00 -1.33078849e+00 -5.31978130e-01
-2.45962173e-01 7.88495958e-01 -5.10567248e-01 3.41847599e-01
-9.66505587e-01 4.17255104e-01 3.35344315e-01 -3.23148131e-01
-3.90870154e-01 -2.31715575e-01 -5.12803257e-01 -3.98434699e-01
-7.77563453e-02 -6.92960322e-01 6.64088368e-01 -8.25499415e-01
-2.40102053e+00 7.20566034e-01 2.19120368e-01 -2.29191273e-01
5.11103868e-01 -1.57484621e-01 -9.27142441e-01 8.81374121e-01
1.13031603e-01 -2.28954814e-02 1.64180309e-01 -1.13000107e+00
-3.11573178e-01 -1.88842475e-01 1.77731186e-01 -3.19097549e-01
-1.29321516e-01 6.76569164e-01 1.19395661e+00 -6.51378691e-01
3.10046434e-01 -1.15364313e+00 -3.33916813e-01 -6.86439335e-01
-8.22391748e-01 -1.72624454e-01 8.49064112e-01 -7.68356502e-01
5.69664896e-01 -1.59016359e+00 -1.40304625e-01 3.15306365e-01
7.68010542e-02 2.98818827e-01 2.13915095e-01 1.27329755e+00
2.95566600e-02 2.97554612e-01 -7.82036632e-02 8.55794966e-01
3.19811255e-02 -1.12906002e-01 1.08423688e-01 5.35609603e-01
-2.91242301e-01 5.36444604e-01 -1.11367762e+00 -1.95699781e-01
4.68832970e-01 3.08676571e-01 -2.98296034e-01 -5.32983005e-01
-4.73487526e-01 7.17906833e-01 -1.08906972e+00 1.01028764e+00
8.84968460e-01 1.30654611e-02 4.69174325e-01 -3.24635923e-01
-6.99882209e-01 1.33702114e-01 -7.72541821e-01 2.51317739e-01
3.32959235e-01 2.31272876e-01 4.55873132e-01 -3.23294222e-01
1.06899273e+00 6.82928205e-01 9.96337712e-01 -1.00170016e+00
2.77409047e-01 5.02509296e-01 -5.79633750e-02 -1.28450549e+00
7.87121713e-01 -1.12996972e+00 5.41399956e-01 6.48598373e-01
-6.15128338e-01 6.35454953e-01 5.31470120e-01 -1.57063469e-01
6.68217540e-01 -1.73554018e-01 -3.30665298e-02 -9.92968738e-01
7.11718202e-01 2.41874546e-01 -2.74643060e-02 2.90891588e-01
2.79709220e-01 5.22645265e-02 3.78265321e-01 -5.70089757e-01
-1.82815063e+00 -8.05722952e-01 -2.63078213e-01 1.17017913e+00
2.57533103e-01 2.13432372e-01 -8.79953921e-01 4.12404388e-01
-2.23873183e-03 5.29619873e-01 -2.53151506e-01 3.36916447e-01
-6.07817650e-01 -7.82010555e-01 1.40198674e-02 3.91360700e-01
7.46976376e-01 -1.17162192e+00 -4.30844247e-01 9.44021344e-01
4.06077772e-01 -1.16160750e+00 1.44429490e-01 -7.13694870e-01
-1.45205295e+00 -9.59083319e-01 -4.79411274e-01 -4.46018994e-01
3.14841092e-01 -1.00712314e-01 5.63078821e-01 -3.94267365e-02
5.37139922e-02 7.75468707e-01 -5.87135732e-01 1.01637490e-01
-1.10096192e+00 -4.31932151e-01 3.43745142e-01 -1.08365595e-01
6.19692728e-02 -1.16201723e+00 -9.50521708e-01 4.87339109e-01
-1.00656867e+00 -1.39963061e-01 3.56773615e-01 3.40113223e-01
6.24425113e-01 -1.29913509e-01 2.38810673e-01 -1.45519829e+00
8.83648336e-01 -2.81490266e-01 -5.00127733e-01 3.28849144e-02
-2.60235608e-01 -1.70086771e-01 6.65187001e-01 -7.18421221e-01
-1.07400119e+00 -8.68527234e-01 -3.93251687e-01 2.88045593e-02
-1.42913967e-01 7.29678035e-01 1.30737111e-01 -9.83061418e-02
3.64980489e-01 6.51909932e-02 1.73425853e-01 -5.05846322e-01
-5.10231137e-01 6.95129633e-01 5.58319569e-01 -1.03797436e+00
4.48744565e-01 9.29244757e-01 1.71455443e-01 -1.08852899e+00
-4.29783076e-01 2.85517126e-01 -2.57198602e-01 -5.92717707e-01
1.13718104e+00 -4.50761139e-01 -1.19976401e+00 3.82383227e-01
-4.20379370e-01 -6.56565055e-02 1.60589874e-01 9.86931086e-01
-8.32075104e-02 7.27396309e-01 -1.46486163e+00 -6.27080560e-01
-7.00385749e-01 -9.91513669e-01 -7.55818887e-03 1.99791282e-01
8.65846202e-02 -1.60726631e+00 2.75741398e-01 8.44486833e-01
8.26949000e-01 1.48723698e+00 8.72989893e-01 -6.25114962e-02
-1.93752512e-01 -4.26896065e-01 8.83046761e-02 3.18560660e-01
1.90208524e-01 4.60912257e-01 -5.11258721e-01 -3.83322090e-01
5.09457648e-01 1.24595910e-01 4.84816492e-01 6.92618489e-01
4.82170314e-01 -8.63312662e-01 -6.08359993e-01 -2.06567094e-01
1.94277489e+00 3.52131277e-01 1.09970200e+00 9.50540781e-01
8.74647021e-01 1.98331684e-01 5.04008710e-01 3.30469072e-01
1.68807395e-02 1.82029214e-02 5.29696822e-01 5.16232252e-01
3.15938890e-01 7.63879418e-02 8.00799191e-01 1.07954681e+00
-1.34012067e+00 -1.36039838e-01 -5.17217398e-01 1.11514725e-01
-8.97215962e-01 -1.24235177e+00 -7.14638770e-01 2.14359641e+00
6.66639388e-01 8.56677964e-02 6.35219097e-01 -3.12905759e-02
1.36777842e+00 -9.36473608e-02 -3.33781272e-01 -7.28707790e-01
4.96204533e-02 1.16901979e-01 7.56384015e-01 5.34488738e-01
-6.53764963e-01 2.73578376e-01 5.52975178e+00 4.02059108e-01
-1.79001820e+00 7.51345605e-02 -8.40560421e-02 1.67278796e-01
-8.70534182e-01 4.14842963e-01 -2.12499902e-01 8.50224018e-01
9.10834670e-01 -1.62294447e-01 2.64711231e-01 4.13414508e-01
5.63374981e-02 -6.28136992e-01 -8.50776017e-01 2.00685412e-01
-8.41278136e-01 -1.44035780e+00 -1.56386822e-01 4.17073399e-01
7.44113505e-01 3.41719449e-01 -3.32026780e-02 -4.24911946e-01
1.08867630e-01 -3.91813576e-01 8.00463021e-01 5.87602079e-01
8.30349267e-01 -3.66177797e-01 3.82376164e-01 -7.81739950e-02
-1.11548328e+00 -2.11463913e-01 -7.06343114e-01 -2.36550376e-01
2.14234263e-01 1.01766741e+00 -3.84979963e-01 1.46286860e-01
7.34525204e-01 1.62493736e-01 3.30010861e-01 2.77965188e-01
3.54198992e-01 6.98206663e-01 -5.07577538e-01 -3.28312725e-01
5.94302595e-01 -8.42994213e-01 1.24372220e+00 7.44191051e-01
5.49607258e-03 7.40610421e-01 -6.14299998e-03 1.00161147e+00
1.50396585e-01 1.12469524e-01 -8.35903212e-02 -6.74672425e-01
5.25604725e-01 1.05567575e+00 -1.40741706e+00 -5.93732655e-01
-5.61196029e-01 -4.30068970e-02 -4.42154318e-01 8.61021429e-02
-4.41644818e-01 -4.35028911e-01 -1.21853864e-02 8.09479952e-01
3.81108671e-01 -2.40191326e-01 1.85384393e-01 -9.70116138e-01
-1.51218668e-01 -3.49818349e-01 6.92268386e-02 -1.82471886e-01
-1.48650956e+00 2.83809036e-01 2.22738720e-02 -1.08060157e+00
5.44789255e-01 -5.99171162e-01 -8.13868999e-01 4.67100024e-01
-9.31792498e-01 -6.17684186e-01 1.80631839e-02 -1.83195528e-02
-1.80744380e-01 2.19557732e-02 4.57519859e-01 1.83875352e-01
-3.23008537e-01 -1.92736879e-01 7.56761849e-01 -1.98656812e-01
1.25217989e-01 -7.23733962e-01 7.56113371e-03 -1.58013374e-01
-1.37682700e+00 4.26482201e-01 1.17000043e+00 -1.13376653e+00
-1.76418722e+00 -8.15784991e-01 6.02379739e-01 1.98402315e-01
8.33676696e-01 2.77738094e-01 -6.77438319e-01 8.66062999e-01
5.46498716e-01 -2.77281195e-01 9.48066115e-01 -7.01325536e-01
3.73101503e-01 5.50680697e-01 -1.72486138e+00 6.40684962e-01
2.84883171e-01 -3.80728513e-01 -3.85261327e-01 6.26237690e-01
3.36643159e-01 -4.91876394e-01 -2.15902758e+00 2.29797199e-01
1.08317995e+00 -1.02721405e+00 7.07709074e-01 -5.98285854e-01
1.02693570e+00 -2.51884609e-02 -2.81924814e-01 -2.02377662e-01
-6.17034674e-01 -3.52190048e-01 5.03706753e-01 9.88536656e-01
1.92482412e-01 -1.05513227e+00 3.48881632e-01 1.04226673e+00
-1.69590846e-01 -6.62726462e-01 -5.16236782e-01 -9.08730984e-01
-3.79683562e-02 5.88846028e-01 5.26576996e-01 1.11361194e+00
4.15701479e-01 -6.87313601e-02 -3.59864920e-01 5.02784476e-02
4.87605900e-01 3.15278262e-01 8.13563466e-02 -1.24223781e+00
-6.15493506e-02 1.35358959e-01 -4.09099013e-01 -1.47109896e-01
-2.05618635e-01 -6.68736398e-01 -1.03115082e+00 -1.71931469e+00
2.09396258e-01 -3.73772979e-01 5.69038242e-02 -1.00151360e-01
3.89059275e-01 -1.76080480e-01 6.48148209e-02 6.53151095e-01
-3.24822426e-01 2.80474782e-01 1.57143474e+00 -1.78704590e-01
-2.83219248e-01 -3.32603484e-01 -1.34702250e-01 5.07043481e-01
9.10497308e-01 -7.73398638e-01 1.61132127e-01 -1.17253982e-01
4.54482704e-01 7.59744525e-01 1.14631966e-01 -1.23844779e+00
-5.74080981e-02 -5.70509851e-01 4.69231140e-03 -2.31381550e-01
8.30619931e-02 -9.21945691e-01 1.15247059e+00 1.01971507e+00
9.50042345e-03 -1.43238947e-01 -3.15349065e-02 2.60287166e-01
1.23490676e-01 -6.36830509e-01 1.02363932e+00 -5.70567250e-01
-1.44607961e-01 -2.48690367e-01 -1.11072540e+00 -5.44803202e-01
1.19392228e+00 -1.02495885e+00 -4.98531103e-01 3.33234817e-01
-1.21321392e+00 -1.65551543e-01 1.11971009e+00 -4.58217472e-01
1.63668588e-01 -1.44919097e+00 -4.09946144e-01 -2.98371136e-01
-7.35536456e-01 -6.59937978e-01 8.30499530e-01 1.43898296e+00
-1.31697118e+00 1.22363679e-01 -1.11189342e+00 -5.00313938e-01
-1.24724603e+00 4.89945471e-01 3.15161049e-01 -1.57309100e-01
-8.08988929e-01 1.74623102e-01 -2.33185887e-01 -1.60638466e-01
-7.54178524e-01 -2.45708063e-01 -4.74810153e-01 -5.43976903e-01
2.98997700e-01 1.11143589e+00 -4.22066450e-01 -4.64531600e-01
-6.70633018e-02 4.80064183e-01 -1.55599549e-01 4.34943661e-02
1.83746815e+00 -1.09812886e-01 -8.52947116e-01 2.65615225e-01
9.32532907e-01 3.55539411e-01 -9.09963191e-01 2.66974241e-01
-2.99915690e-02 -1.13105297e-01 -3.56066406e-01 -1.11964479e-01
-1.05933619e+00 9.54949707e-02 6.18319623e-02 6.37151837e-01
4.32466954e-01 5.60149439e-02 1.27610767e+00 1.68860093e-01
-6.45923242e-02 -1.68872869e+00 1.44341113e-02 3.02858800e-01
6.56268418e-01 -6.09981954e-01 3.49353611e-01 -6.94765210e-01
-3.57709855e-01 1.67994010e+00 1.72334924e-01 -8.65846753e-01
5.35349786e-01 1.78637013e-01 -3.97301108e-01 -5.77488303e-01
-8.04813445e-01 7.23332465e-01 -1.88747987e-01 2.18132854e-01
7.65313029e-01 5.98956943e-01 -7.41242886e-01 5.47854662e-01
-2.72953987e-01 1.20187171e-01 8.10951889e-01 1.36947513e+00
-9.52632070e-01 -9.97589767e-01 -7.68939912e-01 8.92834961e-01
-8.95187557e-01 3.74209017e-01 -4.64208983e-02 6.49829328e-01
-5.49785078e-01 6.98727131e-01 -3.89259565e-03 -3.74589860e-01
2.39640683e-01 -2.98600942e-01 8.42955947e-01 -6.45121792e-03
-3.97664458e-01 4.14929181e-01 6.74599111e-01 -2.18795121e-01
-1.06806278e+00 -7.49086380e-01 -1.45413828e+00 -8.33810449e-01
-4.43994969e-01 4.27988976e-01 7.33543336e-01 1.01600718e+00
-1.46001354e-01 3.68592471e-01 6.77335083e-01 -7.71377623e-01
-5.74111417e-02 -1.12895751e+00 -1.23283517e+00 1.63136914e-01
2.64799856e-02 -4.31311041e-01 -2.72158682e-01 -4.41389345e-02] | [5.1427998542785645, 4.792929649353027] |
acd8b2fd-ea87-4be7-ba75-b45759093a1f | model-agnostic-meta-learning-for-natural | 2303.02841 | null | https://arxiv.org/abs/2303.02841v2 | https://arxiv.org/pdf/2303.02841v2.pdf | Model-Agnostic Meta-Learning for Natural Language Understanding Tasks in Finance | Natural language understanding(NLU) is challenging for finance due to the lack of annotated data and the specialized language in that domain. As a result, researchers have proposed to use pre-trained language model and multi-task learning to learn robust representations. However, aggressive fine-tuning often causes over-fitting and multi-task learning may favor tasks with significantly larger amounts data, etc. To address these problems, in this paper, we investigate model-agnostic meta-learning algorithm(MAML) in low-resource financial NLU tasks. Our contribution includes: 1. we explore the performance of MAML method with multiple types of tasks: GLUE datasets, SNLI, Sci-Tail and Financial PhraseBank; 2. we study the performance of MAML method with multiple single-type tasks: a real scenario stock price prediction problem with twitter text data. Our models achieve the state-of-the-art performance according to the experimental results, which demonstrate that our method can adapt fast and well to low-resource situations. | ['Zhihan Li', 'Yuxuan He', 'Shaoling Chen', 'Bixing Yan'] | 2023-03-06 | null | null | null | null | ['stock-price-prediction'] | ['time-series'] | [-5.23102462e-01 -1.91669479e-01 -4.02513236e-01 -5.13489783e-01
-1.17841303e+00 -3.73545229e-01 6.26835167e-01 4.47028317e-02
-5.53715169e-01 8.98290336e-01 2.70392776e-01 -9.53121185e-02
1.61949411e-01 -7.76421607e-01 -8.18066895e-01 -1.50999576e-01
-3.84978838e-02 7.47800231e-01 1.41845390e-01 -4.07053977e-01
4.56871480e-01 -2.27896161e-02 -1.25081527e+00 5.83315611e-01
6.92152023e-01 1.05358648e+00 3.87778133e-01 4.19038564e-01
-8.55181813e-01 1.13975120e+00 -4.01383400e-01 -5.32135546e-01
3.77094567e-01 4.83057685e-02 -8.09945047e-01 -2.56346345e-01
9.27720591e-02 -1.86936110e-01 2.55626172e-01 8.56236339e-01
6.61639571e-01 1.19501606e-01 6.76593065e-01 -1.05803323e+00
-8.59818101e-01 1.00101101e+00 -9.25300658e-01 4.48307663e-01
-6.36163801e-02 -2.71506488e-01 1.21555793e+00 -1.07342505e+00
3.26690167e-01 1.42239845e+00 6.37686491e-01 6.52444124e-01
-7.16901362e-01 -1.07272136e+00 3.87492031e-01 2.71182120e-01
-9.76708055e-01 -3.59803557e-01 6.74790740e-01 -3.95501912e-01
1.29137421e+00 -3.90866756e-01 3.14395167e-02 1.27621818e+00
4.51901972e-01 9.94360864e-01 1.36464882e+00 -4.58251059e-01
1.34816423e-01 4.38461274e-01 4.39556479e-01 5.14177680e-01
2.27662116e-01 -1.93369895e-01 -7.77230799e-01 -2.37933427e-01
5.72178006e-01 -1.13346763e-02 2.69912511e-01 3.53963554e-01
-1.10838437e+00 1.09891653e+00 -1.84579551e-01 5.24057508e-01
-2.95726746e-01 -8.84033516e-02 6.72309160e-01 5.52768171e-01
1.15231812e+00 2.36802250e-01 -1.13945532e+00 -3.55147392e-01
-1.05702126e+00 3.65418792e-01 1.18116832e+00 8.98456275e-01
6.54934943e-01 1.84071317e-01 1.59658700e-01 1.05609441e+00
5.09038687e-01 1.37257323e-01 1.27753949e+00 -3.31952602e-01
9.20635521e-01 5.18988788e-01 2.28873678e-02 -5.70529044e-01
-6.58587098e-01 -5.53049624e-01 -8.23647559e-01 4.95186634e-02
4.64960694e-01 -4.00567263e-01 -5.86956024e-01 1.51431835e+00
1.22709163e-02 5.29276550e-01 4.68228608e-01 5.62606394e-01
9.83643472e-01 8.97231102e-01 2.87734061e-01 -2.88874328e-01
1.59228635e+00 -1.29358876e+00 -7.44220018e-01 -5.88441491e-01
8.41768146e-01 -9.67477381e-01 1.25927091e+00 3.29197705e-01
-1.08866858e+00 -7.18105614e-01 -8.24031949e-01 3.77650894e-02
-5.62109709e-01 -1.50812775e-01 5.76179862e-01 5.70873678e-01
-4.49235022e-01 4.94113773e-01 -5.45622528e-01 1.65304374e-02
2.70030469e-01 -5.48075363e-02 -4.89887372e-02 2.28010491e-01
-1.56093228e+00 8.72935534e-01 7.50999033e-01 -1.72508687e-01
-4.79346633e-01 -9.91493165e-01 -5.94926596e-01 7.02514201e-02
3.41615528e-01 -5.45774579e-01 1.48230422e+00 -1.01566470e+00
-1.56906307e+00 9.68067288e-01 1.12350816e-02 -8.09167445e-01
6.94049418e-01 -7.12284684e-01 -4.02573943e-01 -4.62122113e-01
4.52059098e-02 3.48889381e-01 7.51338661e-01 -9.22501862e-01
-7.92868674e-01 -9.26073641e-02 -2.54374146e-01 -4.44126315e-02
-4.46776837e-01 3.20483357e-01 -1.06547117e-01 -1.10224831e+00
-3.37494403e-01 -4.12868798e-01 -2.17086300e-01 -8.21258664e-01
1.41620385e-02 -6.70145094e-01 6.39152646e-01 -7.18874991e-01
1.46922874e+00 -1.87780964e+00 -2.74506569e-01 -1.45598859e-01
-1.29477099e-01 2.50778198e-01 -1.99626297e-01 3.28659862e-01
-1.68044224e-01 3.21231544e-01 -3.15672010e-02 -6.89969718e-01
1.45803049e-01 2.10054949e-01 -5.23926497e-01 -1.27865613e-01
6.73314780e-02 1.06387818e+00 -6.44814789e-01 -6.67514682e-01
-2.24417314e-01 -7.60917664e-02 -3.87174338e-01 4.76174891e-01
-6.57837749e-01 2.31207594e-01 -8.29786420e-01 9.60541070e-01
6.25981033e-01 -4.05650437e-01 -1.90317303e-01 -4.81906384e-02
-1.12928040e-01 2.01430291e-01 -1.28732979e+00 1.65977216e+00
-8.06655467e-01 1.45147726e-01 -2.33664438e-01 -1.16894746e+00
1.17567813e+00 3.74887288e-01 5.32954335e-01 -7.77189970e-01
8.19618255e-02 5.60268939e-01 -1.19943529e-01 -7.05660880e-01
2.01703042e-01 -3.83037895e-01 -1.60951108e-01 8.51679146e-01
-4.17195112e-02 3.48340333e-01 8.02255496e-02 -1.21279366e-01
7.15413392e-01 3.16518128e-01 5.41656137e-01 -4.63544488e-01
5.28500915e-01 -1.68327957e-01 6.90963149e-01 8.32970917e-01
-1.75791889e-01 2.75047719e-01 1.86274111e-01 -8.59655976e-01
-8.97778571e-01 -5.77090204e-01 -2.69696981e-01 1.62341714e+00
-6.13483369e-01 -2.15061039e-01 -6.12033963e-01 -6.57241404e-01
2.28106201e-01 9.68762636e-01 -3.32338959e-01 2.31345624e-01
-6.52528226e-01 -1.53226995e+00 3.30289304e-01 3.66741478e-01
8.81085694e-01 -1.26110077e+00 -4.84685063e-01 7.63651133e-01
-9.49967429e-02 -1.32690072e+00 -2.91699678e-01 -2.86780279e-02
-1.06951308e+00 -9.26110566e-01 -8.12422514e-01 -7.01896787e-01
2.66511049e-02 -9.74541530e-02 1.47748172e+00 -2.89812297e-01
5.78204468e-02 1.83873683e-01 -5.49024224e-01 -8.39186966e-01
-3.44376236e-01 4.93527621e-01 -1.35207865e-02 1.56622157e-01
6.76303566e-01 -4.11867887e-01 -1.92305788e-01 1.51943475e-01
-9.48731780e-01 -1.05651170e-02 6.08603120e-01 8.48903656e-01
4.59759772e-01 -1.77602265e-02 1.33852601e+00 -1.35033286e+00
9.36904907e-01 -9.12408233e-01 -5.83158135e-01 6.46578193e-01
-7.37583041e-01 1.78186193e-01 6.08132482e-01 -5.83443999e-01
-1.31328607e+00 -3.60575974e-01 3.83230150e-02 -2.11500779e-01
1.70728058e-01 1.05585301e+00 -8.04860219e-02 2.24995673e-01
5.24434924e-01 1.40455604e-01 -2.75505990e-01 -1.01710713e+00
2.23338693e-01 9.17435408e-01 1.13173008e-01 -8.06083381e-01
5.59400141e-01 1.66624367e-01 -2.18387097e-01 -4.62023318e-01
-1.57733428e+00 -3.37509811e-01 -8.07145059e-01 3.43011245e-02
5.60761690e-01 -1.24002135e+00 -3.66301179e-01 5.72443128e-01
-1.02479839e+00 -4.02065724e-01 1.15739122e-01 5.65729678e-01
-4.83117402e-01 3.51476490e-01 -7.40518570e-01 -9.47050035e-01
-8.78478706e-01 -8.68773580e-01 7.64468491e-01 5.59487753e-02
-1.23374701e-01 -1.41456461e+00 1.82958707e-01 5.47218442e-01
6.18786395e-01 1.54680341e-01 9.46560323e-01 -1.26988423e+00
-3.32251519e-01 -5.33735305e-02 -2.60505736e-01 3.91776443e-01
9.04640034e-02 -3.45012605e-01 -1.12465775e+00 -2.92582095e-01
2.97661841e-01 -8.90862107e-01 1.09220898e+00 3.82208019e-01
1.19878209e+00 -4.71126974e-01 9.55030397e-02 4.15258110e-01
1.31968582e+00 -9.66997519e-02 1.62893042e-01 1.01832211e+00
3.98347557e-01 6.56942368e-01 7.86414504e-01 8.29584837e-01
6.84529245e-01 3.19358230e-01 -6.21829703e-02 1.83969349e-01
3.25408012e-01 -8.07959214e-02 6.53628707e-01 1.14464664e+00
-1.04231521e-01 -1.76018961e-02 -1.08215630e+00 2.17512280e-01
-2.40867710e+00 -8.45000803e-01 1.19433403e-01 1.83794105e+00
1.11204040e+00 4.66744781e-01 1.17705658e-01 -3.98446739e-01
3.67480904e-01 1.82502642e-01 -6.16485178e-01 -3.26422006e-01
-3.09834152e-01 1.51786983e-01 6.25982404e-01 3.26776713e-01
-1.31164038e+00 1.29082382e+00 6.16404009e+00 1.23615563e+00
-8.83085251e-01 6.35551810e-01 9.77765501e-01 -1.72282323e-01
-2.05480576e-01 -3.69638801e-01 -1.36420739e+00 6.07081532e-01
1.42821419e+00 -5.38382292e-01 1.32187784e-01 9.73624289e-01
2.66476899e-01 2.09879234e-01 -1.01226807e+00 1.01174784e+00
9.57746059e-02 -1.37783492e+00 2.91286498e-01 -1.58734217e-01
7.68102765e-01 4.14331138e-01 -5.84972762e-02 8.25677276e-01
2.94300795e-01 -9.27939653e-01 7.89444685e-01 7.65368819e-01
3.10953647e-01 -7.01463938e-01 9.76687789e-01 8.16440880e-01
-1.29449463e+00 -8.27049762e-02 -7.70939171e-01 -1.37205184e-01
2.34935582e-01 6.38168335e-01 -5.63886881e-01 4.94576931e-01
7.24028826e-01 1.00692201e+00 -4.94218886e-01 7.46722758e-01
1.91291109e-01 7.48645127e-01 -2.67056257e-01 -1.47573799e-01
5.20016015e-01 -1.69160157e-01 1.30267173e-01 1.42364132e+00
4.19351429e-01 1.47064790e-01 5.44945419e-01 6.35524869e-01
-2.99038917e-01 6.56235933e-01 -4.45608348e-01 -1.50837094e-01
3.03049564e-01 1.11928630e+00 -2.78701663e-01 -5.52723944e-01
-9.49378967e-01 5.29273033e-01 6.19531691e-01 1.29569799e-01
-6.43327057e-01 1.72841951e-01 4.01005447e-01 3.36190709e-03
-1.32112011e-01 -3.07404339e-01 -4.41808850e-01 -1.51685357e+00
-1.60627544e-01 -7.88062632e-01 9.07866180e-01 -4.04247344e-01
-1.93991482e+00 5.68422914e-01 2.07186460e-01 -1.22082186e+00
-5.51879585e-01 -7.96265841e-01 -7.17567146e-01 8.29972804e-01
-2.37395334e+00 -1.31836188e+00 2.52003372e-01 5.43315351e-01
1.12650156e+00 -1.13294864e+00 7.58097947e-01 4.45078194e-01
-6.64370954e-01 4.66906697e-01 3.34769547e-01 2.33110756e-01
1.00481784e+00 -1.08731008e+00 5.12882769e-01 3.54568839e-01
4.22056317e-02 3.93761516e-01 3.67665231e-01 -6.48162127e-01
-1.02699685e+00 -1.11263657e+00 1.12761772e+00 -4.10970092e-01
1.27174759e+00 -2.13202313e-01 -1.23434651e+00 8.10950935e-01
3.02061677e-01 -1.59809917e-01 9.71898735e-01 1.80943042e-01
-4.65020627e-01 -1.41024441e-01 -1.05772805e+00 2.13599175e-01
5.60801506e-01 -2.17827827e-01 -8.71973455e-01 5.53649664e-01
6.86008573e-01 -2.24518701e-01 -9.80130076e-01 2.58115262e-01
4.41948533e-01 -6.55448318e-01 9.63045239e-01 -1.03717124e+00
5.38018644e-01 3.40632409e-01 -1.94090515e-01 -1.12386751e+00
-2.38383740e-01 -4.50387359e-01 -2.63666034e-01 1.47617579e+00
7.68910170e-01 -8.43007147e-01 6.56181991e-01 7.94784427e-01
1.41108721e-01 -7.35504746e-01 -1.18321919e+00 -7.94209898e-01
4.89471942e-01 -6.09562159e-01 6.09769940e-01 1.09894502e+00
-5.31082973e-02 5.14930308e-01 -7.11450696e-01 -1.38392314e-01
7.62700617e-01 3.31072062e-01 5.87271929e-01 -1.37739873e+00
-5.15461087e-01 -4.89536464e-01 3.30987394e-01 -7.35165298e-01
7.15193629e-01 -7.59170473e-01 -5.34671366e-01 -1.06305993e+00
3.03860784e-01 -3.79409492e-01 -6.31300926e-01 5.92226982e-01
-2.24272415e-01 -3.81620914e-01 2.82130301e-01 6.85292661e-01
-5.59050024e-01 4.71148789e-01 9.82088804e-01 -2.16151655e-01
-1.21548213e-01 2.54254818e-01 -6.25601232e-01 8.75458598e-01
1.14953375e+00 -4.98580933e-01 -8.90264884e-02 -3.83521050e-01
4.63578820e-01 -2.04139858e-01 -9.91342142e-02 -5.08138597e-01
3.26671489e-02 -2.52215594e-01 2.35170826e-01 -5.24066567e-01
8.36787745e-02 -7.10146368e-01 -3.45742017e-01 3.50929409e-01
-3.32583159e-01 3.47215056e-01 3.15723568e-01 4.45012033e-01
-3.70196283e-01 -8.09318244e-01 7.82976389e-01 -8.48519564e-01
-9.78945136e-01 3.10122937e-01 -8.63506868e-02 3.91312659e-01
8.61347079e-01 9.45006534e-02 -2.31072724e-01 -2.14118615e-01
-5.17023683e-01 4.43000287e-01 -2.86251098e-01 6.53971553e-01
4.14271712e-01 -1.19621456e+00 -1.20205045e+00 1.10654101e-01
9.79150366e-03 -1.78685915e-02 9.11407322e-02 4.29185241e-01
-3.06659609e-01 4.76331204e-01 -8.83184001e-02 -9.57640484e-02
-8.85680556e-01 3.24284911e-01 3.16026896e-01 -8.45610619e-01
-4.32095319e-01 8.95986617e-01 1.96826294e-01 -4.43474174e-01
2.68202901e-01 -3.70437920e-01 -6.07973933e-01 5.28634071e-01
7.54359901e-01 1.39027581e-01 -1.25525901e-02 -3.60999167e-01
-5.86834326e-02 6.05858207e-01 -5.56213856e-01 -2.16752253e-02
1.56214988e+00 -7.61049017e-02 -3.71740684e-02 8.63243163e-01
1.12952566e+00 -3.81496072e-01 -1.09386909e+00 -7.58844078e-01
7.84111381e-01 -2.65347898e-01 1.44960538e-01 -6.30865455e-01
-9.51676250e-01 8.17114294e-01 4.31403160e-01 6.68571740e-02
7.74530530e-01 -1.01207405e-01 1.11001253e+00 6.55097067e-01
4.78968918e-01 -1.54647207e+00 3.13688636e-01 8.04913461e-01
9.04447556e-01 -1.64755177e+00 -8.41725767e-02 6.28929138e-02
-9.01373863e-01 1.39243186e+00 6.41139328e-01 -9.46933851e-02
1.20118833e+00 2.97170520e-01 4.05429035e-01 2.93624997e-02
-1.11339331e+00 1.75730214e-02 1.67692840e-01 8.03823322e-02
7.53708601e-01 -2.49714404e-01 -2.53222644e-01 1.20385981e+00
-3.22183907e-01 1.61157638e-01 4.45395112e-01 9.13585305e-01
-7.47703195e-01 -1.33224928e+00 -4.01420623e-01 4.47501987e-01
-8.98136735e-01 -2.68228263e-01 -1.23313695e-01 8.12584817e-01
-1.55918896e-01 7.81024516e-01 5.80116063e-02 1.19584724e-01
4.46887612e-01 5.94206691e-01 -3.50327441e-03 -7.07865596e-01
-8.36448967e-01 3.73270541e-01 1.19380131e-01 -2.68812001e-01
-6.74671471e-01 -7.06409097e-01 -1.16167951e+00 -1.63969010e-01
3.64048295e-02 1.84089988e-01 3.50161761e-01 1.19232762e+00
2.97719508e-01 3.88384134e-01 6.47047043e-01 -5.70059597e-01
-1.09391105e+00 -1.34831214e+00 -7.54468322e-01 3.84394825e-01
5.61422259e-02 -5.91059804e-01 -2.05583349e-01 1.49903432e-01] | [10.74565315246582, 8.243820190429688] |
e3c88798-45d4-4b5e-83c9-6d02f512c0cf | modeling-temporal-dependencies-in-high | 1206.6392 | null | http://arxiv.org/abs/1206.6392v1 | http://arxiv.org/pdf/1206.6392v1.pdf | Modeling Temporal Dependencies in High-Dimensional Sequences: Application to Polyphonic Music Generation and Transcription | We investigate the problem of modeling symbolic sequences of polyphonic music
in a completely general piano-roll representation. We introduce a probabilistic
model based on distribution estimators conditioned on a recurrent neural
network that is able to discover temporal dependencies in high-dimensional
sequences. Our approach outperforms many traditional models of polyphonic music
on a variety of realistic datasets. We show how our musical language model can
serve as a symbolic prior to improve the accuracy of polyphonic transcription. | ['Pascal Vincent', 'Nicolas Boulanger-Lewandowski', 'Yoshua Bengio'] | 2012-06-27 | null | null | null | null | ['music-modeling'] | ['music'] | [ 2.81224549e-01 -2.48500586e-01 -2.64535815e-01 -9.23789814e-02
-8.96736979e-01 -7.30436862e-01 5.81153393e-01 -4.29996163e-01
-6.35205163e-03 7.51704335e-01 4.59691465e-01 -3.64130288e-02
-3.58238816e-01 -3.74786228e-01 -8.03238869e-01 -6.32752955e-01
-3.47180963e-01 6.17374897e-01 -6.75057620e-02 -1.27992453e-02
1.88504875e-01 4.10265833e-01 -1.40440762e+00 4.63101029e-01
5.86348400e-02 7.10795462e-01 3.08428824e-01 1.37499774e+00
1.99603915e-01 9.52367604e-01 -8.44082773e-01 -4.13916409e-02
1.07813023e-01 -8.36558700e-01 -3.53367329e-01 -3.45950335e-01
-6.20705411e-02 4.54261201e-03 -5.20504951e-01 9.99226987e-01
4.49337423e-01 1.79205626e-01 1.06442785e+00 -6.98775947e-01
-2.42427096e-01 1.36446679e+00 -1.35947019e-01 7.27299824e-02
3.15909594e-01 -3.82838547e-01 1.23712194e+00 -6.79308116e-01
3.98764670e-01 1.24504042e+00 1.10624194e+00 8.73644799e-02
-1.54421151e+00 -6.26837552e-01 -3.87689382e-01 4.53654453e-02
-1.66401553e+00 -3.77947539e-01 1.00694799e+00 -4.36207950e-01
7.46043026e-01 1.98373243e-01 8.02509785e-01 1.38895226e+00
2.45602652e-01 1.02065289e+00 6.67486548e-01 -6.52978480e-01
1.36113718e-01 -3.98817509e-01 -1.70223221e-01 2.51069427e-01
-2.10873589e-01 3.62542957e-01 -1.05677807e+00 -7.02317595e-01
1.17994547e+00 -1.52592167e-01 3.09024043e-02 -1.04716502e-01
-1.13033831e+00 6.99485123e-01 -3.37819010e-01 4.45621967e-01
-5.06841838e-01 7.87120521e-01 4.76416558e-01 2.39523590e-01
1.80710867e-01 6.24078870e-01 -5.98836303e-01 -8.77305567e-01
-1.42233038e+00 2.51142204e-01 1.23125994e+00 7.74767756e-01
4.73252721e-02 4.26243603e-01 -9.01865065e-02 9.35771525e-01
1.47693515e-01 6.35680974e-01 5.96355259e-01 -1.27834582e+00
9.83274132e-02 -4.74207580e-01 2.80168504e-01 -7.30731666e-01
-5.41920923e-02 -6.55297160e-01 -7.13740408e-01 -2.50861347e-01
3.23684722e-01 -8.80306289e-02 -4.86880362e-01 1.68806267e+00
-5.74506044e-01 9.36245203e-01 3.18304189e-02 4.19721901e-01
9.17012841e-02 8.09815347e-01 -3.13286752e-01 -4.36170906e-01
7.03175962e-01 -5.32501280e-01 -8.80406976e-01 3.02186519e-01
3.28008160e-02 -1.09250045e+00 7.84025073e-01 1.00884140e+00
-1.07257807e+00 -5.78783631e-01 -9.29272175e-01 2.59820849e-01
3.03137809e-01 5.06042302e-01 4.64949578e-01 5.95295906e-01
-7.48732388e-01 1.37553406e+00 -8.96697521e-01 2.21752793e-01
-1.25759527e-01 2.80254364e-01 3.22676122e-01 4.51223344e-01
-1.20908761e+00 4.26603675e-01 6.19741023e-01 1.95009723e-01
-1.21343911e+00 -5.30720174e-01 -5.14015615e-01 -2.96950713e-03
-6.19433112e-02 -4.42800820e-01 1.76079249e+00 -7.43024111e-01
-1.89911020e+00 3.34875315e-01 -2.03070492e-01 -8.34068716e-01
1.85877696e-01 -5.87345004e-01 -5.79428792e-01 1.21960454e-01
-2.61771321e-01 7.02540874e-02 1.17805004e+00 -8.90370965e-01
-1.68367758e-01 2.13046819e-01 -5.18764019e-01 -9.29556228e-03
2.88313646e-02 -1.42605394e-01 -3.37547958e-01 -1.32073843e+00
-8.72775819e-03 -1.35595441e+00 -3.04669172e-01 -8.12425435e-01
-5.72458506e-01 3.95572297e-02 4.64373559e-01 -6.25777781e-01
1.52301323e+00 -2.30140209e+00 3.81469935e-01 4.13068414e-01
-5.43898821e-01 1.03650447e-02 -1.38583645e-01 7.25618422e-01
-2.11805239e-01 -1.53015882e-01 -2.05798268e-01 -4.30581272e-01
1.42212719e-01 4.23755795e-01 -1.09412646e+00 2.28387713e-01
-1.54201254e-01 8.01127195e-01 -7.59056807e-01 -4.43744808e-02
-1.89713892e-02 6.60776556e-01 -3.14657211e-01 -1.13318972e-02
-5.27691960e-01 4.85820442e-01 -2.03553379e-01 3.97876978e-01
-7.84430131e-02 -2.34374791e-01 3.66352975e-01 1.64586559e-01
7.42344037e-02 6.91687584e-01 -1.16068053e+00 2.12849331e+00
-4.90765750e-01 7.52114832e-01 -3.42419356e-01 -6.08668923e-01
1.17180359e+00 7.66379476e-01 4.39195871e-01 1.35009557e-01
1.99680090e-01 3.17209810e-01 -1.27424393e-03 -8.28819349e-02
5.54421127e-01 -5.59260666e-01 -2.26480722e-01 6.35679960e-01
1.79529190e-01 -3.95000935e-01 -5.57941273e-02 -5.45036383e-02
1.05261850e+00 3.37896764e-01 1.47430763e-01 3.61572788e-03
2.12342381e-01 -3.56874764e-01 4.70312476e-01 1.09174359e+00
3.85455489e-01 7.74260879e-01 4.26591128e-01 -2.20732585e-01
-1.06490409e+00 -1.09155059e+00 2.71617956e-02 1.06707084e+00
-4.33323056e-01 -7.56808162e-01 -1.91872865e-01 1.10501043e-01
-8.12198594e-02 1.00209486e+00 -4.04235601e-01 7.97370914e-03
-6.26289070e-01 -6.33402705e-01 1.02647436e+00 6.63161635e-01
-3.59434992e-01 -1.38728595e+00 -1.97579205e-01 6.26645267e-01
-1.10723741e-01 -9.80762303e-01 -4.32932675e-01 6.84150219e-01
-1.07244802e+00 -6.15488470e-01 -8.58212233e-01 -6.30400002e-01
-3.61446172e-01 -3.76700670e-01 1.23443007e+00 -7.42405236e-01
-2.67212778e-01 5.03930509e-01 -7.81937018e-02 -5.66348851e-01
-7.58112967e-01 9.57947895e-02 2.56160259e-01 -1.03464976e-01
1.80153117e-01 -1.30418980e+00 -1.39284208e-01 8.77659395e-02
-6.92470729e-01 -3.23203176e-01 4.56491023e-01 7.66014397e-01
6.82615995e-01 1.89392328e-01 4.51654285e-01 -7.82702267e-01
7.78139651e-01 -4.22189176e-01 -5.36428154e-01 -5.50193377e-02
-7.87586719e-02 3.26017767e-01 6.16688073e-01 -1.00899529e+00
-7.18801618e-01 4.69174325e-01 1.52607113e-01 -1.06432688e+00
1.73092008e-01 6.92863941e-01 3.90695781e-01 3.04968774e-01
7.31051803e-01 5.02999842e-01 -2.86056310e-01 -7.39891350e-01
5.34829140e-01 3.07673603e-01 1.05780590e+00 -8.44883025e-01
5.71512878e-01 1.91572592e-01 2.07464173e-01 -9.95256841e-01
-7.79861450e-01 -5.96554756e-01 -7.18977749e-01 3.52142043e-02
3.98170710e-01 -1.02858317e+00 -8.08648050e-01 1.88290581e-01
-1.32879364e+00 -4.50138211e-01 -5.74147105e-01 9.70952153e-01
-1.24808061e+00 3.54856461e-01 -1.10533333e+00 -1.39458072e+00
-1.66864581e-02 -4.53777224e-01 1.21309519e+00 -1.69315964e-01
-9.06145811e-01 -1.06071794e+00 8.60327363e-01 -4.43799943e-01
-1.32308900e-01 1.37000471e-01 8.29233706e-01 -4.56567734e-01
-5.85913062e-01 -3.19767237e-01 4.93914425e-01 4.35194790e-01
-1.69871841e-02 1.97563618e-01 -1.02806389e+00 1.45733684e-01
2.21243590e-01 -2.10304871e-01 9.42814708e-01 7.54003942e-01
1.13869381e+00 -8.64553377e-02 -5.27465641e-02 5.18762171e-01
1.15366411e+00 3.21989268e-01 3.95192355e-01 -1.88743770e-01
4.74665046e-01 5.96509576e-02 5.26331723e-01 7.99243867e-01
-3.97382140e-01 5.62668920e-01 -1.99652486e-03 6.81798220e-01
3.98185700e-02 -8.16685617e-01 6.07669413e-01 1.43653250e+00
-2.97105193e-01 -2.23810207e-02 -8.43766749e-01 7.83915401e-01
-2.03455782e+00 -1.21592999e+00 1.00107016e-02 2.10085869e+00
9.98334110e-01 8.88723955e-02 4.24391478e-01 4.95678395e-01
7.59206057e-01 1.35712415e-01 -4.49022621e-01 -3.85649055e-01
-2.26821736e-01 6.23793066e-01 4.68123734e-01 3.76040101e-01
-1.05382681e+00 6.66807234e-01 8.76930141e+00 8.59794915e-01
-7.67657518e-01 -5.62945083e-02 -5.39850816e-02 -2.12640390e-01
-2.44094893e-01 -7.99113214e-02 -7.62054622e-01 3.12747985e-01
1.44761670e+00 -2.01576799e-01 6.01585627e-01 8.25487316e-01
1.27792209e-01 3.39529634e-01 -1.30213356e+00 1.38895237e+00
1.42564654e-01 -1.44896388e+00 1.39477640e-01 -9.44916531e-02
9.43646848e-01 1.41136602e-01 3.17169398e-01 1.65564895e-01
6.61917806e-01 -1.29333007e+00 8.02890539e-01 1.07499206e+00
7.13741243e-01 -9.81509209e-01 3.10693294e-01 5.65360725e-01
-1.13222015e+00 2.64524091e-02 -2.40503266e-01 -2.43630007e-01
3.66713494e-01 6.00724757e-01 -1.14402640e+00 4.05472606e-01
2.70315498e-01 1.17557526e+00 6.77091256e-02 1.12059772e+00
-2.46834755e-01 1.10562718e+00 -5.42002678e-01 1.82088688e-02
1.48369253e-01 -1.60544783e-01 7.89003313e-01 1.37199044e+00
6.65333688e-01 9.37607363e-02 2.08136365e-02 8.35952759e-01
1.56383902e-01 -1.61321312e-01 -7.43169725e-01 -5.91947973e-01
4.26000714e-01 5.34189582e-01 -6.25073433e-01 -3.43894184e-01
2.13939905e-01 9.15197670e-01 -1.22378804e-01 3.75802696e-01
-6.98277235e-01 -1.74696147e-01 2.87380755e-01 -3.01068127e-01
8.06230605e-01 -8.24502885e-01 -2.56345212e-01 -1.12583077e+00
-3.54658544e-01 -8.64615560e-01 2.69542545e-01 -8.36626351e-01
-1.37102592e+00 4.95993316e-01 -1.50446668e-01 -1.37496424e+00
-1.14075541e+00 -3.94752175e-01 -4.80100870e-01 9.85377371e-01
-9.18321669e-01 -6.01223230e-01 6.60950899e-01 6.01823866e-01
4.80360836e-01 -4.35368508e-01 1.38911080e+00 -1.04881660e-03
1.58684969e-01 1.12467565e-01 4.84116584e-01 -1.57092733e-03
5.72598577e-01 -1.30598783e+00 4.63377774e-01 4.12634194e-01
1.22194505e+00 6.66053414e-01 1.05180025e+00 -5.87203383e-01
-1.07544434e+00 -9.03482318e-01 8.66451025e-01 -6.70542300e-01
9.44552958e-01 -2.27312207e-01 -8.20709050e-01 9.45959508e-01
-1.67265162e-01 -3.40675592e-01 9.26687837e-01 4.41984147e-01
-3.68961751e-01 2.91826963e-01 -2.32128993e-01 5.33842802e-01
7.98211753e-01 -1.24752104e+00 -9.35250282e-01 2.49928489e-01
6.28668249e-01 -1.84566259e-01 -8.35276425e-01 2.52556443e-01
9.16390061e-01 -6.93510473e-01 1.15557456e+00 -4.86809701e-01
2.44400829e-01 -2.96575993e-01 -4.45046484e-01 -1.08971131e+00
-2.58792728e-01 -1.41854978e+00 -5.72342515e-01 7.95602620e-01
4.00507420e-01 4.96029668e-02 7.07524478e-01 -1.36231467e-01
2.51101136e-01 -6.94094449e-02 -8.66974831e-01 -1.17671335e+00
-2.53679216e-01 -1.04373074e+00 3.24085444e-01 6.70847535e-01
-1.39700785e-01 3.19856882e-01 -1.10476959e+00 1.66643143e-01
6.14584148e-01 3.00025314e-01 6.42821372e-01 -1.34980679e+00
-1.20839739e+00 -3.98267716e-01 -3.12383503e-01 -1.23877513e+00
3.59318912e-01 -8.65370572e-01 3.01541001e-01 -6.18018329e-01
-1.55021563e-01 -2.09347829e-01 -7.09513724e-01 3.98478545e-02
4.72656250e-01 4.51876849e-01 1.95696682e-01 4.34418052e-01
-4.25368130e-01 7.23108113e-01 7.38726377e-01 5.67245670e-02
-4.10038233e-01 5.16210377e-01 9.86603089e-03 1.08793402e+00
6.64596856e-01 -9.80171919e-01 -4.46415514e-01 1.12512246e-01
5.49923122e-01 7.47099996e-01 2.67796874e-01 -1.15112150e+00
2.58960277e-01 9.11327899e-02 3.87245417e-01 -1.00661826e+00
7.95719445e-01 -4.81904179e-01 5.55927515e-01 4.26714092e-01
-6.84443712e-01 -6.88844323e-02 2.51733303e-01 1.00224459e+00
-5.39150059e-01 -3.73205096e-01 2.18640253e-01 4.46383804e-02
-2.00184554e-01 -9.05009732e-03 -5.69779158e-01 -7.93673620e-02
2.24470526e-01 3.80394727e-01 5.68881273e-01 -8.34500253e-01
-1.34776163e+00 -7.16891348e-01 -2.74688266e-02 3.51784557e-01
2.68424839e-01 -1.59817815e+00 -5.08503199e-01 2.13279184e-02
-3.00226033e-01 -7.14969039e-01 -1.01715818e-01 3.27499539e-01
-3.52956235e-01 6.39473259e-01 6.43116161e-02 -8.01228225e-01
-1.20548034e+00 2.59696573e-01 1.39475718e-01 -3.75173211e-01
-6.05293274e-01 7.19896674e-01 -1.91335559e-01 -1.78358003e-01
3.91710639e-01 -5.57645917e-01 -2.79809032e-02 -1.90882191e-01
4.71518993e-01 1.12343937e-01 -5.22961497e-01 -4.03732628e-01
-1.53915972e-01 5.43536365e-01 4.01848167e-01 -9.02513087e-01
1.32375968e+00 2.65692547e-02 -6.47113249e-02 1.92906582e+00
8.41731191e-01 2.69821495e-01 -1.07397366e+00 -2.30201036e-01
2.93041885e-01 -1.27613798e-01 -3.25359374e-01 -7.46796131e-01
-2.47433513e-01 7.69939184e-01 1.80921033e-01 1.69671133e-01
8.38644445e-01 -1.10707611e-01 6.68868840e-01 7.29413450e-01
4.11149323e-01 -8.35911870e-01 -7.98435323e-03 9.27976668e-01
9.45496440e-01 -4.02000755e-01 -1.98974952e-01 -2.89652646e-02
-5.66745937e-01 1.45305538e+00 -6.61438465e-01 -6.84221089e-01
1.10673535e+00 5.21552086e-01 -8.11451823e-02 7.51812831e-02
-8.85182023e-01 4.64261808e-02 5.86448610e-01 2.47410789e-01
5.53928792e-01 3.39603812e-01 1.19233534e-01 9.47007060e-01
-8.76454115e-01 2.76634574e-01 5.23172200e-01 5.62669754e-01
-1.94327459e-01 -1.28927624e+00 -5.12937486e-01 -1.25547737e-01
-6.59557343e-01 -3.22786927e-01 -2.91711032e-01 4.97483194e-01
-1.20108582e-01 5.03549516e-01 -2.23446805e-02 -3.45260203e-01
-1.35001326e-02 6.27915025e-01 7.20784426e-01 -6.70268714e-01
-3.46606076e-01 8.27424586e-01 2.56950647e-01 -3.64122152e-01
-6.48432732e-01 -7.54160047e-01 -9.34162438e-01 4.72841524e-02
4.03164327e-02 4.55653906e-01 6.59304976e-01 6.85640037e-01
9.67226550e-02 6.09284461e-01 5.22023797e-01 -1.05515921e+00
-6.18144572e-01 -1.03340447e+00 -1.22781789e+00 8.40571895e-02
3.59281570e-01 -2.61658788e-01 -2.78227895e-01 3.39844048e-01] | [15.693521499633789, 5.593790531158447] |
2b264e3c-d5c9-41eb-9bec-7b111b4eb155 | datasets-and-models-for-authorship | 2011.07975 | null | https://arxiv.org/abs/2011.07975v1 | https://arxiv.org/pdf/2011.07975v1.pdf | Datasets and Models for Authorship Attribution on Italian Personal Writings | Existing research on Authorship Attribution (AA) focuses on texts for which a lot of data is available (e.g novels), mainly in English. We approach AA via Authorship Verification on short Italian texts in two novel datasets, and analyze the interaction between genre, topic, gender and length. Results show that AV is feasible even with little data, but more evidence helps. Gender and topic can be indicative clues, and if not controlled for, they might overtake more specific aspects of personal style. | ['Malvina Nissim', 'Albert Gatt', 'Gaetana Ruggiero'] | 2020-11-16 | null | null | null | null | ['authorship-verification'] | ['natural-language-processing'] | [-1.46164834e-01 2.32179426e-02 -7.30006218e-01 -2.36704618e-01
-7.40898252e-02 -1.00797617e+00 1.07552958e+00 6.08262300e-01
-7.31600523e-01 7.16456711e-01 6.99524522e-01 -3.25927168e-01
3.84669080e-02 -3.02890748e-01 -2.77881801e-01 -9.49667394e-02
2.30495840e-01 4.96900141e-01 -1.63866863e-01 -8.15098733e-02
8.81793678e-01 2.45241210e-01 -1.14228106e+00 -2.21642211e-01
8.89778435e-01 2.24429786e-01 -1.57215640e-01 6.48725867e-01
-3.06272149e-01 6.74389482e-01 -1.23629153e+00 -1.06153357e+00
-9.11915451e-02 -5.20068586e-01 -9.42582011e-01 1.60062879e-01
5.01356483e-01 4.88804244e-02 -1.60927713e-01 9.04330671e-01
4.11845475e-01 -1.50144458e-01 9.20108318e-01 -1.22167861e+00
-7.73747265e-01 1.12302208e+00 -5.78262687e-01 5.30358255e-01
4.00324941e-01 9.21377167e-02 1.06072068e+00 -5.25529921e-01
1.11627042e+00 1.42794609e+00 8.02113533e-01 3.87159914e-01
-1.47232568e+00 -9.59291935e-01 -2.91106980e-02 -4.18158285e-02
-1.15552592e+00 -4.65745330e-01 1.04726195e+00 -7.61469960e-01
3.52236360e-01 2.50483632e-01 5.46294332e-01 2.00446057e+00
2.21815497e-01 4.02364552e-01 1.58023643e+00 -6.70361459e-01
-2.24260405e-01 8.11544657e-01 5.05142152e-01 1.39672160e-01
7.14891791e-01 -1.12921461e-01 -9.23662961e-01 -2.83169031e-01
3.42703491e-01 -4.10475045e-01 -2.45333929e-02 3.36026013e-01
-1.48788047e+00 8.25357497e-01 -1.38550460e-01 6.88445985e-01
1.45046618e-02 -1.57030150e-01 8.87033582e-01 6.76113963e-01
5.31460941e-01 1.04904604e+00 -4.37543303e-01 -7.86021531e-01
-1.14518976e+00 5.54203212e-01 1.10943401e+00 8.08265984e-01
3.66535366e-01 -1.84554964e-01 -3.10343832e-01 7.91736782e-01
1.75009921e-01 4.63892460e-01 9.11455393e-01 -8.53931785e-01
5.68583786e-01 5.77996671e-01 2.87032008e-01 -1.24461389e+00
-5.27834058e-01 -2.69476324e-01 -3.30044568e-01 -2.72272587e-01
1.00115252e+00 -1.73404336e-01 -3.66082974e-02 1.72176635e+00
-1.68291807e-01 -5.91153324e-01 -3.34683388e-01 6.63782060e-01
7.84569979e-01 7.54647478e-02 2.51484573e-01 -5.69548905e-01
1.63247275e+00 -6.54793859e-01 -1.02569127e+00 -4.90373731e-01
7.70723522e-01 -1.19509947e+00 1.22449172e+00 4.47303027e-01
-8.62299621e-01 -5.72831690e-01 -7.12465346e-01 -2.70022660e-01
-5.21440148e-01 3.29294175e-01 5.07944345e-01 1.08054078e+00
-4.47518736e-01 7.80202925e-01 -3.36988270e-01 -5.83166599e-01
3.90106380e-01 -1.53214365e-01 -1.03303276e-01 5.53875864e-01
-1.06151533e+00 8.28097343e-01 1.47994637e-01 -3.87243032e-01
-1.21239379e-01 -5.35468280e-01 -4.09782261e-01 -5.11872411e-01
1.65599287e-01 -4.40610439e-01 9.97106373e-01 -1.42254102e+00
-1.36730897e+00 1.29009438e+00 -2.54824847e-01 -2.55189866e-01
9.23553467e-01 -2.22780228e-01 -6.37233198e-01 -9.92919132e-02
3.96721780e-01 4.60917085e-01 9.58036542e-01 -8.76461387e-01
-3.55889082e-01 -5.91429591e-01 -1.17413215e-01 -1.76641658e-01
-8.24013352e-01 8.27644348e-01 2.88689256e-01 -1.36092603e+00
-3.13374490e-01 -9.02194500e-01 3.16822380e-01 -1.73736811e-01
-7.15823829e-01 -6.67175829e-01 5.47493517e-01 -1.09427452e+00
1.60600877e+00 -2.11812782e+00 3.06089520e-01 1.01115972e-01
5.89320779e-01 -3.53968114e-01 5.36720514e-01 3.68445516e-01
2.49899209e-01 7.96503544e-01 2.25404322e-01 -6.70041680e-01
2.46781558e-01 -1.08332500e-01 -2.45691970e-01 8.31509292e-01
-1.18967310e-01 9.25665915e-01 -6.12244546e-01 -7.13550031e-01
-5.19993961e-01 2.90918536e-02 -1.67601496e-01 -5.57526201e-02
1.48253053e-01 4.22016591e-01 -3.80753689e-02 7.47355759e-01
3.83846819e-01 -2.15887964e-01 1.70281172e-01 2.59829313e-01
-6.10755861e-01 6.62714660e-01 -7.50584304e-01 1.20934069e+00
-1.21276807e-02 1.50306571e+00 -2.38235846e-01 -4.01549399e-01
8.01118314e-01 -5.48920743e-02 -3.73429805e-01 -3.95540357e-01
5.60518444e-01 3.14265370e-01 4.38053697e-01 -3.68216395e-01
8.35492909e-01 -1.40821308e-01 -3.37269992e-01 7.28577256e-01
-2.36963287e-01 2.40518004e-01 4.40525204e-01 3.47374856e-01
6.39152825e-01 -1.91881791e-01 3.60090703e-01 -4.70977396e-01
1.04616679e-01 -6.85287714e-02 6.12299740e-01 9.83542860e-01
-3.03299129e-01 4.18029487e-01 9.69157398e-01 8.30365419e-02
-1.27948713e+00 -4.33777690e-01 -7.09090948e-01 1.07343662e+00
-5.21439239e-02 -8.73282254e-01 -6.37323320e-01 -7.24685490e-01
2.36825123e-01 8.30410898e-01 -1.11789334e+00 1.11392751e-01
-5.10803759e-01 -5.73765755e-01 1.07984865e+00 5.81347287e-01
1.37657925e-01 -9.94110763e-01 -7.32832313e-01 -3.36761847e-02
-1.56433448e-01 -1.02593899e+00 -4.49365050e-01 -1.62220746e-01
-9.50647414e-01 -9.76416945e-01 -7.14600682e-01 -1.94239333e-01
4.01844203e-01 -1.67899922e-01 1.11938000e+00 2.22028494e-01
-2.78479047e-02 1.28979698e-01 -3.26978564e-01 -8.22502911e-01
-6.99501812e-01 4.26743984e-01 1.22270763e-01 -2.33123332e-01
4.40280378e-01 -3.23203772e-01 1.07446916e-01 2.41304755e-01
-3.52428138e-01 -3.34100991e-01 4.45319384e-01 1.03290689e+00
-1.53209776e-01 -3.84640753e-01 4.48423535e-01 -1.38120842e+00
7.40838885e-01 -6.22349143e-01 -6.16851039e-02 -1.10138297e-01
-1.00403965e+00 -3.72041970e-01 5.94125390e-01 -8.59234214e-01
-7.52136469e-01 -9.27650273e-01 4.33878064e-01 -1.35790139e-01
-4.78030711e-01 5.18730402e-01 6.53643683e-02 3.92458960e-02
7.50015259e-01 -3.18302304e-01 6.09064512e-02 -8.09572279e-01
-2.59697050e-01 8.85908663e-01 1.56573072e-01 -7.29490221e-01
8.75754476e-01 2.06166789e-01 -2.74339914e-01 -9.41725850e-01
-8.03684473e-01 -1.31365925e-01 -9.12266016e-01 -7.54091963e-02
3.57861638e-01 -7.23816216e-01 -7.36788154e-01 3.69104803e-01
-1.08011138e+00 -1.87059686e-01 -1.86034963e-01 5.95096350e-01
-1.57329351e-01 4.45984423e-01 -5.82201004e-01 -6.99291766e-01
-8.09524208e-02 -7.78066158e-01 6.46794200e-01 2.04461232e-01
-1.35898042e+00 -1.37100637e+00 -1.14994645e-01 3.17942142e-01
1.68438017e-01 2.74218798e-01 8.27822566e-01 -1.07297361e+00
6.18660077e-02 -2.49622509e-01 -7.37509280e-02 -1.80152938e-01
-2.89955381e-02 4.40833598e-01 -7.63030410e-01 -9.62948576e-02
-2.35517919e-01 -4.57597017e-01 6.60610795e-01 -1.54681532e-02
1.15855753e+00 -7.79299915e-01 -4.56489414e-01 4.46026742e-01
8.48003566e-01 -3.35965782e-01 1.05180457e-01 9.13622677e-01
8.52994263e-01 8.20153773e-01 4.81092721e-01 7.19588280e-01
6.09442294e-01 8.53785992e-01 -3.61098915e-01 3.67063642e-01
-1.19209707e-01 -2.33933955e-01 3.70879352e-01 7.25662410e-01
-2.01358050e-01 -3.15522105e-01 -1.02216864e+00 4.57456976e-01
-1.41750693e+00 -9.32473004e-01 -6.01039350e-01 1.91719830e+00
9.66062129e-01 3.37928176e-01 6.93110883e-01 5.13746142e-01
5.56543410e-01 2.22942770e-01 -1.71233192e-01 -7.15724587e-01
-7.50156462e-01 -2.02060238e-01 5.10951817e-01 1.30309418e-01
-7.92203188e-01 6.48036301e-01 7.29835176e+00 5.05590498e-01
-1.03510332e+00 -1.79797150e-02 5.86634517e-01 4.76556309e-02
-5.19654453e-01 1.04322866e-01 -9.25462544e-01 8.44053388e-01
1.12763667e+00 -3.99712682e-01 1.08072728e-01 5.88787973e-01
1.19085647e-01 -1.57396272e-01 -1.27415133e+00 8.89076173e-01
4.96278048e-01 -8.51546645e-01 -2.25590169e-01 3.39286566e-01
4.90291595e-01 -6.06330097e-01 1.38975739e-01 1.08024731e-01
-1.28266299e-02 -9.54384804e-01 1.29500401e+00 3.73753339e-01
5.84862113e-01 -6.10971570e-01 8.00641835e-01 2.82742888e-01
-4.27438170e-02 -2.31828421e-01 -1.57133028e-01 -3.69772136e-01
-9.64281484e-02 3.35568815e-01 -6.08900785e-01 8.05938095e-02
6.85700238e-01 1.10104561e+00 -1.41716194e+00 5.80302238e-01
-4.70301837e-01 1.21738923e+00 -4.15436178e-02 -2.22852498e-01
-1.41004696e-01 7.50784576e-02 1.01332903e+00 1.35682726e+00
6.01566955e-02 -9.09916013e-02 -1.03526264e-01 8.83872151e-01
-4.00959283e-01 5.05243778e-01 -5.19169211e-01 -6.83309972e-01
6.66986346e-01 1.18385887e+00 -1.12129021e+00 -3.54251415e-01
-3.48167747e-01 1.16395211e+00 4.88421261e-01 1.78428158e-01
-5.77921212e-01 -2.74438620e-01 3.98312062e-01 3.96646917e-01
-1.57255858e-01 -1.18062377e-01 -1.13828683e+00 -1.27066994e+00
-8.31949264e-02 -1.06785536e+00 5.14388680e-01 -4.69337285e-01
-1.53867102e+00 -5.89267910e-02 -1.32533982e-01 -7.33835161e-01
-3.54278684e-02 -4.96830076e-01 -5.25869548e-01 8.29664409e-01
-1.07898545e+00 -1.17739582e+00 -1.04769401e-01 1.58368751e-01
4.47523445e-01 -3.16852331e-01 6.60564423e-01 3.76171768e-02
-8.96436572e-01 1.36071599e+00 3.70712369e-03 2.10446298e-01
1.51633143e+00 -1.44945180e+00 2.07017004e-01 4.90139693e-01
6.10456280e-02 9.89164591e-01 1.09779954e+00 -9.04293120e-01
-1.08432400e+00 -4.63186145e-01 1.71784914e+00 -1.16748118e+00
1.02988565e+00 -4.02256459e-01 -7.33709455e-01 8.41643631e-01
3.59844774e-01 -6.50030971e-01 1.11466444e+00 7.83554852e-01
-6.52308583e-01 4.59598243e-01 -1.01878285e+00 6.57971680e-01
1.12256157e+00 -4.48024005e-01 -8.68024766e-01 2.61413962e-01
5.08236885e-01 -3.25225741e-01 -1.12451649e+00 -5.71543463e-02
8.73494029e-01 -7.27648556e-01 6.35706782e-01 -8.50794792e-01
8.78426731e-01 1.92461878e-01 1.74361691e-01 -1.29565477e+00
-4.56152141e-01 -7.75558949e-01 -6.09517060e-02 1.94388533e+00
5.34647226e-01 -7.00572193e-01 3.44416440e-01 7.60478377e-01
5.14329001e-02 -1.64345622e-01 -7.75475025e-01 -8.80808592e-01
4.04850245e-01 -2.94219982e-02 5.33001125e-01 1.71103978e+00
1.79300159e-01 5.05457580e-01 -3.11545461e-01 -4.32844639e-01
5.06472886e-01 2.47332662e-01 9.57818210e-01 -1.82385886e+00
-9.26614627e-02 -1.00150871e+00 -4.47087824e-01 -2.13423610e-01
7.17720211e-01 -6.60259306e-01 -6.22184932e-01 -8.34977567e-01
4.40729886e-01 -4.81901705e-01 2.30886504e-01 3.16989779e-01
-3.47540975e-01 3.05434197e-01 2.43632108e-01 6.04181230e-01
-3.82173568e-01 1.71476290e-01 1.05372274e+00 5.78402635e-03
-3.08279663e-01 -2.87292928e-01 -1.07572544e+00 7.40535259e-01
7.86808968e-01 -5.43959737e-01 2.70299703e-01 5.76443002e-02
4.23779130e-01 -3.04546237e-01 4.01493311e-01 -5.46438694e-01
2.91150641e-02 -3.20156574e-01 7.53052533e-01 -8.00279900e-02
-8.62472206e-02 -4.86162961e-01 -9.91686806e-02 7.81337395e-02
-7.31606901e-01 4.74205196e-01 1.24801539e-01 4.73418742e-01
-2.84230523e-02 -4.14243668e-01 3.32177341e-01 1.22634768e-01
-8.81500468e-02 -2.95532674e-01 -4.57913607e-01 5.75973749e-01
7.47668326e-01 -5.15946567e-01 -4.61679608e-01 -1.28839850e-01
-4.16955024e-01 -1.93077043e-01 1.08585703e+00 6.49943829e-01
-6.05804212e-02 -1.17319942e+00 -7.50152111e-01 -4.51231673e-02
3.64230216e-01 -9.85989571e-01 -1.86777130e-01 1.04785883e+00
-1.84488297e-01 1.48003966e-01 -2.07118019e-01 -1.56086504e-01
-1.33198071e+00 5.89010358e-01 -2.34583542e-01 1.93603814e-01
-6.84259295e-01 7.22765982e-01 -2.29202196e-01 -7.76771531e-02
6.82528168e-02 2.22064987e-01 -6.52890146e-01 1.06303763e+00
4.85642552e-01 7.99298465e-01 -4.67778593e-01 -1.11786401e+00
-3.42692316e-01 2.09435254e-01 -2.70124972e-01 -3.21757883e-01
9.72407699e-01 -5.31406403e-01 -4.06840712e-01 1.31132329e+00
8.45015228e-01 6.96122289e-01 -4.50997531e-01 -7.48980567e-02
2.96811610e-01 -8.78891468e-01 -1.86610579e-01 -7.99028039e-01
-4.08679843e-01 7.08517909e-01 1.56370312e-01 2.92758375e-01
3.11711818e-01 5.66748716e-02 4.45994735e-01 -3.72387841e-03
-5.46021806e-03 -1.49142969e+00 3.92758064e-02 4.02360916e-01
7.92009175e-01 -1.25407434e+00 3.31278533e-01 -2.64030725e-01
-6.66043282e-01 1.36124706e+00 7.49105692e-01 2.57688403e-01
4.97161299e-01 -7.89946020e-02 1.62333399e-02 -2.24124178e-01
-4.65061784e-01 3.82865638e-01 4.05365884e-01 2.38322839e-01
1.01728272e+00 2.32443154e-01 -1.07643235e+00 1.23495972e+00
-1.07628286e+00 -4.68403906e-01 1.05919957e+00 5.21162689e-01
1.36621311e-01 -1.16769564e+00 -6.55016124e-01 7.25557983e-01
-1.10306549e+00 -5.90902343e-02 -1.11922121e+00 8.96633685e-01
1.00478463e-01 7.12809563e-01 3.57950330e-01 -1.76997498e-01
2.66994089e-02 1.97923869e-01 4.52065289e-01 -3.85186285e-01
-1.13494146e+00 -4.70088758e-02 5.54225981e-01 2.49935277e-02
-4.45243835e-01 -1.43064249e+00 -4.54452902e-01 -9.70590472e-01
-3.26147228e-01 1.39550641e-01 4.19959486e-01 9.90166843e-01
1.23445340e-01 3.51601243e-01 4.86571223e-01 -5.38412511e-01
-2.57467747e-01 -1.24332738e+00 -7.49274850e-01 5.32990456e-01
2.36444667e-01 -7.80910432e-01 -7.74473071e-01 1.09622359e-01] | [9.579882621765137, 10.54582691192627] |
79101cac-c8a8-4799-96fa-4e5ccdfc1c23 | physics-informed-machine-learning-simulator | 2012.06825 | null | https://arxiv.org/abs/2012.06825v1 | https://arxiv.org/pdf/2012.06825v1.pdf | Physics-Informed Machine Learning Simulator for Wildfire Propagation | The aim of this work is to evaluate the feasibility of re-implementing some key parts of the widely used Weather Research and Forecasting WRF-SFIRE simulator by replacing its core differential equations numerical solvers with state-of-the-art physics-informed machine learning techniques to solve ODEs and PDEs, in order to transform it into a real-time simulator for wildfire spread prediction. The main programming language used is Julia, a compiled language which offers better perfomance than interpreted ones, providing Just in Time (JIT) compilation with different optimization levels. Moreover, Julia is particularly well suited for numerical computation and for the solution of complex physical models, both considering the syntax and the presence of some specific libraries such as DifferentialEquations.jl and ModellingToolkit.jl. | ['Simone Azeglio', 'Sara Tiengo', 'Martina Scauda', 'Valerio Pagliarino', 'Giovanni Graziano', 'Francesco Calisto', 'Luca Bottero'] | 2020-12-12 | null | null | null | null | ['physics-informed-machine-learning'] | ['graphs'] | [-6.56637132e-01 -3.86069804e-01 1.46218687e-01 1.43664032e-02
-1.19579442e-01 -5.54554760e-01 7.44735122e-01 3.53061706e-01
-5.35837591e-01 8.89838994e-01 -3.40492308e-01 -8.70516181e-01
-3.06852460e-01 -1.01532471e+00 -3.11571628e-01 -7.01560616e-01
-5.53868890e-01 7.96499550e-01 1.50006354e-01 -8.35879862e-01
-1.06279753e-01 1.13099372e+00 -1.56011271e+00 -9.71023217e-02
8.08747292e-01 5.90464175e-01 -2.21328698e-02 1.01443338e+00
-1.15017764e-01 7.40037143e-01 -3.42033118e-01 9.97341648e-02
2.63350308e-01 -1.95361719e-01 -5.24925888e-01 -4.86407280e-01
4.38133217e-02 -1.47679791e-01 -7.20571429e-02 4.69939470e-01
3.69915932e-01 3.03482622e-01 7.73908377e-01 -1.11301279e+00
4.06375706e-01 8.90619755e-02 -4.59638350e-02 3.77940774e-01
1.80177584e-01 5.27751207e-01 5.74258506e-01 -3.12074900e-01
3.97935271e-01 7.98762739e-01 9.58225012e-01 -6.08588569e-02
-1.26474679e+00 -3.32453072e-01 -1.54876530e-01 5.41700907e-02
-1.34204304e+00 -5.49502857e-03 1.74479306e-01 -9.37371612e-01
1.39157891e+00 7.28053987e-01 9.93304014e-01 5.30324936e-01
5.11307061e-01 8.28432888e-02 1.19245374e+00 -3.39571267e-01
7.55439103e-01 1.32185638e-01 4.42132689e-02 4.15925771e-01
1.84660226e-01 6.75501585e-01 -2.10536607e-02 -5.11169136e-01
6.53691053e-01 -3.79427642e-01 -4.04854476e-01 -2.55055539e-02
-8.50999594e-01 8.90537977e-01 4.41629678e-01 5.06271541e-01
-5.62392056e-01 1.01400204e-02 4.56387937e-01 2.87524551e-01
7.20494032e-01 6.83907211e-01 -9.39423084e-01 -2.56581575e-01
-1.25362527e+00 8.93775463e-01 1.24998808e+00 2.29829758e-01
7.02378750e-01 5.31115592e-01 3.65508318e-01 3.80278826e-01
4.17530000e-01 7.90634155e-01 3.14122349e-01 -8.90394032e-01
-1.29088625e-01 3.61554027e-01 2.27459684e-01 -9.10870373e-01
-7.52598166e-01 -5.75205207e-01 -9.62077737e-01 8.57385159e-01
1.64066061e-01 -6.06526852e-01 -4.22440886e-01 9.66220975e-01
4.58457381e-01 4.72164094e-01 3.52771521e-01 1.01042783e+00
6.87714458e-01 1.28608966e+00 2.23854825e-01 -1.87730312e-01
8.01416278e-01 -6.46542370e-01 -3.38113964e-01 -2.45523766e-01
9.46255505e-01 -5.70534527e-01 5.85827172e-01 3.55580419e-01
-7.39914775e-01 -5.14982462e-01 -8.56209457e-01 3.59760135e-01
-8.35890293e-01 2.54783574e-02 8.20954859e-01 4.32932973e-01
-1.10403180e+00 1.09164822e+00 -6.91293061e-01 -7.15812668e-02
-3.47283781e-01 3.12350355e-02 -1.32086009e-01 7.97247410e-01
-1.32509243e+00 1.39162016e+00 3.64726663e-01 3.46294641e-01
-6.14853263e-01 -1.28171766e+00 -7.99757600e-01 1.87038228e-01
-8.43771249e-02 -7.09837973e-01 1.24449253e+00 -7.16906726e-01
-1.77514851e+00 7.95805097e-01 1.97408885e-01 -5.94849944e-01
7.42665648e-01 -2.82133967e-01 -4.68737364e-01 -2.66486496e-01
-2.15258569e-01 2.04863742e-01 4.41289186e-01 -1.05413926e+00
-3.72089893e-01 5.02321543e-03 1.01399437e-01 2.75820717e-02
1.86937556e-01 3.63447033e-02 2.43571341e-01 -5.85094154e-01
-4.11829293e-01 -6.62044287e-01 -5.45199811e-01 1.49976969e-01
6.35229051e-02 2.17747957e-01 6.57174766e-01 -9.25628483e-01
9.80033994e-01 -1.65438771e+00 4.02201712e-01 3.10610920e-01
-2.16676965e-01 7.91676462e-01 2.47975051e-01 9.89304364e-01
-2.89829016e-01 -2.18202308e-01 -5.42062283e-01 -1.38101742e-01
-1.74206033e-01 6.32556140e-01 -5.31475484e-01 6.18691266e-01
-2.80432701e-02 2.33570293e-01 -6.28799081e-01 -1.42201170e-01
6.95657849e-01 5.44829786e-01 -2.49141410e-01 4.14594769e-01
-7.99513459e-01 7.12262392e-01 -3.27569216e-01 7.89163485e-02
8.48091602e-01 3.71545762e-01 2.62604076e-02 2.66949117e-01
-9.77854431e-01 1.59380119e-02 -1.57445264e+00 1.22261333e+00
-9.18627143e-01 2.41520703e-01 3.91245842e-01 -1.05860615e+00
9.44759130e-01 3.92518759e-01 4.34828758e-01 -4.12763149e-01
2.53478944e-01 1.77818909e-01 -2.52047896e-01 -6.49713218e-01
3.03120166e-01 -5.19179516e-02 3.55452269e-01 3.69807243e-01
-1.36957735e-01 -6.48001611e-01 2.55119056e-01 -1.32071897e-01
3.65107507e-01 5.02758324e-01 4.29490596e-01 -8.30938280e-01
9.42049205e-01 6.16661787e-01 2.66009331e-01 3.65574807e-01
3.63703400e-01 2.37204462e-01 4.17881489e-01 -8.77747893e-01
-9.41659987e-01 -6.22700751e-01 -2.86819696e-01 8.15839529e-01
-3.81896496e-01 -2.62709379e-01 -5.23933113e-01 2.55475342e-02
9.11899954e-02 9.15716469e-01 -2.85269409e-01 2.83117741e-01
-4.59907532e-01 -1.05430973e+00 6.28924966e-01 1.36841908e-01
4.43151444e-01 -9.94367361e-01 -1.07326984e+00 3.76937866e-01
4.52927589e-01 -8.48783135e-01 6.01938307e-01 3.40325296e-01
-1.15658760e+00 -9.45551932e-01 -5.58497310e-01 -1.31633148e-01
-1.19482420e-01 -2.65494466e-01 1.37203777e+00 2.85420179e-01
-3.67639601e-01 -5.39901406e-02 -4.13479358e-01 -3.52009237e-01
-7.81853378e-01 1.67235285e-01 -3.51620764e-01 -4.22695249e-01
-3.88057292e-01 -8.57974231e-01 -2.35298127e-01 -7.63826594e-02
-8.82629037e-01 6.96409270e-02 5.79781868e-02 3.54218930e-01
7.70588070e-02 5.01498841e-02 -4.95792069e-02 -6.20739698e-01
3.24145883e-01 -4.61285710e-01 -1.31739843e+00 3.80854346e-02
-4.24615681e-01 -4.20431374e-03 9.70830262e-01 8.22120607e-02
-9.85356092e-01 7.21959919e-02 -6.25325859e-01 1.25150993e-01
-5.96785843e-01 9.17751253e-01 2.58341283e-01 -3.34328294e-01
7.51176417e-01 2.46240124e-01 1.77877024e-01 -6.95399523e-01
2.40367562e-01 3.55541289e-01 4.51026767e-01 -5.04787803e-01
9.78607476e-01 2.99623668e-01 2.79182136e-01 -1.35438502e+00
-2.29204699e-01 -4.64027405e-01 -4.76348370e-01 -2.83292949e-01
4.89125282e-01 -8.30501914e-01 -7.57446945e-01 8.91334832e-01
-1.14728832e+00 -8.74346793e-01 -1.41733177e-02 4.29896325e-01
-2.32835084e-01 2.20043465e-01 -3.95790726e-01 -9.18530762e-01
-4.42310840e-01 -9.30862248e-01 7.57436156e-01 2.94050276e-01
-1.53557837e-01 -1.38946140e+00 9.38719213e-01 -2.79273838e-01
8.52133274e-01 4.08369184e-01 7.68445313e-01 -5.72198331e-02
-1.93443060e-01 -2.26978138e-01 8.12642928e-03 1.86840907e-01
-5.25218248e-01 6.06778920e-01 -1.01142299e+00 -1.00461759e-01
9.01824683e-02 2.06308156e-01 6.54924095e-01 2.61438698e-01
6.68941379e-01 -1.18502207e-01 -1.76213965e-01 9.61914837e-01
1.83624268e+00 -3.58506143e-02 4.95962262e-01 5.74316204e-01
5.40357940e-02 3.20307136e-01 3.12612414e-01 7.63477921e-01
2.08287746e-01 7.74852753e-01 5.89495838e-01 -5.70549846e-01
3.86885852e-01 3.42421830e-01 2.55178005e-01 1.62890568e-01
-3.47353697e-01 3.79329547e-02 -1.29643679e+00 1.35811955e-01
-1.87087941e+00 -8.89908016e-01 -8.07113647e-01 2.02096057e+00
7.57806718e-01 -1.82010368e-01 -1.20117068e-02 1.72505915e-01
-3.20969485e-02 3.77838552e-01 1.68794766e-01 -8.44629765e-01
1.64343029e-01 6.45428240e-01 7.52515078e-01 8.91807258e-01
-1.10105419e+00 6.50039434e-01 6.52497339e+00 4.20794636e-01
-1.59250855e+00 9.63182598e-02 4.88878116e-02 3.53599876e-01
1.18581839e-01 2.02604502e-01 -6.77717328e-01 1.43558145e-01
1.36898875e+00 1.72714993e-01 5.82139373e-01 9.60922599e-01
8.27689111e-01 -5.85012078e-01 -4.58591968e-01 2.35615909e-01
-3.88359010e-01 -1.63079023e+00 -2.97992229e-01 -2.11609736e-01
3.49419206e-01 3.44406545e-01 -6.10948980e-01 3.98943126e-01
9.03993919e-02 -8.71520817e-01 6.38090372e-01 6.67853057e-01
1.09730318e-01 -5.82666278e-01 1.06899500e+00 6.83373392e-01
-1.09264624e+00 2.67424881e-01 -3.91077638e-01 -7.36320376e-01
3.87442917e-01 1.03946662e+00 -3.13336402e-01 9.12078321e-01
7.86432683e-01 4.10045475e-01 -4.13558632e-01 1.33991754e+00
6.60840794e-02 6.32872820e-01 -7.36353934e-01 1.12420529e-01
4.43977267e-01 -5.54343879e-01 5.91070652e-01 1.56444335e+00
3.32607359e-01 3.28289978e-02 2.25953013e-01 6.48755610e-01
9.32967603e-01 3.31525087e-01 -4.00344759e-01 3.16406518e-01
-1.66548312e-01 1.25975490e+00 -4.00107443e-01 -3.49604577e-01
-2.55109034e-02 5.39412856e-01 -1.02074049e-01 3.38830233e-01
-1.07161868e+00 -1.08800195e-01 1.04067016e+00 3.01100463e-01
1.12970896e-01 -6.17738485e-01 -2.91154861e-01 -1.04198384e+00
-4.33667064e-01 -6.46521270e-01 2.23522991e-01 -1.02444267e+00
-1.00240803e+00 6.56860471e-01 4.66998428e-01 -8.37696970e-01
-3.39618623e-01 -9.85279441e-01 -1.02813649e+00 1.31313419e+00
-1.77761805e+00 -8.97446811e-01 -3.59207302e-01 2.31719881e-01
-5.41762002e-02 -1.72298566e-01 1.33898497e+00 1.34631470e-01
-4.34192181e-01 -3.27022403e-01 5.33993006e-01 -2.57722676e-01
1.98407933e-01 -1.28921151e+00 4.55919147e-01 6.75142765e-01
-2.25133494e-01 6.62535653e-02 1.05596411e+00 -5.12374997e-01
-1.42134249e+00 -8.87534320e-01 1.09168887e+00 3.86061877e-01
9.38107133e-01 -5.81836626e-02 -1.08726394e+00 2.48692408e-01
2.85903215e-01 7.44968057e-02 1.63577780e-01 1.08636944e-02
-3.65415812e-02 -2.57737875e-01 -9.31951642e-01 4.71214205e-01
1.34946734e-01 -1.98595241e-01 -4.93891835e-01 6.36342943e-01
-3.46844867e-02 -7.51231611e-01 -9.28267658e-01 4.48511392e-01
2.74383575e-01 -1.24927318e+00 1.03414536e+00 -3.20792288e-01
3.07091594e-01 -4.69780296e-01 2.46316344e-01 -1.36298347e+00
-4.98472489e-02 -5.48365235e-01 -4.59470786e-02 8.59831989e-01
3.37367415e-01 -1.00925314e+00 3.10233325e-01 3.00863504e-01
7.68575221e-02 -4.52958286e-01 -1.02380192e+00 -7.97413230e-01
3.74777943e-01 -4.87235755e-01 4.19392198e-01 8.67894292e-01
-2.38712400e-01 -1.50155380e-01 -2.28294790e-01 6.28955483e-01
1.71445936e-01 3.54579270e-01 9.44057822e-01 -1.50486553e+00
-7.35518456e-01 -4.21685010e-01 -2.02871114e-01 -2.78536558e-01
2.48718396e-01 -6.05457246e-01 1.04316808e-01 -1.23888659e+00
-7.25274920e-01 -5.26553869e-01 4.05005276e-01 3.95784914e-01
1.97490856e-01 -1.29781470e-01 6.10013418e-02 -3.31618786e-02
4.09433573e-01 4.45376545e-01 7.16445565e-01 2.23739788e-01
-3.01782548e-01 1.91408306e-01 2.66966641e-01 6.81975901e-01
9.50805306e-01 -5.90920031e-01 9.07796528e-03 -1.96471930e-01
3.13299239e-01 4.45373245e-02 9.27506447e-01 -1.34462762e+00
-1.74807817e-01 -4.24355865e-01 1.45550922e-01 -3.85610759e-01
1.00886226e-01 -9.22374964e-01 5.88190913e-01 7.68001139e-01
9.66319963e-02 2.21589118e-01 8.66191328e-01 -2.00033680e-01
-1.17275953e-01 -6.89768970e-01 9.66748536e-01 -2.30212465e-01
-7.69700587e-01 2.89676748e-02 -1.10125840e+00 7.64924213e-02
1.07804513e+00 1.60817236e-01 -3.56542259e-01 -1.93005055e-01
-5.70656776e-01 6.86496422e-02 4.67908800e-01 -1.66672707e-01
-1.76908284e-01 -2.89115548e-01 -8.88607264e-01 3.00989240e-01
-3.18278432e-01 -2.40337744e-01 3.87413979e-01 6.01266205e-01
-1.61407638e+00 3.93445522e-01 -3.43175828e-01 -3.46666604e-01
-9.69576597e-01 3.54597837e-01 1.04820085e+00 -4.99414474e-01
-8.56786191e-01 4.17997062e-01 -5.26733637e-01 -7.13527322e-01
-4.82023619e-02 -4.76652861e-01 -1.52961403e-01 -5.32997027e-02
4.61154431e-01 4.08105224e-01 4.18162674e-01 -4.27925944e-01
-2.95541704e-01 5.92778504e-01 9.27124560e-01 -1.90603375e-01
1.57757616e+00 4.85708922e-01 -4.47000623e-01 3.78722101e-01
6.52646780e-01 -2.44349957e-01 -9.58022118e-01 2.91046977e-01
-9.41004083e-02 -1.59475535e-01 8.13868701e-01 -9.85475898e-01
-1.02261698e+00 8.20992351e-01 5.85326374e-01 2.15547964e-01
8.39664042e-01 -7.26762712e-01 5.19968688e-01 4.82842237e-01
1.41037092e-01 -7.87147403e-01 -9.79003787e-01 8.51267099e-01
1.05758560e+00 -7.57570326e-01 3.66116136e-01 -3.82748395e-01
-4.21142012e-01 1.43159151e+00 1.89775214e-01 -3.12853247e-01
1.17734504e+00 8.34102690e-01 3.59776169e-01 -8.42346475e-02
-4.62367684e-01 -2.69778252e-01 2.93074641e-03 5.68860948e-01
3.29519898e-01 5.91944121e-02 -5.20692468e-01 1.34284049e-01
-1.39782488e-01 4.14898425e-01 4.47033644e-01 9.39716220e-01
-3.31501245e-01 -1.02222776e+00 -7.30468452e-01 -8.11903402e-02
4.69054244e-02 -1.65137812e-01 2.12985724e-02 1.01898777e+00
2.65197426e-01 5.00418842e-01 -1.22809941e-02 1.78224131e-01
1.85910657e-01 2.63617069e-01 1.26703665e-01 -3.33497316e-01
-1.42277598e+00 -3.10674608e-01 1.39848351e-01 -2.94996232e-01
-4.16438729e-01 -5.49960852e-01 -1.15964139e+00 -6.67643249e-01
1.16244361e-01 5.07777452e-01 8.56789649e-01 1.25707805e+00
1.62470773e-01 4.65128154e-01 2.55362719e-01 -1.39604259e+00
-3.94721597e-01 -7.75618076e-01 -7.53947318e-01 -4.10138816e-01
1.24648444e-01 -5.16816556e-01 -4.10292268e-01 -5.27580418e-02] | [6.467236042022705, 3.106466293334961] |
3669e4e8-3280-4bfa-9a42-a3133bada64e | peak-piloted-deep-network-for-facial | 1607.06997 | null | http://arxiv.org/abs/1607.06997v2 | http://arxiv.org/pdf/1607.06997v2.pdf | Peak-Piloted Deep Network for Facial Expression Recognition | Objective functions for training of deep networks for face-related
recognition tasks, such as facial expression recognition (FER), usually
consider each sample independently. In this work, we present a novel
peak-piloted deep network (PPDN) that uses a sample with peak expression (easy
sample) to supervise the intermediate feature responses for a sample of
non-peak expression (hard sample) of the same type and from the same subject.
The expression evolving process from non-peak expression to peak expression can
thus be implicitly embedded in the network to achieve the invariance to
expression intensities. A special purpose back-propagation procedure, peak
gradient suppression (PGS), is proposed for network training. It drives the
intermediate-layer feature responses of non-peak expression samples towards
those of the corresponding peak expression samples, while avoiding the inverse.
This avoids degrading the recognition capability for samples of peak expression
due to interference from their non-peak expression counterparts. Extensive
comparisons on two popular FER datasets, Oulu-CASIA and CK+, demonstrate the
superiority of the PPDN over state-ofthe-art FER methods, as well as the
advantages of both the network structure and the optimization strategy.
Moreover, it is shown that PPDN is a general architecture, extensible to other
tasks by proper definition of peak and non-peak samples. This is validated by
experiments that show state-of-the-art performance on pose-invariant face
recognition, using the Multi-PIE dataset. | ['Nuno Vasconcelos', 'Luoqi Liu', 'Yugang Han', 'Xiaodan Liang', 'Teng Li', 'Xiangyun Zhao', 'Shuicheng Yan'] | 2016-07-24 | null | null | null | null | ['robust-face-recognition'] | ['computer-vision'] | [ 3.46739650e-01 -2.24577829e-01 -1.14535935e-01 -7.28317916e-01
-1.91087723e-01 4.82126474e-02 4.86540496e-01 -4.99278843e-01
-5.24642587e-01 7.03383446e-01 -4.08912390e-01 3.47221673e-01
-2.71468490e-01 -4.82547253e-01 -6.09770179e-01 -1.26377904e+00
-3.72961164e-01 8.13501552e-02 -4.05932426e-01 -4.86383677e-01
-2.40157515e-01 1.17263472e+00 -1.87632108e+00 4.43388551e-01
4.73313361e-01 1.69505858e+00 -2.72380322e-01 1.49520040e-01
-5.03510684e-02 6.59585953e-01 -8.48293245e-01 -4.25683290e-01
2.26978943e-01 -3.75037849e-01 -3.60523015e-01 4.73335683e-02
6.60693347e-01 -9.90021005e-02 -1.05829656e-01 1.04728115e+00
7.04917789e-01 2.51028091e-01 6.93057895e-01 -1.42844021e+00
-5.66375911e-01 -1.60497859e-01 -7.50599265e-01 8.11502337e-02
9.47770998e-02 -2.18106002e-01 5.48119485e-01 -1.15717828e+00
5.50445974e-01 1.23305678e+00 7.11734712e-01 8.28026295e-01
-1.25579274e+00 -8.88770163e-01 -8.04608017e-02 2.64350802e-01
-1.59578693e+00 -7.01198816e-01 1.11242855e+00 -3.33634615e-01
8.17246139e-01 2.86557525e-01 5.51554859e-01 1.28262794e+00
5.59512973e-02 7.81389475e-01 1.19825649e+00 -3.42925936e-01
4.84038964e-02 9.19092894e-02 4.97962870e-02 5.21772921e-01
-3.49510550e-01 1.42273158e-01 -6.06617630e-01 1.87705122e-02
4.69802648e-01 -2.86282986e-01 -4.82611090e-01 -2.61884660e-01
-1.76373079e-01 6.15511596e-01 3.50083530e-01 5.29091120e-01
-8.00513506e-01 -6.04208075e-02 3.85309517e-01 3.81348521e-01
6.44576073e-01 9.92279500e-02 -5.10048687e-01 2.79326495e-02
-1.05396461e+00 2.41828680e-01 6.88642681e-01 4.96109635e-01
9.05540943e-01 4.83147830e-01 -5.56151271e-01 1.18837333e+00
1.81774616e-01 5.11847854e-01 3.13643843e-01 -7.20099628e-01
-1.21876083e-01 6.03240192e-01 -5.39379716e-02 -1.02858353e+00
-3.84267151e-01 -6.26959562e-01 -1.04825175e+00 4.59451765e-01
3.30985278e-01 -3.73634160e-01 -9.13574934e-01 2.12168717e+00
3.43811512e-01 9.63123515e-02 6.83297142e-02 9.42539871e-01
7.78257728e-01 5.52050173e-01 2.79814899e-01 -5.55936217e-01
1.22482741e+00 -5.41435242e-01 -7.62050748e-01 1.35036139e-02
2.92350322e-01 -4.53523189e-01 7.39963949e-01 4.26083326e-01
-7.17879713e-01 -7.55512178e-01 -9.16346014e-01 2.77433932e-01
-3.50535512e-01 4.93862391e-01 6.73890412e-01 7.53661990e-01
-1.08819115e+00 6.49583995e-01 -2.98469514e-01 3.01388893e-02
7.36672401e-01 7.45709896e-01 -7.09327638e-01 2.14019418e-01
-1.35443056e+00 5.69940984e-01 1.68936312e-01 6.29848778e-01
-5.22994339e-01 -9.30690885e-01 -6.58309758e-01 1.74457416e-01
-1.17386974e-01 -2.56884769e-02 9.82315123e-01 -2.13178205e+00
-1.85570049e+00 1.08902371e+00 -3.37793261e-01 -2.06743717e-01
3.61708432e-01 -1.43605992e-01 -7.06341147e-01 2.12005228e-01
-3.58977139e-01 6.92963481e-01 1.26764846e+00 -1.06133866e+00
-1.69361711e-01 -5.67989051e-01 -3.59543025e-01 -3.26709859e-02
-4.92372930e-01 3.71919662e-01 -3.31531316e-01 -3.72758716e-01
-3.51755679e-01 -6.34468913e-01 3.37566406e-01 2.49050543e-01
-6.92256540e-02 -4.42263216e-01 1.17105234e+00 -4.80800092e-01
9.82829154e-01 -2.58582997e+00 -2.13699918e-02 4.70009923e-01
-9.29377899e-02 5.99165380e-01 -4.27809000e-01 -9.05937999e-02
-6.49389088e-01 -3.15252542e-01 -7.61380866e-02 -2.17835277e-01
3.84235978e-02 3.05752188e-01 -7.94029459e-02 5.98488629e-01
7.09338307e-01 7.38521039e-01 -5.44714987e-01 -1.36630997e-01
-7.25176409e-02 9.76935148e-01 -2.61597753e-01 2.52811432e-01
4.01794836e-02 3.51097643e-01 -2.13244408e-01 5.92177749e-01
1.09266436e+00 1.85493082e-01 4.51974906e-02 -5.71979880e-01
3.29136811e-02 -5.39573312e-01 -9.15309250e-01 9.24196601e-01
-3.77415895e-01 6.87152624e-01 5.92399776e-01 -1.14754498e+00
1.35702491e+00 4.72973526e-01 6.46254003e-01 -9.41417038e-01
3.31535101e-01 1.85527503e-01 1.60028562e-01 -4.99403059e-01
9.02799219e-02 -3.62269759e-01 4.82021719e-01 -1.44084439e-01
4.77577895e-01 4.41660225e-01 2.88423181e-01 -5.47140658e-01
4.12867159e-01 1.04160592e-01 5.51127940e-02 -4.72791135e-01
1.03648603e+00 -8.73703063e-01 6.66822433e-01 4.21945989e-01
-4.08189148e-01 1.02753244e-01 8.06563914e-01 -3.30854774e-01
-7.07990766e-01 -8.74900699e-01 -5.69309652e-01 1.36514521e+00
-2.91420072e-01 1.42609969e-01 -8.78148615e-01 -5.78637481e-01
5.82933053e-02 5.34857810e-01 -8.99474084e-01 -1.80164799e-01
-5.60802221e-01 -9.46720839e-01 7.79179752e-01 3.41512531e-01
7.70276606e-01 -1.26454210e+00 -4.10390109e-01 2.54755206e-02
1.40070125e-01 -9.80974615e-01 -1.37920916e-01 4.34324771e-01
-3.97692978e-01 -8.53768945e-01 -9.66026604e-01 -7.57546842e-01
7.39574313e-01 -4.54940170e-01 8.11438382e-01 -7.24744126e-02
-1.86766863e-01 2.68736720e-01 -1.68544929e-02 -4.27367866e-01
-2.12786302e-01 -4.13501501e-01 1.78467315e-02 8.42657387e-01
6.30854905e-01 -5.43625474e-01 -3.58730763e-01 3.12640160e-01
-8.91209483e-01 -4.38105017e-01 6.34872615e-01 1.17266536e+00
6.67424798e-01 -1.77715853e-01 6.33482873e-01 -5.11345267e-01
7.03012228e-01 -2.47681439e-01 -3.99556100e-01 2.03335047e-01
-3.46450835e-01 -4.50839810e-02 7.60192692e-01 -7.25193620e-01
-1.33042347e+00 -8.49421043e-03 -4.19398665e-01 -7.85335839e-01
-3.07245404e-01 1.91327527e-01 -3.90404612e-01 -5.08753657e-01
7.01669574e-01 2.85561413e-01 3.60972732e-01 -2.04439282e-01
-1.26253828e-01 6.28062904e-01 4.82409924e-01 -5.90147972e-01
3.49778503e-01 3.06501478e-01 4.09967244e-01 -1.34746873e+00
-7.09655941e-01 -2.09428877e-01 -4.06360626e-01 -5.20830870e-01
5.55939317e-01 -7.33987451e-01 -8.43453228e-01 1.02952898e+00
-1.08663762e+00 -2.73940623e-01 -1.56036273e-01 1.70142531e-01
-3.24336290e-01 1.35233313e-01 -5.22911906e-01 -1.04440463e+00
-6.08908832e-01 -9.44405258e-01 1.15899074e+00 5.91725290e-01
-7.13047432e-03 -8.96113217e-01 -1.96558580e-01 -6.74112514e-02
5.19300997e-01 4.12505239e-01 7.23430753e-01 -7.11725473e-01
9.20606107e-02 -2.65781432e-01 -2.12925360e-01 9.43494141e-01
5.83758131e-02 3.09322506e-01 -1.39484906e+00 -3.52167279e-01
9.45514366e-02 -4.66588944e-01 6.14585280e-01 3.15628529e-01
1.18283033e+00 -2.31277049e-01 -1.96487270e-02 9.02305663e-01
1.22130370e+00 3.69958788e-01 8.36508453e-01 1.16997257e-01
2.09763974e-01 6.49332881e-01 4.24219936e-01 3.52438211e-01
-4.27419424e-01 9.50514257e-01 2.82293469e-01 -4.43926007e-01
1.65860541e-02 2.46594891e-01 3.95401716e-01 6.12707995e-02
-2.30785981e-01 -1.49069905e-01 -4.67275232e-01 2.03359514e-01
-1.42438269e+00 -9.62790608e-01 2.54015368e-03 1.93274546e+00
6.84787333e-01 -3.31814200e-01 -8.76333565e-02 1.40035540e-01
5.68316936e-01 2.03735873e-01 -4.61943805e-01 -7.23092020e-01
-6.14919364e-01 9.21969235e-01 9.86624882e-02 1.36215359e-01
-1.23322618e+00 7.27594912e-01 5.91765928e+00 1.13252509e+00
-1.81998873e+00 -6.14093766e-02 9.31378663e-01 -7.56193399e-02
3.20949197e-01 -7.50483274e-01 -7.58985817e-01 2.96416968e-01
8.48415792e-01 1.05598330e-01 4.14551556e-01 1.09188235e+00
1.92429110e-01 1.09913804e-01 -1.00394559e+00 1.17333281e+00
1.82952076e-01 -7.04124272e-01 -5.25604188e-02 -2.39261851e-01
3.93425941e-01 -2.74393022e-01 3.00434589e-01 5.29442787e-01
-5.18670261e-01 -1.11481595e+00 4.88533914e-01 5.68250000e-01
9.63458776e-01 -9.78220165e-01 9.02413607e-01 -8.79332349e-02
-9.01516676e-01 -1.47869274e-01 -2.59483993e-01 8.13685879e-02
-1.10812686e-01 5.79933703e-01 -5.10971785e-01 6.14190102e-01
6.39079809e-01 4.53147501e-01 -3.35922509e-01 4.62576509e-01
-3.06309849e-01 4.72950190e-01 -3.65929753e-01 -2.84228176e-02
2.03768894e-01 -3.27542335e-01 4.83723074e-01 1.40745711e+00
2.75733560e-01 -1.30908728e-01 -3.81742686e-01 8.69060338e-01
-1.08729511e-01 4.00785893e-01 -3.45652759e-01 -6.98944256e-02
8.40711594e-02 1.57503581e+00 -1.80624321e-01 -2.36060962e-01
-2.03955099e-01 1.00943828e+00 3.01596016e-01 7.08511114e-01
-7.71109879e-01 -5.03578842e-01 1.05880809e+00 -9.33997508e-04
4.74678278e-01 3.11216593e-01 2.40050703e-01 -5.54847598e-01
1.45623550e-01 -9.50998068e-01 1.68669775e-01 -7.66631961e-01
-1.22121024e+00 9.07099605e-01 -3.03569958e-02 -8.40777099e-01
-3.73113811e-01 -9.96661186e-01 -6.23747408e-01 1.04892945e+00
-1.74387419e+00 -1.02978230e+00 -3.33837450e-01 8.62639844e-01
-8.97983238e-02 -1.77030072e-01 8.77629817e-01 6.41876936e-01
-7.17953444e-01 1.17186677e+00 -1.31625235e-01 2.92370617e-01
4.81994927e-01 -7.22600341e-01 -4.71525133e-01 6.97986901e-01
-1.12182051e-01 5.99367797e-01 5.93300641e-01 -8.25086460e-02
-1.16776836e+00 -9.88215029e-01 7.38826573e-01 4.31549162e-01
3.70519459e-01 -2.64307320e-01 -1.00770736e+00 2.98673004e-01
7.49893636e-02 2.79790938e-01 6.97346330e-01 -7.75991008e-02
-3.06869864e-01 -6.75197721e-01 -1.37813330e+00 5.63315153e-01
7.44201303e-01 -4.76744771e-01 -1.64682999e-01 1.37578607e-01
-1.56773537e-01 -2.19139636e-01 -8.83476019e-01 8.01108956e-01
8.21956217e-01 -1.20309126e+00 8.14792931e-01 -5.68889678e-01
2.23912597e-01 -1.35308534e-01 -7.44589046e-02 -1.20610905e+00
-3.29647511e-01 -6.28291368e-01 -8.62834975e-02 1.15287578e+00
6.09929375e-02 -7.24348426e-01 7.76277781e-01 3.80427480e-01
-1.24585116e-02 -1.03445470e+00 -1.25453341e+00 -8.41198623e-01
-2.23214254e-01 -2.05797583e-01 3.20631713e-01 7.88077533e-01
-3.40428591e-01 2.08511069e-01 -4.14857566e-01 1.84964575e-02
3.62904251e-01 -1.17945395e-01 6.25758111e-01 -1.10953784e+00
-2.40476549e-01 -5.26214778e-01 -7.27705002e-01 -8.21019590e-01
5.94009280e-01 -7.71165133e-01 1.62143447e-02 -6.72043443e-01
-1.53305784e-01 -2.84617469e-02 -4.89713401e-01 6.72641695e-01
-1.70660187e-02 4.19975579e-01 1.10243715e-01 -2.41091967e-01
-9.46257859e-02 7.50109732e-01 1.19956231e+00 2.38613877e-02
-1.32498667e-01 7.45363126e-04 -3.50545287e-01 6.61864817e-01
5.74892938e-01 -1.55589193e-01 -2.95056373e-01 1.36026561e-01
-3.09684187e-01 -2.18436807e-01 3.16521615e-01 -1.02687776e+00
-2.00592235e-01 1.63344100e-01 7.93720901e-01 -1.56469524e-01
6.51925862e-01 -8.74622405e-01 2.84543723e-01 2.63700217e-01
-9.36085135e-02 -2.94110477e-01 5.83488762e-01 -1.08722910e-04
-4.77232188e-01 -2.38054693e-01 1.16187716e+00 3.24033290e-01
-8.30449641e-01 4.25406426e-01 -3.27995420e-01 -2.05646172e-01
9.51629102e-01 -4.51337695e-01 -5.51164756e-03 -3.10117573e-01
-7.81021774e-01 -1.61025420e-01 -1.05645530e-01 2.40554497e-01
5.39008617e-01 -1.34017003e+00 -6.73794091e-01 8.10521722e-01
-2.47394685e-02 -4.61395770e-01 4.64314133e-01 1.21375728e+00
-9.97372195e-02 9.93928686e-02 -4.80040818e-01 -6.77060306e-01
-1.55947804e+00 3.16060841e-01 9.29000199e-01 -1.48179725e-01
-1.30177468e-01 9.87407327e-01 4.32677507e-01 -2.61947900e-01
2.09684670e-01 1.82261273e-01 -3.96777451e-01 2.92011380e-01
7.14073062e-01 4.69062738e-02 2.07516089e-01 -9.08079267e-01
-4.02095914e-01 5.52118242e-01 -3.54861245e-02 3.36340010e-01
1.36085355e+00 4.98364836e-01 -4.44387704e-01 1.54430270e-01
1.58285761e+00 -2.32251331e-01 -1.33899093e+00 2.37786360e-02
-7.50442520e-02 -3.21741462e-01 9.17067602e-02 -8.39192092e-01
-1.42199373e+00 7.57014155e-01 1.12103295e+00 -3.12388182e-01
1.68112504e+00 -4.09175694e-01 2.07424968e-01 3.21742564e-01
2.07739007e-02 -1.04449260e+00 9.57549829e-03 5.17208517e-01
1.16114414e+00 -7.32553363e-01 -4.13653404e-01 -2.31357068e-01
-4.43069667e-01 1.42851150e+00 7.16627121e-01 -5.40127717e-02
7.01856494e-01 4.58232194e-01 2.03775361e-01 -2.34259143e-01
-5.15379906e-01 2.94176918e-02 4.15500641e-01 6.09474361e-01
5.22425354e-01 -1.64133891e-01 -3.20874035e-01 5.44527590e-01
1.15568101e-01 3.47384661e-01 -2.59804636e-01 6.23325348e-01
-2.45069414e-02 -8.86745870e-01 -3.40957493e-01 3.40434074e-01
-6.90763652e-01 1.06537841e-01 -4.21225876e-01 1.08937156e+00
3.39475632e-01 4.17947143e-01 3.35568726e-01 -2.32238799e-01
5.46694636e-01 4.97313619e-01 6.95632577e-01 8.39252174e-02
-6.46955371e-01 1.32913515e-01 1.63338020e-01 -5.58956325e-01
-5.77816010e-01 -5.18156171e-01 -1.03138900e+00 -9.99445692e-02
-8.14934447e-03 6.95588216e-02 5.96240103e-01 8.23381484e-01
4.38960820e-01 3.99947405e-01 7.49503732e-01 -9.50650811e-01
-5.06896555e-01 -1.04964519e+00 -8.49318981e-01 7.38342285e-01
3.12105179e-01 -9.41000819e-01 -4.98666316e-01 -7.07042292e-02] | [13.594206809997559, 1.7299158573150635] |
b3fc692a-45e5-4313-95d4-de53b13ab9a9 | on-the-usage-of-continual-learning-for-out-of | 2305.04106 | null | https://arxiv.org/abs/2305.04106v1 | https://arxiv.org/pdf/2305.04106v1.pdf | On the Usage of Continual Learning for Out-of-Distribution Generalization in Pre-trained Language Models of Code | Pre-trained language models (PLMs) have become a prevalent technique in deep learning for code, utilizing a two-stage pre-training and fine-tuning procedure to acquire general knowledge about code and specialize in a variety of downstream tasks. However, the dynamic nature of software codebases poses a challenge to the effectiveness and robustness of PLMs. In particular, world-realistic scenarios potentially lead to significant differences between the distribution of the pre-training and test data, i.e., distribution shift, resulting in a degradation of the PLM's performance on downstream tasks. In this paper, we stress the need for adapting PLMs of code to software data whose distribution changes over time, a crucial problem that has been overlooked in previous works. The motivation of this work is to consider the PLM in a non-stationary environment, where fine-tuning data evolves over time according to a software evolution scenario. Specifically, we design a scenario where the model needs to learn from a stream of programs containing new, unseen APIs over time. We study two widely used PLM architectures, i.e., a GPT2 decoder and a RoBERTa encoder, on two downstream tasks, API call and API usage prediction. We demonstrate that the most commonly used fine-tuning technique from prior work is not robust enough to handle the dynamic nature of APIs, leading to the loss of previously acquired knowledge i.e., catastrophic forgetting. To address these issues, we implement five continual learning approaches, including replay-based and regularization-based methods. Our findings demonstrate that utilizing these straightforward methods effectively mitigates catastrophic forgetting in PLMs across both downstream tasks while achieving comparable or superior performance. | ['Houari Sahraoui', 'David Lo', 'Kisub Kim', 'Xin Zhou', 'Martin Weyssow'] | 2023-05-06 | null | null | null | null | ['general-knowledge'] | ['miscellaneous'] | [ 1.20992884e-01 -2.84616470e-01 -1.94066107e-01 -2.69423604e-01
-4.48597610e-01 -4.59278136e-01 3.56938630e-01 1.93495061e-02
-1.83718920e-01 5.40303707e-01 7.85998702e-02 -5.26944876e-01
-4.26770039e-02 -6.20526254e-01 -1.08638358e+00 -4.94976550e-01
-9.78521258e-02 -2.63726874e-03 3.81399930e-01 -2.78220832e-01
4.01112825e-01 2.11426660e-01 -1.75120401e+00 3.27449113e-01
8.74543011e-01 3.57176244e-01 4.98677313e-01 7.18572378e-01
-3.66877407e-01 8.22854578e-01 -8.82883430e-01 -5.18784702e-01
3.53035592e-02 -1.57424286e-01 -6.08863711e-01 -1.69542298e-01
7.65415877e-02 -1.00477293e-01 -4.58157927e-01 1.02702427e+00
3.59247565e-01 3.50285396e-02 3.97657305e-01 -1.18699741e+00
-7.35372961e-01 8.88997912e-01 -7.44572818e-01 4.99572307e-01
-2.25577876e-02 3.04504305e-01 4.95405555e-01 -5.91579556e-01
4.18216974e-01 1.06368816e+00 1.05235422e+00 7.54913151e-01
-1.21593058e+00 -5.27739167e-01 2.76601672e-01 8.93312618e-02
-1.21911669e+00 -4.98690665e-01 7.10293055e-01 -6.66800559e-01
1.41712272e+00 -8.44649747e-02 2.68899769e-01 1.57728136e+00
8.28423738e-01 6.88881099e-01 5.94944835e-01 -4.67325628e-01
2.60606736e-01 3.38619471e-01 5.94735518e-02 5.39416432e-01
2.58257627e-01 2.24187240e-01 -7.64372110e-01 -4.22816962e-01
3.04226905e-01 1.73186004e-01 -2.53551364e-01 -2.96373367e-01
-8.70748580e-01 5.32678664e-01 -3.72963585e-03 4.56105411e-01
-3.21768194e-01 2.53785491e-01 6.67836785e-01 5.56178927e-01
3.72609377e-01 2.56384790e-01 -9.74798858e-01 -6.53479517e-01
-1.12175202e+00 2.71768565e-03 8.90522540e-01 1.14764071e+00
8.15731943e-01 2.17920005e-01 -2.17657894e-01 6.93259597e-01
3.53622615e-01 2.14951575e-01 1.23881221e+00 -4.35907692e-01
5.09302795e-01 5.77383697e-01 -3.04125398e-01 -7.03510582e-01
-1.29249826e-01 -6.28221214e-01 -4.54562157e-01 6.04841039e-02
7.94690698e-02 -1.13406904e-01 -1.01366246e+00 2.23504972e+00
-3.89400572e-02 3.33391130e-01 5.21836989e-02 2.37978414e-01
1.00901157e-01 4.20607537e-01 1.26758873e-01 -1.10057943e-01
9.45440590e-01 -1.01333487e+00 -4.76462454e-01 -4.72793549e-01
6.30623877e-01 -7.90207326e-01 1.58459175e+00 4.70069945e-01
-7.80316532e-01 -5.78794956e-01 -1.09796405e+00 2.03721151e-01
-3.22512984e-01 -1.34988734e-03 6.62256420e-01 5.97812831e-01
-1.15581143e+00 7.74823248e-01 -1.27885127e+00 -5.77081203e-01
1.90977409e-01 3.34288478e-01 -7.35162720e-02 5.75292408e-02
-8.65887821e-01 7.42504418e-01 3.69312137e-01 -2.84652352e-01
-1.20534158e+00 -1.02089512e+00 -5.69205046e-01 3.38138342e-01
2.86551714e-01 -6.89051867e-01 1.63274956e+00 -9.79259193e-01
-1.55936754e+00 5.27244747e-01 -1.75328389e-01 -5.14670193e-01
4.38850105e-01 -4.29853141e-01 -5.02108216e-01 -6.83635056e-01
2.43457649e-02 -1.44766076e-02 1.17819548e+00 -1.22374713e+00
-5.44821620e-01 -1.78333208e-01 -9.83942021e-03 -2.15917215e-01
-6.57721102e-01 -2.34340504e-01 -5.35983384e-01 -7.25933790e-01
-4.64832246e-01 -8.35979223e-01 1.13956682e-01 -6.02005780e-01
-8.86556506e-02 -6.19506948e-02 8.98815572e-01 -6.46800876e-01
1.90450490e+00 -2.50045323e+00 1.34676084e-01 -1.83761269e-01
1.10587977e-01 3.20041865e-01 -3.28316748e-01 5.66084266e-01
1.45102795e-02 1.66002020e-01 -2.69398719e-01 -5.08012831e-01
-6.18247427e-02 3.88947368e-01 -5.74702024e-01 1.39929309e-01
3.18101823e-01 7.64917552e-01 -9.11902368e-01 8.38118978e-03
-2.81125486e-01 4.49918896e-01 -7.58407652e-01 3.72953773e-01
-5.24791718e-01 3.12522650e-01 -3.45777661e-01 5.74168622e-01
3.23799491e-01 -3.65198582e-01 2.16406621e-02 1.44862145e-01
-1.75012648e-01 3.13139170e-01 -6.30536079e-01 1.96903372e+00
-9.85106289e-01 5.38792133e-01 -2.62240291e-01 -8.08964670e-01
7.89573610e-01 2.27929682e-01 1.46895811e-01 -6.00875020e-01
-1.21765532e-01 2.60793895e-01 1.75555587e-01 -8.44802082e-01
5.02131164e-01 -1.31854713e-01 -1.95278466e-01 6.50941908e-01
2.13323489e-01 2.77388543e-01 1.14625767e-01 1.91365313e-02
1.74194074e+00 2.88245171e-01 1.83253542e-01 -1.44241393e-01
3.57877374e-01 -6.70157149e-02 8.25380981e-01 9.85491335e-01
-1.38002366e-01 3.61984134e-01 4.83787119e-01 -3.84764373e-01
-1.05387366e+00 -9.00960624e-01 1.59620374e-01 1.44325030e+00
-2.13811100e-01 -5.33019722e-01 -7.01978803e-01 -7.89988101e-01
1.60642520e-01 1.07561255e+00 -5.87345839e-01 -1.08541131e+00
-7.67226934e-01 -1.04424393e+00 6.25127316e-01 3.79261136e-01
2.77075231e-01 -9.77550745e-01 -7.97885180e-01 5.42557895e-01
1.63807329e-02 -4.65998650e-01 -7.02973485e-01 4.70019937e-01
-9.91664171e-01 -9.03610528e-01 -3.83248687e-01 -4.74151433e-01
4.79245782e-01 2.55838752e-01 1.45821202e+00 1.78617269e-01
-2.16429070e-01 5.24931073e-01 -3.27528059e-01 -3.21810216e-01
-9.97167587e-01 4.44381058e-01 9.83284190e-02 -1.48089483e-01
6.99217618e-01 -9.02205825e-01 -4.71490920e-01 1.50586665e-02
-1.15303886e+00 -1.55214787e-01 9.28440452e-01 9.41585362e-01
1.91792980e-01 3.40869844e-01 7.17280030e-01 -1.05169475e+00
9.41848099e-01 -9.35224175e-01 -5.02439797e-01 5.81065238e-01
-9.34127331e-01 4.73783463e-01 8.34018409e-01 -9.23769772e-01
-1.40294337e+00 -2.13376820e-01 -2.49198414e-02 -6.11641645e-01
3.87213528e-02 7.90670216e-01 -1.46603324e-02 6.01489621e-04
8.95160913e-01 4.73108172e-01 -1.45112693e-01 -6.75452352e-01
1.62237272e-01 7.93999434e-01 5.49543321e-01 -8.73080313e-01
7.87915587e-01 1.37203500e-01 -5.98446429e-01 -4.09566730e-01
-6.76696360e-01 -2.87307084e-01 -4.30205852e-01 9.27148908e-02
2.46954992e-01 -8.11470091e-01 -2.85038024e-01 6.03562951e-01
-1.11786103e+00 -6.59888327e-01 -2.79201299e-01 1.93566978e-01
-6.17964387e-01 3.67748141e-01 -6.58726513e-01 -5.37121892e-01
-3.06244940e-01 -1.38527906e+00 9.69755590e-01 2.60727197e-01
-3.10643703e-01 -1.10296309e+00 4.35921490e-01 -2.11811513e-02
9.32843804e-01 -3.33788171e-02 1.37674391e+00 -5.03603578e-01
-5.41718423e-01 -6.67081177e-02 2.21453011e-01 4.58636642e-01
3.03480685e-01 1.13042139e-01 -1.01103950e+00 -7.33637691e-01
4.00391698e-01 -1.20489240e-01 7.54464388e-01 -2.05608886e-02
1.12111044e+00 -2.25025669e-01 -3.77370894e-01 7.67652631e-01
1.53831100e+00 3.23067605e-01 2.61078745e-01 4.02862906e-01
3.84542704e-01 1.40934870e-01 3.97351086e-01 6.34962082e-01
3.35532635e-01 4.22621012e-01 2.34251365e-01 5.10709345e-01
-3.20960253e-01 -4.82672304e-01 7.59157658e-01 1.14692521e+00
3.35513741e-01 -2.69043893e-01 -1.22053218e+00 6.93164527e-01
-1.96133947e+00 -5.10337412e-01 3.52055609e-01 2.47952366e+00
1.11118734e+00 1.48864999e-01 -3.51843476e-01 -2.82177746e-01
5.09813726e-01 -1.29413813e-01 -1.06148112e+00 -2.75608361e-01
1.97175890e-01 1.81300834e-01 3.87817204e-01 6.21091425e-02
-6.98637843e-01 9.00347173e-01 5.80732918e+00 5.85572600e-01
-1.36899614e+00 5.10874808e-01 1.54126585e-01 -3.73141989e-02
-3.04423541e-01 1.24895155e-01 -7.86820650e-01 8.53956342e-01
1.52586687e+00 -6.38312638e-01 7.35224843e-01 9.87813294e-01
4.55061719e-02 5.08533083e-02 -1.32249522e+00 7.88290083e-01
8.85245055e-02 -9.40997958e-01 -9.20678303e-02 -1.64155021e-01
7.61057973e-01 6.06630445e-01 2.31695428e-01 9.29499626e-01
3.65939409e-01 -6.12030268e-01 8.04616451e-01 6.10681236e-01
7.08123028e-01 -4.59394783e-01 6.54704034e-01 6.40686452e-01
-8.83374393e-01 -3.38159502e-01 -2.49924719e-01 5.67991771e-02
-3.44552062e-02 6.57179952e-01 -7.53473878e-01 2.28218183e-01
8.39933157e-01 5.21831155e-01 -7.50376642e-01 1.07443035e+00
-5.71733676e-02 7.81004250e-01 -4.24982198e-02 3.64673913e-01
-1.22640148e-01 3.54038686e-01 4.78522539e-01 1.31688237e+00
7.15795636e-01 -6.48735404e-01 -1.79617196e-01 1.09296191e+00
-1.60747379e-01 -2.79347330e-01 -6.94384396e-01 -2.32309550e-01
5.30683994e-01 8.46533895e-01 -2.89319694e-01 -1.52864203e-01
-7.84810305e-01 1.02117109e+00 6.13469124e-01 5.43385446e-01
-9.13266242e-01 -3.65735888e-01 8.22659850e-01 -3.92328054e-02
3.41835827e-01 -2.75273591e-01 -8.97908658e-02 -1.57105696e+00
2.68527120e-01 -1.11232686e+00 2.36017421e-01 -2.70170361e-01
-1.41301823e+00 6.13289118e-01 -1.09627984e-01 -8.83573294e-01
-3.58339816e-01 -3.01771462e-01 -7.72526920e-01 8.24610233e-01
-1.76807809e+00 -8.83339942e-01 -2.61576831e-01 4.66369718e-01
8.14008176e-01 -4.27849203e-01 5.17762184e-01 5.84333658e-01
-6.79571271e-01 8.50249827e-01 2.59475142e-01 -4.22303826e-01
8.91474903e-01 -1.16917837e+00 5.69566011e-01 1.09251297e+00
-1.02135234e-01 1.10849488e+00 7.80164838e-01 -7.93867648e-01
-1.66014671e+00 -1.34617424e+00 6.86846018e-01 -6.64601266e-01
7.73720920e-01 -4.25523430e-01 -1.51588893e+00 8.93822253e-01
7.33299553e-02 -2.31785536e-01 6.63130939e-01 5.09528778e-02
-5.20576000e-01 -6.55370057e-02 -8.58732522e-01 5.06467760e-01
9.79508579e-01 -6.47466838e-01 -6.92162097e-01 1.16929017e-01
9.34019744e-01 -4.26230073e-01 -8.22594762e-01 3.78102571e-01
2.96798617e-01 -9.65403438e-01 6.27439976e-01 -7.92006969e-01
3.00734729e-01 -2.47370929e-01 -1.11641772e-01 -1.45090365e+00
-2.86164820e-01 -8.89657795e-01 -5.95890045e-01 1.38269603e+00
3.78248036e-01 -5.44323623e-01 5.22227526e-01 5.57268023e-01
-2.74935156e-01 -5.24069548e-01 -7.65103221e-01 -8.56368005e-01
1.62539199e-01 -3.67763132e-01 7.18547761e-01 7.64470220e-01
-1.35311782e-01 3.04633111e-01 -2.73022979e-01 2.27121636e-01
2.66722053e-01 3.04283611e-02 8.08264852e-01 -9.61259544e-01
-7.40681946e-01 -4.94075269e-01 -1.05801865e-01 -9.09134388e-01
3.39627653e-01 -7.88182020e-01 1.98330313e-01 -7.99342573e-01
4.52392489e-01 -3.49040240e-01 -5.22714615e-01 4.94477570e-01
-2.27481127e-01 -5.04615128e-01 -7.92031735e-02 4.69702423e-01
-4.13970232e-01 6.44846559e-01 5.79360306e-01 -9.57424343e-02
-5.29620767e-01 3.00346136e-01 -7.23981440e-01 5.31782269e-01
8.61486256e-01 -7.24646986e-01 -7.87845433e-01 -8.00392210e-01
5.32529652e-01 -1.58638284e-01 1.13770096e-02 -1.09523094e+00
3.42219383e-01 -5.58805391e-02 -1.60642713e-01 -5.99500090e-02
-2.36859083e-01 -6.98146462e-01 3.30729753e-01 5.60977817e-01
-2.17324868e-01 4.29467857e-01 5.91491461e-01 9.05976653e-01
-1.91464454e-01 -4.37395364e-01 6.32758975e-01 -2.53834963e-01
-8.94673109e-01 2.39479482e-01 -2.93157130e-01 5.53107038e-02
9.31512713e-01 1.38830706e-01 -5.02341270e-01 1.07289981e-02
-4.70588923e-01 2.71317475e-02 6.52119517e-01 1.00436831e+00
3.19411188e-01 -1.02802074e+00 -3.35595340e-01 4.48492736e-01
3.19066256e-01 -1.68856785e-01 3.10415924e-01 9.26388085e-01
-1.85902759e-01 2.56333143e-01 -4.76824120e-02 -5.53009808e-01
-9.16656315e-01 8.28757644e-01 3.25513542e-01 -4.21936572e-01
-6.27098680e-01 7.33295619e-01 2.37247109e-01 -4.50795680e-01
3.84004086e-01 -4.75311488e-01 2.29409456e-01 -3.28614414e-01
3.96573573e-01 -4.29983102e-02 3.82003754e-01 -7.69498944e-02
-1.58059493e-01 2.42702514e-01 -5.47076344e-01 3.42087895e-01
1.46972764e+00 -2.38576591e-01 -2.26270631e-01 8.61440539e-01
1.09525359e+00 -2.00983603e-03 -1.56817305e+00 -4.36470628e-01
5.34479380e-01 -3.69178474e-01 -6.44796342e-03 -8.95553172e-01
-9.11699295e-01 8.72587442e-01 5.21970689e-01 1.20550558e-01
1.19856811e+00 -6.67302012e-02 8.86532128e-01 2.61701465e-01
7.04547942e-01 -8.97777617e-01 2.17427835e-01 6.54459834e-01
5.82077563e-01 -1.11563265e+00 -4.54501331e-01 3.09472531e-01
-2.98656911e-01 1.13380587e+00 9.01102901e-01 3.19665670e-01
5.57227612e-01 6.07185423e-01 -2.23860834e-02 2.05252469e-02
-1.34549916e+00 2.93685883e-01 -9.13821533e-02 4.36796337e-01
5.49770832e-01 -1.99861944e-01 1.49574745e-02 7.52167642e-01
-7.21343085e-02 3.01332742e-01 6.08760595e-01 1.27717316e+00
-1.36473045e-01 -1.25692940e+00 -5.27340546e-02 4.93931085e-01
-4.82760102e-01 -1.62943080e-01 1.19061269e-01 6.10527873e-01
1.15738571e-01 4.89426672e-01 -1.99861139e-01 -6.18545055e-01
3.47488433e-01 3.87464523e-01 2.99082100e-01 -9.64314342e-01
-6.86198652e-01 -2.92961091e-01 -3.88356805e-01 -5.06971955e-01
-4.80790474e-02 -6.48898542e-01 -9.50995505e-01 -2.45620504e-01
-3.20882916e-01 -1.60926990e-02 6.04633689e-01 7.71948040e-01
9.37662125e-01 9.01606381e-01 3.22867751e-01 -4.73075360e-01
-1.09962499e+00 -9.38952208e-01 -2.41065443e-01 4.87938702e-01
5.74042976e-01 -6.87769115e-01 -5.11666477e-01 3.07033628e-01] | [7.699618816375732, 7.870054721832275] |
b627ce30-fefb-43b1-a1a0-4ca9efdd580f | gaussigan-controllable-image-synthesis-with | 2106.13215 | null | https://arxiv.org/abs/2106.13215v1 | https://arxiv.org/pdf/2106.13215v1.pdf | GaussiGAN: Controllable Image Synthesis with 3D Gaussians from Unposed Silhouettes | We present an algorithm that learns a coarse 3D representation of objects from unposed multi-view 2D mask supervision, then uses it to generate detailed mask and image texture. In contrast to existing voxel-based methods for unposed object reconstruction, our approach learns to represent the generated shape and pose with a set of self-supervised canonical 3D anisotropic Gaussians via a perspective camera, and a set of per-image transforms. We show that this approach can robustly estimate a 3D space for the camera and object, while recent baselines sometimes struggle to reconstruct coherent 3D spaces in this setting. We show results on synthetic datasets with realistic lighting, and demonstrate object insertion with interactive posing. With our work, we help move towards structured representations that handle more real-world variation in learning-based object reconstruction. | ['James Tompkin', 'Kwang In Kim', 'Oliver Wang', 'Aaron Gokaslan', 'Isa Milefchik', 'Youssef A. Mejjati'] | 2021-06-24 | null | null | null | null | ['object-reconstruction'] | ['computer-vision'] | [ 5.67933619e-01 4.05945569e-01 1.51683375e-01 -4.12010252e-01
-1.29450893e+00 -8.50698292e-01 8.45597684e-01 -5.65052629e-01
7.90015236e-02 2.09575281e-01 2.47654438e-01 1.99055701e-01
1.43482924e-01 -5.39242089e-01 -1.13986850e+00 -5.59753180e-01
2.82073617e-01 1.29040480e+00 4.29289103e-01 1.34500772e-01
2.95767277e-01 9.10054326e-01 -1.60165811e+00 5.58444858e-01
3.25921029e-01 6.88237011e-01 2.85732418e-01 8.17223430e-01
1.59957092e-02 4.74211276e-01 -1.05897859e-01 -2.96093434e-01
6.59973741e-01 -2.21051291e-01 -7.12939143e-01 1.10366929e+00
1.19959962e+00 -6.81257844e-01 -2.53649913e-02 6.77849650e-01
1.45071149e-01 2.20773020e-03 1.05930233e+00 -8.63314152e-01
-3.42594296e-01 2.17390433e-01 -7.86089778e-01 -2.96414196e-01
6.95634425e-01 1.72305048e-01 7.82303035e-01 -1.23687673e+00
1.15567160e+00 1.57893240e+00 6.19088352e-01 4.71784532e-01
-1.87015271e+00 -1.75007701e-01 2.40483433e-01 -4.32661206e-01
-1.08422184e+00 -4.48858231e-01 1.10641730e+00 -6.19304836e-01
7.21754849e-01 3.94000947e-01 7.23041952e-01 1.19513392e+00
1.16441086e-01 7.97438264e-01 1.39325809e+00 -5.31880915e-01
7.12867901e-02 2.45337158e-01 -4.70784664e-01 8.43178093e-01
-1.66998789e-01 2.18625128e-01 -5.47063351e-01 -4.55188096e-01
1.48989987e+00 -8.92405026e-03 -2.33162940e-01 -1.21417725e+00
-1.47380793e+00 6.76384091e-01 1.99134454e-01 -3.08637649e-01
-5.09076715e-01 3.87958944e-01 -2.24225029e-01 -2.68840402e-01
8.76401603e-01 3.46354574e-01 -4.08453703e-01 2.40409568e-01
-9.01469469e-01 6.57060444e-01 6.96280777e-01 1.19231606e+00
8.60682845e-01 1.78740695e-01 1.32467195e-01 6.38894856e-01
4.73734260e-01 6.64254427e-01 -8.41350704e-02 -1.57740247e+00
2.32645512e-01 2.92144418e-01 2.71825939e-01 -6.85998738e-01
-5.82571365e-02 -1.65655464e-01 -1.49778843e-01 7.10765243e-01
3.03797752e-01 3.53988409e-01 -1.31522191e+00 1.26921821e+00
7.01082110e-01 3.23883563e-01 -2.70928532e-01 8.64659429e-01
5.87109089e-01 3.57985854e-01 -4.97048616e-01 5.73958829e-02
9.87580895e-01 -8.75240088e-01 -3.28546554e-01 -4.22095299e-01
-3.99228819e-02 -9.99696493e-01 8.82712662e-01 5.57923913e-01
-1.69970965e+00 -2.88474381e-01 -7.69389570e-01 -2.25310400e-01
1.58933207e-01 -1.71282142e-01 5.30917168e-01 4.13263351e-01
-1.01542318e+00 5.21851063e-01 -1.06291771e+00 -9.53078121e-02
6.84389830e-01 1.94649026e-01 -6.35078907e-01 -3.18979383e-01
-1.13995343e-01 8.02350402e-01 2.36076176e-01 -2.00331360e-01
-1.46121275e+00 -9.40458715e-01 -9.42161560e-01 -5.06062210e-01
3.47723514e-01 -1.13198781e+00 1.25805891e+00 -7.51755118e-01
-1.44298255e+00 1.45010376e+00 1.46227283e-02 -1.04865938e-01
8.39438021e-01 -3.39027762e-01 4.61947560e-01 3.85529637e-01
2.75011063e-02 8.59153092e-01 1.34157777e+00 -2.31451941e+00
-1.69846416e-01 -6.28601372e-01 5.78657649e-02 4.88235116e-01
3.54122162e-01 -9.42234099e-02 -6.12582564e-01 -7.46295631e-01
7.45266318e-01 -9.72461283e-01 -4.82457906e-01 4.88662630e-01
-5.53400517e-01 3.67085367e-01 9.20333743e-01 -6.03678882e-01
1.17408462e-01 -1.80352914e+00 6.72015548e-01 1.65545672e-01
2.31137723e-01 -4.40066963e-01 -4.68691364e-02 1.84136197e-01
-1.07085273e-01 -8.32787156e-02 -5.51128864e-01 -1.07419538e+00
-7.08532184e-02 3.70715141e-01 -3.76558721e-01 8.54848921e-01
1.14037707e-01 8.21781456e-01 -6.90481603e-01 -3.80438656e-01
6.31466866e-01 8.02089632e-01 -8.44621122e-01 4.57405776e-01
-6.75781786e-01 6.56024396e-01 -4.37828511e-01 5.88022411e-01
8.29417408e-01 -1.58941701e-01 -1.28366396e-01 -3.53713155e-01
1.20582901e-01 -5.80482408e-02 -1.34979451e+00 2.14802504e+00
-4.03425425e-01 1.81240082e-01 6.06808841e-01 -7.24036276e-01
6.50776505e-01 2.38355249e-01 6.84728026e-01 8.53192583e-02
-1.50588304e-02 5.34192435e-02 -6.79742217e-01 -3.33246648e-01
1.84565887e-01 -3.88735861e-01 2.97483683e-01 6.82461977e-01
7.91226029e-02 -1.43021512e+00 -6.00928068e-01 3.32648814e-01
7.53285050e-01 9.19776618e-01 1.32967811e-02 2.99277250e-02
5.77041805e-02 -3.17436531e-02 7.99992010e-02 3.83437097e-01
4.03228432e-01 1.63327289e+00 1.05131440e-01 -4.81422991e-01
-1.50103390e+00 -1.53559506e+00 -3.68382156e-01 4.63924170e-01
1.07028512e-02 -1.55660555e-01 -7.37687051e-01 -7.49678254e-01
2.16301098e-01 7.90673137e-01 -6.70308650e-01 3.87076825e-01
-7.26358712e-01 -4.72030222e-01 -4.65827696e-02 3.70701432e-01
5.99982142e-02 -7.98128247e-01 -5.74892640e-01 4.43242192e-02
-9.28816721e-02 -1.23812962e+00 -4.75511789e-01 1.97073236e-01
-1.09501326e+00 -1.09177351e+00 -6.12121701e-01 -6.75664723e-01
1.01874161e+00 3.05280775e-01 1.41850019e+00 -9.47495997e-02
-4.37207609e-01 1.07400095e+00 -8.34421515e-02 -3.60609531e-01
-6.98285222e-01 -5.18993497e-01 -1.12718921e-02 1.25843883e-01
-3.95110339e-01 -8.11080575e-01 -5.09394407e-01 3.24034363e-01
-1.05280924e+00 5.18480957e-01 3.29760551e-01 5.58803499e-01
1.02100646e+00 -1.81161925e-01 -2.62629300e-01 -1.10108244e+00
-1.29598528e-01 -2.06833526e-01 -6.20292842e-01 -2.31530145e-02
-9.52914357e-02 5.21311350e-02 1.41178608e-01 -4.72434908e-01
-1.36324334e+00 6.23906910e-01 -1.34880887e-02 -9.37017620e-01
-3.88979346e-01 -3.29658419e-01 -1.75465539e-01 -2.34786764e-01
8.05225074e-01 2.89076239e-01 2.58228123e-01 -7.19488978e-01
7.78204739e-01 1.79220274e-01 4.03826505e-01 -8.73183310e-01
1.29198182e+00 1.20781231e+00 -3.01915947e-02 -6.14036918e-01
-8.63048851e-01 -2.13550612e-01 -1.17584288e+00 -1.60285681e-01
8.59313071e-01 -9.76452649e-01 -1.90997660e-01 2.76053369e-01
-1.27144098e+00 -4.95523840e-01 -5.43932796e-01 3.42017174e-01
-1.38511992e+00 3.71282548e-01 -4.65170473e-01 -7.46952951e-01
3.84870395e-02 -1.28372383e+00 2.00318480e+00 -4.32581335e-01
-2.16985315e-01 -9.83878076e-01 -1.24821216e-01 8.31068456e-01
7.36135542e-02 6.66951180e-01 7.32158065e-01 -6.34902567e-02
-1.24411190e+00 -3.24600004e-03 1.09501578e-01 4.77633923e-01
7.17451051e-02 -6.66712299e-02 -1.20334280e+00 -1.81163654e-01
3.58797193e-01 -4.55775559e-01 5.44601500e-01 5.73175728e-01
1.21009934e+00 -2.39775971e-01 -4.25755739e-01 1.01903069e+00
1.40441108e+00 -3.34458053e-01 4.46053207e-01 1.33462071e-01
1.04641223e+00 7.07364380e-01 4.58055615e-01 3.01142812e-01
2.33173758e-01 8.78915071e-01 7.57455230e-01 -8.07907805e-02
-4.09734607e-01 -4.62033063e-01 1.72753483e-01 2.85323262e-01
-2.60906368e-01 5.79999723e-02 -7.57785678e-01 2.79171735e-01
-1.45363903e+00 -7.00471342e-01 1.21497558e-02 2.08102202e+00
7.67028630e-01 1.47002831e-01 7.77711272e-02 -2.51121938e-01
3.07610482e-01 1.78540707e-01 -7.18472242e-01 1.39573276e-01
-6.57383725e-02 1.22885212e-01 4.50636953e-01 9.21027303e-01
-7.83423841e-01 1.02201664e+00 7.35645914e+00 5.43180585e-01
-8.36907923e-01 1.17355347e-01 6.22201800e-01 -1.78488538e-01
-9.86877441e-01 2.44768411e-02 -5.57538092e-01 -1.24610282e-01
3.06349397e-01 2.95507371e-01 5.39841592e-01 8.01061988e-01
4.02728468e-02 -5.64487129e-02 -1.46157873e+00 1.07552803e+00
6.03244901e-01 -1.57857668e+00 2.52196521e-01 2.38334879e-01
9.67596114e-01 1.28931463e-01 1.42155573e-01 -4.28233325e-01
5.08185744e-01 -1.09045684e+00 1.16793787e+00 4.82861340e-01
8.89010787e-01 -2.91688651e-01 -1.14769809e-01 5.58299839e-01
-7.63778210e-01 3.24525356e-01 -1.06378317e-01 3.61095905e-01
4.05664742e-01 3.12645137e-01 -1.04124951e+00 4.86620694e-01
5.65178514e-01 7.51885176e-01 -4.47532088e-01 7.17689395e-01
-1.04882576e-01 1.96179569e-01 -5.24875939e-01 6.91438854e-01
-7.69293904e-02 -3.85339469e-01 8.65801275e-01 7.62722850e-01
8.07449371e-02 9.42310542e-02 3.58663917e-01 1.17711735e+00
1.72095418e-01 -1.84445396e-01 -8.30119312e-01 3.70108813e-01
1.30893812e-01 1.22579038e+00 -1.05030453e+00 -2.06794962e-01
-1.22447416e-01 1.22040415e+00 3.53489161e-01 5.02020419e-01
-7.00202405e-01 6.10312521e-01 4.54778373e-01 8.01507592e-01
3.55243117e-01 -5.31729519e-01 -4.66276020e-01 -1.27378130e+00
8.49846900e-02 -8.30322087e-01 -5.73863424e-02 -1.25896955e+00
-1.22086632e+00 4.19477075e-01 5.55939078e-01 -1.23129010e+00
-2.98421919e-01 -7.15128005e-01 -2.86671549e-01 7.50300229e-01
-1.01842177e+00 -1.64081275e+00 -7.23683015e-02 3.58417749e-01
1.02312493e+00 1.33184209e-01 8.85146558e-01 -3.36188644e-01
2.07513645e-01 -5.24403751e-02 -1.02559596e-01 -3.53322595e-01
4.69212413e-01 -1.41458476e+00 5.13391912e-01 4.09937859e-01
6.44519746e-01 4.57251400e-01 8.21513712e-01 -6.35990858e-01
-1.75111532e+00 -8.67054284e-01 -8.16363394e-02 -1.50743294e+00
1.54398948e-01 -7.82923341e-01 -6.73096299e-01 1.26569390e+00
1.34155869e-01 4.21086609e-01 2.02965930e-01 -3.16269934e-01
-4.17853743e-01 2.75688320e-01 -1.38961136e+00 4.70808387e-01
1.29745376e+00 -5.13872087e-01 -6.50837183e-01 8.37922692e-01
5.81129968e-01 -1.10142279e+00 -7.60192275e-01 3.17172050e-01
6.10567868e-01 -1.19515169e+00 1.57054377e+00 -5.41471481e-01
4.30683404e-01 -3.46320242e-01 -3.03283364e-01 -1.41274035e+00
-1.49896711e-01 -5.96586168e-01 -8.10224339e-02 7.35163271e-01
9.27983373e-02 -2.34917209e-01 1.11074460e+00 6.15087509e-01
-3.35914791e-01 -8.71719182e-01 -8.16506803e-01 -4.16806608e-01
3.00445911e-02 -6.47994280e-01 2.60621071e-01 6.65171146e-01
-8.87398601e-01 4.96869646e-02 -2.19450846e-01 3.18871230e-01
1.14559364e+00 3.41306955e-01 1.04182506e+00 -1.07718635e+00
-4.76292402e-01 -1.56209350e-01 -4.11215633e-01 -1.21722174e+00
3.56599182e-01 -8.01461637e-01 1.42836124e-01 -1.43024707e+00
2.77504832e-01 -4.55386162e-01 5.36930680e-01 6.20951578e-02
1.43565461e-01 5.80159187e-01 -6.98585361e-02 1.23744227e-01
-1.90269694e-01 5.99117398e-01 1.58028221e+00 5.71576543e-02
7.07466826e-02 -8.11530575e-02 -3.96289319e-01 1.11415339e+00
1.47951454e-01 -4.19377148e-01 -5.61458886e-01 -7.24991739e-01
-9.87485200e-02 -7.19816461e-02 6.99123085e-01 -5.47545075e-01
-2.24932045e-01 -3.46262872e-01 8.59126806e-01 -8.48940372e-01
1.08273184e+00 -1.01742578e+00 4.65258211e-01 -1.01225920e-01
-2.48656794e-01 -1.12503216e-01 7.62422979e-02 7.70976365e-01
2.78569013e-01 -4.04026406e-03 7.74321616e-01 -7.16325045e-01
-1.90854013e-01 5.14461935e-01 3.32534388e-02 -1.43413339e-02
1.14486754e+00 -5.42299747e-01 2.19215527e-01 -3.15349191e-01
-1.02563417e+00 -7.91223422e-02 1.18929517e+00 3.68771166e-01
1.00673044e+00 -1.28945637e+00 -8.15001607e-01 6.68468773e-01
-1.06498785e-01 8.01485896e-01 1.54835433e-01 2.82257020e-01
-7.99536109e-01 -4.23343144e-02 5.27426116e-02 -1.26144469e+00
-1.30722070e+00 3.65634799e-01 6.10102117e-01 2.70455927e-01
-1.22755742e+00 1.12400472e+00 6.08944952e-01 -8.05515707e-01
1.92893520e-01 -2.12720484e-01 3.99433821e-01 -3.18804145e-01
2.78101802e-01 1.56599954e-02 1.18638072e-02 -9.15972710e-01
-7.95323625e-02 1.01929963e+00 1.24794450e-02 -7.06848979e-01
1.64937139e+00 -1.68997154e-01 -5.83224930e-02 5.78481674e-01
1.04416692e+00 3.02062213e-01 -1.99242175e+00 -2.50106663e-01
-4.10315186e-01 -9.85526741e-01 1.12914652e-01 -3.84281218e-01
-9.13599014e-01 6.76143944e-01 3.08216333e-01 -1.51822060e-01
7.27405190e-01 5.99491954e-01 4.12600309e-01 8.65233466e-02
6.00876570e-01 -6.64613664e-01 4.87364173e-01 7.82964453e-02
1.32867706e+00 -1.16136026e+00 4.15129513e-01 -9.16677296e-01
-5.59468031e-01 1.12407196e+00 6.39029980e-01 -4.52923089e-01
7.09282517e-01 5.66793799e-01 1.18717998e-01 -5.43099940e-01
-5.41580319e-01 2.04506174e-01 4.77225065e-01 8.48681509e-01
1.73665881e-01 -7.11366832e-02 7.01399028e-01 -1.31407827e-01
-3.23449790e-01 -4.76985335e-01 5.37457407e-01 9.97028768e-01
-1.98943451e-01 -1.07487011e+00 -8.20910156e-01 3.37789595e-01
-1.04581073e-01 2.26820067e-01 -4.17367071e-01 6.50633633e-01
3.20067108e-02 2.60242015e-01 2.25660533e-01 -2.34951247e-02
4.18330461e-01 6.45748228e-02 1.29716337e+00 -1.25343716e+00
-1.25485554e-01 5.91822565e-01 -1.38841346e-02 -8.81582677e-01
-6.42166853e-01 -8.87981534e-01 -1.07349145e+00 2.62355268e-01
-1.42437026e-01 -3.77673745e-01 6.01826966e-01 7.76307702e-01
4.54623066e-02 1.03328370e-01 6.43724561e-01 -1.74297249e+00
-4.20022428e-01 -6.33527458e-01 -6.42805636e-01 6.91701055e-01
4.53257829e-01 -8.60007048e-01 -6.14991367e-01 4.10405755e-01] | [8.634632110595703, -3.029369831085205] |
3cc1d4f9-89a6-46dd-aad9-d7dcf87651ff | learning-efficient-online-3d-bin-packing-on | null | null | https://openreview.net/forum?id=bfuGjlCwAq | https://openreview.net/pdf?id=bfuGjlCwAq | Learning Efficient Online 3D Bin Packing on Packing Configuration Trees | Online 3D Bin Packing Problem (3D-BPP) has widespread applications in industrial automation and has aroused enthusiastic research interest recently. Existing methods usually solve the problem with limited resolution of spatial discretization, and/or cannot deal with complex practical constraints well. We propose to enhance the practical applicability of online 3D-BPP via learning on a novel hierarchical representation – packing configuration tree (PCT). PCT is a full-fledged description of the state and action space of bin packing which can support packing policy learning based on deep reinforcement learning (DRL). The size of the packing action space is proportional to the number of leaf nodes, making the DRL model easy to train and well-performing even with continuous solution space. During training, PCT expands based on heuristic rules, however, the DRL model learns a much more effective and robust packing policy than heuristic methods. Through extensive evaluation, we demonstrate that our method outperforms all existing online BPP methods and is versatile in terms of incorporating various practical constraints. | ['Kai Xu', 'Yang Yu', 'Hang Zhao'] | 2021-09-29 | null | null | null | iclr-2022-4 | ['3d-bin-packing'] | ['miscellaneous'] | [-2.43557274e-01 1.52238309e-01 -5.53208053e-01 -6.73903199e-03
-3.54327500e-01 -5.89021802e-01 -8.59796107e-02 3.41814220e-01
1.10040590e-01 1.02801275e+00 -1.83601931e-01 -8.07764173e-01
-4.45822090e-01 -9.31727529e-01 -9.81156707e-01 -7.15680122e-01
-5.44082880e-01 1.03040385e+00 2.34711349e-01 8.22861344e-02
9.53813940e-02 7.84312665e-01 -1.09411860e+00 1.37383252e-01
8.24779749e-01 1.46175885e+00 6.62592769e-01 4.44944352e-01
-2.94939503e-02 4.57951903e-01 -7.21970201e-01 1.38117388e-01
6.89966619e-01 -4.52548079e-02 -9.16915596e-01 6.41022444e-01
-2.00412929e-01 -3.13831121e-01 -6.56424105e-01 7.49305129e-01
3.07346046e-01 3.80520880e-01 3.31307620e-01 -1.17860818e+00
-6.34942412e-01 6.57793283e-01 -4.94995236e-01 3.46085906e-01
1.03007212e-01 4.39086467e-01 8.33140671e-01 -1.21098295e-01
3.19327176e-01 1.15295720e+00 4.95745778e-01 1.44881457e-01
-1.32364070e+00 -1.52507991e-01 5.46647012e-01 3.57852042e-01
-9.55570579e-01 2.61711299e-01 4.41320360e-01 -1.94015354e-01
1.50124466e+00 -1.75846472e-01 1.20891809e+00 8.63459826e-01
6.13812506e-01 8.90896559e-01 9.79306459e-01 -1.97677210e-01
8.46534312e-01 -5.21268010e-01 -1.44879967e-01 5.45501351e-01
5.20297468e-01 3.92526150e-01 -1.52506411e-01 -5.33496477e-02
1.19528270e+00 1.66443929e-01 -8.47752467e-02 -1.19777620e+00
-7.78591216e-01 9.17670488e-01 5.49866080e-01 -1.57017231e-01
-3.12022716e-01 1.10236481e-01 5.49186409e-01 4.17757690e-01
1.11219056e-01 8.79709005e-01 -8.17858756e-01 -3.17832291e-01
-7.15577185e-01 6.16529882e-01 1.20051873e+00 1.64565480e+00
5.04355073e-01 1.48776313e-02 -3.37352216e-01 5.74121952e-01
-1.85640559e-01 5.19140005e-01 1.19315624e-01 -9.91015553e-01
8.30308020e-01 3.16518843e-01 6.42722964e-01 -4.92251754e-01
-6.16844177e-01 -2.57482648e-01 -7.99485743e-01 1.49308555e-02
2.63817012e-01 -2.73040444e-01 -1.16938353e+00 1.13952160e+00
3.65657300e-01 -2.30741650e-01 -5.96506745e-02 8.16136718e-01
-1.34686142e-01 9.00976479e-01 -1.85493842e-01 -5.01606703e-01
9.80865598e-01 -1.25937593e+00 -7.03327179e-01 -3.76630932e-01
4.84608114e-01 -1.97935179e-01 6.37559175e-01 3.75343025e-01
-1.21610534e+00 -4.73568916e-01 -1.19203889e+00 3.88659954e-01
-4.76297557e-01 -3.61553490e-01 9.34974790e-01 4.99928981e-01
-7.04670787e-01 1.03942895e+00 -1.16254437e+00 -2.17898320e-02
7.44859934e-01 5.39835095e-01 -1.00569233e-01 -5.44158101e-01
-9.50381637e-01 1.27622581e+00 9.20468688e-01 2.24727482e-01
-9.94163513e-01 -3.84597242e-01 -1.09454465e+00 3.37025434e-01
9.49289024e-01 -4.61720586e-01 1.67949975e+00 2.35151738e-01
-1.78775930e+00 4.30080295e-02 2.63277978e-01 -8.76704276e-01
3.15746635e-01 -7.49513581e-02 -7.59124458e-02 2.64189273e-01
-1.11593924e-01 4.29387152e-01 7.01139867e-01 -9.11636591e-01
-8.39570999e-01 -1.96714878e-01 4.32398736e-01 2.12873191e-01
1.18114196e-01 -7.54647136e-01 -1.83069497e-01 -9.35862437e-02
1.05523393e-01 -7.82920957e-01 -7.27704644e-01 -2.68241376e-01
-1.67446464e-01 -3.27276349e-01 7.79884875e-01 -4.23231751e-01
1.07523179e+00 -2.00516248e+00 1.40718281e-01 1.87046900e-01
-1.53969973e-01 4.75337625e-01 -8.94744247e-02 6.21315598e-01
2.26033211e-01 -1.22176573e-01 7.53033757e-02 1.47529140e-01
5.25701165e-01 9.30423737e-01 -3.34859639e-01 4.47782844e-01
2.21550927e-01 1.00213122e+00 -1.11197102e+00 -2.63875455e-01
6.30795836e-01 -3.49740177e-01 -7.62429595e-01 2.53622115e-01
-7.34319687e-01 4.87263277e-02 -6.39317572e-01 7.09553719e-01
9.09457743e-01 -4.57231313e-01 4.79763806e-01 1.75783008e-01
-1.53786791e-02 3.94897968e-01 -1.06052887e+00 1.57830131e+00
-5.05834281e-01 2.48150695e-02 2.21022740e-01 -1.52527261e+00
8.37402463e-01 5.81056736e-02 6.20886207e-01 -8.20202172e-01
1.21167615e-01 4.65060398e-02 -3.76255549e-02 -4.62539136e-01
2.31572628e-01 -5.74347898e-02 -2.31040508e-01 1.18138105e-01
1.76311806e-01 -7.04883873e-01 7.16728568e-01 -5.05589783e-01
1.27977133e+00 1.11787722e-01 4.98763740e-01 -2.13392287e-01
-2.95874458e-02 9.90088135e-02 7.58267403e-01 1.06123614e+00
-2.66682267e-01 -1.60575539e-01 7.52430618e-01 -6.15388632e-01
-1.09108269e+00 -1.13847518e+00 -1.86840564e-01 6.20384514e-01
6.75103724e-01 -1.48571029e-01 -4.68901664e-01 -7.60102928e-01
7.71213233e-01 7.48275697e-01 -2.97744811e-01 -1.30106106e-01
-5.88758767e-01 -4.06097293e-01 -3.93492430e-01 9.51627493e-01
5.50441206e-01 -1.20047498e+00 -9.07347023e-01 7.51472473e-01
2.57743150e-01 -1.38875508e+00 -3.69087100e-01 9.60589647e-01
-1.07331967e+00 -1.08546257e+00 -4.07668859e-01 -1.03947306e+00
5.51033318e-01 4.05286908e-01 9.55716848e-01 -4.82786953e-01
-5.36523163e-01 2.34013766e-01 -5.85536003e-01 -4.02790219e-01
-1.54024050e-01 2.75179952e-01 2.72693913e-02 -8.86553526e-01
2.43942246e-01 -5.31409860e-01 -3.89334530e-01 4.76795137e-01
-5.00599623e-01 -6.17664605e-02 7.83898890e-01 9.77423787e-01
6.94016695e-01 7.12531507e-01 4.06126469e-01 -5.57875872e-01
5.52509427e-01 -1.29629046e-01 -1.16112256e+00 1.28461540e-01
-3.53209943e-01 7.27267563e-02 8.78831744e-01 -4.35376912e-01
-6.40045404e-01 -1.19856901e-01 1.58995073e-02 -6.93223357e-01
-1.12822145e-01 5.21195471e-01 -1.08624540e-01 1.41841665e-01
-3.24848033e-02 -7.90894032e-02 -4.79906723e-02 -3.10271055e-01
2.43248329e-01 3.37628663e-01 8.84752646e-02 -9.52663541e-01
5.92809558e-01 -8.25776309e-02 1.33908093e-01 -4.25464571e-01
-1.17308068e+00 -4.37951297e-01 -6.87823951e-01 2.18723193e-01
6.12977147e-01 -5.31905591e-01 -1.05082142e+00 1.11678772e-01
-8.84037971e-01 -9.77378845e-01 -7.39114165e-01 2.91894615e-01
-1.34978747e+00 2.05945417e-01 -9.29805100e-01 -9.84218657e-01
6.42280374e-03 -1.15716124e+00 9.90456343e-01 7.30813965e-02
2.07042694e-01 -7.73185432e-01 -4.09676909e-01 8.89613479e-02
3.13936204e-01 3.23629320e-01 1.30534697e+00 -2.90689111e-01
-7.07075953e-01 -1.45606428e-01 -1.32022277e-01 3.15928161e-01
2.46846050e-01 -5.88148713e-01 -1.61427677e-01 -7.09980309e-01
5.10478280e-02 -5.66385329e-01 4.83313352e-01 7.05123246e-01
1.87296104e+00 -5.32461584e-01 -5.53522468e-01 3.04806709e-01
1.35571909e+00 5.24778903e-01 3.08004826e-01 4.73246723e-01
1.72183052e-01 5.93788102e-02 1.12262452e+00 7.64272273e-01
6.41632527e-02 3.34506482e-01 8.07436049e-01 1.14019409e-01
3.26300770e-01 -4.00788099e-01 4.28541452e-02 5.96544027e-01
1.97972715e-01 -4.36192989e-01 -7.46782959e-01 4.24944133e-01
-2.20357108e+00 -8.80582869e-01 8.20947945e-01 2.15667892e+00
7.30200350e-01 7.34899938e-01 8.58752355e-02 1.10509761e-01
7.41483569e-01 5.71566634e-02 -1.05340374e+00 -8.51474345e-01
3.95286620e-01 3.28201562e-01 9.30356085e-01 3.40666920e-01
-1.40578091e+00 7.66602755e-01 6.80749750e+00 7.13546753e-01
-5.28845549e-01 -2.73351878e-01 4.22280848e-01 -7.21242428e-02
5.50427198e-01 -3.42856854e-01 -1.05047214e+00 5.76776147e-01
8.39077950e-01 -1.82592958e-01 8.59688103e-01 1.27599633e+00
8.88997987e-02 -8.15654546e-02 -1.45060480e+00 8.81026030e-01
-4.69568700e-01 -1.57488275e+00 -1.87483758e-01 4.27949905e-01
9.46622431e-01 -1.43355593e-01 -4.51349979e-03 7.83570051e-01
7.54432797e-01 -8.81895006e-01 5.88403821e-01 -1.39881760e-01
4.79940355e-01 -1.06093740e+00 9.47384298e-01 5.31075776e-01
-1.16005492e+00 -9.89953458e-01 -9.55016673e-01 -1.94940671e-01
4.02682602e-01 4.98886108e-01 -9.23741102e-01 6.56794488e-01
6.69366717e-01 5.29665887e-01 1.82836384e-01 1.43216527e+00
-2.20825270e-01 1.98652178e-01 -4.39363539e-01 -3.26839805e-01
7.51575828e-01 -2.22129256e-01 2.11153060e-01 8.26381743e-01
2.25399017e-01 9.57175642e-02 1.11025763e+00 8.53935361e-01
1.90134928e-01 -4.70987260e-01 -5.01029611e-01 -3.33328456e-01
6.77700460e-01 7.64886141e-01 -7.00581312e-01 -6.04077019e-02
-2.26415187e-01 5.76016307e-01 5.28505206e-01 1.79330051e-01
-9.99216199e-01 -2.16820955e-01 6.04545832e-01 -6.40593916e-02
1.35407519e+00 -6.63279951e-01 -4.98677939e-02 -5.12334287e-01
-2.73096934e-02 -7.12613940e-01 2.93307364e-01 -4.56457764e-01
-1.56968951e+00 2.20101103e-02 2.12766081e-01 -1.10758424e+00
-1.35256186e-01 -1.21635807e+00 -2.43446141e-01 4.64203179e-01
-1.64837849e+00 -5.87882638e-01 9.08560678e-02 2.31360465e-01
7.93689609e-01 1.09023370e-01 8.90480161e-01 -1.90908134e-01
-5.39972842e-01 2.81127959e-01 2.70062119e-01 -2.44462550e-01
-1.13030514e-02 -1.50610793e+00 4.62434560e-01 3.66472095e-01
-4.29572612e-01 1.63920626e-01 7.36880362e-01 -5.52670121e-01
-1.82086170e+00 -1.28062010e+00 1.68340534e-01 2.88295783e-02
8.04608881e-01 -3.65667850e-01 -4.86633092e-01 5.96968889e-01
2.48464867e-02 1.33989424e-01 3.29630584e-01 2.24094570e-01
9.57002267e-02 -2.66379416e-01 -1.23563647e+00 3.69897783e-01
1.24750900e+00 3.92893218e-02 -6.63149297e-01 7.89263308e-01
1.09265697e+00 -1.08591342e+00 -1.06120086e+00 4.87381488e-01
6.52756765e-02 -6.17134690e-01 6.61279619e-01 -6.88572764e-01
2.36125365e-01 -4.97249365e-02 -1.11615270e-01 -1.51962411e+00
-8.29378843e-01 -7.99510241e-01 -9.06924307e-01 4.86826837e-01
2.13101611e-01 -5.94755828e-01 8.95882249e-01 4.61913049e-01
-6.64491355e-01 -1.52910316e+00 -1.07008636e+00 -1.31190813e+00
2.35219076e-01 1.42773613e-01 9.56447482e-01 4.19687241e-01
5.85010767e-01 4.48604822e-02 -4.56790887e-02 4.13264811e-01
3.88902307e-01 8.55843306e-01 4.82447028e-01 -8.67070913e-01
-5.47496617e-01 -4.11654770e-01 -2.42970929e-01 -1.68846011e+00
5.68239056e-02 -6.13517404e-01 3.58316869e-01 -1.88786566e+00
-1.19691901e-01 -6.70924902e-01 -3.81557316e-01 6.63007736e-01
3.55828941e-01 -6.05599701e-01 2.29659483e-01 -3.27841550e-01
-8.74388933e-01 7.21995413e-01 1.92084908e+00 -3.02681953e-01
-2.97831208e-01 2.81354904e-01 -3.85496169e-01 2.29403436e-01
9.42796946e-01 1.51596397e-01 -4.29014891e-01 -3.39104354e-01
-1.49385393e-01 5.48040807e-01 -2.01842859e-01 -8.89379382e-01
8.28484371e-02 -4.44015145e-01 6.00880742e-01 -9.12975371e-01
2.48979017e-01 -1.14115870e+00 -4.21440601e-01 6.08713090e-01
-2.78717503e-02 1.06951706e-01 5.21699667e-01 8.88088763e-01
9.83142480e-02 -2.25447863e-01 6.58824801e-01 -4.38155204e-01
-6.27129257e-01 7.89757490e-01 -4.68780994e-01 5.66107454e-03
1.33362985e+00 -4.18461978e-01 -2.34924570e-01 3.19599286e-02
-8.95382941e-01 9.68832314e-01 2.61584878e-01 2.87604600e-01
3.32563370e-01 -1.25443852e+00 2.59284601e-02 3.88776422e-01
-3.32962155e-01 6.49863660e-01 1.63481861e-01 3.89612705e-01
-5.93458056e-01 8.45725477e-01 -3.85580212e-01 -5.59679449e-01
-3.88165206e-01 1.35680461e+00 2.79359788e-01 -8.38780284e-01
-1.19149101e+00 5.42801619e-01 -2.64568895e-01 -1.59846410e-01
6.30065322e-01 -7.16829002e-01 2.64088839e-01 -2.87483573e-01
1.44978821e-01 3.56058478e-01 1.70926496e-01 4.45306510e-01
-9.01534185e-02 -4.84687723e-02 -3.06808114e-01 5.29623330e-01
1.62566698e+00 1.33580506e-01 2.33663946e-01 1.20650619e-01
6.79936647e-01 -9.94921684e-01 -1.86338627e+00 -7.96313658e-02
-7.42319450e-02 -6.23058736e-01 -9.69140679e-02 -7.45408356e-01
-8.35569382e-01 7.18701601e-01 2.10342586e-01 6.59901977e-01
7.07609594e-01 -1.46329880e-01 1.00116611e+00 8.31613660e-01
1.12121892e+00 -1.66391611e+00 2.31130674e-01 9.22684431e-01
6.64685130e-01 -8.04734170e-01 1.35325447e-01 -5.85434198e-01
-4.23114240e-01 1.08865845e+00 8.34132731e-01 -4.58730400e-01
5.83370566e-01 5.63981473e-01 -6.38325453e-01 9.07403752e-02
-9.14067864e-01 -1.66179508e-01 -3.50881159e-01 8.85019600e-01
-3.48259568e-01 2.78974026e-01 3.82719524e-02 4.16621417e-01
-3.34561974e-01 -2.83008814e-01 2.15702146e-01 1.29673207e+00
-7.44270623e-01 -9.30776954e-01 -1.22629769e-01 6.90854073e-01
1.61454201e-01 4.45979148e-01 3.02553117e-01 1.00874722e+00
8.38736352e-03 8.59382033e-01 4.71237689e-01 6.18736353e-03
3.94787788e-01 -5.57622267e-03 1.17051780e+00 -8.30117583e-01
1.01523876e-01 -7.08914548e-02 -1.08806472e-02 -8.07513177e-01
1.31497383e-01 -6.17932916e-01 -1.07701659e+00 -3.08873773e-01
-6.88302875e-01 3.07389647e-01 2.94729799e-01 1.05797398e+00
3.29946697e-01 7.66019702e-01 7.78795660e-01 -1.32645094e+00
-1.30221009e+00 -6.37831807e-01 -1.04427254e+00 -1.31585319e-02
2.88862526e-01 -1.25157213e+00 -2.96204686e-02 -5.21062613e-01] | [4.958523273468018, 2.6823055744171143] |
f81fb7c7-7f70-41ef-aba6-e5b149dade66 | analyzing-bert-cross-lingual-transfer | null | null | https://aclanthology.org/2022.mmmpie-1.3 | https://aclanthology.org/2022.mmmpie-1.3.pdf | Analyzing BERT Cross-lingual Transfer Capabilities in Continual Sequence Labeling | Knowledge transfer between neural language models is a widely used technique that has proven to improve performance in a multitude of natural language tasks, in particular with the recent rise of large pre-trained language models like BERT. Similarly, high cross-lingual transfer has been shown to occur in multilingual language models. Hence, it is of great importance to better understand this phenomenon as well as its limits. While most studies about cross-lingual transfer focus on training on independent and identically distributed (i.e. i.i.d.) samples, in this paper we study cross-lingual transfer in a continual learning setting on two sequence labeling tasks: slot-filling and named entity recognition. We investigate this by training multilingual BERT on sequences of 9 languages, one language at a time, on the MultiATIS++ and MultiCoNER corpora. Our first findings are that forward transfer between languages is retained although forgetting is present. Additional experiments show that lost performance can be recovered with as little as a single training epoch even if forgetting was high, which can be explained by a progressive shift of model parameters towards a better multilingual initialization. We also find that commonly used metrics might be insufficient to assess continual learning performance. | ['Sophie Rosset', 'Olivier Galibert', 'Hervé Bredin', 'Guillaume Bernard', 'Sahar Ghannay', 'Mathilde Veron', 'Juan Manuel Coria'] | null | null | null | null | mmmpie-coling-2022-10 | ['slot-filling'] | ['natural-language-processing'] | [ 4.73239496e-02 -2.09425092e-02 -1.57527164e-01 -2.78952658e-01
-7.75056899e-01 -7.85920322e-01 7.49128342e-01 2.22646505e-01
-1.20907581e+00 1.12403727e+00 -4.70994477e-04 -6.01855338e-01
1.10581115e-01 -4.47170466e-01 -1.08088505e+00 -4.76409137e-01
-1.08859345e-01 7.38645613e-01 3.87036651e-01 -1.84016615e-01
-2.03978419e-01 2.82050520e-01 -1.17600441e+00 2.89545029e-01
9.69017088e-01 8.91387239e-02 4.95338708e-01 4.05978411e-01
-3.70113701e-01 5.21630645e-01 -5.88054478e-01 -6.18363738e-01
4.05413173e-02 -5.19400477e-01 -1.02334499e+00 -1.95772275e-01
2.05601633e-01 1.34658381e-01 5.93150444e-02 8.74479115e-01
4.68501449e-01 1.71214059e-01 5.50720572e-01 -8.90097380e-01
-5.96392214e-01 8.96289408e-01 -5.69491863e-01 3.33335191e-01
-5.70976660e-02 3.96491848e-02 9.21387076e-01 -8.30723345e-01
9.69713092e-01 1.22905803e+00 6.93133354e-01 4.58479553e-01
-1.56416166e+00 -7.76805758e-01 1.95995107e-01 2.15371102e-01
-1.36484802e+00 -5.25885642e-01 3.24332416e-01 -3.65782231e-01
1.26431370e+00 -2.88576275e-01 4.41898644e-01 1.05305350e+00
2.33876541e-01 1.00066686e+00 1.41867959e+00 -9.07970369e-01
2.72963289e-02 6.08073533e-01 -2.18419526e-02 3.85108799e-01
3.39606851e-01 2.28131458e-01 -5.96522391e-01 1.86608374e-01
6.26813114e-01 -6.53758585e-01 -1.20192833e-01 -4.07104343e-01
-1.18975198e+00 8.93654406e-01 2.31521502e-01 9.47683334e-01
-2.41570368e-01 -1.84408262e-01 7.51705945e-01 7.97793865e-01
6.10551953e-01 3.86372030e-01 -9.33854043e-01 -3.65344852e-01
-8.94620955e-01 -2.13499129e-01 7.00741947e-01 7.12626636e-01
8.95303726e-01 8.29148665e-02 2.20004350e-01 1.27264988e+00
-7.79829919e-03 4.08816159e-01 8.76247406e-01 -3.89625967e-01
4.89974529e-01 2.85352439e-01 -1.40806451e-01 -3.91569316e-01
-3.15313131e-01 -5.30752242e-01 -6.21053159e-01 1.48853257e-01
7.14453459e-01 -3.00825864e-01 -7.55556047e-01 2.35158324e+00
1.28584370e-01 7.93253407e-02 2.90396452e-01 4.60627049e-01
3.76860052e-02 5.08677959e-01 6.04447663e-01 -2.71059453e-01
1.37241018e+00 -7.55663633e-01 -5.66027939e-01 -4.82608467e-01
1.16561711e+00 -8.89388859e-01 1.12488306e+00 3.15290779e-01
-8.90011728e-01 -5.45398831e-01 -8.35656703e-01 -1.71663631e-02
-6.76070392e-01 -1.59557983e-01 6.25142395e-01 9.47475672e-01
-1.14815891e+00 6.54970825e-01 -9.24070835e-01 -7.43223250e-01
9.08155963e-02 2.90875435e-01 -6.64443076e-01 -2.82889813e-01
-1.58177042e+00 1.33818007e+00 8.66993845e-01 -2.57918507e-01
-4.82363105e-01 -6.52216315e-01 -6.72071874e-01 -1.32705018e-01
1.10273473e-01 -4.97660935e-01 1.24488008e+00 -1.27803826e+00
-1.19057214e+00 1.18433607e+00 -2.01623902e-01 -6.32207096e-01
5.49672008e-01 -1.17313877e-01 -5.48861325e-01 -3.54735762e-01
1.73898131e-01 7.12088346e-01 3.86148036e-01 -1.17932272e+00
-8.26283336e-01 -4.84373838e-01 -1.97161779e-01 3.54290336e-01
-3.62020224e-01 1.57609165e-01 -1.52841404e-01 -5.25572598e-01
-3.54879200e-01 -1.10681546e+00 1.17700763e-01 -6.11415386e-01
1.60043120e-01 -2.28862777e-01 2.95484602e-01 -9.11986351e-01
1.15817559e+00 -1.82832873e+00 7.74114206e-02 -1.70457482e-01
-5.22908032e-01 6.09256744e-01 -2.71041811e-01 5.44825792e-01
-2.65127271e-01 1.34609461e-01 -3.57199073e-01 -5.26702046e-01
-1.06922433e-01 4.76066738e-01 -1.42363548e-01 4.69119042e-01
2.20788494e-01 9.17084277e-01 -8.52783680e-01 -3.38618577e-01
7.44853020e-02 6.28761232e-01 -4.47940022e-01 -2.21333236e-01
-3.74226868e-02 6.70925200e-01 2.45059803e-01 1.37893051e-01
4.23248112e-01 -4.76855375e-02 3.35645378e-01 2.40848169e-01
-2.70341486e-01 4.11224782e-01 -9.78435993e-01 1.94742990e+00
-7.09037960e-01 6.15484297e-01 -2.06902593e-01 -9.65833485e-01
6.05499029e-01 6.89201653e-01 1.04106791e-01 -1.08473527e+00
8.07181874e-04 6.74620211e-01 5.56580961e-01 -4.74570468e-02
5.49055338e-01 -8.12881887e-01 1.60889048e-02 6.34729981e-01
4.22600567e-01 2.25554690e-01 3.80534291e-01 -1.03002992e-02
7.64134169e-01 1.73888564e-01 4.53305840e-01 -3.99183720e-01
6.03531599e-01 3.61921862e-02 4.33109522e-01 6.47898674e-01
-2.30494455e-01 1.15086399e-01 1.70460075e-01 4.57130186e-02
-1.12795377e+00 -8.66897047e-01 -3.76734406e-01 1.52965772e+00
-3.48623186e-01 1.09066553e-02 -6.83077037e-01 -6.81249022e-01
-1.08101666e-01 1.18219709e+00 -5.35815179e-01 -3.02082598e-01
-7.94888139e-01 -8.83187771e-01 6.82831645e-01 3.43380481e-01
2.02400520e-01 -1.43475461e+00 -3.37324083e-01 4.85239953e-01
5.32244407e-02 -1.00507259e+00 -2.60041147e-01 5.05910814e-01
-1.11194479e+00 -7.02368855e-01 -1.05352044e+00 -9.88223433e-01
3.79625142e-01 2.58528851e-02 1.39228368e+00 -9.51460227e-02
9.45829079e-02 4.04033989e-01 -2.86199421e-01 -3.87913883e-01
-7.21879005e-01 6.41186118e-01 2.87413448e-01 -2.04147682e-01
6.38325632e-01 -5.30022264e-01 -9.30147618e-02 2.49460369e-01
-1.09222734e+00 -5.21263108e-02 8.54245424e-01 9.97990012e-01
3.75788420e-01 -1.78860947e-01 8.03669631e-01 -1.09045756e+00
6.80846274e-01 -5.02304137e-01 -3.85748386e-01 4.14936185e-01
-6.70700848e-01 4.69550878e-01 5.02470493e-01 -5.19130290e-01
-1.31500781e+00 -2.35027686e-01 -2.99210817e-01 -9.82009321e-02
-4.54401910e-01 7.05827057e-01 -4.22717333e-02 1.03918880e-01
8.46643507e-01 2.98659801e-01 -2.34917983e-01 -5.76464057e-01
5.84606469e-01 5.81165791e-01 2.09356830e-01 -7.63333738e-01
6.23857200e-01 1.35628328e-01 -6.46910787e-01 -1.03006923e+00
-7.26908326e-01 -5.79784214e-01 -9.55927610e-01 4.30597290e-02
6.73600018e-01 -9.66416836e-01 -2.46503934e-01 5.27457356e-01
-1.13504505e+00 -8.81871760e-01 -3.10052037e-01 6.96203113e-01
-4.19404060e-01 9.28606316e-02 -6.93919122e-01 -6.87458694e-01
-1.74584314e-02 -9.46020603e-01 6.42522395e-01 4.36023772e-02
-2.71016508e-01 -1.63793802e+00 4.48615938e-01 -6.65508397e-03
4.25283283e-01 -3.32090259e-01 1.12911081e+00 -9.36278820e-01
-1.88608930e-01 1.12522505e-01 3.63816619e-02 4.79831487e-01
2.60936975e-01 -4.80647504e-01 -9.64951754e-01 -7.10962355e-01
-5.29735945e-02 -4.97565627e-01 8.22450817e-01 8.32584351e-02
2.60801703e-01 1.47628188e-01 -3.40373605e-01 1.39093637e-01
1.48037779e+00 4.64298755e-01 5.31405449e-01 5.90068698e-01
3.92236710e-01 6.90526783e-01 3.53514165e-01 -3.21150839e-01
3.16081047e-01 6.02005064e-01 -2.41501689e-01 -9.55486074e-02
-2.42864951e-01 -3.30344200e-01 5.32208562e-01 1.36962605e+00
1.50558710e-01 -4.81721573e-02 -1.17232263e+00 9.07593668e-01
-1.42930090e+00 -6.19363904e-01 9.62637290e-02 2.52254605e+00
1.14904177e+00 2.81328499e-01 2.54905690e-02 -1.86940040e-02
7.52066851e-01 -1.37101293e-01 -4.76040453e-01 -4.33357894e-01
-3.06261331e-01 3.75929803e-01 5.05679071e-01 6.79772794e-01
-7.01959431e-01 1.34984767e+00 5.51673603e+00 8.54465902e-01
-1.42856312e+00 4.96533632e-01 5.91128170e-01 3.33058983e-01
-1.96814448e-01 1.93229821e-02 -9.64326560e-01 3.26252013e-01
1.47714424e+00 -3.77759278e-01 2.93999106e-01 5.16505480e-01
-1.95746124e-01 -3.11072707e-01 -1.05007064e+00 6.76175892e-01
1.72533337e-02 -7.02088296e-01 -4.30986919e-02 -7.64608681e-02
8.28850806e-01 4.16571498e-01 -3.78320925e-02 8.11825156e-01
4.16313440e-01 -8.82694125e-01 6.43481433e-01 1.30053997e-01
8.48536372e-01 -8.69994164e-01 6.62512481e-01 6.62413418e-01
-9.34018075e-01 3.56952816e-01 -3.08166295e-01 1.09768860e-01
4.15197313e-01 3.66176009e-01 -1.08013928e+00 6.27296984e-01
4.08536345e-01 3.96484524e-01 -6.77274883e-01 1.19885564e+00
-2.39758015e-01 8.42170119e-01 -3.22671294e-01 2.73553818e-01
2.50077009e-01 -1.91410422e-01 2.63317287e-01 1.44448471e+00
3.52473974e-01 -3.77492279e-01 -2.02889353e-01 3.49262297e-01
-8.02277699e-02 5.30764639e-01 -6.36823833e-01 -1.42674610e-01
1.81012258e-01 8.74350071e-01 -8.46036792e-01 -3.77230465e-01
-6.85897827e-01 1.19299507e+00 7.23923743e-01 3.20120424e-01
-6.02176726e-01 -2.50278831e-01 3.83749545e-01 1.77936871e-02
2.59596109e-01 -4.61823851e-01 -1.20932646e-01 -1.16540241e+00
-2.54108042e-01 -7.59896636e-01 4.12028909e-01 -3.24279994e-01
-1.25072384e+00 7.58122563e-01 -4.01564986e-02 -7.09314942e-01
-4.13122147e-01 -6.22138619e-01 -1.40521914e-01 1.15895998e+00
-1.76029801e+00 -9.79600012e-01 3.89310300e-01 6.18155658e-01
5.21161854e-01 -9.46324691e-02 1.00288570e+00 7.11516619e-01
-4.11646098e-01 6.92716062e-01 3.05192560e-01 1.74552158e-01
1.14790273e+00 -1.15703392e+00 5.29376507e-01 8.33386719e-01
5.80256402e-01 8.70315611e-01 7.10686862e-01 -5.81334114e-01
-8.38691533e-01 -9.29509282e-01 1.33443511e+00 -4.40384835e-01
8.36879373e-01 -4.44233418e-01 -1.46233809e+00 1.07701313e+00
5.78356564e-01 -2.54173517e-01 7.66737223e-01 4.67881441e-01
-3.60093445e-01 1.10659339e-01 -7.19575584e-01 6.32297039e-01
1.00999379e+00 -6.71131313e-01 -8.47079575e-01 2.05548361e-01
5.56432903e-01 -7.87669718e-02 -7.78759181e-01 1.87717080e-01
2.47375771e-01 -9.88088965e-01 7.23805845e-01 -6.81878269e-01
-1.43076017e-01 -1.05118826e-01 1.20047018e-01 -1.80335259e+00
-1.00527287e-01 -3.01670432e-01 4.63805586e-01 1.41970062e+00
7.65382469e-01 -9.36829746e-01 5.06780684e-01 2.80860573e-01
-4.88338806e-02 -1.95057020e-01 -1.13602614e+00 -1.20616889e+00
7.70740509e-01 -5.05357623e-01 1.11225694e-01 1.31725824e+00
-4.39139530e-02 7.88648844e-01 -2.47766703e-01 -1.97342589e-01
2.62485474e-01 -3.76838893e-01 5.61290443e-01 -1.25190067e+00
-2.58168012e-01 -3.91694665e-01 -4.92863543e-02 -8.38884950e-01
5.16764104e-01 -1.32072878e+00 1.66428804e-01 -1.10444701e+00
8.41254592e-02 -7.71114826e-01 -6.06202841e-01 4.50411320e-01
-2.51127124e-01 8.33613425e-02 1.49600551e-01 1.81731582e-01
-4.82255459e-01 4.78857636e-01 1.00092864e+00 3.29618156e-01
-3.78050625e-01 -1.55992523e-01 -4.89461452e-01 4.46788430e-01
8.89073372e-01 -7.20785141e-01 -3.49055856e-01 -6.32429361e-01
1.80944562e-01 -2.85013337e-02 -2.01996103e-01 -1.09282219e+00
1.02682658e-01 2.48656854e-01 7.70003200e-02 -7.32476935e-02
8.80043581e-02 -7.61176527e-01 1.91095039e-01 5.28952241e-01
-2.74731129e-01 7.28140995e-02 7.51109004e-01 2.85703361e-01
-3.08074385e-01 -3.57353300e-01 9.61412013e-01 -2.89379865e-01
-8.67123306e-01 -4.14461223e-03 -3.91250193e-01 2.77590781e-01
7.36533940e-01 5.92070483e-02 -2.86685526e-02 -7.01119900e-02
-7.54757524e-01 -2.27197278e-02 5.22712946e-01 7.62847006e-01
-2.25159213e-01 -1.26227582e+00 -7.62497723e-01 1.47859082e-01
1.49836853e-01 -3.90181988e-01 2.88462639e-01 9.33420956e-01
-2.87776411e-01 8.66355777e-01 -2.56135046e-01 -6.54012322e-01
-1.04137754e+00 5.65915942e-01 2.82657295e-01 -6.59506023e-01
-2.81776488e-01 8.34632576e-01 2.76188105e-01 -8.08162808e-01
1.42218731e-02 -1.87744617e-01 -8.53132829e-02 4.61624533e-01
3.18461448e-01 4.68185134e-02 3.31713647e-01 -8.54328156e-01
-2.56618083e-01 3.93920630e-01 -4.68496233e-01 -5.30038714e-01
1.15672779e+00 -2.78522313e-01 9.13607106e-02 1.05150974e+00
1.16186011e+00 9.26526710e-02 -1.07383931e+00 -5.37452817e-01
5.78079283e-01 -2.55446713e-02 -4.11699474e-01 -9.35657561e-01
-7.37460732e-01 1.06890523e+00 6.31366789e-01 -3.16714495e-02
7.32316494e-01 -7.85538033e-02 6.11280978e-01 2.14577049e-01
7.31725872e-01 -1.06367898e+00 -2.90154099e-01 9.25051510e-01
6.65889263e-01 -1.32857680e+00 -4.59395021e-01 5.72926365e-02
-7.03899324e-01 6.59027219e-01 3.83840352e-01 1.00003593e-01
4.96954322e-01 9.70665365e-02 2.91540802e-01 2.45745733e-01
-6.49164975e-01 -3.04912955e-01 -1.58887003e-02 4.93166506e-01
9.61792111e-01 6.04725331e-02 -4.55665559e-01 2.36838162e-01
-3.53381932e-01 7.32529238e-02 2.35776722e-01 9.81402516e-01
-2.80767739e-01 -1.65559602e+00 -2.87100166e-01 6.87991381e-02
-6.50556326e-01 -3.44227493e-01 1.59782004e-02 1.16832864e+00
7.28456900e-02 6.07377648e-01 1.02738567e-01 7.43551776e-02
4.43560690e-01 8.93110991e-01 5.26145339e-01 -7.77733862e-01
-6.77950144e-01 -3.90105657e-02 4.50050022e-04 -8.16633999e-02
-5.88139772e-01 -9.14014518e-01 -1.07239497e+00 -6.74210489e-02
-3.79457891e-01 3.75178397e-01 7.21893668e-01 9.80696976e-01
5.65143712e-02 5.58319211e-01 -4.43907827e-02 -5.37294447e-01
-3.46743852e-01 -1.22601295e+00 -6.61110580e-01 4.74672437e-01
1.18609726e-01 -6.69398665e-01 -4.29830313e-01 2.65103817e-01] | [10.931459426879883, 9.93690013885498] |
84ddfa76-3f69-4b77-9e61-fa232e97d4eb | learnable-sampling-3d-convolution-for-video | 2011.10974 | null | https://arxiv.org/abs/2011.10974v1 | https://arxiv.org/pdf/2011.10974v1.pdf | Learnable Sampling 3D Convolution for Video Enhancement and Action Recognition | A key challenge in video enhancement and action recognition is to fuse useful information from neighboring frames. Recent works suggest establishing accurate correspondences between neighboring frames before fusing temporal information. However, the generated results heavily depend on the quality of correspondence estimation. In this paper, we propose a more robust solution: \emph{sampling and fusing multi-level features} across neighborhood frames to generate the results. Based on this idea, we introduce a new module to improve the capability of 3D convolution, namely, learnable sampling 3D convolution (\emph{LS3D-Conv}). We add learnable 2D offsets to 3D convolution which aims to sample locations on spatial feature maps across frames. The offsets can be learned for specific tasks. The \emph{LS3D-Conv} can flexibly replace 3D convolution layers in existing 3D networks and get new architectures, which learns the sampling at multiple feature levels. The experiments on video interpolation, video super-resolution, video denoising, and action recognition demonstrate the effectiveness of our approach. | ['Dong Chen', 'Jianmin Bao', 'Shuyang Gu'] | 2020-11-22 | null | null | null | null | ['video-denoising', 'video-enhancement'] | ['computer-vision', 'computer-vision'] | [ 1.31453827e-01 -4.16075289e-01 1.78770977e-03 -5.78666687e-01
-6.31678283e-01 -1.16916083e-01 2.93111920e-01 -6.56823575e-01
-3.55441362e-01 5.71099639e-01 6.49042964e-01 2.09269032e-01
7.65692666e-02 -7.37763464e-01 -8.95654798e-01 -6.24006450e-01
-6.55266941e-02 -5.70799172e-01 4.45537060e-01 -9.50626358e-02
2.09564239e-01 6.75691485e-01 -1.46049845e+00 7.71254182e-01
7.79790282e-01 1.27313399e+00 3.79894257e-01 5.47933638e-01
-2.52516121e-01 8.35508585e-01 -5.08001328e-01 1.16393425e-01
5.32430887e-01 -5.02067685e-01 -5.30025125e-01 3.33696246e-01
5.29756248e-01 -1.15329397e+00 -6.33909702e-01 9.58916783e-01
4.88518924e-01 1.41505182e-01 2.73804426e-01 -1.07302308e+00
-1.06567156e+00 4.10633713e-01 -6.62122071e-01 5.05646884e-01
3.75641376e-01 3.39088470e-01 3.50879103e-01 -1.10509706e+00
4.80440676e-01 1.50922096e+00 8.95749211e-01 6.34374440e-01
-8.71486425e-01 -6.06611490e-01 4.34108794e-01 3.97398740e-01
-1.19736922e+00 -5.24794996e-01 8.27688694e-01 -3.98884982e-01
6.85894489e-01 1.75142810e-01 6.35080218e-01 1.18459260e+00
-3.91361453e-02 8.32439780e-01 1.12482893e+00 -4.89239879e-02
-7.13065714e-02 -2.81459868e-01 -4.84319888e-02 6.23705804e-01
-1.75064087e-01 2.14414194e-01 -7.05124021e-01 2.33587384e-01
1.73160255e+00 3.61332864e-01 -6.44002318e-01 8.55571851e-02
-1.38485658e+00 3.44614506e-01 4.63612944e-01 2.82382697e-01
-3.90792906e-01 3.12916845e-01 3.08166653e-01 3.22853655e-01
6.56664848e-01 -7.02349981e-03 -4.62243289e-01 -4.55287434e-02
-8.85577083e-01 1.05727106e-01 -3.17287706e-02 1.10400176e+00
8.28320920e-01 2.83752799e-01 -6.55340075e-01 8.44786882e-01
8.75303969e-02 2.60537565e-01 4.03625607e-01 -1.45856559e+00
4.23675597e-01 5.20860076e-01 3.72825980e-01 -8.81869018e-01
-2.71123946e-01 -2.75681049e-01 -1.02093446e+00 3.23837459e-01
3.81557345e-01 -1.75662637e-01 -9.19783771e-01 1.61177015e+00
4.20499653e-01 7.74337471e-01 -1.45392850e-01 1.20699942e+00
8.86833608e-01 6.56022906e-01 -1.42188758e-01 -9.03471559e-02
8.99897575e-01 -1.00832760e+00 -7.90921330e-01 2.43501410e-01
2.73787349e-01 -6.66344523e-01 9.26690161e-01 1.57644525e-01
-1.18924916e+00 -1.22394609e+00 -9.18857157e-01 -4.61458951e-01
-9.21949744e-02 2.29804963e-01 4.64020312e-01 1.68795884e-02
-1.17819226e+00 9.44230735e-01 -8.48550916e-01 9.02715921e-02
6.31387234e-01 2.96663105e-01 -4.90301430e-01 -2.53410697e-01
-1.31050670e+00 6.47892416e-01 3.04867506e-01 4.49287623e-01
-9.24235821e-01 -7.70552456e-01 -8.81698489e-01 -8.60043988e-02
1.44871026e-01 -7.22729206e-01 9.29324269e-01 -1.17845798e+00
-1.41636193e+00 5.58039308e-01 -1.70290709e-01 -3.61438215e-01
6.32451057e-01 -3.58223200e-01 -4.16758567e-01 2.80401736e-01
5.26455268e-02 8.16655517e-01 1.08064198e+00 -9.03839588e-01
-8.09545994e-01 -3.02584678e-01 3.48998517e-01 8.87424797e-02
-4.15034533e-01 8.50028694e-02 -5.07953227e-01 -1.10674977e+00
2.03866690e-01 -3.80839735e-01 -1.95126131e-01 5.99902332e-01
1.02837272e-01 -2.67509222e-01 1.05798411e+00 -1.06477284e+00
1.14272177e+00 -2.32830215e+00 1.87100604e-01 -2.72114992e-01
1.58210114e-01 4.09045458e-01 -3.72031748e-01 -6.49627820e-02
-1.67601496e-01 -9.20022577e-02 -1.20455518e-01 -2.12626949e-01
-2.18386307e-01 1.95046976e-01 -9.67475995e-02 2.63294727e-01
7.71357417e-01 8.27383101e-01 -8.89582098e-01 -3.68625134e-01
4.40469384e-01 9.30214584e-01 -7.07517385e-01 1.79522634e-01
-9.89133939e-02 7.91921675e-01 -7.27280676e-01 5.49731255e-01
1.00832534e+00 -1.19529553e-01 -3.44236851e-01 -7.04327404e-01
-2.71135837e-01 8.70669410e-02 -1.46566546e+00 2.05601335e+00
-3.91326249e-01 5.17235219e-01 1.22133680e-01 -1.06062794e+00
1.04605126e+00 3.08896601e-01 6.86231375e-01 -5.64034164e-01
-3.61364186e-02 6.51251823e-02 -3.93762350e-01 -7.21625984e-01
3.33493859e-01 2.95392126e-01 3.42815042e-01 1.46438956e-01
1.46407753e-01 3.54047060e-01 -1.72073379e-01 -1.88895717e-01
1.01459396e+00 6.94357157e-01 -9.08710584e-02 -1.05796859e-01
7.28029191e-01 -3.52530718e-01 8.62647653e-01 4.77775544e-01
-2.91922808e-01 8.94776523e-01 2.96010435e-01 -8.52845252e-01
-1.07971263e+00 -9.06717002e-01 -6.47978410e-02 7.55421937e-01
1.66816071e-01 -1.62595287e-01 -8.29183757e-01 -7.25005388e-01
-1.06895752e-01 1.49110049e-01 -6.96851909e-01 -1.25737458e-01
-9.40573514e-01 -2.93442607e-01 3.37325186e-01 9.99356091e-01
1.20586407e+00 -8.31592500e-01 -3.47348928e-01 3.58226627e-01
-2.84303695e-01 -1.17953610e+00 -9.50258076e-01 -1.07754230e-01
-9.90607619e-01 -8.59857500e-01 -1.03902566e+00 -7.60854900e-01
6.53034627e-01 5.76881826e-01 7.35558271e-01 9.45864618e-02
-4.66138460e-02 1.19424596e-01 -3.71212065e-01 1.02496065e-01
-1.16893435e-02 -4.09817159e-01 3.21195982e-02 2.68753290e-01
3.92658561e-01 -8.58922362e-01 -8.30299377e-01 4.62238818e-01
-1.02554882e+00 2.33255297e-01 4.71566409e-01 7.73216963e-01
7.29845643e-01 -3.22099891e-03 5.61584055e-01 -3.87357682e-01
4.39944655e-01 -2.12315217e-01 -3.49839926e-01 8.38590413e-02
1.45551607e-01 1.58150226e-01 6.51430368e-01 -6.25658333e-01
-1.05549455e+00 1.93876952e-01 -3.11659962e-01 -1.05513704e+00
-3.10152620e-01 1.16833083e-01 -2.96792865e-01 -6.64607510e-02
4.03704882e-01 3.49283397e-01 5.93980886e-02 -7.91680098e-01
3.03125203e-01 6.69372618e-01 6.34807348e-01 -3.97044927e-01
6.49623811e-01 6.12844050e-01 -1.72968447e-01 -5.37878454e-01
-1.04363334e+00 -1.17916107e-01 -8.57809603e-01 -3.18420410e-01
1.00797975e+00 -1.20420051e+00 -5.18711329e-01 7.50804543e-01
-1.31292558e+00 -4.44442809e-01 -3.89116049e-01 5.84614933e-01
-6.05722547e-01 4.43358719e-01 -6.45593464e-01 -4.18137103e-01
-1.76834315e-01 -1.25496387e+00 1.22238076e+00 2.34625146e-01
1.68806106e-01 -7.82136738e-01 -4.25021976e-01 2.35929593e-01
4.91683245e-01 4.05761361e-01 3.95246476e-01 6.50093481e-02
-8.31284285e-01 2.03421950e-01 -4.02848512e-01 9.13396299e-01
4.57081646e-01 -7.35911205e-02 -8.91994417e-01 -4.19555753e-02
1.58089280e-01 -6.12327009e-02 1.14102614e+00 6.88843369e-01
1.71152353e+00 -4.31032807e-01 -2.19246186e-02 9.90405083e-01
1.08015382e+00 6.42493814e-02 8.17434490e-01 1.33254424e-01
7.59155393e-01 3.99467885e-01 7.36527145e-01 6.06194258e-01
2.52045751e-01 7.54767060e-01 3.40349495e-01 -1.81761399e-01
-5.43955982e-01 -1.36743307e-01 5.93432903e-01 5.24556041e-01
-5.18200278e-01 3.58674854e-01 -2.84535289e-01 3.37993562e-01
-1.86127865e+00 -1.25294101e+00 3.03463894e-03 1.88455117e+00
9.26647067e-01 -4.59784344e-02 7.59979635e-02 3.06889862e-02
9.29927886e-01 3.04585665e-01 -6.21311247e-01 8.89393464e-02
-1.53846100e-01 2.38524526e-01 3.00053924e-01 5.36832035e-01
-1.17048657e+00 7.36453235e-01 5.87097073e+00 9.76777732e-01
-1.18746054e+00 1.36903420e-01 6.04914546e-01 -9.84296948e-02
-3.20165455e-01 -2.53764093e-01 -9.58729029e-01 6.80196643e-01
3.73722434e-01 1.97551832e-01 3.87958586e-01 5.95813692e-01
6.74954832e-01 1.42673939e-01 -1.14198327e+00 1.19341576e+00
7.13553838e-03 -1.69721174e+00 2.44321644e-01 -2.21312284e-01
8.53033245e-01 -3.31038624e-01 5.42212203e-02 1.47446260e-01
7.28476718e-02 -9.77672219e-01 7.52440035e-01 9.32803035e-01
9.11193073e-01 -6.69623494e-01 3.93670470e-01 2.72543937e-01
-1.40470326e+00 -1.62815943e-01 -5.04411221e-01 -1.41959175e-01
2.17437580e-01 5.69766581e-01 -2.24293873e-01 5.64651489e-01
1.07048786e+00 1.37461805e+00 -2.90313393e-01 9.54384804e-01
-2.10437104e-01 1.61848962e-01 -1.14926629e-01 3.78690183e-01
2.53223091e-01 -1.80804700e-01 3.74914885e-01 1.03669572e+00
5.70546448e-01 3.56033325e-01 1.30915493e-01 9.88699555e-01
-4.11043838e-02 -3.56840312e-01 -4.46294963e-01 3.95818472e-01
2.83321917e-01 1.08324432e+00 -2.39136413e-01 -3.83109331e-01
-6.22684121e-01 1.19405985e+00 1.95457965e-01 4.98364300e-01
-1.09181857e+00 -1.43672183e-01 9.90278006e-01 2.15625390e-01
6.71151817e-01 -5.19123256e-01 -5.57823330e-02 -1.28576338e+00
3.04067135e-01 -8.25203776e-01 1.81078598e-01 -1.04935026e+00
-1.28387880e+00 5.08301795e-01 -1.93535481e-02 -1.60173261e+00
1.00628994e-01 -6.22743785e-01 -4.88447756e-01 9.91462290e-01
-1.63493216e+00 -1.04819191e+00 -6.71633303e-01 1.16628850e+00
8.55748475e-01 1.07422322e-01 4.02287632e-01 6.52858973e-01
-3.87514234e-01 4.11445767e-01 -8.28193575e-02 3.82716209e-01
7.70902336e-01 -7.76505828e-01 4.35927689e-01 8.50313842e-01
-1.95584819e-01 3.61102700e-01 1.71496272e-01 -4.42716062e-01
-1.28454876e+00 -1.47202921e+00 3.99412632e-01 -2.38024473e-01
2.71496564e-01 -9.01957825e-02 -9.40044284e-01 6.69346511e-01
-3.26655828e-03 5.27277291e-01 1.92366630e-01 -5.21037817e-01
-2.81177372e-01 -4.46396589e-01 -1.23203611e+00 5.22756279e-01
1.59995174e+00 -4.10248667e-01 -3.70497316e-01 1.56653211e-01
9.89430130e-01 -5.05139172e-01 -1.14405465e+00 5.16111434e-01
4.28663492e-01 -1.17671609e+00 1.24412334e+00 -4.88908261e-01
5.79545379e-01 -7.52175391e-01 -2.26696104e-01 -1.00783348e+00
-3.72576565e-01 -5.77477455e-01 -2.66498357e-01 1.18836439e+00
-4.25353050e-02 -3.80362034e-01 5.84283769e-01 5.29298425e-01
-4.10108685e-01 -7.09750473e-01 -9.47085857e-01 -6.56405747e-01
-9.37661082e-02 -2.82131761e-01 7.80142725e-01 9.93257940e-01
-4.88670766e-01 -1.59748793e-01 -7.43560553e-01 3.59026790e-01
5.96130967e-01 9.68684535e-03 6.25140309e-01 -8.66190612e-01
-2.86936522e-01 -1.97538987e-01 -4.89524305e-01 -1.80360985e+00
-4.79465760e-02 -5.07393599e-01 -1.96769506e-01 -1.55182004e+00
-4.42916118e-02 -2.95011967e-01 -3.61107409e-01 5.82687020e-01
-2.83163756e-01 3.89485776e-01 3.60686854e-02 4.12064269e-02
-4.54512835e-01 7.29460239e-01 1.89764309e+00 -4.27885987e-02
-1.09828442e-01 -1.46807790e-01 -5.11169255e-01 7.39185572e-01
6.42662048e-01 -3.06541733e-02 -3.34365159e-01 -9.27614331e-01
-4.57606941e-01 2.22483188e-01 7.32772112e-01 -1.12195468e+00
9.71090421e-02 -2.66199589e-01 1.02270591e+00 -8.42657149e-01
4.61937666e-01 -9.37226474e-01 2.35684305e-01 6.70606568e-02
-3.49775106e-01 1.87034830e-02 4.00808975e-02 4.12315369e-01
-3.54928851e-01 7.96840787e-02 7.88893998e-01 -3.54329348e-01
-1.09324348e+00 6.52499437e-01 -3.72783244e-02 -8.15959126e-02
9.99298334e-01 -5.52900076e-01 -8.65534320e-02 -2.57254571e-01
-8.92639101e-01 1.16502814e-01 3.44225824e-01 4.80614573e-01
9.91239130e-01 -1.77381361e+00 -6.98459029e-01 4.56607491e-01
-4.59145159e-01 2.31141463e-01 5.33519030e-01 8.71311247e-01
-4.28008229e-01 3.48205715e-02 -6.09708726e-01 -7.30119526e-01
-1.03444648e+00 5.16594172e-01 5.59892476e-01 1.40612200e-01
-7.19823360e-01 8.09455216e-01 7.28802979e-02 -4.90097366e-02
3.17407101e-01 -6.72331512e-01 -1.75406113e-01 -1.20012730e-01
8.57803345e-01 4.04231012e-01 -2.99214631e-01 -4.24717814e-01
-8.50729346e-02 7.49198973e-01 -8.21541324e-02 2.12894365e-01
1.67451334e+00 -3.17491204e-01 -1.03304774e-01 1.60002589e-01
1.33546913e+00 -3.52176279e-01 -2.22040629e+00 -3.40923607e-01
-3.99411321e-01 -8.85720253e-01 9.11185443e-02 -3.44604611e-01
-1.45418680e+00 8.03221524e-01 6.68840706e-01 -8.15432742e-02
1.50429690e+00 -2.73466736e-01 8.98509324e-01 -1.77243594e-02
2.85142958e-01 -9.73912776e-01 2.89391726e-01 4.35481012e-01
1.03405440e+00 -1.09966779e+00 -1.66035563e-01 -3.56150776e-01
-3.50813717e-01 1.42781186e+00 8.55775893e-01 -4.41727012e-01
7.22342789e-01 3.31574768e-01 -4.37943749e-02 2.13500947e-01
-4.86192077e-01 -1.46923840e-01 2.23414183e-01 7.06099331e-01
4.42288250e-01 -5.32075524e-01 -2.71908522e-01 5.66207588e-01
3.81181419e-01 3.71083975e-01 3.86505544e-01 7.43037879e-01
-3.71600032e-01 -1.10135520e+00 -2.20143750e-01 1.52962998e-01
-2.86892980e-01 -1.72579139e-01 8.91041607e-02 5.40865779e-01
5.43708324e-01 6.87558472e-01 1.70232266e-01 -4.60209459e-01
4.20544237e-01 -1.99213386e-01 4.99932766e-01 -2.15954080e-01
-3.33197713e-01 1.38810262e-01 -9.70482975e-02 -9.41927791e-01
-9.93471086e-01 -6.01430714e-01 -1.18372536e+00 -2.55715817e-01
-1.31863400e-01 -1.92686766e-01 1.31419823e-01 6.92561746e-01
5.48626482e-01 6.75806880e-01 7.60519922e-01 -1.14449751e+00
-5.29168844e-01 -9.36836839e-01 -5.21951556e-01 4.54944283e-01
5.73760808e-01 -6.87491775e-01 -2.31992707e-01 4.61650789e-01] | [10.796847343444824, -1.5258222818374634] |
111851bb-e050-4906-b93e-6ca87c2180c1 | exit-chart-aided-near-capacity-irregular-bit | null | null | https://ieeexplore.ieee.org/abstract/document/4786476/references#references | https://ieeexplore.ieee.org/abstract/document/4786476/references#references | EXIT-chart aided near-capacity Irregular Bit-Interleaved Coded Modulation design | A near-capacity irregular bit-interleaved coded modulation based iterative decoding (Ir-BICM-ID) aided scheme is proposed. The irregular design of the scheme pervades the three basic components of BICM-ID, namely the encoder, the unity-rate precoder and the bit-to-symbol mapper. As a result, irregular BICM-ID schemes constituted by irregular components are created, which are capable of approaching the capacity of coded modulation. This is achieved by creating a marginally open extrinsic information transfer (EXIT) chart tunnel, and exploiting the theorem that the open tunnel's area is characteristic of how closely the scheme operates to the channel's capacity. The proposed Ir-BICM-ID scheme employs irregular convolutional codes (IrCC), irregular unity-rate codes (IrURC) and irregular mappers (IrMapper). | ['Ronald Y. S. Tee; Robert G. Maunder; Lajos Hanzo'] | 2009-01-01 | null | null | null | http-ieeexplore-ieee-org-stamp-stamp-jsp-tp | ['unity'] | ['computer-vision'] | [ 7.76272595e-01 7.53291547e-01 -4.31374192e-01 1.68636903e-01
-3.41432273e-01 -4.57285158e-03 1.01663160e+00 -4.93738621e-01
-1.20884806e-01 8.66331995e-01 2.02609360e-01 -1.00838268e+00
-2.42919728e-01 -3.62362742e-01 -5.61506450e-01 -7.35799253e-01
-9.19504225e-01 -2.03932241e-01 8.80851820e-02 -1.84088379e-01
7.07941592e-01 1.02615289e-01 -1.19930542e+00 2.38091961e-01
1.01558292e+00 1.09743726e+00 2.57983476e-01 1.38610029e+00
3.21834460e-02 1.10640168e+00 1.01635881e-01 -4.57382798e-01
2.08979234e-01 -8.97942901e-01 -9.08429563e-01 4.87175696e-02
-4.14688885e-01 -4.47960258e-01 -9.12293077e-01 7.23371625e-01
-1.35468747e-02 -5.35940528e-01 1.21356130e+00 -9.42277014e-01
-5.65001070e-01 5.38084745e-01 -2.80525237e-01 -8.71794447e-02
2.98392951e-01 -5.19564033e-01 3.82779032e-01 -8.68951559e-01
3.51544410e-01 1.17227793e+00 8.34754050e-01 2.27845326e-01
-1.11534762e+00 -4.64798093e-01 -1.06435108e+00 4.23547626e-01
-1.87192655e+00 -8.36313665e-01 8.89727846e-02 -2.54056752e-01
8.07798028e-01 4.60324466e-01 2.52934724e-01 2.02438638e-01
7.70138383e-01 2.18500793e-01 8.37823689e-01 -1.18952763e+00
3.20514709e-01 1.88988686e-01 -6.52100325e-01 6.64536476e-01
6.15475059e-01 5.17442167e-01 -1.87106997e-01 1.72328092e-02
1.02985418e+00 -7.43476272e-01 -3.93313229e-01 -3.30189347e-01
-1.05774307e+00 5.76886058e-01 1.62183106e-01 6.42577410e-01
4.95076030e-02 5.31490862e-01 1.66457534e-01 5.92601418e-01
-1.68269217e-01 -2.40282282e-01 4.45665605e-02 -3.59086663e-01
-7.92803407e-01 -3.96050483e-01 1.03799832e+00 1.66081655e+00
8.36581171e-01 8.59078467e-02 -4.36122268e-02 6.51820421e-01
8.37256432e-01 4.15232092e-01 2.60526866e-01 -8.94026399e-01
5.38973749e-01 2.92907834e-01 -1.55711904e-01 -6.02464497e-01
-2.48049870e-01 -3.67336720e-01 -1.40355980e+00 8.65745246e-02
1.58964291e-01 -1.27832353e-01 -6.57886446e-01 1.12993252e+00
-1.90890089e-01 -2.37414706e-02 6.49615645e-01 3.51155549e-01
2.12199435e-01 7.76653349e-01 -3.20387065e-01 -3.34082454e-01
1.26515353e+00 -5.22271395e-01 -8.09942782e-01 2.78570764e-02
1.29754078e+00 -8.14124167e-01 -3.37477803e-01 1.80512652e-01
-1.20315182e+00 -5.11287570e-01 -1.44811416e+00 1.04902573e-01
9.07781646e-02 4.11049157e-01 2.16254905e-01 1.11304998e+00
-1.22410309e+00 3.55062217e-01 -7.12560713e-02 1.93283573e-01
4.18633193e-01 5.07716596e-01 -3.79593760e-01 -1.18028902e-01
-1.22630298e+00 1.03462493e+00 8.46383035e-01 -1.81974694e-02
-8.23718488e-01 -5.45002595e-02 -8.73889983e-01 -5.36605343e-02
-4.11918163e-01 -2.90195525e-01 1.06323504e+00 -9.75756586e-01
-1.46251106e+00 5.23524106e-01 -4.31439489e-01 -7.69667029e-01
1.42111987e-01 5.65419912e-01 -8.02808404e-01 5.23237169e-01
-2.95655847e-01 3.79312634e-01 1.01467621e+00 -1.27831578e+00
-6.44747138e-01 9.36821997e-02 -5.56897640e-01 5.60892373e-02
6.66403845e-02 -2.46102378e-01 -3.09931159e-01 -6.10034525e-01
4.40008253e-01 -6.74648762e-01 -2.83151120e-01 -5.92983216e-02
-2.10834458e-01 3.30269009e-01 1.11223495e+00 -7.25615442e-01
1.77869487e+00 -2.45149231e+00 1.62812881e-02 4.03574854e-01
-5.92887029e-02 4.47322965e-01 1.50070950e-01 9.22521174e-01
-1.18183129e-01 1.93191022e-02 -6.89481199e-01 -1.18760258e-01
-3.47734898e-01 4.61013526e-01 1.72949061e-01 6.90994799e-01
-5.20703532e-02 7.81542778e-01 -6.67211354e-01 -6.04415238e-01
-2.34469376e-03 4.15032297e-01 -5.36233723e-01 -1.03300922e-02
2.60730147e-01 6.08545363e-01 -6.17774166e-02 6.10025883e-01
1.14554870e+00 -3.60835083e-02 3.12198669e-01 6.08266369e-02
-6.57088459e-01 2.25620210e-01 -1.29626954e+00 1.30148613e+00
-4.78852272e-01 9.51182306e-01 2.57968783e-01 -1.14706004e+00
1.18655145e+00 8.98514807e-01 5.72446845e-02 -8.46160710e-01
1.45378530e-01 8.45292866e-01 1.41347989e-01 -2.93687373e-01
8.61893058e-01 1.20354826e-02 2.58287638e-01 1.11238152e-01
2.98592389e-01 1.61827818e-01 -3.32945306e-03 4.66393828e-01
9.34914529e-01 -1.83443069e-01 7.10570693e-01 -6.17247939e-01
1.29771948e+00 -3.63244802e-01 2.31791973e-01 7.33348668e-01
-2.70015677e-03 2.18563706e-01 5.74238241e-01 1.04675770e-01
-1.77936316e+00 -8.47904384e-01 -1.03084385e+00 1.19966269e-02
5.63566267e-01 -2.90351897e-01 -5.76341033e-01 1.56397820e-01
-3.70790482e-01 6.01891994e-01 -2.56837979e-02 -1.40681073e-01
-1.94385275e-01 -6.52809501e-01 9.65438485e-01 -1.63034663e-01
1.15070879e+00 -3.76952261e-01 9.25541222e-02 6.08660936e-01
-5.89212179e-02 -1.04623330e+00 -3.09573263e-01 5.05387664e-01
-9.19580996e-01 -7.78160274e-01 -5.26801050e-01 -1.20646393e+00
8.61744106e-01 1.90858021e-01 5.71920156e-01 1.39006004e-01
-1.09682657e-01 1.72471985e-01 -7.20466077e-01 2.29794934e-01
-1.22890139e+00 -3.48493606e-01 -1.99279457e-01 2.81069279e-01
1.35423839e-02 -4.60292459e-01 -6.59838259e-01 5.15228629e-01
-7.49221265e-01 4.29582685e-01 7.59335876e-01 8.03268075e-01
-3.32821645e-02 3.56012173e-02 7.70831704e-01 -1.17367774e-01
2.77190786e-02 -7.42998719e-01 -4.38751042e-01 -1.20017402e-01
-6.91464365e-01 3.59855920e-01 7.62053728e-01 1.58865556e-01
-9.49750006e-01 7.12884143e-02 -5.05962729e-01 5.92252135e-01
2.21477732e-01 -1.98885072e-02 3.20670642e-02 -9.77194369e-01
6.76483333e-01 6.33209944e-01 1.83836848e-01 -2.46208623e-01
3.82545173e-01 1.87999845e+00 6.61507845e-01 -4.70988601e-02
8.01747262e-01 3.36088955e-01 4.66178805e-01 -1.26809931e+00
1.51277289e-01 -4.76321101e-01 -6.73933804e-01 -3.27741623e-01
6.89471185e-01 -1.20040655e+00 -6.28220558e-01 5.73858857e-01
-1.31608117e+00 -1.66880593e-01 -1.66581184e-01 7.43083179e-01
-9.93559897e-01 7.61730015e-01 -5.94176054e-01 -1.23857439e+00
-5.54364696e-02 -1.08154249e+00 6.50238693e-01 -3.76500413e-02
2.14348927e-01 -1.01610291e+00 -2.09871247e-01 -1.21733427e-01
6.89128339e-01 -2.11549029e-01 8.93738687e-01 -1.23322364e-02
-8.32685649e-01 -1.84761852e-01 -5.59084177e-01 5.09610593e-01
3.88888977e-02 -6.56620800e-01 -7.96112657e-01 -2.60826886e-01
-4.54171821e-02 4.61535034e-04 6.24725759e-01 1.76051602e-01
1.00232911e+00 -6.69633687e-01 -3.03535789e-01 9.49625552e-01
1.67359507e+00 6.39888704e-01 1.66888011e+00 1.36526823e-01
2.35014483e-01 -1.72115996e-01 7.13021383e-02 6.33393109e-01
3.47917438e-01 7.30738580e-01 4.39276099e-01 1.19926028e-01
-2.62763709e-01 -2.04139754e-01 5.17220259e-01 1.20463705e+00
-2.20013916e-01 -1.33123383e-01 -3.87266934e-01 4.23919499e-01
-1.36592531e+00 -1.11322093e+00 -9.50571418e-01 2.37487602e+00
1.05470705e+00 5.80842756e-02 -5.53344309e-01 9.05784607e-01
7.37071097e-01 -3.28468569e-02 2.15644374e-01 -1.31129730e+00
-4.15393189e-02 8.67070258e-02 1.38308752e+00 6.84123755e-01
-8.99875998e-01 2.18253091e-01 6.62032127e+00 1.32848489e+00
-3.16314965e-01 5.00415675e-02 3.94242972e-01 1.26703799e+00
-3.66790175e-01 5.64216614e-01 -8.70860338e-01 7.16957211e-01
1.52570236e+00 -1.16509525e-02 7.44779229e-01 4.26588148e-01
3.00032040e-03 -5.45954823e-01 -6.95935011e-01 9.56900179e-01
-1.27451524e-01 -1.46718943e+00 4.93010879e-03 3.48178834e-01
8.15884829e-01 -4.67479527e-01 -3.31081152e-01 4.49409224e-02
-2.64640828e-03 -9.83374953e-01 5.24537981e-01 6.77986085e-01
1.72272670e+00 -6.90123498e-01 4.35315043e-01 4.60586995e-01
-1.68833506e+00 -5.61219037e-01 -3.93173873e-01 -1.58055857e-01
2.22550541e-01 4.49457169e-01 -6.63742304e-01 8.72076392e-01
4.87853996e-02 4.68373954e-01 -1.49115786e-01 1.15010095e+00
3.00981611e-01 6.98035002e-01 -3.28561030e-02 9.01978612e-02
4.89688009e-01 -3.72295171e-01 6.54248834e-01 1.49360085e+00
7.78442800e-01 4.11525704e-02 -6.73900187e-01 3.19693446e-01
-7.68083706e-02 -1.21083714e-01 -7.99410284e-01 8.08577389e-02
5.72880805e-01 6.92946672e-01 -1.14094019e-01 -3.68192405e-01
-6.28095865e-01 1.06609607e+00 -4.27870154e-01 1.03563718e-01
-2.68710971e-01 -7.72534907e-01 4.32901859e-01 -1.52290948e-02
7.00506747e-01 -6.24731183e-01 -6.84780300e-01 -6.26459897e-01
-4.28264886e-01 -6.75442517e-01 -1.62777275e-01 -4.51871961e-01
-2.67338783e-01 1.75536260e-01 -3.05891424e-01 -1.87667549e+00
-1.86489686e-01 -3.53009045e-01 -3.22796851e-01 1.07992721e+00
-1.72435868e+00 -7.51720250e-01 -5.51875569e-02 4.51656073e-01
7.62555972e-02 -4.40661520e-01 9.78951454e-01 5.14349043e-01
1.33316621e-01 7.66959250e-01 1.10503590e+00 5.14607653e-02
-5.97951524e-02 -8.70780230e-01 4.05546695e-01 7.60065258e-01
-3.00617993e-01 1.00047104e-01 3.15342695e-01 -4.03078049e-01
-1.85109854e+00 -1.11785841e+00 1.19294918e+00 2.54482299e-01
3.58802229e-01 -1.80423185e-01 -5.37213862e-01 2.93215662e-01
2.58330375e-01 -2.92213440e-01 4.55502123e-01 -8.65135193e-01
-5.65969152e-04 1.47223338e-01 -1.20340550e+00 4.92708862e-01
9.57325876e-01 -5.28477132e-01 -1.89026177e-01 1.09602720e-01
4.08824056e-01 -3.43298972e-01 -1.06863260e+00 3.15772146e-01
6.54179454e-01 -9.72753882e-01 8.40010405e-01 4.71814424e-01
4.47392821e-01 -4.76258010e-01 -4.30631042e-01 -7.56493390e-01
-5.18641055e-01 -1.02471793e+00 -2.67995805e-01 6.35483205e-01
4.01988387e-01 -6.33856237e-01 2.43424237e-01 -4.41476822e-01
-5.03231049e-01 -4.89007086e-01 -1.59691381e+00 -9.19026732e-01
-1.02228686e-01 -5.12426376e-01 2.89176673e-01 2.71423042e-01
5.17406523e-01 -4.42313738e-02 -8.55954945e-01 1.68900788e-01
9.40751851e-01 -8.46867740e-01 4.40635473e-01 -9.04482901e-01
-2.77113497e-01 -1.10774346e-01 -8.99864018e-01 -1.77552557e+00
-6.54266119e-01 -1.24714553e+00 -3.80937271e-02 -1.30280507e+00
-1.66254967e-01 -1.04032981e+00 3.57854843e-01 -7.81384781e-02
5.58432639e-01 5.73725760e-01 -8.69229808e-02 4.27246451e-01
-2.53103137e-01 3.82471681e-01 1.25321245e+00 1.90544233e-01
-2.13700421e-02 -9.53509361e-02 -4.02340889e-01 2.32191265e-01
8.51541162e-01 -1.67598262e-01 -7.76858330e-02 2.46162027e-01
1.29027665e-01 6.98015928e-01 3.88759583e-01 -1.32029295e+00
2.54832983e-01 4.18634862e-01 9.85323861e-02 -5.32102525e-01
1.92441329e-01 -9.29572999e-01 4.77903485e-01 9.62486863e-01
-2.56504249e-02 -6.17493629e-01 -1.39512300e-01 7.04473376e-01
-1.18981175e-01 -8.43193948e-01 1.04684842e+00 2.69059300e-01
-7.63740540e-01 -1.30905166e-01 -9.86005843e-01 -2.18854427e-01
1.15021336e+00 -8.60308588e-01 -1.86487049e-01 -3.53064865e-01
-3.31134439e-01 -7.67016709e-02 3.58389586e-01 -2.27422148e-01
6.37141466e-01 -1.50499046e+00 -7.89423048e-01 7.80380189e-01
1.25015169e-01 -4.02495295e-01 2.72930190e-02 1.07013309e+00
-9.97228980e-01 9.91692960e-01 -2.59199310e-02 -7.53817558e-01
-7.39749730e-01 3.24074961e-02 3.69518548e-01 9.58873518e-03
-7.41453528e-01 4.19015288e-01 -2.41519451e-01 1.30766243e-01
2.63260812e-01 2.73544014e-01 -1.02784827e-01 -7.06922710e-01
7.61765063e-01 3.62061411e-01 4.37346175e-02 -1.06025064e+00
5.13710766e-05 6.24689460e-01 3.92868310e-01 5.17738611e-02
5.36894858e-01 -6.92680299e-01 -4.03447062e-01 -1.22048311e-01
1.48743093e+00 -2.43374974e-01 -9.99219179e-01 -2.95553748e-02
3.92375104e-02 -4.89360541e-01 2.91753523e-02 -1.90799043e-01
-2.70722479e-01 4.96853858e-01 6.89607859e-01 2.67005682e-01
1.20413899e+00 -3.39552730e-01 5.11312902e-01 6.27436116e-02
7.31821120e-01 -1.00246727e+00 -4.05757815e-01 8.20809424e-01
5.38887084e-01 -6.10783219e-01 -2.22925574e-01 -4.34446454e-01
-3.85220081e-01 1.54895747e+00 -3.43842715e-01 2.27180824e-01
8.99246573e-01 4.42869842e-01 -6.15702271e-01 3.59454423e-01
-7.48183072e-01 -2.38994986e-01 9.99859497e-02 7.85664797e-01
5.35944581e-01 7.49191344e-02 -1.08311808e+00 -2.29563713e-01
1.01720147e-01 9.90858003e-02 8.92069221e-01 7.79134929e-01
-9.90702808e-01 -1.30555713e+00 -8.42318356e-01 5.60984790e-01
-1.91176832e-01 -2.94949681e-01 4.92805302e-01 3.63767117e-01
3.58764142e-01 1.20462286e+00 2.78415561e-01 -6.21210039e-01
-2.93085784e-01 -1.60452798e-01 3.56153995e-01 1.37518987e-01
1.45989463e-01 -3.65817159e-01 4.48489368e-01 -1.16626292e-01
-1.89862251e-01 -3.11390728e-01 -1.35390460e+00 -5.41337311e-01
-4.91261691e-01 4.00879383e-01 1.02073109e+00 9.74300385e-01
2.33655348e-01 1.88782930e-01 1.42814016e+00 -6.52964294e-01
-2.98342675e-01 -8.86432827e-01 -9.13081408e-01 -3.84369403e-01
5.87875903e-01 -2.41029128e-01 -5.24038434e-01 3.23196679e-01] | [6.470767498016357, 1.5187091827392578] |
1551d151-a39d-432f-a87b-94131364654e | modality-transferable-emotion-embeddings-for | 2009.09629 | null | https://arxiv.org/abs/2009.09629v3 | https://arxiv.org/pdf/2009.09629v3.pdf | Modality-Transferable Emotion Embeddings for Low-Resource Multimodal Emotion Recognition | Despite the recent achievements made in the multi-modal emotion recognition task, two problems still exist and have not been well investigated: 1) the relationship between different emotion categories are not utilized, which leads to sub-optimal performance; and 2) current models fail to cope well with low-resource emotions, especially for unseen emotions. In this paper, we propose a modality-transferable model with emotion embeddings to tackle the aforementioned issues. We use pre-trained word embeddings to represent emotion categories for textual data. Then, two mapping functions are learned to transfer these embeddings into visual and acoustic spaces. For each modality, the model calculates the representation distance between the input sequence and target emotions and makes predictions based on the distances. By doing so, our model can directly adapt to the unseen emotions in any modality since we have their pre-trained embeddings and modality mapping functions. Experiments show that our model achieves state-of-the-art performance on most of the emotion categories. In addition, our model also outperforms existing baselines in the zero-shot and few-shot scenarios for unseen emotions. | ['Pascale Fung', 'Tiezheng Yu', 'Zihan Liu', 'Wenliang Dai'] | 2020-09-21 | null | https://aclanthology.org/2020.aacl-main.30 | https://aclanthology.org/2020.aacl-main.30.pdf | asian-chapter-of-the-association-for | ['multimodal-emotion-recognition', 'multimodal-emotion-recognition'] | ['computer-vision', 'speech'] | [ 6.86359778e-02 -3.99942756e-01 3.33879367e-02 -5.55237889e-01
-7.40961134e-01 -3.98795873e-01 4.50440764e-01 6.34441711e-03
-7.48084426e-01 3.11675996e-01 4.95400846e-01 2.48717055e-01
2.83424169e-01 -5.07648051e-01 -3.47813100e-01 -5.07179320e-01
2.62585580e-01 2.29963154e-01 -1.58772338e-02 -3.15603524e-01
9.97983068e-02 5.87535882e-03 -1.83551466e+00 3.45801115e-01
6.36002123e-01 1.30233884e+00 9.56744477e-02 6.52844727e-01
-3.88108701e-01 5.57620704e-01 -3.91196579e-01 -4.67897683e-01
-1.82737947e-01 -5.81642091e-01 -6.21609688e-01 5.50632691e-03
2.74529476e-02 -1.50135353e-01 -2.08426908e-01 9.35967803e-01
6.73831880e-01 4.51499999e-01 9.20679271e-01 -1.43506646e+00
-1.08143210e+00 1.86993122e-01 -5.04696906e-01 -6.85842261e-02
3.97581369e-01 -2.42038414e-01 1.13007462e+00 -1.30653775e+00
4.62007701e-01 1.36205494e+00 5.24789870e-01 8.61193120e-01
-8.01468015e-01 -5.68280995e-01 2.24284753e-01 5.56799591e-01
-1.28366804e+00 -3.48514080e-01 7.53276765e-01 -2.43974358e-01
1.07899451e+00 -1.31230623e-01 4.27118033e-01 1.45966196e+00
-1.70252454e-02 7.61936605e-01 8.21333170e-01 -4.91207719e-01
2.30474889e-01 3.60682249e-01 -9.27363187e-02 3.11688900e-01
-6.38847828e-01 -2.08233535e-01 -6.47281468e-01 6.95375958e-03
3.34305525e-01 2.22824320e-01 -2.25098208e-01 -3.27485681e-01
-1.20975280e+00 9.33324039e-01 3.42482328e-01 5.50032139e-01
-3.90047014e-01 -6.94508664e-03 8.54854167e-01 4.01668549e-01
5.91652036e-01 1.74089238e-01 -5.21380246e-01 -5.23064196e-01
-4.07637656e-01 -2.90238678e-01 4.63166565e-01 6.72052383e-01
7.33236611e-01 2.61384906e-04 2.81467456e-02 1.32231593e+00
3.13555390e-01 3.13035429e-01 7.46084154e-01 -3.07013929e-01
3.47444832e-01 5.55336595e-01 -2.07728334e-02 -1.16862512e+00
-4.74532276e-01 1.10796414e-01 -7.95005322e-01 -1.49817526e-01
7.84422383e-02 -2.42386073e-01 -8.51131260e-01 1.85233390e+00
2.69834220e-01 4.79098231e-01 4.30633277e-01 9.71919596e-01
9.87895429e-01 1.04732633e+00 3.53940487e-01 -8.04753527e-02
1.47569239e+00 -1.24206102e+00 -1.01642585e+00 -5.16225517e-01
5.93506694e-01 -8.24442565e-01 1.40037823e+00 2.39401534e-01
-6.39267981e-01 -6.79890096e-01 -8.32571983e-01 -1.65550545e-01
-7.83798695e-01 7.68672898e-02 3.68174314e-01 5.44775069e-01
-7.56282210e-01 1.10223822e-01 -4.84916955e-01 -6.24014437e-01
1.34999186e-01 -1.51743889e-02 -4.31294322e-01 -2.18683988e-01
-1.54594779e+00 1.09443414e+00 3.70290786e-01 1.53834283e-01
-7.48199582e-01 -5.50613940e-01 -1.22806251e+00 1.05997048e-01
1.62670717e-01 -3.71017098e-01 9.28484738e-01 -1.22257650e+00
-1.59173512e+00 6.93570018e-01 -2.37777263e-01 1.61580339e-01
-2.41611451e-02 -3.28136832e-01 -9.10471559e-01 2.02232778e-01
-1.94127008e-01 6.92718148e-01 7.62145102e-01 -1.26109171e+00
-6.58033490e-01 -3.74938548e-01 -2.79003739e-01 4.51344848e-01
-1.25083685e+00 2.71607250e-01 -6.35453939e-01 -5.56584954e-01
-2.34731272e-01 -7.03489780e-01 -4.42541689e-02 -4.62471657e-02
-1.06304288e-01 -4.35538977e-01 9.17840540e-01 -4.04030949e-01
1.26445162e+00 -2.49062967e+00 5.08800030e-01 -8.86878520e-02
-1.87011957e-01 4.73330542e-02 -6.63792372e-01 5.26773036e-01
5.89270554e-02 4.54861149e-02 -2.13291403e-03 -6.24510348e-01
3.25414717e-01 2.98566669e-01 -3.83720428e-01 2.06322089e-01
2.26574600e-01 8.33180308e-01 -9.33981538e-01 -5.84863365e-01
3.11051577e-01 8.19308698e-01 -3.42919230e-01 6.32979393e-01
6.92721084e-02 3.37859303e-01 -2.86469042e-01 6.31796479e-01
4.46437865e-01 -4.78594527e-02 1.74826771e-01 -3.85225922e-01
1.06455483e-01 -2.47638389e-01 -9.87385035e-01 1.98710346e+00
-7.58492053e-01 4.30093199e-01 -2.12269366e-01 -1.09820187e+00
1.10371566e+00 6.51033163e-01 3.27279776e-01 -6.52924657e-01
3.86441350e-01 5.88422641e-02 -3.22398752e-01 -8.35467935e-01
6.86681986e-01 -6.55519903e-01 -5.68890035e-01 4.29082096e-01
5.29226243e-01 8.83903820e-03 -1.10281967e-01 -8.53386819e-02
7.48271704e-01 5.23894615e-02 2.76906013e-01 3.67005825e-01
3.88943791e-01 -3.37499678e-01 6.57762885e-01 3.21381360e-01
-3.82119745e-01 6.87620223e-01 2.88190454e-01 -2.91496962e-01
-7.54076481e-01 -9.25186276e-01 8.49368200e-02 1.70791733e+00
3.30889285e-01 -3.08878720e-01 -3.17048162e-01 -6.98179305e-01
-2.24762365e-01 6.61462367e-01 -8.19695532e-01 -3.54397327e-01
3.81388776e-02 -4.70950663e-01 5.53178251e-01 6.58164501e-01
6.55677021e-02 -1.35457027e+00 -5.03992081e-01 2.41355866e-01
-4.18965787e-01 -1.27182615e+00 -3.68914843e-01 1.58624083e-01
-5.02574205e-01 -7.15802014e-01 -7.92367220e-01 -1.02002382e+00
4.14816439e-01 1.50646836e-01 9.10593212e-01 -2.53803730e-01
-1.72800079e-01 6.18061841e-01 -8.92681420e-01 -4.90883976e-01
-2.08154134e-03 -1.92575589e-01 4.53222841e-02 4.13933724e-01
8.06006670e-01 -4.69908386e-01 -3.77683610e-01 3.36667091e-01
-1.05240977e+00 -3.23883772e-01 3.76080215e-01 9.94781852e-01
5.78530669e-01 -6.91707283e-02 9.19279277e-01 -4.30819273e-01
7.52169251e-01 -6.22114003e-01 9.62162018e-02 5.24039388e-01
-1.17941283e-01 -1.58804953e-01 7.99166381e-01 -8.78742874e-01
-1.06461358e+00 3.11630871e-02 -2.62141317e-01 -7.40305543e-01
-4.24535722e-01 6.10340714e-01 -2.48075917e-01 1.78551197e-01
2.26990208e-01 1.88121378e-01 -1.80913150e-01 -3.21332514e-01
7.43221104e-01 9.74587500e-01 3.77883047e-01 -4.65892136e-01
4.38550472e-01 3.38988125e-01 -6.27569675e-01 -7.86195517e-01
-8.80694747e-01 -5.31239092e-01 -4.75407958e-01 -3.75228941e-01
1.07780099e+00 -9.12991405e-01 -4.32433218e-01 3.71402234e-01
-1.13594925e+00 1.50106307e-02 -4.20058817e-02 7.14050710e-01
-6.11113131e-01 4.20183927e-01 -7.19787717e-01 -9.87762570e-01
-2.95616686e-01 -1.13334727e+00 1.03619468e+00 4.07498538e-01
-2.59021491e-01 -1.26952350e+00 2.04158232e-01 1.13183416e-01
6.16333544e-01 1.09119996e-01 1.11368608e+00 -7.85880089e-01
3.04794252e-01 -2.06354365e-01 -1.74581662e-01 4.03906196e-01
6.14111088e-02 -1.09055974e-01 -1.08001018e+00 -9.15909708e-02
-1.77359372e-01 -1.06266773e+00 7.83630550e-01 1.73598691e-03
1.12190402e+00 1.34985745e-01 -1.47736356e-01 3.17609251e-01
1.35856557e+00 1.70393303e-01 6.46576226e-01 1.94689363e-01
6.21940553e-01 6.89948440e-01 8.38657796e-01 6.24457002e-01
6.40945911e-01 5.62094867e-01 5.51263392e-01 -3.16118598e-02
3.49045962e-01 -2.72561729e-01 4.35238600e-01 1.25004423e+00
1.59803584e-01 -4.43926364e-01 -6.66424930e-01 7.63329148e-01
-1.96339130e+00 -8.90845537e-01 3.37339103e-01 1.94178236e+00
8.06862950e-01 -1.80050910e-01 -5.77038601e-02 -1.39875978e-01
6.79107368e-01 4.16871816e-01 -6.50554121e-01 -8.81418169e-01
-9.32049081e-02 1.26856878e-01 -2.15119436e-01 1.70296505e-01
-1.16492987e+00 1.14348412e+00 5.97301054e+00 6.91126347e-01
-1.23755860e+00 4.06024635e-01 3.42882901e-01 -2.24342003e-01
-3.17177683e-01 -3.32804024e-01 -4.63548481e-01 3.08726013e-01
1.00052202e+00 -1.15440302e-01 3.44560802e-01 8.04301798e-01
-1.85535505e-01 2.70769060e-01 -9.31305230e-01 1.32176280e+00
6.73136473e-01 -6.33212686e-01 5.81451282e-02 -3.37758541e-01
5.06396294e-01 -1.99548617e-01 1.18771628e-01 9.13580358e-01
-8.40681791e-02 -1.02643728e+00 5.22601485e-01 5.55567086e-01
7.66691446e-01 -1.11249065e+00 8.21222603e-01 2.02165022e-01
-1.23495817e+00 -1.24466255e-01 -7.28382826e-01 1.12867065e-01
4.62357640e-01 3.32780808e-01 -3.25292289e-01 5.23663282e-01
6.06273055e-01 8.43703330e-01 -2.81605244e-01 7.00364947e-01
-1.95030183e-01 3.23260069e-01 -6.33185655e-02 -2.40022644e-01
2.89277315e-01 -1.55392900e-01 9.65686291e-02 1.33424938e+00
6.47245407e-01 -1.00273289e-01 2.01674864e-01 4.77657825e-01
-1.66261628e-01 5.12642145e-01 -4.86644834e-01 -3.38749081e-01
4.68576252e-01 1.52934694e+00 -4.01827633e-01 -2.82694250e-01
-7.87615180e-01 1.31037283e+00 6.17477000e-01 5.23518145e-01
-9.53512132e-01 -6.67372406e-01 9.49080169e-01 -7.94488966e-01
4.63220060e-01 8.07547048e-02 2.41036750e-02 -1.28046930e+00
-2.17019498e-01 -7.75267482e-01 6.21851325e-01 -9.28221703e-01
-1.76759005e+00 8.72065783e-01 -4.12739873e-01 -1.24260378e+00
-2.66734600e-01 -6.76747024e-01 -5.79781532e-01 6.50327802e-01
-1.58027208e+00 -1.20722139e+00 -1.72879905e-01 7.40368545e-01
6.70280397e-01 -1.33761778e-01 1.36874533e+00 5.41021168e-01
-5.32981873e-01 7.14558244e-01 1.22998990e-01 1.02023736e-01
1.23440051e+00 -1.06089914e+00 -1.94527432e-01 5.02097249e-01
1.32889047e-01 3.10164660e-01 6.81907117e-01 -2.65591174e-01
-1.39069796e+00 -9.26284075e-01 9.62890148e-01 -2.82674849e-01
9.15290833e-01 -3.70267659e-01 -1.01395118e+00 5.72850764e-01
5.24898648e-01 2.26307049e-01 1.25362277e+00 5.54258645e-01
-8.05137277e-01 7.97901154e-02 -8.51556599e-01 5.04964888e-01
7.19977796e-01 -7.85793006e-01 -7.88935483e-01 2.76493840e-02
6.77466869e-01 -1.09433517e-01 -9.11220670e-01 3.30486745e-01
7.26535022e-01 -7.43606746e-01 8.46775472e-01 -9.17826414e-01
7.62771308e-01 -1.94232911e-02 -4.81610477e-01 -1.87485576e+00
-1.76950350e-01 2.73937378e-02 -1.13990799e-01 1.38079119e+00
3.73152107e-01 -4.89103436e-01 1.47517174e-01 3.75353962e-01
-9.02142078e-02 -8.70177686e-01 -1.00385213e+00 -5.39871335e-01
2.12406293e-02 -5.24790108e-01 5.66242635e-01 1.16191804e+00
3.19933861e-01 6.89082325e-01 -8.36781085e-01 8.53340402e-02
1.91580355e-01 2.96841413e-01 4.99958694e-01 -9.25138891e-01
-4.00870740e-02 -4.21207279e-01 -5.05914450e-01 -8.79953742e-01
5.65137208e-01 -6.05457187e-01 3.33642006e-01 -1.55640984e+00
3.18916529e-01 -1.21572226e-01 -8.38109791e-01 5.79838395e-01
-4.47764695e-01 4.17610824e-01 1.83354169e-01 -1.47078916e-01
-1.02811694e+00 1.00724530e+00 1.01445639e+00 1.27882473e-02
4.86530270e-03 -6.05958700e-01 -6.71886981e-01 6.73379004e-01
8.28076959e-01 -3.34750623e-01 -4.16626811e-01 -5.69694698e-01
3.03925484e-01 -4.45295312e-02 1.35748625e-01 -7.51512766e-01
1.39887765e-01 -2.58885086e-01 3.14132094e-01 -5.77035725e-01
7.30408967e-01 -1.00115263e+00 -3.27563167e-01 -1.87236860e-01
-4.88209903e-01 1.24775536e-01 2.05640301e-01 7.31365323e-01
-5.58537126e-01 -2.50852972e-01 6.15685701e-01 2.86420017e-01
-1.22553933e+00 2.58618563e-01 -3.71198922e-01 1.36692747e-01
1.06811821e+00 -2.77452301e-02 -2.25505039e-01 -6.17830873e-01
-8.03459227e-01 3.61718386e-01 2.70880818e-01 1.02007782e+00
8.25027823e-01 -1.71696949e+00 -4.16820347e-01 -6.52964860e-02
7.59306192e-01 -5.16633987e-01 7.01580703e-01 6.70661390e-01
2.73611993e-01 1.68775872e-03 -4.16474760e-01 -3.09001297e-01
-1.23543358e+00 8.30116868e-01 1.41906723e-01 -9.01163965e-02
-1.74756020e-01 7.86940813e-01 1.07156396e-01 -8.69593441e-01
3.39255303e-01 1.20217696e-01 -5.06105185e-01 5.51614046e-01
8.25068176e-01 -7.67317191e-02 -2.46264935e-01 -1.03201210e+00
-3.79376322e-01 7.60650754e-01 8.23942646e-02 -1.13043658e-01
1.43133020e+00 -3.29927862e-01 6.55620694e-02 1.04290700e+00
1.69927371e+00 -3.03993970e-01 -1.01232243e+00 -4.00437593e-01
-2.54126400e-01 -4.03453350e-01 -5.78198992e-02 -7.54096389e-01
-9.66056347e-01 1.40371704e+00 7.31614649e-01 6.63749725e-02
1.37091124e+00 5.93835711e-02 9.72705066e-01 1.07665628e-01
1.00480847e-01 -1.32626092e+00 4.64428455e-01 8.48091662e-01
7.04436660e-01 -1.22997749e+00 -4.84077871e-01 -3.40715349e-02
-1.23857844e+00 1.21775615e+00 6.82155728e-01 1.78420812e-01
5.25266469e-01 -2.10997052e-02 5.17924666e-01 -9.89538506e-02
-8.27696681e-01 -4.44351643e-01 3.45710546e-01 6.16306901e-01
5.51011145e-01 5.00374585e-02 -1.54562220e-01 7.55468547e-01
2.23174155e-01 -6.05716147e-02 1.51856810e-01 7.92478681e-01
-5.90460300e-01 -1.09936535e+00 -2.95591325e-01 1.99339822e-01
-2.11231917e-01 -1.04969116e-02 -2.24691927e-01 4.98426139e-01
-3.24730538e-02 1.12659919e+00 1.34009540e-01 -6.52380705e-01
5.10766566e-01 4.85459983e-01 2.41704345e-01 -5.81424057e-01
-3.53646696e-01 8.47955644e-02 -1.23250959e-02 -4.31937814e-01
-4.74540859e-01 -2.76666731e-01 -1.20420933e+00 1.15360208e-01
-2.81826645e-01 2.13285089e-01 6.65203273e-01 9.74640965e-01
5.49650729e-01 5.57982087e-01 7.06072450e-01 -8.96816432e-01
-5.72734892e-01 -1.04990220e+00 -7.05512822e-01 8.16854954e-01
-1.27061401e-02 -8.03912878e-01 -4.28221405e-01 -6.96267858e-02] | [13.159917831420898, 5.343306064605713] |
68d21f47-ae04-452b-ba9e-7772ffebc2e4 | plane-representation-learning-over-planar | 2307.01180 | null | https://arxiv.org/abs/2307.01180v1 | https://arxiv.org/pdf/2307.01180v1.pdf | PlanE: Representation Learning over Planar Graphs | Graph neural networks are prominent models for representation learning over graphs, where the idea is to iteratively compute representations of nodes of an input graph through a series of transformations in such a way that the learned graph function is isomorphism invariant on graphs, which makes the learned representations graph invariants. On the other hand, it is well-known that graph invariants learned by these class of models are incomplete: there are pairs of non-isomorphic graphs which cannot be distinguished by standard graph neural networks. This is unsurprising given the computational difficulty of graph isomorphism testing on general graphs, but the situation begs to differ for special graph classes, for which efficient graph isomorphism testing algorithms are known, such as planar graphs. The goal of this work is to design architectures for efficiently learning complete invariants of planar graphs. Inspired by the classical planar graph isomorphism algorithm of Hopcroft and Tarjan, we propose PlanE as a framework for planar representation learning. PlanE includes architectures which can learn complete invariants over planar graphs while remaining practically scalable. We empirically validate the strong performance of the resulting model architectures on well-known planar graph benchmarks, achieving multiple state-of-the-art results. | ['İsmail İlkan Ceylan', 'Ralph Abboud', 'Zeyang Zhao', 'Radoslav Dimitrov'] | 2023-07-03 | null | null | null | null | ['graph-regression'] | ['graphs'] | [ 2.73093373e-01 7.14198291e-01 -4.16990399e-01 -2.09139690e-01
-3.35945070e-01 -8.16375971e-01 4.51466650e-01 5.24218261e-01
1.06568173e-01 2.73658544e-01 -7.60427490e-02 -7.32265055e-01
-4.41216707e-01 -1.30908298e+00 -1.17357457e+00 -4.77859288e-01
-8.62648129e-01 8.10796142e-01 1.66340798e-01 -3.79460603e-01
-2.89708413e-02 7.22547889e-01 -1.37273872e+00 8.18360820e-02
5.32947183e-01 5.69434524e-01 -3.67338926e-01 9.40586865e-01
-1.24892659e-01 6.68621182e-01 -1.46533355e-01 -4.92991865e-01
4.93975610e-01 -2.13538393e-01 -1.43044019e+00 5.10418154e-02
9.59360600e-01 -9.67773423e-02 -8.12750459e-01 1.28544521e+00
-2.41763309e-01 5.37268259e-02 6.55349612e-01 -1.43730581e+00
-8.57049227e-01 9.16645825e-01 -3.75883609e-01 -6.89896494e-02
5.27727127e-01 -2.11864501e-01 1.66035771e+00 -1.84485465e-01
6.60813034e-01 1.35740077e+00 6.96129620e-01 2.58789361e-01
-1.44247890e+00 -2.80392647e-01 2.81088650e-01 3.25642943e-01
-1.19276440e+00 5.34002334e-02 7.85678864e-01 -2.95404613e-01
1.04932666e+00 3.05809855e-01 6.97807908e-01 6.40695393e-01
3.26133192e-01 4.54867065e-01 6.51447833e-01 -4.66277212e-01
5.92354313e-02 -4.95421112e-01 4.86349374e-01 1.25834370e+00
6.86865211e-01 1.84138566e-01 1.49306133e-01 -2.42624823e-02
7.35935986e-01 6.77213073e-02 -2.94594586e-01 -1.11576295e+00
-1.07624674e+00 9.55027044e-01 1.11905301e+00 4.55172509e-01
-3.95205170e-02 5.59176326e-01 5.71626425e-01 8.07503760e-01
-4.52970378e-02 7.02784061e-01 -2.02663288e-01 5.52821219e-01
-3.59271467e-01 1.57476947e-01 1.15834415e+00 1.01107657e+00
1.20500493e+00 9.95528996e-02 2.21741810e-01 1.98260605e-01
1.05121501e-01 3.06003481e-01 2.58786008e-02 -5.73846340e-01
4.33319420e-01 8.97215545e-01 -6.79485083e-01 -1.47087228e+00
-4.25025493e-01 -2.73406714e-01 -9.96861160e-01 4.34882082e-02
6.07692957e-01 4.29310024e-01 -9.68547821e-01 1.87066257e+00
-7.37306625e-02 2.62669951e-01 1.17552899e-01 5.27597308e-01
8.22318792e-01 3.76324087e-01 -2.48108864e-01 2.36616775e-01
1.02695441e+00 -7.88120270e-01 -1.61536112e-01 -2.64166117e-01
1.00996149e+00 -7.01818690e-02 6.47900105e-01 1.82143986e-01
-8.85695517e-01 -4.68625486e-01 -1.25884497e+00 1.54895177e-02
-6.14041269e-01 -5.54092646e-01 1.24973392e+00 6.88844860e-01
-1.37273335e+00 9.47628021e-01 -6.98670268e-01 -5.25579691e-01
3.31407130e-01 7.03042150e-01 -9.10067618e-01 -4.43844616e-01
-9.51233149e-01 5.39662600e-01 7.40794599e-01 3.01040933e-02
-9.60822403e-01 -4.33372557e-01 -1.39535427e+00 3.05476248e-01
4.92351145e-01 -5.86538494e-01 9.15762186e-01 -1.18752944e+00
-9.21360016e-01 1.06127477e+00 1.02997988e-01 -7.42213607e-01
6.30746111e-02 4.04677153e-01 -3.80399048e-01 2.71495193e-01
-9.68203023e-02 3.98815095e-01 7.27800012e-01 -1.04051626e+00
-7.07097203e-02 -5.25327325e-01 7.76916087e-01 -2.07256451e-01
-1.09522410e-01 -3.14734399e-01 -2.85887271e-01 -1.23824760e-01
3.70957732e-01 -1.14761436e+00 -3.95819634e-01 -2.52232820e-01
-7.24213421e-01 -3.00492942e-01 6.63401008e-01 -4.32550371e-01
7.76435733e-01 -1.91746271e+00 2.86315829e-01 7.33391583e-01
8.42302501e-01 3.82393450e-02 -5.67684770e-01 5.42323411e-01
-7.36103714e-01 2.63261706e-01 -1.54752463e-01 4.61239189e-01
1.52004659e-01 4.03292030e-01 -4.10976142e-01 7.13681400e-01
3.05652231e-01 1.23861194e+00 -9.08600450e-01 -1.94566920e-01
1.12046555e-01 1.54001102e-01 -7.85569131e-01 2.85535127e-01
-3.68044108e-01 2.16343738e-02 -3.64427745e-01 4.98699695e-01
7.40418136e-01 -6.41035199e-01 5.86884797e-01 -5.99200428e-02
4.22221005e-01 3.90883088e-01 -1.21332240e+00 1.59490156e+00
-8.01553428e-02 5.20581186e-01 -3.05606097e-01 -1.57800210e+00
8.79911900e-01 1.65351838e-01 3.39535952e-01 -4.89028245e-01
5.37318140e-02 1.27462037e-02 2.89747298e-01 -6.46570027e-02
3.48223716e-01 1.60995811e-01 -2.38748983e-01 3.65243316e-01
2.87554055e-01 -1.79959655e-01 4.81724650e-01 4.21615511e-01
1.58302283e+00 -1.91421863e-02 3.51433843e-01 -2.89109647e-01
6.11361265e-01 -1.61281422e-01 1.68412060e-01 7.37740695e-01
1.41178131e-01 3.39572817e-01 1.10028028e+00 -1.04811811e+00
-9.08035100e-01 -1.35119987e+00 3.53773475e-01 9.73814547e-01
1.14001736e-01 -7.27150202e-01 -6.97111726e-01 -8.48339498e-01
2.19556130e-02 1.70003444e-01 -8.55774164e-01 -5.25003731e-01
-6.56670511e-01 -3.90577242e-02 5.87238669e-01 5.39328516e-01
9.57128406e-02 -9.83740032e-01 -8.65944326e-02 2.25412101e-02
2.79167950e-01 -1.14047456e+00 -9.30637643e-02 4.55647439e-01
-9.83097851e-01 -1.80704129e+00 -2.04785526e-01 -1.20963681e+00
1.10637176e+00 4.88730252e-01 1.60384226e+00 7.40410328e-01
-2.44579256e-01 7.32142210e-01 -1.38642803e-01 -2.25461777e-02
-5.15008032e-01 4.50183898e-01 -2.39980057e-01 -2.37776980e-01
4.78888862e-02 -8.53852749e-01 -1.00334570e-01 1.23854458e-01
-1.06169760e+00 -1.53435245e-01 5.26799619e-01 7.85640001e-01
5.91007233e-01 1.96922466e-01 1.18641995e-01 -1.18529677e+00
3.17451805e-01 -4.71806347e-01 -9.33225572e-01 4.32350218e-01
-3.38848710e-01 5.13740361e-01 1.08367813e+00 -2.33540952e-01
-9.95470062e-02 2.98455469e-02 1.15857363e-01 -4.97083992e-01
-1.52793854e-01 6.81592584e-01 -1.61733136e-01 -5.46646476e-01
5.91019332e-01 1.62841603e-01 -1.16880320e-01 1.40726445e-02
5.33005178e-01 -3.06524664e-01 6.68564856e-01 -8.54185820e-01
1.33599901e+00 3.61036003e-01 6.20611429e-01 -8.98143053e-01
-5.59115589e-01 -3.13435167e-01 -7.12885737e-01 1.68015867e-01
6.73913121e-01 -5.17937541e-01 -8.35853696e-01 -7.56609887e-02
-1.10812163e+00 -2.56877452e-01 -2.66786546e-01 1.13873988e-01
-7.96849608e-01 7.16995478e-01 -4.99976784e-01 -2.32994065e-01
-5.87123744e-02 -1.10464358e+00 8.68432403e-01 -1.21230491e-01
-1.56425744e-01 -1.31466973e+00 4.66315150e-01 -1.07207052e-01
1.94763616e-01 5.42602360e-01 1.64961922e+00 -9.32705283e-01
-9.85307038e-01 -3.97133291e-01 -3.02626878e-01 9.29978043e-02
-9.43004116e-02 -8.46326575e-02 -6.21683896e-01 -7.20156789e-01
-7.39585996e-01 -4.28117782e-01 8.67343903e-01 1.45878375e-01
1.29724109e+00 -2.26285279e-01 -3.89318854e-01 9.08900201e-01
1.55511820e+00 -3.23906153e-01 7.71413863e-01 1.18238539e-01
9.72042263e-01 4.61245507e-01 -2.83851981e-01 -2.50436246e-01
3.56768489e-01 3.59334350e-01 8.82827699e-01 -1.23714432e-01
-2.47479090e-03 -4.86478895e-01 1.37571022e-01 6.92688406e-01
-2.99966484e-01 -1.15244061e-01 -9.59185123e-01 4.32434469e-01
-1.59867167e+00 -8.02499294e-01 -1.58832520e-01 2.19732070e+00
1.23408429e-01 1.18114062e-01 6.76208511e-02 2.49046594e-01
6.63857162e-01 5.50933957e-01 -2.89990395e-01 -6.52362227e-01
2.13723518e-02 7.58801937e-01 7.09012270e-01 4.15672243e-01
-1.08898437e+00 9.57400620e-01 6.19126940e+00 8.24237615e-02
-8.77550185e-01 -4.29307163e-01 3.96941155e-01 6.50929689e-01
-5.59061825e-01 4.63451713e-01 -1.97498947e-01 -3.21767420e-01
9.85273838e-01 -4.51332033e-01 9.05732751e-01 9.69149351e-01
-7.96251833e-01 5.12045503e-01 -1.64155316e+00 7.92769015e-01
9.57066193e-02 -1.36944711e+00 1.88256249e-01 1.83141336e-01
6.48626089e-01 9.20122564e-02 -9.26531572e-03 7.44357944e-01
7.61992514e-01 -1.42977655e+00 1.29942879e-01 1.35772660e-01
6.33480489e-01 -8.87892246e-01 5.56803346e-01 -9.56628919e-02
-1.71047711e+00 9.65381972e-03 -7.05100417e-01 6.33125659e-03
-4.66770560e-01 1.36801273e-01 -8.95053744e-01 9.87052023e-01
4.43249941e-01 8.40816498e-01 -9.29493308e-01 9.50540364e-01
-4.93479103e-01 4.07659501e-01 -4.41429541e-02 2.75945574e-01
3.86672020e-01 -2.91703790e-01 3.19645643e-01 9.16404605e-01
1.98708341e-01 -2.66194105e-01 4.64540482e-01 9.12280798e-01
-5.38890660e-01 7.14396909e-02 -1.50255740e+00 -5.43589413e-01
-4.70393687e-04 1.29634833e+00 -8.27204287e-01 8.31353758e-03
-4.97351527e-01 6.76526308e-01 9.17082429e-01 3.97467852e-01
-6.76662207e-01 -4.00192857e-01 5.09803176e-01 1.47665381e-01
4.22277719e-01 -3.74974608e-01 3.57458830e-01 -1.12302220e+00
-7.51525313e-02 -9.90607679e-01 7.94320405e-01 -4.42818016e-01
-1.14878070e+00 6.02791429e-01 -6.75753653e-02 -9.23560202e-01
-3.29640031e-01 -9.66549575e-01 -8.10884833e-01 4.17313129e-01
-1.44330645e+00 -1.32434464e+00 -3.48535180e-01 8.36540163e-01
-1.15055449e-01 -1.34130731e-01 1.12730587e+00 -1.19017713e-01
-1.44717485e-01 6.57649517e-01 -3.19579154e-01 5.43286562e-01
1.36935011e-01 -1.57696986e+00 9.15179074e-01 7.98036635e-01
8.29861462e-01 6.56554520e-01 3.67729187e-01 -4.22590375e-01
-2.35629058e+00 -1.11490417e+00 5.95763385e-01 -1.84840009e-01
9.52598989e-01 -3.68643254e-01 -1.03988886e+00 1.27788270e+00
1.57446638e-02 5.40485919e-01 3.26922059e-01 5.12155712e-01
-1.11410022e+00 8.88880808e-03 -7.05115616e-01 7.24763334e-01
1.31176412e+00 -7.95298100e-01 -4.69280958e-01 5.02593279e-01
9.11591768e-01 -4.17227507e-01 -8.66847575e-01 3.96219373e-01
3.35575104e-01 -8.94962192e-01 1.07530951e+00 -1.24772036e+00
3.11759412e-01 -2.76888490e-01 -2.69466162e-01 -1.37645543e+00
-5.50450027e-01 -6.35697484e-01 -8.07701796e-02 5.83001852e-01
4.43373114e-01 -8.71458769e-01 9.31012452e-01 1.52381271e-01
-7.27415234e-02 -4.68325049e-01 -7.78969526e-01 -9.92000818e-01
1.21516854e-01 -4.12334740e-01 7.55902231e-01 1.13186145e+00
1.08041927e-01 4.21259433e-01 1.74108334e-02 4.74392772e-01
5.85330963e-01 4.39346582e-01 1.12734962e+00 -1.57061207e+00
-5.00366628e-01 -5.23687243e-01 -1.23915029e+00 -7.96400785e-01
8.13058615e-01 -1.67725265e+00 -3.39166909e-01 -1.55435109e+00
1.65376067e-01 -8.63392502e-02 -3.59112501e-01 5.22067189e-01
2.52677441e-01 -6.37371987e-02 3.44698876e-02 -3.37405622e-01
-8.36841166e-01 3.08878899e-01 1.17054760e+00 -6.41244769e-01
1.64993316e-01 -1.69645339e-01 -9.06062365e-01 6.09561920e-01
6.04610860e-01 -4.10838485e-01 -6.06923699e-01 -3.95423889e-01
5.22002399e-01 1.00448765e-01 4.97303724e-01 -9.96943831e-01
1.84293419e-01 -4.48718034e-02 7.39077404e-02 -3.19389403e-01
-5.41820265e-02 -9.44395185e-01 1.98298454e-01 5.99924028e-01
-3.96047860e-01 4.53764796e-01 1.12570450e-01 6.96516097e-01
-1.80287570e-01 -9.83765721e-02 7.46055543e-01 -2.02076688e-01
-6.65438294e-01 8.02299023e-01 1.92575067e-01 3.14093292e-01
8.86523843e-01 -3.04797530e-01 -5.14634252e-01 -6.34906709e-01
-5.03998518e-01 1.52963437e-02 5.91849089e-01 3.71199429e-01
6.62569761e-01 -1.41222012e+00 -5.52426159e-01 4.13114816e-01
4.25438583e-01 8.07852075e-02 -6.81276247e-02 6.67129695e-01
-7.63159990e-01 3.63128781e-01 -3.30485314e-01 -6.19795322e-01
-1.13014352e+00 1.02162445e+00 4.75122660e-01 -6.83045447e-01
-8.00521195e-01 3.85701478e-01 5.30340075e-01 -8.50118697e-01
1.17614143e-01 -5.94816267e-01 -3.38094682e-02 -6.69518709e-01
3.28611076e-01 1.53788745e-01 1.87163576e-01 -5.61435819e-01
-1.09285749e-01 5.92519999e-01 -3.27507138e-01 6.45955980e-01
1.40686452e+00 6.37934327e-01 -5.53548217e-01 5.01676165e-02
1.46217716e+00 -1.61316603e-01 -6.17719114e-01 -2.53294319e-01
3.00854802e-01 -2.80687779e-01 -4.12896574e-01 -1.11475028e-02
-1.11926794e+00 7.20176697e-01 1.13606475e-01 5.60770094e-01
8.60948741e-01 2.40047827e-01 5.27545035e-01 1.15042853e+00
5.15958309e-01 -4.83863443e-01 6.72668517e-02 9.05449986e-01
8.06379795e-01 -9.99340594e-01 -4.42122389e-03 -5.88068485e-01
4.17184494e-02 1.42138791e+00 5.01649976e-01 -7.77604580e-01
3.76236290e-01 7.40419794e-03 -6.35731280e-01 -5.49904943e-01
-6.81689799e-01 -2.84563601e-01 5.59099376e-01 8.81301105e-01
2.81548291e-01 3.26156169e-01 3.34382623e-01 1.72362521e-01
-4.00755703e-01 -5.64877868e-01 5.28672099e-01 6.46424413e-01
-2.03971803e-01 -1.10665369e+00 -1.62577242e-01 5.02412081e-01
-7.19279945e-02 4.55349386e-02 -6.96669459e-01 1.25640786e+00
-3.58071595e-01 5.40290892e-01 -5.33440858e-02 -4.04024333e-01
3.34833860e-01 -9.42531377e-02 8.53216529e-01 -7.62024820e-01
-3.43563259e-01 -5.50730348e-01 2.00520322e-01 -8.52157354e-01
-5.32826819e-02 -3.75709124e-02 -1.25725281e+00 -3.85962665e-01
-1.01880029e-01 2.07559839e-01 2.72680372e-01 8.61707807e-01
8.09853300e-02 5.91080487e-01 5.95927536e-01 -7.11234629e-01
-4.76715058e-01 -4.50533211e-01 -7.69279122e-01 6.87869966e-01
3.24678093e-01 -4.87369001e-01 -4.28177267e-01 -2.15886921e-01] | [6.916382789611816, 6.223355770111084] |
c4c12f55-73b1-41d0-ac30-733014631f84 | deep-versus-wide-an-analysis-of-student | 2207.06867 | null | https://arxiv.org/abs/2207.06867v2 | https://arxiv.org/pdf/2207.06867v2.pdf | Deep versus Wide: An Analysis of Student Architectures for Task-Agnostic Knowledge Distillation of Self-Supervised Speech Models | Self-supervised learning (SSL) is seen as a very promising approach with high performance for several speech downstream tasks. Since the parameters of SSL models are generally so large that training and inference require a lot of memory and computational cost, it is desirable to produce compact SSL models without a significant performance degradation by applying compression methods such as knowledge distillation (KD). Although the KD approach is able to shrink the depth and/or width of SSL model structures, there has been little research on how varying the depth and width impacts the internal representation of the small-footprint model. This paper provides an empirical study that addresses the question. We investigate the performance on SUPERB while varying the structure and KD methods so as to keep the number of parameters constant; this allows us to analyze the contribution of the representation introduced by varying the model architecture. Experiments demonstrate that a certain depth is essential for solving content-oriented tasks (e.g. automatic speech recognition) accurately, whereas a certain width is necessary for achieving high performance on several speaker-oriented tasks (e.g. speaker identification). Based on these observations, we identify, for SUPERB, a more compressed model with better performance than previous studies. | ['Tomohiro Tanaka', 'Kohei Matsuura', 'Takafumi Moriya', 'Takanori Ashihara'] | 2022-07-14 | null | null | null | null | ['speaker-identification'] | ['speech'] | [ 1.06066398e-01 2.32508838e-01 -4.04371202e-01 -4.05854583e-01
-5.36749840e-01 -3.02749872e-01 4.51658487e-01 3.91566902e-01
-5.15082121e-01 5.57094693e-01 8.09635743e-02 -5.97758949e-01
-2.60756940e-01 -5.45541227e-01 -6.81142032e-01 -6.97093189e-01
-8.00917372e-02 5.39501250e-01 4.52599317e-01 7.64450803e-03
3.12875330e-01 6.02046609e-01 -1.88424337e+00 9.47047323e-02
6.68965399e-01 7.80037761e-01 4.73956406e-01 4.91078973e-01
-3.54703367e-01 4.52607423e-01 -7.48369038e-01 -8.67382288e-02
1.06655121e-01 -4.30451959e-01 -9.26403224e-01 4.07373458e-02
3.43122512e-01 -4.23679985e-02 -3.03483475e-02 9.25358176e-01
5.33694744e-01 2.10239872e-01 4.65052694e-01 -9.51454580e-01
1.42946944e-01 1.08519804e+00 -2.14940518e-01 4.06550199e-01
5.27348332e-02 -1.34141415e-01 8.36949289e-01 -5.72588027e-01
3.85201991e-01 1.41193974e+00 3.89471084e-01 4.08727884e-01
-1.29201579e+00 -5.97999275e-01 1.58834293e-01 2.48966679e-01
-1.51993942e+00 -1.03364015e+00 6.00884795e-01 -6.68326169e-02
1.09201050e+00 4.08707261e-01 3.97817880e-01 8.06773007e-01
-2.19605520e-01 4.55206156e-01 9.87916768e-01 -9.60986912e-01
6.42150819e-01 6.43292010e-01 3.39759797e-01 5.11089087e-01
4.64385629e-01 -1.48477899e-02 -8.36801410e-01 -2.06885636e-01
3.85037243e-01 -5.44886649e-01 -2.89024293e-01 -2.37047181e-01
-8.86937320e-01 8.74063790e-01 -4.81513105e-02 5.22383928e-01
2.30683181e-02 4.06277329e-02 5.28169513e-01 5.87471664e-01
3.62101346e-01 5.94108403e-01 -5.35192072e-01 -4.32337314e-01
-1.32987285e+00 1.15398183e-01 1.07888460e+00 8.56845975e-01
6.25953913e-01 3.02299142e-01 8.73903632e-02 1.13179111e+00
2.62690961e-01 2.79919744e-01 8.50994825e-01 -8.29290569e-01
4.94909644e-01 4.27540034e-01 -2.14632422e-01 -7.60955274e-01
-4.00191844e-01 -7.05661952e-01 -6.00936234e-01 -1.10409208e-01
4.69361722e-01 -8.94950852e-02 -6.40651047e-01 1.84789395e+00
2.30439663e-01 5.48396930e-02 8.76180604e-02 6.06953442e-01
6.26778603e-01 8.01016212e-01 1.74084213e-02 -5.64362526e-01
1.37352169e+00 -8.20193827e-01 -6.98027611e-01 -5.10949016e-01
1.00510180e+00 -7.54510403e-01 1.20254838e+00 4.06713337e-01
-1.20710766e+00 -6.03271067e-01 -1.22034240e+00 1.47431597e-01
-2.41663381e-01 1.89327970e-02 3.89570326e-01 1.12209964e+00
-1.01891875e+00 5.38575292e-01 -8.28458071e-01 -4.78509277e-01
1.72739133e-01 3.52976054e-01 -1.57656983e-01 1.66107431e-01
-1.12238193e+00 1.16434240e+00 6.99045300e-01 -2.31703982e-01
-4.63126212e-01 -6.07864618e-01 -7.50072300e-01 3.47069263e-01
3.46917719e-01 -4.31287348e-01 1.12847114e+00 -6.22343481e-01
-1.46858084e+00 6.14078403e-01 -3.38234395e-01 -1.04543817e+00
2.95145929e-01 1.65396789e-03 -2.64931649e-01 6.19435161e-02
-5.42117357e-01 6.11401796e-01 1.07184708e+00 -1.14052510e+00
-4.48755711e-01 -4.60600227e-01 8.92548710e-02 1.34901091e-01
-9.37749863e-01 3.39581072e-02 -4.46014315e-01 -6.06878400e-01
3.44513237e-01 -1.05088663e+00 -1.35438859e-01 -3.50402325e-01
-1.86760753e-01 -2.43733570e-01 8.32817078e-01 -5.94335139e-01
1.75865316e+00 -2.32375479e+00 1.00512363e-01 3.74097854e-01
-8.26746896e-02 6.16396904e-01 -2.57064458e-02 5.30309260e-01
9.45244431e-02 1.29872411e-01 -2.25302190e-01 -5.22636294e-01
-1.01416886e-01 3.76459569e-01 -2.55342275e-01 1.52268305e-01
-2.60933101e-01 5.22777200e-01 -4.12208050e-01 -5.75919151e-01
1.39995143e-01 4.94494766e-01 -7.22524524e-01 8.05028453e-02
-2.58784175e-01 6.52243569e-02 -2.40542382e-01 1.91535518e-01
4.59911138e-01 -4.54826988e-02 2.87047416e-01 -1.57590628e-01
8.34055804e-03 7.46173501e-01 -1.40706158e+00 1.43305552e+00
-7.85427988e-01 7.21294940e-01 2.71815434e-02 -1.26820517e+00
1.08603251e+00 3.44139278e-01 7.46038109e-02 -6.07963264e-01
-6.82646856e-02 1.38893992e-01 1.91942289e-01 -3.85017633e-01
4.62594151e-01 -2.86414236e-01 1.75637454e-01 4.77142692e-01
-1.43410519e-01 -1.49903402e-01 3.38740677e-01 1.21077210e-01
8.26166570e-01 -4.75864679e-01 2.80213147e-01 -3.85195732e-01
6.16053522e-01 -1.52395621e-01 4.19800401e-01 7.06904709e-01
6.80686384e-02 9.62605774e-02 4.48447436e-01 -1.16818048e-01
-1.06815767e+00 -5.92775643e-01 -2.80285984e-01 1.01941228e+00
-1.52299345e-01 -6.66808844e-01 -1.08920300e+00 -1.60130575e-01
-1.09277688e-01 9.55832839e-01 -4.50242497e-02 -4.75885868e-01
-5.89081168e-01 -7.35566020e-01 7.27514923e-01 4.33492780e-01
6.22669280e-01 -9.17580009e-01 -7.76539147e-01 2.65421540e-01
-5.64360656e-02 -1.29882264e+00 -1.72589675e-01 4.29906696e-01
-1.33326209e+00 -6.11681998e-01 -3.29460979e-01 -7.97720432e-01
5.25713444e-01 2.58264393e-01 9.60193396e-01 4.25748900e-02
-2.52402388e-02 1.75636917e-01 -5.46838164e-01 -3.63930851e-01
-9.04068112e-01 4.58758503e-01 1.81573316e-01 -1.19531408e-01
2.94202238e-01 -6.27596974e-01 -1.53628632e-01 2.13971481e-01
-9.56104100e-01 1.05787866e-01 5.85282743e-01 7.27129161e-01
3.55462462e-01 5.80478549e-01 7.21072555e-01 -8.76722753e-01
7.50732064e-01 -3.51924449e-02 -6.89284682e-01 2.32414126e-01
-7.95206189e-01 5.17411113e-01 6.70169532e-01 -4.68184322e-01
-9.55188036e-01 -8.27320591e-02 -2.88709849e-01 -2.95888215e-01
-1.65848300e-01 5.92496932e-01 -2.07299724e-01 9.50969160e-02
6.33063078e-01 3.09873104e-01 1.86825573e-01 -7.29497969e-01
2.41500646e-01 9.63102818e-01 -3.12057231e-02 -4.16841775e-01
5.51886618e-01 1.23441078e-01 -1.78867742e-01 -1.35876131e+00
-6.81807339e-01 -3.04770231e-01 -5.53268313e-01 1.26936466e-01
3.91073883e-01 -7.23405719e-01 -4.39464837e-01 1.01474822e-01
-8.53935063e-01 -4.05590802e-01 -3.13388735e-01 6.18007541e-01
-5.36700845e-01 3.35312396e-01 -4.15162206e-01 -9.27364647e-01
-2.77007759e-01 -1.12957418e+00 6.90459847e-01 -1.94243863e-01
-4.73748028e-01 -8.69910359e-01 -2.82408416e-01 6.33546174e-01
5.32988131e-01 -5.03986537e-01 1.24907410e+00 -7.72975206e-01
-3.51132482e-01 5.31951301e-02 9.33077559e-02 6.36635482e-01
-1.28164843e-01 -3.66687924e-01 -1.04688406e+00 -5.96271276e-01
2.26122692e-01 -1.71710655e-01 8.69115889e-01 2.35408619e-01
1.30525911e+00 -4.72389221e-01 -1.53531387e-01 3.67180228e-01
1.18753254e+00 1.77152529e-01 5.68862438e-01 1.70222178e-01
2.92373478e-01 7.60020733e-01 5.47510684e-01 4.33171064e-01
2.88977176e-02 9.85674798e-01 8.91460776e-02 2.65606999e-01
-3.65421712e-01 -1.60868675e-01 4.72033143e-01 1.20431376e+00
1.97924912e-01 -2.24183742e-02 -9.28681910e-01 4.10062283e-01
-1.51889277e+00 -8.31112266e-01 2.21642002e-01 2.57702899e+00
9.94110703e-01 5.57868898e-01 1.01134740e-01 7.65430331e-01
4.87218648e-01 2.40456641e-01 -3.08258146e-01 -7.61545777e-01
-1.69042461e-02 2.07996279e-01 4.36162800e-01 5.90623081e-01
-6.41559660e-01 8.39765966e-01 6.27756119e+00 1.15481198e+00
-1.18378484e+00 1.32990956e-01 4.98784721e-01 -2.37073302e-01
-1.85201630e-01 1.23413824e-01 -1.18733573e+00 5.20627201e-01
1.49041283e+00 -2.20399827e-01 4.41066027e-01 8.00937951e-01
4.11230952e-01 -2.24171370e-01 -1.05490410e+00 9.36637700e-01
3.31289880e-02 -1.19082284e+00 1.69437498e-01 1.37646273e-01
3.26693743e-01 -1.40192613e-01 -1.36881351e-01 2.83066601e-01
-3.13880384e-01 -7.68388748e-01 6.82803035e-01 5.33656143e-02
5.55909932e-01 -7.81212866e-01 6.04572713e-01 7.13030815e-01
-9.00710642e-01 -1.89443365e-01 -5.01277804e-01 1.66113023e-02
3.25140357e-02 7.85883367e-01 -1.04649770e+00 2.90382430e-02
5.35311699e-01 1.92000300e-01 -5.66954315e-01 8.45472753e-01
3.02983653e-02 1.12681913e+00 -4.92677838e-01 -9.17900354e-02
1.33133575e-01 3.34088877e-02 4.80747133e-01 1.46653664e+00
2.87267029e-01 -1.81322023e-01 -1.98015451e-01 7.19291627e-01
8.65049586e-02 2.29530320e-01 -5.86184978e-01 -1.41879976e-01
9.09479916e-01 6.66377306e-01 -5.66513538e-01 -3.21338505e-01
-2.81249344e-01 5.90165198e-01 2.71045983e-01 2.13348910e-01
-6.17553353e-01 -1.11732237e-01 4.70929891e-01 5.02993882e-01
4.00543869e-01 -4.08401579e-01 -3.32385123e-01 -7.38370419e-01
9.76493359e-02 -9.75726247e-01 2.21770525e-01 -3.84213954e-01
-6.32244229e-01 6.47120357e-01 3.05938244e-01 -9.67333496e-01
-4.43399519e-01 -2.39155069e-01 -2.71216422e-01 6.51304722e-01
-1.48352826e+00 -7.66447723e-01 -1.57869384e-02 3.47885251e-01
7.20432758e-01 -3.56840223e-01 9.37029839e-01 2.66344517e-01
-7.26712584e-01 8.80520642e-01 9.11463946e-02 -3.05500627e-01
4.33114141e-01 -8.18711817e-01 1.37157500e-01 5.70544600e-01
3.49346071e-01 7.33252406e-01 8.64308357e-01 -2.96574414e-01
-1.40128601e+00 -8.66501093e-01 1.17469692e+00 -2.30167247e-02
3.06600302e-01 -4.07867193e-01 -1.06246305e+00 3.78656030e-01
-1.30756602e-01 -2.91658968e-01 6.73774779e-01 2.22941056e-01
-3.27327073e-01 -3.28850001e-01 -9.98341262e-01 5.05571842e-01
9.39180076e-01 -5.59171677e-01 -6.00642562e-01 1.68525040e-01
7.48274386e-01 -1.30635813e-01 -8.37832808e-01 2.31316641e-01
2.63521552e-01 -1.08465171e+00 1.01472068e+00 -3.25872540e-01
4.26167995e-02 -5.09172603e-02 -1.78135812e-01 -1.19656992e+00
-1.28507791e-02 -4.07792747e-01 -4.94595408e-01 1.30629265e+00
6.19906306e-01 -7.42403209e-01 8.54125619e-01 5.18322349e-01
-4.16295268e-02 -8.21528077e-01 -1.08983386e+00 -1.10763025e+00
-1.02512635e-01 -5.77004790e-01 5.56540906e-01 7.00004101e-01
7.32193142e-02 4.58840400e-01 -1.29364058e-01 1.50319310e-02
3.67463946e-01 1.45116802e-02 7.58924425e-01 -1.10192800e+00
-3.47710371e-01 -4.50679988e-01 -2.71313041e-01 -1.14559758e+00
1.91890806e-01 -8.38251591e-01 -2.58840978e-01 -1.27779031e+00
-4.20656130e-02 -8.81601095e-01 -2.49368459e-01 3.75425249e-01
2.22776130e-01 -1.52723745e-01 3.23827356e-01 2.45055377e-01
-2.73380101e-01 3.46019655e-01 8.43468904e-01 2.53729392e-02
-4.25240368e-01 2.31124386e-01 -6.09635472e-01 6.15439534e-01
9.99536514e-01 -3.84982854e-01 -8.20393145e-01 -3.66041839e-01
5.27379960e-02 1.03195034e-01 7.01091215e-02 -1.06430328e+00
4.41857308e-01 8.82854015e-02 -2.03097537e-01 -4.84509736e-01
6.10147595e-01 -8.41839492e-01 -5.78790670e-03 5.73471963e-01
-5.16154170e-01 -2.15982601e-01 4.12797809e-01 3.16735178e-01
-4.23476666e-01 -7.81984270e-01 9.53959703e-01 3.25063020e-02
-5.89778483e-01 -2.51665354e-01 -3.17453623e-01 -2.25527138e-02
7.44072199e-01 -3.53055537e-01 1.08579360e-01 -4.15054798e-01
-7.00185180e-01 -1.37303978e-01 2.17503518e-01 4.57993388e-01
4.90307927e-01 -8.88806999e-01 -5.20724714e-01 5.96236765e-01
3.05725690e-02 -7.77688324e-02 6.02903478e-02 7.09932745e-01
-4.75353301e-01 7.77133882e-01 1.72959000e-01 -5.11599898e-01
-1.56266296e+00 4.10974026e-01 2.23647639e-01 -2.28564948e-01
-6.47365808e-01 7.33537734e-01 -2.74628133e-01 -1.75560459e-01
7.86615491e-01 -3.89577538e-01 -7.56784305e-02 1.94547176e-01
5.82228720e-01 5.18037796e-01 4.11119223e-01 -3.86025459e-01
-2.31092349e-01 3.57353687e-01 -1.70511231e-01 -7.83803686e-02
1.20190954e+00 -1.50166482e-01 7.80552765e-03 4.30001110e-01
1.09519470e+00 -9.87006053e-02 -9.02383208e-01 -3.81462216e-01
2.56706268e-01 -2.81897932e-01 5.23115337e-01 -4.12398666e-01
-9.51440275e-01 1.00629318e+00 4.94632751e-01 3.54888171e-01
1.05361915e+00 2.74886601e-02 7.55605638e-01 5.20611584e-01
5.75047314e-01 -1.31774354e+00 -8.79683122e-02 4.49535072e-01
7.64520049e-01 -9.07285929e-01 1.42894521e-01 -5.70895433e-01
-3.94302547e-01 7.92107522e-01 2.94947356e-01 4.54253078e-01
7.82833815e-01 5.75313449e-01 -2.35369071e-01 8.27490017e-02
-8.18252802e-01 -2.36132517e-02 1.58527955e-01 2.65138596e-01
3.63207698e-01 7.72455111e-02 -4.39162046e-01 3.48093688e-01
-6.33265078e-01 -2.04402030e-01 2.79407591e-01 9.10658419e-01
-8.22543859e-01 -1.36183381e+00 -4.74386871e-01 5.44774771e-01
-2.12791443e-01 -9.00551304e-02 -9.98160839e-02 6.36124313e-01
-6.94735274e-02 1.18158221e+00 4.70485911e-02 -2.66649008e-01
2.96914458e-01 4.13006097e-01 5.03830791e-01 -7.21824110e-01
-5.18625855e-01 7.47201070e-02 3.23556572e-01 -2.34818593e-01
-4.09651607e-01 -5.09181380e-01 -1.21382737e+00 -4.49521869e-01
-7.09889710e-01 3.32741022e-01 9.81560826e-01 9.57332313e-01
3.40053111e-01 4.22080368e-01 4.50179577e-01 -3.62132072e-01
-9.21234369e-01 -1.06007695e+00 -6.77910328e-01 2.28940636e-01
2.24841177e-03 -5.82040429e-01 -4.64953870e-01 5.82146347e-02] | [14.12549877166748, 6.5279927253723145] |
083fc393-cd5d-45a8-925d-b1b24ee87b2d | fused-text-segmentation-networks-for-multi | 1709.03272 | null | http://arxiv.org/abs/1709.03272v4 | http://arxiv.org/pdf/1709.03272v4.pdf | Fused Text Segmentation Networks for Multi-oriented Scene Text Detection | In this paper, we introduce a novel end-end framework for multi-oriented
scene text detection from an instance-aware semantic segmentation perspective.
We present Fused Text Segmentation Networks, which combine multi-level features
during the feature extracting as text instance may rely on finer feature
expression compared to general objects. It detects and segments the text
instance jointly and simultaneously, leveraging merits from both semantic
segmentation task and region proposal based object detection task. Not
involving any extra pipelines, our approach surpasses the current state of the
art on multi-oriented scene text detection benchmarks: ICDAR2015 Incidental
Scene Text and MSRA-TD500 reaching Hmean 84.1% and 82.0% respectively. Morever,
we report a baseline on total-text containing curved text which suggests
effectiveness of the proposed approach. | ['Weidong Qiu', 'Jie Guo', 'Kai Chen', 'Youxuan Xu', 'Yuting Gao', 'Zheng Huang', 'Yuchen Dai'] | 2017-09-11 | null | null | null | null | ['multi-oriented-scene-text-detection'] | ['computer-vision'] | [ 5.01402915e-01 1.78484142e-01 1.35348007e-01 -3.40994000e-01
-1.08373821e+00 -6.78003848e-01 7.87604392e-01 1.81795090e-01
-5.34014761e-01 1.62953958e-01 -3.05936169e-02 1.90403163e-02
2.30784655e-01 -5.00504851e-01 -7.49499798e-01 -4.14877504e-01
6.32912219e-01 7.19452083e-01 9.07357693e-01 2.16647666e-02
5.31644225e-01 1.67845875e-01 -1.23009574e+00 6.67509556e-01
8.88783216e-01 1.04899287e+00 4.55461681e-01 8.69156182e-01
-3.71320695e-01 6.35350168e-01 -5.17960250e-01 -4.03099805e-01
3.54337513e-01 2.89170407e-02 -8.85299981e-01 5.90933800e-01
1.19247103e+00 -3.39933634e-01 -1.58677414e-01 8.22931826e-01
4.25550342e-01 -1.17657423e-01 8.80128086e-01 -8.17791641e-01
-3.36460054e-01 7.86475003e-01 -1.22346938e+00 4.98849541e-01
8.01669583e-02 7.37121627e-02 1.18687499e+00 -1.13620126e+00
6.67548656e-01 1.38158298e+00 6.46953404e-01 1.56123728e-01
-7.67248213e-01 -3.59165281e-01 6.38447404e-01 -2.67610043e-01
-1.13575506e+00 -1.66859135e-01 4.82232660e-01 -4.25092250e-01
1.24372721e+00 2.67469645e-01 4.58318204e-01 9.14852798e-01
1.94319203e-01 1.70494866e+00 8.23750794e-01 -4.31721002e-01
-1.63729727e-01 1.18068300e-01 4.51488137e-01 7.21607208e-01
5.70247233e-01 -5.35088778e-01 -5.30939162e-01 2.12770015e-01
3.69458526e-01 -1.46231338e-01 9.48557854e-02 -3.26440990e-01
-1.36051071e+00 5.80416858e-01 1.74778655e-01 3.85239393e-01
-2.30567798e-01 3.27849686e-01 6.67409956e-01 -3.86049360e-01
7.30633318e-01 1.03284642e-01 -7.88823545e-01 1.84397444e-01
-1.46591175e+00 1.50462478e-01 3.22888464e-01 1.20689082e+00
4.25408095e-01 5.97918481e-02 -6.63075686e-01 9.28645492e-01
4.70013618e-01 6.49876356e-01 3.36058974e-01 -3.52524757e-01
9.34272170e-01 1.13203025e+00 -2.91255563e-01 -5.98335981e-01
-5.74943841e-01 -5.11739910e-01 -4.08784628e-01 -3.06998521e-01
4.51634824e-01 -7.16776103e-02 -1.72176778e+00 7.84672618e-01
5.26558042e-01 -6.10662326e-02 -7.30348676e-02 7.46693611e-01
1.12802029e+00 4.83346939e-01 3.01228344e-01 3.40330392e-01
1.79946804e+00 -1.33733964e+00 -6.66959345e-01 -5.06627500e-01
8.27589571e-01 -1.05636299e+00 1.03584874e+00 4.71964657e-01
-1.14908302e+00 -2.93262839e-01 -8.95337760e-01 -3.71782720e-01
-6.23100579e-01 5.99028885e-01 4.55436200e-01 7.83701181e-01
-1.03322065e+00 7.76214823e-02 -8.35733414e-01 -9.77218151e-01
8.92487109e-01 3.25782150e-01 1.18106946e-01 7.72396550e-02
-3.63093287e-01 5.22648036e-01 7.69592166e-01 -1.04324020e-01
-8.94833267e-01 -7.01018393e-01 -8.03820252e-01 -2.17949301e-02
8.20483506e-01 -8.45602036e-01 1.19406092e+00 -6.71705723e-01
-1.07574213e+00 1.11521292e+00 -1.08426604e-02 -6.85167611e-01
9.51374054e-01 -5.93298137e-01 -6.21354133e-02 4.71206933e-01
4.77184266e-01 1.07944107e+00 1.05184066e+00 -1.15479469e+00
-1.14782786e+00 -4.63803113e-01 -3.01767558e-01 4.79965895e-01
-1.66379780e-01 8.55132937e-02 -9.45980370e-01 -9.03416693e-01
3.58674318e-01 -6.11030638e-01 -1.59542277e-01 1.63275059e-02
-1.15223289e+00 -4.46961224e-01 1.49753082e+00 -4.63966787e-01
8.43279302e-01 -1.75337398e+00 -1.99564964e-01 -1.93751082e-01
3.24490666e-01 4.80029061e-02 1.63248349e-02 2.55743325e-01
2.94885635e-01 3.28050792e-01 -4.00111347e-01 -6.42711222e-01
2.59578049e-01 -3.65868509e-01 -2.53190607e-01 6.33141637e-01
2.38576621e-01 1.40468812e+00 -2.50874430e-01 -1.07848585e+00
7.90668070e-01 2.35757858e-01 -3.81208807e-01 -2.31406465e-01
-6.10563219e-01 -4.21605781e-02 -7.95804322e-01 1.19273484e+00
8.79310191e-01 -3.19081575e-01 -2.81481922e-01 -3.17892641e-01
-9.27459821e-02 -1.19097322e-01 -1.19029891e+00 1.84466386e+00
9.05296113e-03 9.52286243e-01 5.25445640e-02 -8.25536907e-01
4.73378688e-01 7.35582858e-02 4.44036067e-01 -6.41357541e-01
4.14169580e-01 -2.68327072e-02 -4.54699248e-01 -4.50130492e-01
1.07202220e+00 2.83050627e-01 -1.99470565e-01 1.41504422e-01
2.24868972e-02 -3.16951692e-01 2.85811543e-01 4.92423147e-01
8.04348409e-01 4.25531089e-01 2.93365002e-01 -5.18702209e-01
4.46319818e-01 3.70227247e-01 8.16624239e-02 1.12513423e+00
-3.06253046e-01 7.86531091e-01 3.39180738e-01 -2.22081646e-01
-9.00748789e-01 -9.17031586e-01 -5.13393939e-01 1.54570925e+00
3.21904868e-01 -3.32544267e-01 -9.79024529e-01 -1.10581350e+00
-2.57400163e-02 6.60756230e-01 -6.88232064e-01 5.89229047e-01
-5.94036937e-01 -1.10177636e+00 9.00974870e-01 6.61191821e-01
8.75245690e-01 -8.27389598e-01 -7.62846410e-01 -9.38925892e-03
-1.34984031e-01 -1.83750999e+00 -6.82619333e-01 3.64848822e-01
-1.02797282e+00 -8.91978860e-01 -8.79500866e-01 -7.90426195e-01
3.96320552e-01 4.48408842e-01 1.17903650e+00 -2.56267339e-01
-8.12633514e-01 6.72189891e-01 -4.33405191e-01 -5.20372033e-01
-6.47261143e-02 2.90850520e-01 -7.12664962e-01 -1.62068635e-01
1.97574481e-01 2.24344984e-01 -7.50940979e-01 1.73895434e-01
-8.99281919e-01 3.03203732e-01 7.81268060e-01 4.15195674e-01
5.90432227e-01 -2.40800027e-02 2.84999877e-01 -1.00198007e+00
1.78390294e-01 -1.62097961e-01 -5.10522723e-01 4.00166184e-01
-2.92527735e-01 -3.88846755e-01 1.88344717e-01 -2.89846566e-02
-1.12573016e+00 5.25089145e-01 1.03289239e-01 -6.77164793e-02
-6.61336839e-01 -8.12669247e-02 -7.35800713e-02 1.07182175e-01
3.80892992e-01 5.97664118e-01 -8.77817333e-01 -4.31315482e-01
5.26558220e-01 8.44743550e-01 3.44038785e-01 -4.69009131e-01
5.82077980e-01 9.01804626e-01 -3.37344557e-01 -1.11347413e+00
-9.66690183e-01 -1.07080388e+00 -8.91550303e-01 5.29067703e-02
1.34809780e+00 -1.19970787e+00 -4.01813030e-01 6.68506145e-01
-1.08227026e+00 -3.67646605e-01 -6.23597652e-02 -8.49071294e-02
-5.26106536e-01 6.68166161e-01 -5.89452744e-01 -9.34808552e-01
-7.83588231e-01 -9.65229034e-01 2.22435570e+00 4.86662902e-04
1.80542424e-01 -8.33223403e-01 -5.19389808e-01 7.64810383e-01
-1.58714235e-03 2.93404222e-01 5.31531870e-01 -8.88567209e-01
-9.05455828e-01 -2.56695658e-01 -6.69331491e-01 -3.00448596e-01
-2.46598765e-01 1.27760455e-01 -1.30313540e+00 -1.11804873e-01
-4.83570397e-01 -2.08607391e-01 1.34706914e+00 7.89198339e-01
1.08631051e+00 1.43441051e-01 -6.20840073e-01 4.82210368e-01
1.64459705e+00 -3.92471589e-02 2.39673495e-01 4.49955434e-01
1.13649666e+00 4.86166060e-01 8.26179385e-01 4.94271487e-01
5.50886393e-01 5.13083100e-01 6.40558958e-01 -5.34245849e-01
-2.79678851e-01 4.65975404e-02 1.35215133e-01 7.39630759e-02
5.12206852e-01 -8.98272693e-01 -1.10275912e+00 8.36499035e-01
-1.83258820e+00 -4.91492540e-01 -7.53675997e-01 1.45719647e+00
1.77101210e-01 3.65245700e-01 4.13295478e-01 4.28189524e-02
9.33578312e-01 8.16628411e-02 -8.39035511e-01 -9.73733962e-02
-3.32091510e-01 8.28599557e-02 8.46233845e-01 1.44471526e-01
-1.73271430e+00 1.79314625e+00 5.74969292e+00 1.26089549e+00
-6.80668056e-01 2.21714824e-01 6.83219492e-01 -3.44560742e-02
2.24959031e-01 -3.77303064e-01 -1.14467180e+00 -6.15030229e-02
3.67427677e-01 3.06043625e-01 -3.09286565e-01 9.50348735e-01
2.93987900e-01 -5.04924893e-01 -8.94310057e-01 9.31593120e-01
1.70864537e-01 -1.23221135e+00 4.37581718e-01 1.24926940e-02
8.24965775e-01 4.51752275e-01 1.19678333e-01 1.73278332e-01
2.22065777e-01 -9.04235482e-01 1.08170164e+00 -2.46481895e-02
4.69722688e-01 -4.54313606e-01 5.30716419e-01 2.97432005e-01
-1.68079782e+00 8.07836950e-02 -1.51874438e-01 5.84178686e-01
1.28333077e-01 5.61251163e-01 -1.02327609e+00 6.57381117e-01
8.11363816e-01 9.71793830e-01 -1.03625548e+00 1.08150923e+00
2.00128838e-01 7.16539085e-01 -4.68604714e-01 -1.07032731e-01
7.59764493e-01 3.07333797e-01 5.85059524e-01 1.84884775e+00
8.90702195e-03 -1.79203272e-01 5.40070295e-01 1.00010157e+00
-8.35245103e-02 5.21549046e-01 -3.08824539e-01 -7.67133897e-03
-1.42409874e-03 1.53246605e+00 -1.69270813e+00 -6.59924030e-01
-3.21905047e-01 1.32637012e+00 -4.98942807e-02 2.74202734e-01
-1.05425513e+00 -2.03523308e-01 -7.38193393e-02 -1.36838377e-01
7.85648763e-01 5.89499921e-02 -7.95369089e-01 -1.17708349e+00
1.73474297e-01 -6.34086132e-01 6.21394455e-01 -7.74495423e-01
-9.37999010e-01 4.00223643e-01 -1.71119049e-01 -9.58797038e-01
3.13875496e-01 -8.11767876e-01 -4.36226219e-01 4.36160088e-01
-1.60431468e+00 -1.85733712e+00 -3.78983110e-01 5.74977815e-01
1.51930583e+00 1.53291136e-01 9.04570371e-02 5.58807589e-02
-9.65135992e-01 6.90668464e-01 -3.03803906e-02 2.14690834e-01
5.21977723e-01 -1.59778512e+00 7.88961172e-01 9.82402742e-01
2.11709127e-01 5.73898889e-02 5.46461165e-01 -7.81474710e-01
-1.19180608e+00 -1.48627245e+00 3.12500745e-01 -8.36111069e-01
5.04321635e-01 -6.19220018e-01 -5.59136748e-01 7.69753635e-01
4.09917474e-01 -9.58105028e-02 -3.88442464e-02 -3.02107573e-01
-1.64419226e-02 4.51287985e-01 -1.32267749e+00 6.44142389e-01
9.94080126e-01 -8.05620253e-02 -4.48561102e-01 6.69519246e-01
7.28933930e-01 -6.87804222e-01 -5.88647187e-01 5.28453648e-01
2.12614164e-01 -9.94341135e-01 1.03066254e+00 -1.41702130e-01
2.95487225e-01 -2.42346197e-01 -3.28286082e-01 -4.00275081e-01
9.77227166e-02 -3.05300593e-01 -4.09430526e-02 1.26615310e+00
3.94544482e-01 -2.94708282e-01 1.12549222e+00 3.02515761e-03
-5.24989009e-01 -6.03580773e-01 -8.18654120e-01 -5.57678103e-01
3.61601800e-01 -8.29195976e-01 2.07494557e-01 7.83199787e-01
-4.20211136e-01 3.76172394e-01 -9.47371945e-02 2.88599223e-01
6.60375297e-01 9.63589251e-02 7.32051671e-01 -1.07239056e+00
9.40518007e-02 -6.83750987e-01 -2.98727751e-01 -1.43000793e+00
-8.78680274e-02 -1.00048435e+00 2.06645459e-01 -1.92992735e+00
5.81238329e-01 -1.27676707e-02 -6.75535053e-02 3.87591839e-01
-4.29011315e-01 5.01717329e-01 3.61682862e-01 -6.08255565e-02
-1.23853171e+00 3.05121809e-01 1.26848137e+00 -2.34321132e-01
-9.84128863e-02 -3.05101216e-01 -5.41364729e-01 1.07203114e+00
7.80574799e-01 -5.27401388e-01 -1.84995338e-01 -4.41098750e-01
-1.42181680e-01 -2.63171136e-01 4.68531370e-01 -1.05751908e+00
1.46211714e-01 5.02913222e-02 4.74581361e-01 -1.52512705e+00
2.34268010e-01 -6.93968177e-01 -6.46759748e-01 2.78602242e-01
-2.46469676e-01 -3.68764609e-01 5.55101573e-01 9.57042158e-01
2.98158318e-01 -1.35907173e-01 9.36966240e-01 -1.38453692e-01
-1.02684474e+00 1.31511465e-01 -5.94234943e-01 3.77312481e-01
1.13286459e+00 -7.50519931e-01 -4.71158475e-01 2.49186248e-01
-4.93404418e-01 4.57894236e-01 3.44769746e-01 5.04295528e-01
5.07815063e-01 -4.55758005e-01 -7.78845966e-01 -2.71411717e-01
4.37276125e-01 2.90448993e-01 2.65618980e-01 1.00069904e+00
-5.66590905e-01 8.86070788e-01 5.12731016e-01 -1.28614664e+00
-1.45723832e+00 2.78790832e-01 4.44998950e-01 -4.54154283e-01
-1.07188618e+00 8.58164072e-01 7.56127954e-01 -4.36364830e-01
3.71295303e-01 -8.13398719e-01 -2.47401651e-02 -3.95503528e-02
6.10405244e-02 4.24395978e-01 2.04529271e-01 -6.88417494e-01
-5.68412423e-01 1.04673910e+00 -5.01280546e-01 7.25887641e-02
9.33355689e-01 -4.36211884e-01 2.90149897e-01 3.12278748e-01
7.62528658e-01 -2.12053835e-01 -1.14588583e+00 -5.09609804e-02
3.00988168e-01 -8.40609595e-02 3.56368542e-01 -1.40243268e+00
-1.02801704e+00 7.76457071e-01 7.48671830e-01 9.96780396e-02
8.85660648e-01 3.51045996e-01 8.86324048e-01 4.73838300e-01
-7.42365718e-02 -1.33075058e+00 4.18998957e-01 4.16078836e-01
5.31261206e-01 -1.64868832e+00 1.66648865e-01 -9.13093150e-01
-7.07602561e-01 1.08322418e+00 7.27665663e-01 -1.45842418e-01
2.80759126e-01 4.26094115e-01 4.96503673e-02 -5.48040748e-01
-5.18532932e-01 -7.19119608e-01 6.13637030e-01 2.74562776e-01
4.68560338e-01 3.63720767e-02 -9.26592723e-02 3.23201209e-01
2.07087874e-01 -4.16228503e-01 5.07651567e-01 8.52751017e-01
-8.73783827e-01 -3.85245055e-01 -5.24406672e-01 8.06516230e-01
-9.21609879e-01 -4.66306269e-01 -7.57891655e-01 1.24275076e+00
9.06368345e-02 1.01947784e+00 1.37877658e-01 1.26867950e-01
3.39710027e-01 -2.49808002e-02 2.43920669e-01 -7.86490977e-01
-8.80908370e-01 7.24002898e-01 1.11960404e-01 -4.11494166e-01
-4.24471855e-01 -8.23487461e-01 -1.56133592e+00 1.30568460e-01
-8.36041093e-01 -5.75971186e-01 7.03864098e-01 1.08744621e+00
3.05331379e-01 8.00544441e-01 5.78642115e-02 -9.61325526e-01
-3.79713923e-02 -1.10629046e+00 -5.46081722e-01 8.81270543e-02
1.70969486e-01 -2.84812450e-01 -1.30171359e-01 2.84746259e-01] | [12.060420036315918, 2.2882165908813477] |
22ca1608-0f9b-4b1a-b09e-be572a376ebb | reinforcement-learning-based-minimum-state | 2304.04950 | null | https://arxiv.org/abs/2304.04950v1 | https://arxiv.org/pdf/2304.04950v1.pdf | Reinforcement Learning Based Minimum State-flipped Control for the Reachability of Boolean Control Networks | To realize reachability as well as reduce control costs of Boolean Control Networks (BCNs) with state-flipped control, a reinforcement learning based method is proposed to obtain flip kernels and the optimal policy with minimal flipping actions to realize reachability. The method proposed is model-free and of low computational complexity. In particular, Q-learning (QL), fast QL, and small memory QL are proposed to find flip kernels. Fast QL and small memory QL are two novel algorithms. Specifically, fast QL, namely, QL combined with transfer-learning and special initial states, is of higher efficiency, and small memory QL is applicable to large-scale systems. Meanwhile, we present a novel reward setting, under which the optimal policy with minimal flipping actions to realize reachability is the one of the highest returns. Then, to obtain the optimal policy, we propose QL, and fast small memory QL for large-scale systems. Specifically, on the basis of the small memory QL mentioned before, the fast small memory QL uses a changeable reward setting to speed up the learning efficiency while ensuring the optimality of the policy. For parameter settings, we give some system properties for reference. Finally, two examples, which are a small-scale system and a large-scale one, are considered to verify the proposed method. | ['Fangfei Li', 'Jingjie Ni'] | 2023-04-11 | null | null | null | null | ['q-learning'] | ['methodology'] | [-1.09227166e-01 -4.76601571e-02 -5.83322167e-01 1.60658136e-01
-6.22525871e-01 -3.58117223e-01 1.37360618e-01 -1.41735762e-01
-3.78264189e-01 1.21880984e+00 -2.99913764e-01 -4.63593513e-01
-6.10405147e-01 -9.81303513e-01 -7.93859541e-01 -1.06449354e+00
-2.39064209e-02 8.78932178e-02 4.38537329e-01 -2.84322888e-01
3.51720303e-01 4.24198210e-01 -1.07827556e+00 -2.21232772e-01
9.27032053e-01 1.11275613e+00 4.34122652e-01 3.05745482e-01
2.92729467e-01 6.96341038e-01 -2.88046688e-01 3.53066236e-01
3.19817692e-01 -5.31121910e-01 -6.94112897e-01 -5.86128794e-02
-1.02951497e-01 -3.99197578e-01 -4.22406852e-01 9.77914333e-01
5.03007472e-01 2.64791697e-01 3.04894477e-01 -1.27078354e+00
-4.29350942e-01 5.55834770e-01 -3.84681553e-01 1.69414535e-01
-7.43383318e-02 5.80957830e-01 9.74510252e-01 -3.81531119e-01
3.90805423e-01 1.43576252e+00 3.00751328e-01 6.83283389e-01
-1.12853563e+00 -8.33108068e-01 4.21533525e-01 3.59440446e-01
-1.25719535e+00 -2.04472721e-01 5.85754156e-01 -8.60015079e-02
7.76824117e-01 2.64096130e-02 9.92127955e-01 6.92322850e-01
6.97931349e-01 6.62142873e-01 1.32020974e+00 -3.08602512e-01
6.16858006e-01 -7.78196752e-02 -1.51976347e-01 9.20233071e-01
2.96604276e-01 4.21667397e-01 4.24592122e-02 -6.94869757e-02
1.16429830e+00 2.20850348e-01 -1.56326503e-01 -3.75433445e-01
-1.26339984e+00 6.47763908e-01 5.94399452e-01 -2.48330981e-02
-2.96869069e-01 5.38942397e-01 2.96617746e-01 5.14487982e-01
-1.55398414e-01 3.73278528e-01 -4.45498228e-01 -6.66294098e-02
-3.49835455e-01 1.21625982e-01 5.64138651e-01 9.03285623e-01
1.05922639e+00 4.29919332e-01 -5.23277462e-01 3.79392505e-01
1.89808026e-01 8.55125725e-01 2.92538613e-01 -1.08182847e+00
4.98030812e-01 6.20423436e-01 2.74700314e-01 -7.63128698e-01
-3.26592833e-01 -1.87793702e-01 -9.70279813e-01 5.92649393e-02
5.65159917e-02 -4.32937592e-01 -8.11887920e-01 1.93894386e+00
2.92731225e-01 4.39985543e-01 2.08882064e-01 7.08560467e-01
1.04214907e-01 1.11741841e+00 -3.09819788e-01 -7.13653326e-01
9.95424807e-01 -8.41655076e-01 -6.72378957e-01 -8.12752694e-02
5.05528688e-01 -1.12626001e-01 1.24784839e+00 1.54401451e-01
-9.12234724e-01 -2.32265547e-01 -1.01094270e+00 7.41997659e-01
-7.39822686e-02 1.46514341e-01 5.10147154e-01 3.76802206e-01
-9.04918134e-01 7.20840573e-01 -8.48804176e-01 -2.34774232e-01
2.23043606e-01 6.29009187e-01 5.47209531e-02 3.33590806e-02
-1.48593438e+00 9.82268691e-01 5.91258705e-01 3.18227798e-01
-1.37727296e+00 -4.43737894e-01 -5.73167384e-01 1.20499611e-01
9.33086812e-01 -4.45948303e-01 1.20626700e+00 -7.45480120e-01
-2.00173426e+00 -1.39988169e-01 5.10855354e-02 -3.38478297e-01
1.54391155e-01 1.41012073e-01 -2.15865970e-01 2.50402808e-01
-5.86274602e-02 5.20073712e-01 1.01714778e+00 -8.31738591e-01
-7.23051667e-01 -4.06711847e-02 2.39287138e-01 1.80163518e-01
-6.88197255e-01 -3.64259332e-01 -3.90301049e-02 3.21828164e-02
-2.35304624e-01 -8.67580950e-01 -4.96853739e-01 -2.05152154e-01
-2.24076107e-01 -3.01999629e-01 7.27401197e-01 -2.36794010e-01
1.27361405e+00 -1.94248831e+00 2.79262751e-01 4.33910400e-01
1.42035848e-02 3.84694368e-01 -2.63779223e-01 4.89060909e-01
4.08824325e-01 -3.51901017e-02 -4.59940545e-02 5.74122310e-01
-5.79071753e-02 3.29734117e-01 -2.75241762e-01 2.31177270e-01
4.60145056e-01 9.22501385e-01 -1.00254166e+00 -4.95233059e-01
1.91310689e-01 -7.26780444e-02 -6.10516310e-01 1.46271229e-01
-2.91197449e-01 3.46524149e-01 -8.36129785e-01 4.26559240e-01
4.02644396e-01 -2.81537831e-01 3.33252370e-01 -2.83462644e-01
-3.28583598e-01 -1.86058402e-01 -1.35471177e+00 8.47614408e-01
-6.01629913e-01 3.02107371e-02 3.05760503e-01 -1.17934740e+00
1.15978491e+00 1.72294304e-01 3.28823090e-01 -9.10975575e-01
3.71256441e-01 3.63394827e-01 2.81448048e-02 -3.75215858e-01
1.75592884e-01 -2.46312067e-01 -1.96638122e-01 2.89496839e-01
-1.41422763e-01 -3.04833204e-01 2.74342716e-01 3.32504474e-02
1.22673297e+00 -1.69433564e-01 4.34319943e-01 -4.58916247e-01
9.05090451e-01 -2.82278836e-01 9.52118158e-01 4.72075850e-01
-5.16136646e-01 -3.51552844e-01 9.27132905e-01 -1.46063939e-01
-7.49439120e-01 -8.34534049e-01 3.53432596e-01 6.55190468e-01
6.33545697e-01 -1.50672749e-01 -5.38363039e-01 -4.88803536e-01
1.21203050e-01 3.68275642e-01 -4.59319949e-01 -7.88855135e-01
-7.03152835e-01 -3.94020259e-01 1.97208151e-01 4.21897978e-01
7.29300022e-01 -1.29277241e+00 -5.50919890e-01 1.80053428e-01
1.57546639e-01 -5.99697590e-01 -6.51333272e-01 1.91093415e-01
-9.01401877e-01 -1.07098544e+00 -5.91158271e-01 -1.02661693e+00
8.29805076e-01 1.25158697e-01 3.09539914e-01 -6.20951764e-02
6.04249537e-02 2.44786367e-01 4.45584673e-03 1.56878993e-01
-1.87670007e-01 1.12189569e-01 2.74767935e-01 -1.30907193e-01
-3.54160726e-01 -3.10144037e-01 -4.78713006e-01 4.89354014e-01
-9.33568418e-01 -4.15672474e-02 1.11922181e+00 1.22495782e+00
7.92010546e-01 7.72933364e-02 1.06317854e+00 -4.70198393e-01
9.17582512e-01 -2.45299652e-01 -1.10035706e+00 3.92445624e-01
-8.69406700e-01 3.34036767e-01 1.29023302e+00 -6.94668829e-01
-8.29232752e-01 -7.86683112e-02 1.58328027e-01 -5.90519428e-01
3.64966542e-01 2.50780165e-01 -2.77049243e-01 -2.25372404e-01
2.76642829e-01 3.86817038e-01 3.26802820e-01 2.68746447e-02
3.75852078e-01 5.12484729e-01 -8.99716467e-02 -8.14249516e-01
5.69608927e-01 -1.89790707e-02 3.22027862e-01 -3.98034662e-01
-6.26984894e-01 8.49293917e-02 -2.78183848e-01 -2.66804904e-01
3.55348170e-01 -6.12753808e-01 -1.32814097e+00 4.09970939e-01
-6.97687089e-01 -6.39392197e-01 -3.39962333e-01 3.55061978e-01
-9.83288348e-01 -1.67674071e-03 -8.14070225e-01 -9.45075154e-01
-4.21296507e-01 -1.22340071e+00 5.22914410e-01 5.62626719e-01
5.29642820e-01 -7.56570458e-01 1.30064085e-01 -2.99731314e-01
6.25575364e-01 5.47627173e-02 1.11140239e+00 -2.44571287e-02
-7.74840176e-01 1.48110420e-01 -2.49927357e-01 2.83850133e-01
2.63045758e-01 2.29831710e-02 -7.90967271e-02 -8.38369489e-01
-3.56061876e-01 -4.75468427e-01 6.02814555e-01 4.10103351e-01
9.52727437e-01 -5.78230500e-01 -5.09502590e-01 1.24842785e-01
1.56026006e+00 6.95443451e-01 5.28025329e-01 3.53776872e-01
4.99536037e-01 1.19975343e-01 1.04826236e+00 5.44231832e-01
1.93548694e-01 3.40926498e-01 6.21469736e-01 1.80081442e-01
2.95396239e-01 -4.66795027e-01 7.49511302e-01 1.07161331e+00
2.44082257e-01 1.55125812e-01 -5.59602916e-01 4.71541613e-01
-2.07713723e+00 -8.16955388e-01 5.88498175e-01 2.42492867e+00
9.33258772e-01 1.97579786e-01 1.11029923e-01 -1.32371932e-02
9.53923821e-01 1.26436865e-03 -1.10801697e+00 -3.37786436e-01
2.47306183e-01 8.59708861e-02 4.40113872e-01 5.54198384e-01
-6.87669277e-01 1.09736907e+00 6.29700804e+00 1.24729979e+00
-1.46771514e+00 -8.72220397e-02 4.70536560e-01 6.77814558e-02
-1.06934533e-01 2.75890440e-01 -1.00275409e+00 5.73634803e-01
9.77677822e-01 -4.68379110e-01 7.83648849e-01 8.49292994e-01
5.21535814e-01 1.20530883e-03 -7.53704369e-01 6.49271071e-01
-5.65308511e-01 -1.22146368e+00 8.06025136e-03 -1.72401458e-01
6.67056262e-01 -4.77437288e-01 -7.17969984e-02 6.89119399e-01
3.63826424e-01 -6.35398149e-01 5.35807788e-01 5.43456912e-01
7.23723233e-01 -1.12193692e+00 6.54582977e-01 5.51398873e-01
-1.41533422e+00 -7.37287462e-01 -5.63967824e-01 -1.20182775e-01
4.50497381e-02 4.71699804e-01 -3.80357683e-01 4.29163992e-01
3.01900983e-01 8.96118045e-01 -2.70560265e-01 9.44597960e-01
-3.56590599e-01 5.78239739e-01 -2.27314815e-01 -8.88258219e-01
3.44984919e-01 -2.37333924e-01 3.24873984e-01 6.02617621e-01
3.63870770e-01 -3.16836946e-02 5.95723093e-01 8.32327366e-01
1.34107620e-01 4.25348692e-02 -4.66425508e-01 -8.68240893e-02
7.48406649e-01 1.47385728e+00 -4.67319310e-01 -2.22411439e-01
3.56626101e-02 3.93465251e-01 5.50213158e-01 2.63829619e-01
-9.26591277e-01 -8.02385211e-01 3.52011442e-01 -5.16619198e-02
3.17813993e-01 -2.09409788e-01 3.01901996e-01 -7.84847260e-01
-2.57292777e-01 -7.35539436e-01 1.91031277e-01 -5.25964916e-01
-1.01678360e+00 2.79682070e-01 -1.91262472e-04 -1.37137508e+00
-1.70177355e-01 -4.69939560e-01 -6.86079919e-01 4.60341364e-01
-1.57039511e+00 -6.59124434e-01 -4.92503904e-02 8.05260777e-01
1.68077573e-01 -1.63077608e-01 4.28972274e-01 2.06396341e-01
-1.00481009e+00 4.29216623e-01 4.19885248e-01 -3.41837466e-01
5.35662770e-01 -9.48099315e-01 -4.28598613e-01 5.43172956e-01
-6.58441365e-01 4.63862240e-01 3.51518542e-01 -5.56294978e-01
-1.91173613e+00 -1.24166286e+00 8.68570507e-02 4.61191267e-01
9.20484781e-01 -3.83802988e-02 -6.27433956e-01 4.11140054e-01
-2.95928270e-02 9.51287299e-02 -2.06814378e-01 -4.99624729e-01
1.74203411e-01 -8.64640713e-01 -1.18404770e+00 7.92592764e-01
7.07460105e-01 -1.25347212e-01 -2.13916108e-01 2.54465789e-01
9.91919339e-01 -2.49087542e-01 -8.54897499e-01 3.70003641e-01
3.62545282e-01 -4.11368608e-01 6.11838222e-01 -4.82980311e-01
2.36143634e-01 -5.13889015e-01 1.33410588e-01 -1.45901418e+00
-5.83196104e-01 -7.91030526e-01 -4.02422577e-01 9.51711118e-01
2.84430146e-01 -1.05398214e+00 5.69066763e-01 9.58362073e-02
3.21981497e-02 -1.13348603e+00 -1.03139365e+00 -1.13580465e+00
1.79221630e-01 3.04096282e-01 4.92295772e-01 3.96696568e-01
2.62849391e-01 4.49776590e-01 -5.42027175e-01 2.23563071e-02
3.26544017e-01 4.37119573e-01 5.09851754e-01 -7.37818062e-01
-1.43062204e-01 -2.87881941e-01 -4.85903174e-02 -1.19094777e+00
2.81000584e-01 -6.15774274e-01 3.60181928e-01 -1.43778145e+00
3.04151177e-01 -5.04990220e-01 -6.64576530e-01 7.81050563e-01
-1.51122302e-01 -4.43198621e-01 2.06333980e-01 3.63120526e-01
-8.73659551e-01 9.80638981e-01 1.76082289e+00 -1.95343718e-01
-3.65901828e-01 2.53072083e-01 -5.12871444e-01 1.68447644e-01
1.04564309e+00 -3.49360764e-01 -5.86435020e-01 1.28337041e-01
1.18956799e-02 6.51956975e-01 2.09718958e-01 -8.34453464e-01
1.69041485e-01 -7.81282485e-01 -7.02734068e-02 -3.51939142e-01
1.98635653e-01 -8.01345766e-01 -2.06666946e-01 1.34904945e+00
-3.64691079e-01 1.35037042e-02 1.08348496e-01 7.88658381e-01
-5.20745479e-02 -2.07180932e-01 9.44337785e-01 -1.85049716e-02
-7.03065276e-01 6.05723023e-01 -5.39049864e-01 1.96431670e-02
1.41931581e+00 1.83227077e-01 -4.40951198e-01 -3.96524608e-01
-5.02013445e-01 8.73465478e-01 -1.07101854e-02 3.08167756e-01
6.70424581e-01 -1.57950294e+00 -1.41028464e-01 1.21042460e-01
-3.36273730e-01 -2.29260340e-01 1.76001400e-01 9.94749725e-01
-1.42846271e-01 7.36959755e-01 -4.69668299e-01 -4.29114103e-01
-8.81518781e-01 5.76104820e-01 4.73359704e-01 -5.96887767e-01
-2.45232582e-01 1.76083282e-01 -1.96370348e-01 -5.41211426e-01
-1.19089223e-01 -5.38394272e-01 -2.18615066e-02 -3.41458291e-01
2.73251623e-01 5.81069887e-01 -2.44257092e-01 1.21635832e-01
-3.59665185e-01 6.78273141e-01 8.94812346e-02 -9.61725116e-02
1.13307190e+00 -3.51851806e-02 -1.53070375e-01 1.97592214e-01
8.29138160e-01 -3.56968433e-01 -1.63121974e+00 -2.04421535e-01
-2.21740380e-01 -3.23853314e-01 4.20272201e-02 -4.33651835e-01
-1.36518657e+00 8.88639510e-01 5.20885348e-01 1.75638646e-01
1.23459733e+00 -4.26977187e-01 7.76540101e-01 8.08133304e-01
8.22563946e-01 -1.23673141e+00 4.02059585e-01 7.71801293e-01
5.53802013e-01 -7.85829246e-01 -4.74162884e-02 -1.38172328e-01
-5.40071130e-01 1.11895883e+00 1.29123557e+00 -3.69255960e-01
5.25835395e-01 2.34599039e-01 -6.22655928e-01 2.84034878e-01
-1.18881094e+00 -1.91565976e-01 -2.74576753e-01 3.18648249e-01
-2.98320681e-01 1.14484623e-01 -6.38304055e-01 4.33550179e-01
4.25815314e-01 1.97921172e-01 4.46924448e-01 1.02526093e+00
-8.60956490e-01 -9.53936934e-01 -1.61441252e-01 5.65805018e-01
1.10478289e-01 2.52490431e-01 8.06682836e-03 6.22939229e-01
-1.88753515e-01 7.47962177e-01 -4.59591597e-02 -4.64263797e-01
3.84903312e-01 -4.34698611e-01 5.62240541e-01 -3.43128115e-01
-3.73297215e-01 -9.99860615e-02 -2.37925634e-01 -7.35696614e-01
-1.41722262e-01 -1.39122844e-01 -1.56670141e+00 -4.04628664e-01
-8.18337560e-01 9.59407240e-02 1.08139545e-01 1.08001828e+00
5.62265038e-01 6.01566195e-01 1.06366193e+00 -5.34238458e-01
-1.16082168e+00 -8.08716834e-01 -7.78945386e-01 -2.12336555e-01
1.30743772e-01 -7.59831607e-01 -2.84557223e-01 -5.79879582e-01] | [4.283729076385498, 2.032435655593872] |
b023e2f6-ddda-430c-9139-ad8d61bdc460 | graphon-based-clustering-and-testing-of | 2110.02722 | null | https://arxiv.org/abs/2110.02722v2 | https://arxiv.org/pdf/2110.02722v2.pdf | Graphon based Clustering and Testing of Networks: Algorithms and Theory | Network-valued data are encountered in a wide range of applications and pose challenges in learning due to their complex structure and absence of vertex correspondence. Typical examples of such problems include classification or grouping of protein structures and social networks. Various methods, ranging from graph kernels to graph neural networks, have been proposed that achieve some success in graph classification problems. However, most methods have limited theoretical justification, and their applicability beyond classification remains unexplored. In this work, we propose methods for clustering multiple graphs, without vertex correspondence, that are inspired by the recent literature on estimating graphons -- symmetric functions corresponding to infinite vertex limit of graphs. We propose a novel graph distance based on sorting-and-smoothing graphon estimators. Using the proposed graph distance, we present two clustering algorithms and show that they achieve state-of-the-art results. We prove the statistical consistency of both algorithms under Lipschitz assumptions on the graph degrees. We further study the applicability of the proposed distance for graph two-sample testing problems. | ['Debarghya Ghoshdastidar', 'Leena Chennuru Vankadara', 'Mahalakshmi Sabanayagam'] | 2021-10-06 | graphon-based-clustering-and-testing-of-1 | https://openreview.net/forum?id=sTNHCrIKDQc | https://openreview.net/pdf?id=sTNHCrIKDQc | iclr-2022-4 | ['hypothesis-testing', 'hypothesis-testing'] | ['methodology', 'miscellaneous'] | [ 2.03349724e-01 1.82073385e-01 -2.90836453e-01 -5.16812503e-01
-3.11245441e-01 -4.99952912e-01 3.91207099e-01 7.28342891e-01
-2.11664140e-01 7.63260067e-01 -4.90491688e-01 -4.03547525e-01
-7.83135772e-01 -8.49053085e-01 -8.09180081e-01 -8.28534722e-01
-7.18438864e-01 6.78615510e-01 4.17707473e-01 8.75891522e-02
2.47258365e-01 5.15039086e-01 -1.41411841e+00 -1.70709267e-01
9.97799397e-01 6.51167452e-01 -7.02151284e-02 6.28100932e-01
-5.30471243e-02 4.70992327e-01 -3.72562259e-01 -4.65786934e-01
2.29393840e-01 -4.00454909e-01 -6.47529542e-01 2.91719586e-01
4.64691818e-01 4.04839605e-01 -3.42141718e-01 1.46512556e+00
4.62240189e-01 2.27413863e-01 9.69138622e-01 -1.61019742e+00
-8.01005960e-01 5.85567296e-01 -6.71037853e-01 7.05353543e-02
1.63411096e-01 -4.56141114e-01 1.21950078e+00 -4.75426286e-01
3.92475784e-01 1.19206738e+00 8.61443996e-01 2.39335805e-01
-1.45327723e+00 -3.75635356e-01 2.93879751e-02 4.92975742e-01
-1.51992869e+00 7.49098063e-02 9.24548030e-01 -5.56350529e-01
5.81914365e-01 2.40285888e-01 4.74514753e-01 7.32664645e-01
3.18674445e-02 4.22519445e-01 1.24823236e+00 -6.16131127e-01
3.84164631e-01 3.91900679e-03 5.03923118e-01 1.06937957e+00
7.54558325e-01 -4.66686845e-01 -7.80394599e-02 -3.57944071e-01
5.71376204e-01 8.46942235e-03 -2.89415300e-01 -9.21753824e-01
-1.01070535e+00 1.14233720e+00 3.70056480e-01 3.32049161e-01
-6.59129843e-02 4.66741137e-02 2.29623079e-01 2.80974597e-01
7.00486720e-01 2.38511235e-01 -5.62677942e-02 4.08996850e-01
-6.34202480e-01 -8.45216215e-02 1.16021740e+00 1.07954109e+00
8.56029749e-01 -3.75962526e-01 1.57034338e-01 6.40636265e-01
2.28949934e-01 2.29058415e-01 3.60980704e-02 -4.68065709e-01
2.23673716e-01 6.97841227e-01 -4.69288021e-01 -1.35272491e+00
-7.09803760e-01 -3.24436039e-01 -1.28587699e+00 -4.68119048e-02
7.49187648e-01 1.69004843e-01 -6.54274583e-01 1.58206558e+00
4.51301426e-01 4.80566800e-01 -1.21476620e-01 6.01480365e-01
7.36078978e-01 1.78225070e-01 -2.46121183e-01 -4.14008409e-01
9.40608859e-01 -7.56693184e-01 -5.64002514e-01 1.71661526e-01
7.54599869e-01 -3.89093012e-01 8.82901430e-01 4.79503661e-01
-8.13481450e-01 -2.54820168e-01 -9.52014327e-01 1.44924432e-01
-4.27663356e-01 -1.70327976e-01 7.39646018e-01 9.16422784e-01
-1.14267480e+00 9.74447727e-01 -8.24653864e-01 -6.53078079e-01
2.87223071e-01 5.07995725e-01 -3.75191569e-01 -1.40236020e-01
-9.54334915e-01 5.63642263e-01 4.53283131e-01 1.14172868e-01
-3.37871671e-01 -2.72839397e-01 -8.42850864e-01 -3.03442273e-02
5.60183823e-01 -3.12179148e-01 6.41102850e-01 -5.79448402e-01
-1.05796099e+00 8.77944291e-01 1.49958566e-01 -4.79320556e-01
4.78965670e-01 4.09985751e-01 -3.27117056e-01 3.34171116e-01
-3.18545736e-02 -1.32103008e-03 7.31786907e-01 -8.53613973e-01
-1.91489831e-01 -6.37771368e-01 -6.80623064e-03 -4.12735231e-02
-5.56461573e-01 -3.30436856e-01 -2.59201378e-01 -4.42221135e-01
2.31126234e-01 -1.06526399e+00 -3.74586552e-01 -4.85157073e-02
-7.18343973e-01 -3.91808450e-01 6.59464180e-01 -2.21781909e-01
1.10003209e+00 -1.95339870e+00 1.63305134e-01 7.15738893e-01
6.49989605e-01 7.72143453e-02 -1.29829720e-01 6.77783608e-01
-1.51460320e-01 2.00753316e-01 -3.67362291e-01 -9.98309702e-02
7.43402466e-02 1.43165171e-01 2.25355089e-01 1.07087672e+00
8.19216818e-02 6.83020115e-01 -8.78707290e-01 -4.05816168e-01
2.14862660e-01 4.33870077e-01 -3.63326699e-01 1.06540494e-01
-7.21416101e-02 3.11908752e-01 -4.60065603e-01 4.28299040e-01
8.39883804e-01 -7.51589715e-01 6.68939471e-01 -1.81987554e-01
3.18402499e-01 -2.42762879e-01 -1.38593471e+00 1.24412465e+00
1.60108626e-01 4.39768046e-01 1.04511537e-01 -1.77955437e+00
9.32006896e-01 -7.93334246e-02 4.97106880e-01 -8.06374475e-02
1.39651388e-01 3.00794691e-02 2.94828385e-01 -3.76491547e-01
-3.59585509e-02 -5.30000553e-02 2.09569648e-01 3.26878309e-01
-6.03616461e-02 8.19804147e-02 4.58851874e-01 3.28005135e-01
1.40162241e+00 -4.14028436e-01 5.00120938e-01 -6.11205995e-01
6.68483913e-01 -3.52016151e-01 2.13656381e-01 6.88734472e-01
-7.84715116e-02 4.11984473e-01 9.94341850e-01 -1.53995097e-01
-7.96874404e-01 -1.03716230e+00 -1.37465641e-01 8.10475886e-01
2.57489771e-01 -3.84063035e-01 -9.14722085e-01 -7.21777678e-01
2.86398381e-01 1.53107479e-01 -6.76877618e-01 -2.82620788e-01
-3.07555676e-01 -9.20085430e-01 4.26033616e-01 3.26810718e-01
1.66765153e-01 -6.51235461e-01 2.67791506e-02 -5.39218783e-02
1.79316714e-01 -1.34982967e+00 -3.42426866e-01 1.69359997e-01
-9.80829835e-01 -1.63525510e+00 -5.48051596e-01 -1.10467458e+00
9.63301241e-01 4.55878943e-01 1.02013218e+00 3.12362850e-01
-3.25381666e-01 5.31771362e-01 -2.97414571e-01 3.06205805e-02
-4.85221714e-01 2.20567971e-01 1.89756244e-01 3.28844994e-01
3.28554094e-01 -6.89516246e-01 -2.89487898e-01 4.89217103e-01
-9.87932801e-01 -1.64081410e-01 6.07688963e-01 8.10789287e-01
6.71128690e-01 2.57299036e-01 5.93104839e-01 -1.38721287e+00
9.38342988e-01 -5.30692816e-01 -7.76515722e-01 5.33614814e-01
-8.84894669e-01 2.81551838e-01 9.16157126e-01 -4.95939463e-01
-3.13921779e-01 -6.02388680e-02 2.62108743e-01 -4.16429341e-01
-1.62422985e-01 7.62803435e-01 -2.50545412e-01 -6.65182710e-01
5.26471078e-01 -1.06334269e-01 3.59721392e-01 -2.25680903e-01
4.70082790e-01 3.58975857e-01 2.67815948e-01 -5.66729426e-01
7.56105959e-01 4.31385905e-01 5.92460573e-01 -1.23069823e+00
-5.43595433e-01 -7.06407726e-01 -6.13558412e-01 -1.72810644e-01
8.12730670e-01 -4.45075810e-01 -1.01292241e+00 3.04800212e-01
-8.69576871e-01 -1.63827330e-01 1.58179328e-01 7.13216484e-01
-6.09961629e-01 1.03269541e+00 -6.17009699e-01 -8.80058944e-01
3.87873985e-02 -9.69421208e-01 7.17399120e-01 -4.27896380e-02
3.30011845e-01 -1.40085471e+00 1.63208529e-01 9.30786431e-02
4.48535308e-02 4.58676696e-01 1.30162179e+00 -1.03334928e+00
-6.17478251e-01 -1.28899977e-01 -3.74481738e-01 3.05213541e-01
2.76965141e-01 -1.94083408e-01 -6.02288187e-01 -6.68103218e-01
-2.11038604e-01 -1.98471114e-01 8.60342026e-01 6.14694715e-01
1.27610040e+00 -1.61165819e-01 -5.37133396e-01 5.68599343e-01
1.46721745e+00 -1.24867380e-01 3.35351348e-01 -1.51116714e-01
9.06963229e-01 7.34365582e-01 3.13915431e-01 3.25855255e-01
1.21900156e-01 4.23195630e-01 4.33767706e-01 -1.77602336e-01
2.27969006e-01 7.50412866e-02 -8.40895995e-02 1.18025124e+00
-1.89089552e-01 -4.59232628e-01 -7.38207817e-01 3.75234932e-01
-2.10979700e+00 -7.37808049e-01 -6.09250188e-01 2.47430778e+00
3.89699548e-01 1.32864609e-01 3.48649740e-01 1.18194900e-01
1.40749216e+00 2.31112614e-02 -5.94836950e-01 -8.86435360e-02
-1.74977258e-01 2.53229260e-01 7.55245268e-01 5.15007973e-01
-1.02021611e+00 5.69809854e-01 6.31206989e+00 7.96207070e-01
-6.15655065e-01 -9.71885324e-02 5.72615206e-01 5.03346682e-01
-7.15834647e-02 9.05345976e-02 -4.92001742e-01 2.30322957e-01
8.44513416e-01 -1.76017642e-01 5.05006194e-01 8.40609789e-01
-1.28965870e-01 1.41645089e-01 -1.11143959e+00 1.03692436e+00
2.10504398e-01 -1.01542449e+00 -2.33044118e-01 3.12341481e-01
7.07033217e-01 -2.96029299e-01 -1.09108187e-01 7.46381134e-02
1.53740317e-01 -1.00808787e+00 -4.04379107e-02 4.38597500e-01
6.87927306e-01 -9.38830674e-01 6.25993133e-01 3.00848663e-01
-1.41285419e+00 1.71611175e-01 -5.60746431e-01 -1.87480778e-01
-1.01580396e-01 1.02904701e+00 -9.31854546e-01 6.65693521e-01
3.37957352e-01 7.57943630e-01 -8.14069569e-01 1.10770011e+00
-1.04618698e-01 7.12645173e-01 -2.07166463e-01 -4.73830849e-01
-7.35711530e-02 -7.73442388e-01 4.35224742e-01 9.56467628e-01
1.66002214e-01 -1.35098621e-01 4.49205995e-01 6.75461769e-01
-2.48288140e-01 3.73160988e-01 -8.70212734e-01 -3.76373023e-01
2.97914952e-01 1.42317820e+00 -1.33488297e+00 -1.46469414e-01
-6.49474502e-01 7.65495062e-01 8.06598186e-01 3.45091313e-01
-8.10717225e-01 -6.65688694e-01 4.51718360e-01 1.74293369e-01
3.50068361e-01 -4.28532451e-01 2.52035648e-01 -1.15791333e+00
1.89429522e-01 -7.54956067e-01 5.12543023e-01 -7.66369626e-02
-1.83111989e+00 3.52644742e-01 1.54117465e-01 -9.96622384e-01
3.67146917e-03 -9.72296059e-01 -5.25134444e-01 3.36028159e-01
-1.19813514e+00 -8.71534586e-01 -3.52505893e-01 6.87912941e-01
-1.41840845e-01 -1.48476720e-01 6.26920462e-01 3.17210138e-01
-2.71458358e-01 4.97996002e-01 4.65730995e-01 1.90507308e-01
7.14226723e-01 -1.58618104e+00 3.75627548e-01 6.31022573e-01
3.28101963e-01 5.67780852e-01 5.33840299e-01 -8.06630075e-01
-1.68294370e+00 -1.07037306e+00 5.69593132e-01 -1.25283167e-01
9.22134101e-01 -6.51467502e-01 -1.10976326e+00 5.87277830e-01
-1.50027424e-01 3.86068553e-01 6.87783599e-01 3.56089234e-01
-2.88624525e-01 -4.80964556e-02 -1.19504333e+00 3.93875957e-01
1.27281582e+00 -4.31409538e-01 -1.42634869e-01 7.24540770e-01
4.41298425e-01 5.05453721e-02 -1.01454639e+00 4.29821432e-01
3.50579679e-01 -1.13061988e+00 8.55981886e-01 -8.16644847e-01
-1.80185437e-01 -2.90353596e-01 -1.14068147e-02 -1.30633724e+00
-4.24802572e-01 -6.62085176e-01 -8.28415081e-02 9.50755060e-01
1.75665468e-01 -9.95189607e-01 9.97900605e-01 1.61756739e-01
7.84873888e-02 -7.10761845e-01 -8.67363632e-01 -1.06161118e+00
-1.18540652e-01 -1.10394306e-01 1.81421578e-01 1.30136609e+00
1.19777456e-01 5.73477089e-01 -2.38403827e-01 2.43661672e-01
1.10880828e+00 1.19614385e-01 6.69639051e-01 -1.74011171e+00
-6.05392158e-01 -5.92392385e-01 -1.12525177e+00 -7.60901213e-01
5.23457050e-01 -1.30198336e+00 -1.29414603e-01 -1.50289130e+00
2.85689563e-01 -3.43949407e-01 -3.44417430e-02 1.57121401e-02
-1.59442782e-01 1.10941030e-01 -2.04702750e-01 -1.51762873e-01
-7.26409733e-01 3.03857684e-01 9.76542234e-01 -1.42111540e-01
-6.23529311e-03 1.87645763e-01 -3.93009990e-01 7.11511433e-01
8.07336986e-01 -3.73691887e-01 -5.73215067e-01 1.61748946e-01
2.88820744e-01 -2.60212943e-02 3.20325643e-01 -1.01317906e+00
4.04772669e-01 6.24268204e-02 6.15209714e-02 -3.83186907e-01
5.25201857e-02 -7.08486676e-01 1.21881038e-01 5.59769750e-01
-3.68944466e-01 2.36326501e-01 -3.24966967e-01 1.16810060e+00
-4.96095531e-02 -3.05223912e-01 8.72058511e-01 9.70521290e-03
-3.16729903e-01 5.28282523e-01 -7.57247061e-02 3.80273193e-01
1.08456349e+00 -1.18638888e-01 -3.47076714e-01 -5.08624732e-01
-8.00814211e-01 1.82624366e-02 3.88037980e-01 1.95745789e-02
6.30829990e-01 -1.28394401e+00 -6.15147948e-01 9.77924615e-02
1.24032862e-01 -3.32990587e-01 -6.65116459e-02 1.17367101e+00
-4.59593505e-01 1.73896238e-01 3.13921422e-02 -7.95400620e-01
-1.38986444e+00 8.77178669e-01 2.35140440e-03 -1.83540687e-01
-3.90640646e-01 4.74560648e-01 2.19172448e-01 -5.14145553e-01
2.98414588e-01 -4.81650591e-01 -8.13145842e-03 -1.24313526e-01
1.33010864e-01 5.56766510e-01 2.13674620e-01 -4.77458835e-01
-3.89616907e-01 6.74123049e-01 4.45857085e-02 3.60216498e-01
1.16380215e+00 1.35637060e-01 -4.47828770e-01 6.23352528e-01
1.36223304e+00 2.25506630e-02 -7.77777016e-01 -2.58618236e-01
3.48392397e-01 -3.49549711e-01 -4.60306257e-01 7.63689540e-03
-9.20241117e-01 8.76352668e-01 3.91868889e-01 1.04763448e+00
9.11195934e-01 3.13514024e-01 2.61955082e-01 6.39014006e-01
3.88258994e-01 -9.06960964e-01 9.64666605e-02 3.77283692e-01
3.32931697e-01 -1.31868207e+00 1.69335768e-01 -8.87827635e-01
-1.77027613e-01 1.29909086e+00 3.95516187e-01 -3.52728695e-01
8.80685985e-01 -1.17566809e-01 -4.60893512e-01 -2.43812546e-01
-3.86871576e-01 -5.03654599e-01 6.20909572e-01 8.88132632e-01
5.76297998e-01 2.48002931e-01 -5.98745167e-01 1.94149092e-01
1.46603048e-01 -5.25616169e-01 5.24653077e-01 5.38926780e-01
-4.87548679e-01 -1.11988807e+00 -1.02958687e-01 8.22378635e-01
-2.86795020e-01 -4.00175229e-02 -6.87817931e-01 9.32734132e-01
-3.53592157e-01 9.44473445e-01 -1.97666958e-01 -3.13646048e-01
3.29723805e-02 -2.49252334e-01 7.98251331e-01 -4.98379469e-01
-1.59063358e-02 -7.02299848e-02 9.59742963e-02 -2.65839577e-01
-7.49506891e-01 -4.55597758e-01 -1.13608027e+00 -3.75697076e-01
-6.90637648e-01 5.12451828e-01 5.24002850e-01 8.88478577e-01
2.94004709e-01 2.14244619e-01 5.90476930e-01 -5.73064983e-01
-6.06270134e-01 -8.97294283e-01 -1.13596892e+00 5.43000221e-01
8.59384462e-02 -7.91536987e-01 -6.39478326e-01 -3.99304539e-01] | [6.964508533477783, 5.4276885986328125] |
3c2de03e-2cdd-45bf-93e3-a7d50d6140ed | exploring-the-role-of-the-bottleneck-in-slot | 2306.02577 | null | https://arxiv.org/abs/2306.02577v1 | https://arxiv.org/pdf/2306.02577v1.pdf | Exploring the Role of the Bottleneck in Slot-Based Models Through Covariance Regularization | In this project we attempt to make slot-based models with an image reconstruction objective competitive with those that use a feature reconstruction objective on real world datasets. We propose a loss-based approach to constricting the bottleneck of slot-based models, allowing larger-capacity encoder networks to be used with Slot Attention without producing degenerate stripe-shaped masks. We find that our proposed method offers an improvement over the baseline Slot Attention model but does not reach the performance of \dinosaur on the COCO2017 dataset. Throughout this project, we confirm the superiority of a feature reconstruction objective over an image reconstruction objective and explore the role of the architectural bottleneck in slot-based models. | ['Kousik Rajesh', 'Abishek Sridhar', 'Robert Lo', 'Andrew Stange'] | 2023-06-05 | null | null | null | null | ['image-reconstruction'] | ['computer-vision'] | [ 1.85184330e-01 7.18189418e-01 -3.32538456e-01 -2.81926632e-01
-7.44993389e-01 8.68369732e-03 8.54618371e-01 -5.68371892e-01
-5.20564079e-01 5.87575436e-01 4.24037635e-01 -6.93322480e-01
-8.29617605e-02 -6.15569472e-01 -9.39927280e-01 -3.37983191e-01
3.08062255e-01 5.40735066e-01 2.55693048e-01 -2.00760946e-01
5.37566125e-01 1.21151306e-01 -1.40310752e+00 6.54841244e-01
3.50683600e-01 1.05529809e+00 6.31940842e-01 3.96274090e-01
-4.22962725e-01 8.77354026e-01 -3.57059747e-01 -4.43332821e-01
5.39256632e-01 -3.92626852e-01 -1.10244977e+00 4.00224000e-01
8.11210930e-01 -3.32341373e-01 -5.33934772e-01 6.14486396e-01
4.47474182e-01 1.72618758e-02 2.56204367e-01 -1.22198641e+00
-6.94148242e-01 6.33115411e-01 -3.98154706e-01 3.19468409e-01
-2.18660042e-01 2.12384835e-02 1.35558295e+00 -1.02039874e+00
9.73186076e-01 1.35540688e+00 7.54515350e-01 4.84099895e-01
-1.43494892e+00 -2.88246483e-01 1.84236288e-01 2.58428633e-01
-1.45612168e+00 -7.84755707e-01 5.25778353e-01 -1.83126256e-02
1.64073241e+00 2.39328310e-01 4.34417695e-01 8.80795956e-01
2.63915896e-01 1.16078556e+00 8.75127733e-01 -4.80316162e-01
-4.33417633e-02 2.84087062e-01 -4.58066076e-01 8.10851157e-01
-1.92457009e-02 6.03410192e-02 -7.88007736e-01 2.18047291e-01
9.26992536e-01 -1.61356688e-01 -9.75841358e-02 -4.09610331e-01
-1.10594547e+00 8.98812711e-01 8.06253493e-01 3.50961328e-01
-2.17162207e-01 6.99537933e-01 5.16086102e-01 3.51770610e-01
7.35934973e-01 8.50040734e-01 -6.16836786e-01 -4.64018583e-02
-1.42937970e+00 1.37058824e-01 2.73603290e-01 1.06636775e+00
7.91239142e-01 2.54688561e-01 -3.50454301e-01 9.43398356e-01
3.93786073e-01 -5.24694622e-02 6.19380116e-01 -1.28426027e+00
4.42417383e-01 2.03522786e-01 1.04725495e-01 -3.45964640e-01
-4.43991959e-01 -5.89012861e-01 -2.10243478e-01 2.07435131e-01
1.27857938e-01 3.30816269e-01 -1.15230167e+00 1.67617178e+00
-1.37373179e-01 2.82706730e-02 -9.89097580e-02 9.88683999e-01
5.92523515e-01 6.58634424e-01 1.32783651e-01 2.53214747e-01
1.21339476e+00 -1.58251345e+00 -5.70872009e-01 -3.09867650e-01
8.52254987e-01 -8.64326417e-01 1.31006169e+00 1.67847976e-01
-1.24479842e+00 -6.60642982e-01 -1.19740283e+00 -4.82604086e-01
-5.29071614e-02 1.10060669e-01 9.25337732e-01 6.72838151e-01
-1.51686573e+00 6.00958407e-01 -6.13333166e-01 -4.62208211e-01
5.19578278e-01 3.21544707e-01 -3.89316410e-01 4.16335873e-02
-6.30532503e-01 1.12041366e+00 2.68113941e-01 -1.47806168e-01
-1.02215505e+00 -8.59786928e-01 -1.02626479e+00 1.24675475e-01
2.56928742e-01 -7.28136718e-01 1.26844776e+00 -1.24378562e+00
-1.29266608e+00 1.01046562e+00 -2.43053302e-01 -8.55875552e-01
3.79319638e-01 -3.24873984e-01 2.17208918e-02 1.25218600e-01
2.75067508e-01 1.27900100e+00 6.32452846e-01 -1.11465955e+00
-4.78167504e-01 8.97845402e-02 3.95315289e-01 2.62961805e-01
-1.06240332e-01 -3.43471974e-01 -5.43503344e-01 -6.23929739e-01
2.08203018e-01 -1.02215707e+00 -3.22893977e-01 5.53085029e-01
-3.86897773e-01 -1.68637291e-01 8.73779833e-01 -4.58342671e-01
8.25500607e-01 -2.47551370e+00 3.52766775e-02 -3.14902753e-01
1.70878634e-01 1.44037098e-01 -7.28597820e-01 4.38544184e-01
-1.75792783e-01 5.66111088e-01 -1.40032962e-01 -8.89978886e-01
1.04723535e-01 4.92518902e-01 -3.72895032e-01 5.03559649e-01
3.99553001e-01 1.03887510e+00 -2.29840666e-01 -5.40862620e-01
1.90879121e-01 3.46032411e-01 -1.01004326e+00 1.20916046e-01
-4.69852805e-01 -1.02008246e-01 -6.88822567e-02 5.10925710e-01
5.71880043e-01 -3.95848125e-01 1.90832525e-01 -8.35859105e-02
-3.74504626e-01 6.79912865e-01 -5.86860716e-01 2.19917154e+00
-8.46159101e-01 1.14283800e+00 2.46680126e-01 -1.16187465e+00
6.47957802e-01 3.48921090e-01 4.43408817e-01 -1.26984334e+00
-5.79047613e-02 2.73203164e-01 1.04896709e-01 -3.85837816e-02
9.52326000e-01 -3.73998642e-01 1.46848783e-01 4.29798186e-01
4.18768734e-01 -2.38245502e-01 5.20563722e-02 1.61807731e-01
9.20332551e-01 5.64531446e-01 -3.43166113e-01 -7.55542219e-01
1.03912398e-01 2.27574538e-02 4.76642966e-01 8.92030239e-01
-1.74710661e-01 1.14683974e+00 6.29167438e-01 -7.49714494e-01
-1.81853843e+00 -1.07008922e+00 -4.66643691e-01 1.11568081e+00
1.81751311e-01 -6.64626300e-01 -5.31237423e-01 -8.10199797e-01
-1.23111866e-01 7.20425308e-01 -8.49701345e-01 -2.77063828e-02
-4.00034368e-01 -5.26177704e-01 6.85755908e-01 4.96599197e-01
7.24167347e-01 -8.21806908e-01 -6.89066768e-01 2.12046444e-01
-2.70338297e-01 -8.68781030e-01 -4.69798952e-01 4.15190935e-01
-9.10579562e-01 -9.22585785e-01 -8.09850574e-01 -6.46550238e-01
3.77765417e-01 2.87513793e-01 1.45933282e+00 3.39681357e-01
-3.74749929e-01 2.50599980e-01 -3.01894724e-01 -2.22673461e-01
1.85975045e-01 2.91920006e-01 -3.03158641e-01 -3.19749862e-01
4.31984700e-02 -2.42485911e-01 -6.74866617e-01 4.59428996e-01
-1.23611856e+00 3.90966386e-01 5.04695833e-01 1.15109277e+00
3.78521770e-01 -2.46527404e-01 3.97522181e-01 -8.16205382e-01
1.60392776e-01 -5.01916170e-01 -4.30418402e-01 5.04989214e-02
-9.07517850e-01 5.15627682e-01 1.95148274e-01 -5.74592426e-02
-1.03451252e+00 9.42059699e-03 -4.80727136e-01 -4.36154217e-01
3.88799012e-01 2.69590586e-01 7.70077854e-02 1.24073863e-01
5.06066680e-01 2.28382304e-01 -1.70188900e-02 -5.62113941e-01
3.26640278e-01 3.35961968e-01 1.24267176e-01 -4.89797264e-01
3.99241507e-01 6.60836577e-01 -2.02614695e-01 -6.94742978e-01
-8.82602990e-01 -3.57565671e-01 -2.48290509e-01 2.66622066e-01
7.28753388e-01 -1.10293829e+00 -6.26048073e-02 1.06150836e-01
-1.20962572e+00 -6.20408893e-01 -6.34809494e-01 2.41204873e-01
-9.98643517e-01 -4.55893390e-03 -6.44004881e-01 -5.62793732e-01
-9.55382138e-02 -1.15438092e+00 1.26051915e+00 1.47703616e-02
-1.54859141e-01 -8.48739684e-01 -9.41736177e-02 4.59036231e-01
9.56615031e-01 -4.03357536e-01 6.46431088e-01 -2.79283345e-01
-9.08752859e-01 4.48891640e-01 -6.77413404e-01 -4.13294248e-02
-3.48319292e-01 -3.97532493e-01 -1.00564504e+00 -1.84134454e-01
-1.82209432e-01 -5.10919392e-01 1.26860130e+00 4.27357882e-01
1.28565884e+00 -1.38044551e-01 -1.38353199e-01 8.95767987e-01
1.69383836e+00 -8.65196437e-02 1.12005258e+00 7.11099982e-01
3.70920092e-01 3.79496843e-01 3.95506918e-01 2.77996153e-01
4.64792252e-01 8.82133067e-01 5.67282915e-01 -3.65195811e-01
-7.45699942e-01 -4.08584148e-01 9.98509079e-02 2.12647691e-01
2.37383038e-01 -3.65207642e-01 -6.41479611e-01 1.01353192e+00
-1.93880630e+00 -7.47739136e-01 1.49149865e-01 1.70614898e+00
5.97651482e-01 2.25139156e-01 -2.32322752e-01 -9.96950790e-02
3.96343648e-01 5.54189444e-01 -3.71543914e-01 -9.09681797e-01
-2.65885353e-01 4.14165050e-01 8.04623425e-01 5.00998795e-01
-1.03727674e+00 1.36216927e+00 7.77853155e+00 7.71212816e-01
-1.11248851e+00 4.88163203e-01 8.48276496e-01 -3.58840138e-01
-7.27491975e-01 1.62930533e-01 -5.94185054e-01 2.49168083e-01
1.01769936e+00 2.28464425e-01 3.82976443e-01 9.11135018e-01
-1.78599302e-02 -3.81449878e-01 -9.90766287e-01 8.11415851e-01
8.53457525e-02 -1.68528068e+00 -1.39322430e-01 3.00655931e-01
6.07663393e-01 4.51966882e-01 4.49109763e-01 3.06841731e-01
1.39815971e-01 -1.40233552e+00 1.22338688e+00 3.12400818e-01
8.22681963e-01 -4.10887241e-01 5.38405895e-01 -1.10121995e-01
-1.02290666e+00 -7.51555115e-02 -7.09066570e-01 -2.88219810e-01
2.01383859e-01 3.64844501e-01 -8.22440922e-01 2.04909578e-01
7.01376736e-01 4.83227462e-01 -5.22457182e-01 1.21793532e+00
7.43716061e-02 5.96458614e-01 -4.08003867e-01 3.66844684e-01
7.85689175e-01 2.49300271e-01 2.68959790e-01 1.06300271e+00
3.68673682e-01 -5.09334028e-01 -8.12292472e-02 1.29081476e+00
-1.10203832e-01 -9.80368704e-02 -7.18843758e-01 1.54062919e-02
6.88858703e-02 7.87038803e-01 -8.72010291e-01 -2.04418495e-01
-6.07102454e-01 1.22814846e+00 3.68505538e-01 1.64280862e-01
-9.16952848e-01 -1.61933862e-02 6.36986375e-01 3.41758788e-01
7.18669116e-01 -3.54093999e-01 -5.43417394e-01 -1.08946359e+00
-1.14631921e-01 -5.25677919e-01 -1.68473590e-02 -1.00387967e+00
-6.39937520e-01 7.12438345e-01 -2.07086548e-01 -7.90513992e-01
4.03691567e-02 -6.28293216e-01 -4.18946534e-01 9.62739825e-01
-1.78972268e+00 -1.28879404e+00 3.26709896e-01 3.08051765e-01
9.43611443e-01 -1.49998620e-01 6.44594789e-01 3.96311879e-01
-3.79010320e-01 6.10294759e-01 -1.15461558e-01 -1.93405703e-01
4.51136172e-01 -1.14686191e+00 6.65793538e-01 6.41285777e-01
6.28231287e-01 4.98843461e-01 7.19216883e-01 -3.05476755e-01
-1.17891979e+00 -8.76735926e-01 9.07041490e-01 -3.90650719e-01
2.92186230e-01 -3.50499213e-01 -3.58967215e-01 8.51192594e-01
5.77141762e-01 1.56593576e-01 4.95852053e-01 3.81459981e-01
-6.44162714e-01 2.21361980e-01 -1.02237070e+00 4.33892459e-01
1.28633463e+00 -6.97754920e-01 -5.10402620e-01 2.14542925e-01
8.95675600e-01 -1.81714982e-01 -4.01036233e-01 2.26642981e-01
5.48215032e-01 -1.14612508e+00 8.88523757e-01 -4.74334151e-01
6.77545011e-01 -1.79193959e-01 -3.70410442e-01 -1.08795166e+00
-5.93808055e-01 -3.04508150e-01 3.89202923e-01 9.38146412e-01
8.04760873e-01 -4.55157548e-01 1.13137841e+00 5.66621482e-01
-4.90139365e-01 -6.89496815e-01 -1.35614300e+00 -7.60072231e-01
2.65374929e-01 -4.60174084e-01 3.18425387e-01 5.90790391e-01
-2.77645409e-01 1.77925438e-01 -6.09200776e-01 -3.60762507e-01
2.36603558e-01 1.46841304e-02 4.44116712e-01 -6.64214373e-01
-4.04490143e-01 -5.32700241e-01 -3.13077271e-01 -1.27861667e+00
2.21194446e-01 -9.79541957e-01 1.49341524e-01 -1.67101693e+00
1.77436918e-01 -8.43689919e-01 -3.96942258e-01 5.80502450e-01
3.47021461e-01 8.32118154e-01 7.06669390e-01 2.59951919e-01
-9.55965459e-01 6.97441757e-01 1.13596737e+00 -1.50159821e-01
2.52577543e-01 -5.79768717e-01 -9.29872513e-01 4.19863820e-01
6.79812789e-01 -4.90903705e-01 -5.42732060e-01 -9.60515797e-01
2.88567305e-01 -1.14007421e-01 3.37209672e-01 -1.00627923e+00
-6.08315989e-02 9.57686156e-02 3.18948001e-01 -3.71816784e-01
6.41095042e-01 -7.44220138e-01 1.01703368e-01 5.13513386e-01
-4.06975895e-01 4.74847704e-02 4.04125005e-01 1.65906221e-01
-2.05445945e-01 -3.02803487e-01 7.39521205e-01 -4.13486063e-01
-7.99912751e-01 1.52858928e-01 -4.53514874e-01 -8.00539404e-02
6.60657227e-01 -1.78821430e-01 -3.91089797e-01 -2.74707258e-01
-6.28650725e-01 1.30336538e-01 7.06441343e-01 7.11700022e-01
5.28895378e-01 -1.34627259e+00 -6.51202381e-01 2.38615334e-01
5.80098294e-02 -2.92641014e-01 1.74997821e-01 7.00783670e-01
-7.44023681e-01 8.61000121e-01 -4.44733888e-01 -5.62524319e-01
-7.33214974e-01 5.30320644e-01 5.26299834e-01 -4.54937488e-01
-7.17101336e-01 1.13547182e+00 5.34659743e-01 -5.29239416e-01
1.26616314e-01 -1.02697231e-01 3.37694049e-01 -1.85780719e-01
3.32497299e-01 -1.69561699e-01 -4.98963185e-02 -5.32440245e-01
-2.77434736e-01 2.09193975e-01 7.80555559e-03 -4.21057314e-01
1.31780958e+00 -2.90780187e-01 1.16162211e-01 1.91422477e-01
1.25817871e+00 -3.79453808e-01 -1.59639573e+00 -1.56350642e-01
-3.80529687e-02 -6.25504017e-01 4.04700726e-01 -8.65961015e-01
-1.32176745e+00 8.79745901e-01 5.95494270e-01 1.20312227e-02
8.63204241e-01 1.62788928e-01 8.62287402e-01 1.99818805e-01
2.76058733e-01 -1.31045365e+00 2.09143624e-01 4.96889889e-01
1.00844574e+00 -1.15852344e+00 -1.24299526e-01 -1.98199809e-01
-5.31746328e-01 8.22509229e-01 6.98707998e-01 -3.71866822e-01
6.09452546e-01 4.41690475e-01 -9.61335599e-02 -1.52064085e-01
-1.03458750e+00 -4.96442467e-01 8.10779929e-02 4.32775915e-01
6.36256874e-01 -1.67163298e-01 -7.21527934e-01 3.54280740e-01
-1.86542228e-01 -8.90304074e-02 4.29014504e-01 8.53765070e-01
-4.11660403e-01 -1.35091698e+00 -1.68196857e-01 3.62076014e-01
-7.50076354e-01 -5.90693653e-01 -4.96814772e-02 8.17450106e-01
2.71373838e-01 7.27032304e-01 5.65022469e-01 -2.26126313e-01
-3.62223946e-04 2.97536314e-01 5.02688229e-01 -7.32120872e-01
-4.82096314e-01 1.67273164e-01 5.35713673e-01 -9.44927931e-01
-3.56813163e-01 -4.65763390e-01 -9.36393142e-01 -2.71080017e-01
-3.39143306e-01 3.09134610e-02 9.21178341e-01 9.32580113e-01
5.50691485e-01 6.32692993e-01 3.18730295e-01 -8.49086761e-01
-4.52542230e-02 -9.03445482e-01 -5.89973748e-01 2.87405759e-01
2.93036878e-01 -7.48981476e-01 -3.78813371e-02 -1.59897551e-01] | [10.763866424560547, 0.10896164923906326] |
d4c0ee13-62ca-45ac-8f1b-6d85e53821a4 | the-influence-of-the-other-race-effect-on | 2204.12591 | null | https://arxiv.org/abs/2204.12591v1 | https://arxiv.org/pdf/2204.12591v1.pdf | The Influence of the Other-Race Effect on Susceptibility to Face Morphing Attacks | Facial morphs created between two identities resemble both of the faces used to create the morph. Consequently, humans and machines are prone to mistake morphs made from two identities for either of the faces used to create the morph. This vulnerability has been exploited in "morph attacks" in security scenarios. Here, we asked whether the "other-race effect" (ORE) -- the human advantage for identifying own- vs. other-race faces -- exacerbates morph attack susceptibility for humans. We also asked whether face-identification performance in a deep convolutional neural network (DCNN) is affected by the race of morphed faces. Caucasian (CA) and East-Asian (EA) participants performed a face-identity matching task on pairs of CA and EA face images in two conditions. In the morph condition, different-identity pairs consisted of an image of identity "A" and a 50/50 morph between images of identity "A" and "B". In the baseline condition, morphs of different identities never appeared. As expected, morphs were identified mistakenly more often than original face images. Moreover, CA participants showed an advantage for CA faces in comparison to EA faces (a partial ORE). Of primary interest, morph identification was substantially worse for cross-race faces than for own-race faces. Similar to humans, the DCNN performed more accurately for original face images than for morphed image pairs. Notably, the deep network proved substantially more accurate than humans in both cases. The results point to the possibility that DCNNs might be useful for improving face identification accuracy when morphed faces are presented. They also indicate the significance of the ORE in morph attack susceptibility in applied settings. | ["Alice J. O'Toole", 'Carlos D. Castillo', 'Connor J. Parde', 'Geraldine Jeckeln', 'Snipta Mallick'] | 2022-04-26 | null | null | null | null | ['face-identification'] | ['computer-vision'] | [ 1.41174823e-01 -2.96775214e-02 4.01219815e-01 -6.39608443e-01
-2.23025978e-01 -7.64722824e-01 4.59385335e-01 -1.34238452e-01
-6.57331944e-01 2.27424115e-01 -1.82038680e-01 -2.87841052e-01
3.02237511e-01 -7.60664523e-01 -6.79990828e-01 -5.35456419e-01
-1.44294053e-01 7.31464773e-02 -5.52946329e-01 -2.58784384e-01
2.17006385e-01 8.91242981e-01 -1.76154780e+00 5.28302789e-01
1.59499794e-01 1.05856168e+00 -3.56179535e-01 4.03255045e-01
1.09409675e-01 1.82663307e-01 -8.09400380e-01 -1.12032616e+00
8.51107419e-01 -8.15011024e-01 -6.61051393e-01 -1.79298759e-01
1.44328856e+00 -4.51676786e-01 -1.42668024e-01 1.28421450e+00
6.68815851e-01 -6.58620372e-02 5.01682460e-01 -1.32530689e+00
-8.83335292e-01 3.08067530e-01 -7.56300449e-01 1.97796196e-01
5.74196398e-01 5.35362840e-01 5.26080489e-01 -8.05662990e-01
4.78116065e-01 1.55962694e+00 9.88972425e-01 1.04275393e+00
-1.51851642e+00 -1.48543334e+00 -2.07232222e-01 -1.96008151e-03
-1.67972815e+00 -9.77127016e-01 5.10955215e-01 -5.13080895e-01
6.44583464e-01 3.02862734e-01 8.03603709e-01 1.23590100e+00
2.05140442e-01 -1.23971820e-01 1.40910482e+00 -1.58408001e-01
-2.52881736e-01 4.65017200e-01 -1.23343691e-01 3.94696265e-01
4.14837837e-01 6.54761851e-01 -4.74836409e-01 -3.54209185e-01
6.90508068e-01 -4.23287481e-01 -1.91218585e-01 2.09074795e-01
-6.62858427e-01 7.01506138e-01 4.67202097e-01 5.51326096e-01
-3.23545516e-01 -1.44820169e-01 2.22581238e-01 6.88110530e-01
9.64715704e-02 9.65917945e-01 -1.12073481e-01 2.53978133e-01
-8.71036053e-01 2.39747524e-01 6.13034606e-01 3.00979137e-01
9.35510457e-01 1.86507300e-01 -7.89722130e-02 8.70556116e-01
-1.31904110e-01 3.37190002e-01 5.70523679e-01 -7.80273438e-01
1.35734439e-01 4.41851616e-01 7.25834146e-02 -1.37160432e+00
-4.96597767e-01 9.16415080e-02 -7.96811163e-01 8.85584474e-01
8.95167232e-01 -1.79679424e-01 -6.81326389e-01 2.18649364e+00
2.24892259e-01 -3.11943203e-01 -7.16107264e-02 9.26344097e-01
5.09130239e-01 5.36052547e-02 4.80511338e-01 -3.68452221e-02
1.73290694e+00 -8.72936547e-02 -4.43910033e-01 -3.77919167e-01
3.62389266e-01 -8.98480654e-01 1.26556361e+00 -1.11225381e-01
-1.04627228e+00 -9.27909315e-01 -1.09487927e+00 2.70568132e-01
-6.65970683e-01 7.60118365e-02 3.72173458e-01 1.65632904e+00
-1.39436615e+00 8.54272544e-01 2.10162179e-04 -5.64172208e-01
6.92500293e-01 6.06431007e-01 -1.14123774e+00 2.33249456e-01
-1.29007101e+00 1.09578907e+00 3.35550010e-02 1.98149458e-01
-7.14088321e-01 -1.00028467e+00 -8.86108041e-01 -8.43179822e-02
-4.42823499e-01 -3.68535042e-01 8.95564318e-01 -2.14777231e+00
-1.17196965e+00 1.80564284e+00 -1.35256782e-01 -2.75625527e-01
4.96824056e-01 5.05658574e-02 -8.99532318e-01 1.71082839e-01
9.74872559e-02 8.89163136e-01 1.26501834e+00 -1.15700006e+00
-1.45679533e-01 -8.64964962e-01 -1.73803940e-02 -5.03498428e-02
-9.67069417e-02 6.08348846e-01 6.16232634e-01 -5.67538440e-01
-3.37817758e-01 -8.08071077e-01 4.08033997e-01 3.99385393e-01
-2.17026949e-01 1.19009025e-01 7.30163753e-01 -8.26365948e-01
8.45116854e-01 -2.47492361e+00 -5.71788907e-01 3.91479015e-01
3.10908228e-01 6.60284579e-01 -4.05263454e-01 2.52182394e-01
-1.00787973e+00 5.90348721e-01 2.15428937e-02 -1.78541422e-01
-1.28557369e-01 -2.71833211e-01 -1.36932479e-02 7.79231250e-01
4.63551521e-01 7.43388772e-01 -4.90924120e-01 -6.44674748e-02
-3.44995856e-01 4.38492298e-01 -2.15459406e-01 2.29138672e-01
4.57437485e-01 4.81113195e-01 3.42581630e-01 4.72381949e-01
1.20757842e+00 4.29091752e-01 9.17369872e-02 -4.10384834e-01
-3.26526016e-02 -2.25764394e-01 -9.99691486e-01 6.68744802e-01
-1.93545401e-01 7.81966865e-01 3.38552982e-01 -2.43562877e-01
8.53633285e-01 2.83780277e-01 -1.09960333e-01 -9.09843743e-01
3.43511552e-01 2.69468337e-01 6.53826118e-01 -4.02141809e-01
3.63654792e-01 -7.11293519e-01 3.97674561e-01 6.20346725e-01
-1.94274157e-01 2.48138588e-02 -2.69285180e-02 -1.00220941e-01
5.52541018e-01 -4.94994402e-01 2.11066917e-01 -4.13385183e-01
5.66411257e-01 -5.25359690e-01 4.63398844e-01 7.14602411e-01
-6.95420682e-01 6.87582493e-01 6.99289978e-01 -6.94439173e-01
-8.28983128e-01 -1.10204816e+00 -1.84839487e-01 8.36306036e-01
-1.30818561e-01 -1.15680583e-01 -1.06073153e+00 -6.24082983e-01
3.62059742e-01 6.51918352e-01 -1.17808414e+00 -5.27335346e-01
-5.63446045e-01 -6.25923932e-01 1.36681354e+00 3.45265895e-01
5.86499512e-01 -9.75818157e-01 -5.80412447e-01 -4.76029426e-01
9.03169289e-02 -6.83248222e-01 -9.09354985e-01 -5.30604780e-01
-2.06512034e-01 -1.46409154e+00 -7.39709854e-01 -6.99852824e-01
7.63309717e-01 1.90000802e-01 9.72273052e-01 3.48678559e-01
-5.07500410e-01 4.50290591e-01 -1.24452628e-01 -3.69068891e-01
-7.07186759e-01 -6.45628572e-01 5.31543791e-01 6.17755413e-01
5.60836792e-01 -4.62526709e-01 -5.95205009e-01 5.48921764e-01
-9.36539173e-01 -4.01348978e-01 5.57110965e-01 6.19477272e-01
-1.49632767e-01 -5.40069081e-02 5.38734794e-01 -8.33843708e-01
5.33803284e-01 -1.80322215e-01 -3.95219892e-01 2.43073091e-01
-5.22650898e-01 -2.96108335e-01 5.60435891e-01 -7.41675496e-01
-1.00176466e+00 -3.08194179e-02 -9.21101570e-02 -4.22113210e-01
-5.56236565e-01 -1.08240411e-01 -3.58595222e-01 -7.29946852e-01
9.32981968e-01 -1.40081957e-01 5.82971513e-01 -2.67990649e-01
-1.27335727e-01 5.17543077e-01 4.84441727e-01 -4.78244603e-01
1.02898729e+00 3.45128715e-01 2.48307101e-02 -8.76628578e-01
-3.59289423e-02 1.21681325e-01 -5.39584994e-01 -3.02271247e-01
9.64423954e-01 -7.38779128e-01 -1.07435644e+00 1.15082872e+00
-1.08577847e+00 -3.12582940e-01 -3.89204584e-02 3.28061432e-01
-5.40416613e-02 3.46371204e-01 -4.55007434e-01 -7.41247833e-01
-1.27330035e-01 -1.12446249e+00 5.35930395e-01 3.91165644e-01
-5.42862535e-01 -6.87631905e-01 -9.06054303e-02 1.25644937e-01
6.00384414e-01 2.37799898e-01 1.04568648e+00 -8.01422179e-01
2.35332981e-01 -4.72882777e-01 -4.95099217e-01 4.91238058e-01
5.72329223e-01 1.84817821e-01 -1.47109079e+00 -3.90408456e-01
1.53418750e-01 -4.08161998e-01 4.71572340e-01 -6.60062060e-02
7.41805613e-01 -5.60730875e-01 -1.47184497e-02 5.48568308e-01
1.04051757e+00 4.11713272e-01 8.38172615e-01 -2.15799529e-02
4.98402655e-01 1.13350260e+00 2.00909171e-02 -1.25376284e-02
-1.20601788e-01 7.15871036e-01 1.32052928e-01 -1.71797916e-01
-2.48318747e-01 -2.44460002e-01 5.52435756e-01 -3.51120204e-01
1.63576126e-01 1.28139138e-01 -7.52432227e-01 2.97632188e-01
-7.50686884e-01 -1.20064843e+00 -1.18793070e-01 2.56954741e+00
7.04314709e-01 -4.26802754e-01 3.72742057e-01 -4.93751466e-02
1.24384415e+00 1.85874388e-01 -4.97584820e-01 -5.78957021e-01
-5.67799091e-01 4.77083415e-01 3.33241969e-01 3.60324085e-01
-9.85599399e-01 7.49198377e-01 6.58093452e+00 5.27241290e-01
-1.33241498e+00 -4.48108949e-02 1.14670229e+00 -8.19191039e-02
4.62727621e-04 -1.79978520e-01 -5.00860453e-01 3.25958580e-01
1.05848444e+00 -2.21687779e-01 6.32640719e-01 4.55559820e-01
-1.82568431e-02 -1.87830657e-01 -1.36877048e+00 1.20833123e+00
3.08991939e-01 -9.69428003e-01 2.63906986e-01 1.39225617e-01
3.65267456e-01 -7.56165802e-01 7.13597655e-01 -6.12419620e-02
6.66333139e-02 -1.48810983e+00 8.84040654e-01 2.96598487e-02
1.37343109e+00 -6.70462370e-01 5.77314317e-01 -3.17737788e-01
-1.03771472e+00 -1.04554623e-01 -4.09190208e-01 -2.62757629e-01
-1.94232404e-01 -1.58502739e-02 -5.37408292e-01 2.19963983e-01
8.82162452e-01 -1.78908743e-02 -9.94428456e-01 6.19278431e-01
9.28311571e-02 1.77461624e-01 -1.27789244e-01 5.21707356e-01
-6.15934022e-02 -2.76627745e-02 4.23363030e-01 1.00917518e+00
1.15860127e-01 9.85760018e-02 -6.77049100e-01 1.20380437e+00
-1.67787060e-01 1.12567097e-01 -7.85116255e-01 -3.98039371e-01
4.72847879e-01 1.38951516e+00 -6.09660387e-01 -1.11540422e-01
-2.61170447e-01 1.18631148e+00 2.45099500e-01 3.81522894e-01
-4.66091514e-01 -4.00255412e-01 1.23251653e+00 3.22777361e-01
-2.45783880e-01 3.97698164e-01 -2.61872590e-01 -6.10718310e-01
-2.04954413e-03 -1.31775022e+00 4.51199561e-01 -7.07621515e-01
-1.52796793e+00 9.58856046e-01 -5.48784770e-02 -8.11693609e-01
-1.40056470e-02 -8.10616136e-01 -7.49632537e-01 1.52513695e+00
-8.35282385e-01 -1.03814471e+00 -3.91413510e-01 6.17065668e-01
-1.12680219e-01 -1.56089112e-01 9.37949181e-01 2.21891493e-01
-6.87756538e-01 1.32563615e+00 -5.27413011e-01 6.82694733e-01
1.10772061e+00 -7.86380112e-01 7.03861475e-01 8.32918763e-01
8.26449692e-02 1.02168989e+00 4.70287383e-01 -6.35495424e-01
-9.72515166e-01 -7.08492815e-01 9.25134838e-01 -4.59064960e-01
2.75252372e-01 -3.37658107e-01 -8.25116038e-01 5.42121530e-01
2.19499409e-01 -2.14038864e-01 9.61163759e-01 -2.87710965e-01
-1.18208754e+00 -4.28672284e-02 -1.83808839e+00 7.05189884e-01
9.42507446e-01 -9.89450455e-01 -1.93114817e-01 -1.19931363e-01
1.76249206e-01 1.06601663e-01 -7.03337550e-01 1.40159845e-01
1.10105598e+00 -1.55068004e+00 1.01226485e+00 -7.87442505e-01
1.37556896e-01 -1.13809370e-01 -6.81541115e-02 -1.30434573e+00
-3.89007032e-01 -6.22058570e-01 6.83543622e-01 1.38952339e+00
2.37742409e-01 -1.10192561e+00 5.23076773e-01 1.09492409e+00
4.16956991e-01 3.29209417e-02 -1.20055175e+00 -9.29627359e-01
4.74035859e-01 1.09284200e-01 8.58629465e-01 1.18060005e+00
-2.90793031e-01 -1.72629997e-01 -7.33033642e-02 2.33184353e-01
4.26424116e-01 -4.62282389e-01 6.66443765e-01 -1.26709020e+00
-6.28150925e-02 -9.57057774e-01 -6.52818501e-01 1.97649926e-01
7.24081099e-01 -6.95941865e-01 -3.78512889e-01 -3.52911949e-01
-3.28146778e-02 -2.83739746e-01 -1.18790008e-02 5.41046977e-01
-4.79896873e-01 6.56493247e-01 4.79482800e-01 4.48733903e-02
7.08191872e-01 5.75622134e-02 8.18947792e-01 -1.17907010e-01
-5.08223437e-02 -9.95696113e-02 -9.75523770e-01 4.51232702e-01
8.55944633e-01 -4.31559235e-01 -9.78013221e-03 -1.54893205e-01
1.14575808e-03 -1.92382455e-01 7.82790005e-01 -1.06970990e+00
-5.13932928e-02 6.76295385e-02 9.42117989e-01 3.24053288e-01
4.18261766e-01 -7.75305510e-01 7.74059534e-01 5.44110417e-01
-1.55440539e-01 6.49439931e-01 7.44644105e-01 2.25906186e-02
4.49038818e-02 -2.88187444e-01 1.39458370e+00 -3.19943756e-01
-3.99148703e-01 1.70620456e-01 -4.11374599e-01 -3.44854109e-02
9.39992368e-01 -5.86930454e-01 -4.02538002e-01 -4.35711443e-01
-5.35853744e-01 -5.54875135e-01 9.49989200e-01 3.59771639e-01
5.02456844e-01 -1.33715153e+00 -7.75853813e-01 8.25351000e-01
8.48338008e-02 -1.08464122e+00 4.54505235e-01 8.67760599e-01
-5.55746138e-01 -1.03093684e-01 -9.12638962e-01 1.89123914e-01
-1.51627302e+00 8.02016973e-01 9.05276120e-01 4.56858665e-01
3.99402343e-02 1.21553993e+00 6.25556588e-01 -2.86022753e-01
-1.20727450e-01 4.98276860e-01 3.56199667e-02 3.05284321e-01
1.06913638e+00 4.26218152e-01 -1.46718472e-01 -1.43089926e+00
-4.90024775e-01 4.79884446e-01 -2.74927557e-01 -1.44694999e-01
4.92212176e-01 2.09821358e-01 -3.53581429e-01 2.14054763e-01
1.41852844e+00 2.61902064e-01 -7.42271006e-01 2.98448116e-01
-4.21046197e-01 -9.88390028e-01 -4.83830273e-01 -8.18958580e-01
-1.30541146e+00 8.67274165e-01 1.04799199e+00 -5.52581139e-02
9.35576797e-01 -4.04318452e-01 3.19376349e-01 -1.31564945e-01
2.71032751e-01 -5.67438126e-01 -6.45037442e-02 -9.40175354e-02
9.55615938e-01 -1.12561870e+00 -3.81469935e-01 -7.46364072e-02
-6.54418468e-01 1.13199818e+00 1.02192473e+00 1.26134083e-01
4.07556146e-01 -1.48773775e-01 5.51841736e-01 -2.55940408e-01
-6.02090135e-02 -5.14612608e-02 3.11012059e-01 1.19758582e+00
4.68120813e-01 -5.07388338e-02 1.64842624e-02 3.84914666e-01
-6.03289902e-01 -4.76716548e-01 3.69247049e-01 4.68087554e-01
1.33310065e-01 -1.08689225e+00 -8.67119372e-01 5.17583549e-01
-4.59134400e-01 -1.50566041e-01 -1.16219223e+00 9.15069520e-01
2.81501293e-01 1.00295639e+00 5.03343999e-01 -7.23742843e-01
2.16957033e-01 4.28012848e-01 5.47429383e-01 -3.41623902e-01
-1.33314073e+00 -6.53306544e-01 -5.12301475e-02 -6.35143936e-01
-1.93069920e-01 -8.37485135e-01 -5.21337330e-01 -8.69522393e-01
8.57347324e-02 -2.40972694e-02 2.70784229e-01 5.42659581e-01
3.81594896e-01 -3.17469716e-01 6.01231277e-01 -9.36223209e-01
-2.96371639e-01 -9.18325245e-01 -8.75788212e-01 8.21880281e-01
2.78505147e-01 -6.18830442e-01 -6.15693629e-01 -2.38050044e-01] | [12.98065185546875, 1.1717380285263062] |
d6109170-780c-43c1-b8f5-550b5927fa28 | a-3d-mesh-based-lifting-and-projection | 2109.11719 | null | https://arxiv.org/abs/2109.11719v1 | https://arxiv.org/pdf/2109.11719v1.pdf | A 3D Mesh-based Lifting-and-Projection Network for Human Pose Transfer | Human pose transfer has typically been modeled as a 2D image-to-image translation problem. This formulation ignores the human body shape prior in 3D space and inevitably causes implausible artifacts, especially when facing occlusion. To address this issue, we propose a lifting-and-projection framework to perform pose transfer in the 3D mesh space. The core of our framework is a foreground generation module, that consists of two novel networks: a lifting-and-projection network (LPNet) and an appearance detail compensating network (ADCNet). To leverage the human body shape prior, LPNet exploits the topological information of the body mesh to learn an expressive visual representation for the target person in the 3D mesh space. To preserve texture details, ADCNet is further introduced to enhance the feature produced by LPNet with the source foreground image. Such design of the foreground generation module enables the model to better handle difficult cases such as those with occlusions. Experiments on the iPER and Fashion datasets empirically demonstrate that the proposed lifting-and-projection framework is effective and outperforms the existing image-to-image-based and mesh-based methods on human pose transfer task in both self-transfer and cross-transfer settings. | ['Ya zhang', 'Siheng Chen', 'Yangheng Zhao', 'Jinxiang Liu'] | 2021-09-24 | null | null | null | null | ['pose-transfer'] | ['computer-vision'] | [ 3.75016391e-01 1.62289605e-01 -1.51765838e-01 -2.35978246e-01
-2.60096073e-01 -1.20024182e-01 3.81982118e-01 -5.27357340e-01
-1.26871288e-01 6.24703228e-01 2.79172987e-01 1.71583340e-01
2.62525916e-01 -7.98214078e-01 -1.00171542e+00 -5.34572780e-01
2.23150879e-01 4.65000540e-01 4.20349091e-01 -2.83308744e-01
-1.65379480e-01 3.39432567e-01 -1.46413112e+00 1.61860451e-01
8.08001578e-01 1.00409460e+00 -3.52792116e-03 1.70181334e-01
7.42610395e-02 2.55772620e-01 -2.46004790e-01 -4.38729912e-01
6.67091787e-01 -4.57333773e-01 -4.59219784e-01 4.53972936e-01
8.63613725e-01 -5.14026165e-01 -4.64267135e-01 8.83212268e-01
7.28927493e-01 3.69789340e-02 4.88697529e-01 -1.24965036e+00
-7.10889816e-01 -1.05847009e-01 -1.11339068e+00 -1.01461403e-01
5.68284273e-01 2.32012615e-01 4.27780330e-01 -8.92750263e-01
9.53420937e-01 1.78306544e+00 7.99404621e-01 5.93405366e-01
-1.34104621e+00 -8.51692617e-01 1.80269182e-01 -2.36876950e-01
-1.26550388e+00 -5.05043790e-02 1.13429487e+00 -4.61948037e-01
7.83401355e-02 1.35942861e-01 1.11403418e+00 1.22233677e+00
2.83784389e-01 8.50192070e-01 1.20165789e+00 -3.94124746e-01
-3.32050145e-01 3.38819288e-02 -3.48742723e-01 8.88690472e-01
2.30221704e-01 2.77600467e-01 -6.85500741e-01 1.01616455e-03
1.53642404e+00 1.45929918e-01 -3.68650287e-01 -9.21304822e-01
-1.11171246e+00 5.62781990e-01 6.89925492e-01 -2.05008492e-01
-3.12280655e-01 1.30588055e-01 3.05804044e-01 1.12048686e-01
7.71874607e-01 7.90272653e-02 -2.65966564e-01 2.89121658e-01
-7.59757817e-01 5.38617194e-01 5.36187470e-01 9.57719803e-01
6.37808561e-01 -8.45897943e-02 -3.81840199e-01 7.68161058e-01
4.17234421e-01 5.26471198e-01 1.53120697e-01 -7.89047837e-01
5.68315387e-01 8.17596495e-01 1.12623766e-01 -1.28214812e+00
-2.76871622e-01 -5.09684920e-01 -8.48853290e-01 1.86789796e-01
3.88535380e-01 5.16584143e-02 -9.74260926e-01 1.94743490e+00
1.04237342e+00 1.39499232e-02 -5.84847867e-01 1.33083284e+00
7.73685813e-01 2.87738264e-01 2.47419685e-01 6.38021827e-02
1.50742149e+00 -1.10311115e+00 -6.02161229e-01 -2.20784515e-01
1.32004008e-01 -7.13156879e-01 1.35304832e+00 -3.79169099e-02
-1.34732413e+00 -7.63371050e-01 -9.20122087e-01 -2.84309953e-01
1.48054073e-02 7.15360716e-02 3.53795528e-01 4.16595578e-01
-6.35092676e-01 5.68720937e-01 -5.93266547e-01 -4.92615283e-01
4.57318753e-01 3.62688571e-01 -5.66412270e-01 -3.07392399e-03
-1.19838500e+00 7.38244891e-01 2.50873208e-01 2.28960723e-01
-5.65379083e-01 -8.61363530e-01 -1.03206658e+00 -2.87043005e-01
4.53007489e-01 -1.49150693e+00 7.80681014e-01 -1.10086596e+00
-1.60613143e+00 1.17483234e+00 5.63914478e-02 5.84970303e-02
1.04014456e+00 -4.82287496e-01 -6.91037551e-02 2.84738332e-01
1.35605738e-01 7.94758558e-01 1.28953969e+00 -1.38110769e+00
-3.19890410e-01 -5.52553475e-01 -1.52106984e-02 6.20878339e-01
-1.76602006e-01 -1.34436741e-01 -7.61064529e-01 -1.14743280e+00
2.06397846e-01 -1.14586365e+00 7.75043592e-02 7.65065730e-01
-4.08681571e-01 1.81948155e-01 9.35899079e-01 -9.59275842e-01
9.35591578e-01 -2.10459805e+00 4.12465572e-01 1.71193749e-01
2.35803038e-01 1.63309351e-01 -1.72246307e-01 1.84673354e-01
-1.78000495e-01 -3.50662112e-01 -7.12718219e-02 -4.29954529e-01
-1.11725792e-01 1.46573916e-01 -1.62269939e-02 6.21285319e-01
2.44898558e-01 1.09819841e+00 -8.78469348e-01 -8.28426003e-01
2.91306138e-01 8.10327768e-01 -7.74199724e-01 3.58669460e-01
-4.79642861e-03 1.02441442e+00 -4.76255119e-01 5.67843258e-01
8.27544570e-01 -2.26811141e-01 -5.79720736e-02 -5.41068673e-01
1.41628996e-01 -1.18734457e-01 -1.03663409e+00 2.00105786e+00
-2.80214727e-01 -6.55609509e-03 3.18659097e-01 -5.39968133e-01
7.79093146e-01 3.31236303e-01 5.09864688e-01 -6.29596472e-01
3.29015017e-01 -1.28590120e-02 -2.70729929e-01 -4.51736897e-01
1.05293617e-01 -4.13002342e-01 1.41271219e-01 1.10691845e-01
-9.90041345e-02 -1.43722584e-02 -2.10191742e-01 -1.52418837e-01
4.85780299e-01 7.26021826e-01 -8.13246444e-02 -3.28853577e-01
5.94136953e-01 -4.40896273e-01 6.84725225e-01 7.01126456e-02
-2.45052338e-01 8.19567680e-01 2.31146172e-01 -6.87461197e-01
-1.08160532e+00 -1.30492306e+00 6.40045758e-03 1.12838459e+00
4.81869340e-01 -1.75332159e-01 -8.13979328e-01 -6.41036928e-01
2.49482319e-01 1.23738572e-01 -8.44599485e-01 -3.96944851e-01
-8.01061809e-01 -3.12989384e-01 1.66960657e-01 5.77794075e-01
8.78472149e-01 -1.01965952e+00 -6.91830873e-01 1.36517882e-01
-4.61666614e-01 -1.07903862e+00 -1.10286343e+00 -5.60952365e-01
-7.98545599e-01 -9.58163142e-01 -1.10050690e+00 -8.78710389e-01
9.46921527e-01 8.69858488e-02 8.11732948e-01 2.69910451e-02
-4.43257630e-01 2.99327463e-01 -1.77708507e-01 -1.04199633e-01
-5.84367141e-02 -1.88718051e-01 1.64069474e-01 3.77211094e-01
-1.84737489e-01 -8.83826733e-01 -1.09361041e+00 6.13562405e-01
-8.17313969e-01 6.61770880e-01 6.32550657e-01 9.06430900e-01
5.67315221e-01 -7.23699182e-02 1.84869580e-02 -5.86530924e-01
3.27309281e-01 -1.86579507e-02 -1.94753870e-01 1.52651414e-01
-1.11280926e-01 -2.33377427e-01 1.51470855e-01 -9.13422525e-01
-1.19662440e+00 2.37882271e-01 4.36695181e-02 -7.50640392e-01
7.27050379e-02 -6.72145933e-02 -4.71977413e-01 -3.69967520e-01
3.46148580e-01 1.77382573e-01 2.53739595e-01 -6.49523139e-01
3.50373000e-01 1.36500031e-01 7.34205723e-01 -8.95734608e-01
1.16961467e+00 6.78525686e-01 2.62308538e-01 -5.81706107e-01
-9.44518447e-01 -1.86843440e-01 -9.47082400e-01 -4.68229383e-01
1.01698208e+00 -1.03004694e+00 -6.67774737e-01 4.61826980e-01
-1.11255777e+00 -2.24821463e-01 -2.72052795e-01 2.72798300e-01
-6.51901722e-01 3.98615479e-01 -7.44701684e-01 -5.46658754e-01
-5.88118851e-01 -1.00963962e+00 1.46121299e+00 2.24211380e-01
-2.98744738e-01 -8.39691103e-01 -1.10588014e-01 5.92862308e-01
2.57820159e-01 8.26655746e-01 9.18594182e-01 7.89538547e-02
-3.43119621e-01 -1.48580924e-01 -3.25635582e-01 2.98152119e-01
1.09950729e-01 -4.70470130e-01 -7.27598667e-01 -5.77096224e-01
-1.14386953e-01 -3.60791475e-01 4.70138013e-01 4.32677776e-01
1.08426535e+00 -4.02120292e-01 -3.08065891e-01 8.25798094e-01
1.06152391e+00 -4.34423029e-01 5.73824227e-01 7.31714889e-02
1.12566578e+00 8.22471201e-01 4.86586154e-01 2.02421933e-01
4.35523778e-01 1.04412222e+00 1.35666981e-01 -6.57775521e-01
-6.70138776e-01 -8.87398481e-01 2.00024635e-01 5.46081066e-01
-4.06060636e-01 3.46575081e-01 -5.82399309e-01 1.47525847e-01
-1.75479889e+00 -5.20408094e-01 1.70448437e-01 2.19983149e+00
9.42256570e-01 1.95798650e-02 2.80614048e-01 -1.77247420e-01
8.64964843e-01 -9.99632701e-02 -5.17521381e-01 2.60020584e-01
1.19043224e-01 1.39967233e-01 1.55321717e-01 2.32463837e-01
-1.13588488e+00 9.31707263e-01 5.58867693e+00 7.53275931e-01
-1.02187347e+00 1.94844395e-01 3.63490731e-01 3.99589259e-03
-4.67737354e-02 -2.27873668e-01 -5.12887657e-01 4.53847736e-01
6.02550767e-02 1.22527599e-01 1.32645398e-01 7.05636084e-01
1.24172896e-01 1.54089153e-01 -1.03989327e+00 1.07808030e+00
6.04734533e-02 -7.51937330e-01 3.61281663e-01 2.24180415e-01
6.85635209e-01 -7.23204255e-01 2.94971198e-01 1.25411659e-01
-1.16951630e-01 -7.39925146e-01 9.20609713e-01 4.47928667e-01
1.02674592e+00 -5.86305439e-01 4.00513858e-01 1.28780186e-01
-1.40566564e+00 7.47712553e-02 -1.68480784e-01 -3.15848701e-02
3.33452046e-01 4.47721452e-01 -3.05762947e-01 5.97740829e-01
8.08356166e-01 5.72570443e-01 -4.16045457e-01 9.08233821e-01
-2.97626376e-01 4.72496636e-02 -2.44785801e-01 6.29436016e-01
-6.58979341e-02 -2.38925785e-01 7.72944748e-01 8.98539364e-01
9.11603644e-02 6.46126494e-02 5.65575719e-01 1.04774749e+00
-6.70980662e-02 2.28534937e-01 -4.43871051e-01 4.66516733e-01
1.09061807e-01 1.17117167e+00 -6.63724303e-01 -2.50691324e-01
-2.35616073e-01 1.29433572e+00 2.24899113e-01 4.28339660e-01
-8.59962821e-01 -5.34970611e-02 4.34752911e-01 8.81142676e-01
2.66846865e-01 -1.06717467e-01 5.68084884e-03 -1.13806748e+00
2.04267591e-01 -9.33955491e-01 2.24745527e-01 -8.77247930e-01
-1.34576142e+00 3.06842417e-01 2.30760455e-01 -1.29534173e+00
2.54620939e-01 -4.17237878e-01 -4.77682948e-01 9.09311831e-01
-1.22200418e+00 -1.80439889e+00 -5.80762386e-01 7.51271844e-01
4.84830856e-01 3.26921523e-01 3.62686515e-01 5.12836576e-01
-4.18959886e-01 6.90058649e-01 -6.37534499e-01 1.62782446e-01
1.07304990e+00 -9.20325279e-01 2.67617077e-01 5.19618750e-01
-4.37976420e-01 7.47610867e-01 5.22330880e-01 -1.00528443e+00
-1.43164527e+00 -1.35221660e+00 4.56498355e-01 -5.22449553e-01
2.02958927e-01 -4.90311712e-01 -8.40533972e-01 6.62230849e-01
-1.00406729e-01 2.25583971e-01 3.70267808e-01 -1.61908790e-01
-4.09247309e-01 -9.41186175e-02 -1.14893532e+00 7.70421922e-01
1.44858563e+00 -2.87999839e-01 -6.81879461e-01 2.99725622e-01
7.25442767e-01 -8.60269070e-01 -9.18782651e-01 6.30465984e-01
9.10457373e-01 -7.75775492e-01 1.39108825e+00 -4.58898574e-01
6.26075804e-01 -3.62817615e-01 8.42113420e-02 -1.03420806e+00
-5.04779816e-01 -7.52268672e-01 -1.81060612e-01 1.00226271e+00
-1.71308696e-01 -4.51786071e-01 7.06755161e-01 5.60177922e-01
5.17684445e-02 -7.68613636e-01 -9.65441644e-01 -7.93427467e-01
5.74435443e-02 7.69291669e-02 4.47112173e-01 8.09210181e-01
-3.26467723e-01 3.22693795e-01 -7.37756133e-01 -4.52970304e-02
8.85603905e-01 1.82321891e-01 1.16105759e+00 -1.22957897e+00
-2.92130321e-01 -9.87415686e-02 -4.22526062e-01 -1.26053345e+00
5.27750626e-02 -7.90734172e-01 -6.48611784e-02 -1.16558444e+00
3.16695005e-01 -2.42028624e-01 3.89805399e-02 4.02319610e-01
-2.72760481e-01 6.50376737e-01 3.41643095e-01 3.09566736e-01
-3.26027691e-01 9.15690184e-01 2.16879153e+00 8.75512064e-02
-2.16699615e-01 -1.25998691e-01 -4.41591084e-01 9.01286721e-01
3.53811294e-01 -2.09230095e-01 -3.46090734e-01 -3.03315312e-01
-1.40571937e-01 -7.36862868e-02 8.05009723e-01 -8.95312905e-01
-6.87286854e-02 -4.04359177e-02 8.37055802e-01 -6.58105254e-01
5.39647043e-01 -9.04428840e-01 4.44133252e-01 6.27204478e-01
-1.43662587e-01 -5.25496043e-02 1.43684626e-01 6.66658580e-01
1.20869406e-01 6.17012441e-01 9.61653650e-01 -1.63829118e-01
-2.63634741e-01 7.22546458e-01 3.71106565e-01 1.24412045e-01
1.00363684e+00 -4.42639083e-01 8.81612599e-02 -3.07276070e-01
-7.42075324e-01 2.34491408e-01 7.61900783e-01 6.00630283e-01
7.24644840e-01 -1.73279524e+00 -6.41924262e-01 5.46052158e-01
6.40714020e-02 1.62594661e-01 4.15074885e-01 1.16878104e+00
-4.99860078e-01 3.56261954e-02 -5.11796653e-01 -7.27151811e-01
-1.29020846e+00 5.94790936e-01 3.22091490e-01 -2.74765253e-01
-1.18169916e+00 6.65076852e-01 9.29286838e-01 -5.97421110e-01
2.97804058e-01 -9.76052657e-02 1.97674900e-01 -3.51607740e-01
2.85766542e-01 3.53479117e-01 -4.92583901e-01 -9.22668099e-01
-1.73113182e-01 1.16570783e+00 1.22541204e-01 -3.07677966e-02
1.11959505e+00 -3.02596420e-01 -6.21114112e-02 1.46062404e-01
1.10954535e+00 -1.94841325e-01 -1.58510649e+00 -3.45475286e-01
-7.41083086e-01 -7.86273360e-01 -1.59778982e-01 -6.06535733e-01
-1.15972376e+00 9.13211763e-01 7.59509742e-01 -4.84244049e-01
1.10106504e+00 -1.58686325e-01 1.08153176e+00 -2.89082676e-01
5.58867395e-01 -1.00835860e+00 3.90011877e-01 1.74521849e-01
1.30092680e+00 -9.82502937e-01 1.45583451e-01 -9.10564125e-01
-4.32169050e-01 7.93992579e-01 1.00098217e+00 -1.80006266e-01
7.34271586e-01 -1.22646026e-01 -7.74598196e-02 -2.88488299e-01
-1.68651074e-01 2.50130240e-02 8.10193062e-01 7.55121469e-01
2.34495059e-01 -5.86614981e-02 -2.94488460e-01 4.62486356e-01
-2.01459348e-01 1.36051401e-01 -2.49025226e-01 9.88307178e-01
-1.68446839e-01 -1.04290652e+00 -6.48890495e-01 6.03598729e-02
-2.33796000e-01 2.04579324e-01 -2.85998404e-01 9.55156446e-01
4.90608931e-01 3.43361586e-01 1.94099192e-02 -3.03093106e-01
6.39624357e-01 -8.74554217e-02 9.08364832e-01 -5.58892667e-01
-4.58980620e-01 4.49599475e-01 -1.94511980e-01 -7.97211587e-01
-4.73666549e-01 -2.87710100e-01 -8.75431836e-01 -3.91442865e-01
-2.60433465e-01 -4.40760911e-01 2.96362072e-01 6.20102584e-01
3.29156667e-01 5.43358564e-01 3.10484469e-01 -1.37186480e+00
-3.47692043e-01 -8.16521883e-01 -7.57455707e-01 9.05697227e-01
1.85851544e-01 -1.23844755e+00 -1.75773520e-02 2.71726042e-01] | [11.934256553649902, -0.8773653507232666] |
d11d5bfd-7776-4525-bc47-1dfb0b1cd68e | sequential-query-encoding-for-complex-query | 2302.13114 | null | https://arxiv.org/abs/2302.13114v3 | https://arxiv.org/pdf/2302.13114v3.pdf | Sequential Query Encoding For Complex Query Answering on Knowledge Graphs | Complex Query Answering (CQA) is an important and fundamental task for knowledge graph (KG) reasoning. Query encoding (QE) is proposed as a fast and robust solution to CQA. In the encoding process, most existing QE methods first parse the logical query into an executable computational direct-acyclic graph (DAG), then use neural networks to parameterize the operators, and finally, recursively execute these neuralized operators. However, the parameterization-and-execution paradigm may be potentially over-complicated, as it can be structurally simplified by a single neural network encoder. Meanwhile, sequence encoders, like LSTM and Transformer, proved to be effective for encoding semantic graphs in related tasks. Motivated by this, we propose sequential query encoding (SQE) as an alternative to encode queries for CQA. Instead of parameterizing and executing the computational graph, SQE first uses a search-based algorithm to linearize the computational graph to a sequence of tokens and then uses a sequence encoder to compute its vector representation. Then this vector representation is used as a query embedding to retrieve answers from the embedding space according to similarity scores. Despite its simplicity, SQE demonstrates state-of-the-art neural query encoding performance on FB15k, FB15k-237, and NELL on an extended benchmark including twenty-nine types of in-distribution queries. Further experiment shows that SQE also demonstrates comparable knowledge inference capability on out-of-distribution queries, whose query types are not observed during the training process. | ['Yangqiu Song', 'Tianshi Zheng', 'Jiaxin Bai'] | 2023-02-25 | null | null | null | null | ['complex-query-answering'] | ['knowledge-base'] | [ 1.45834655e-01 4.43727151e-02 -3.21001917e-01 -3.52752119e-01
-6.22037172e-01 -6.25598192e-01 4.45732653e-01 3.52749348e-01
-5.72574496e-01 3.84852588e-01 3.36662605e-02 -7.00976670e-01
-2.99028337e-01 -1.47341084e+00 -8.88147473e-01 -1.79624870e-01
-4.49470654e-02 7.46634543e-01 5.25628388e-01 -4.11424369e-01
1.09512657e-01 4.00270909e-01 -1.49840593e+00 3.87536794e-01
8.19250405e-01 1.16983283e+00 2.56906927e-01 5.74232578e-01
-6.86852813e-01 1.10913944e+00 -6.31847441e-01 -8.64303350e-01
-2.96575665e-01 -2.67745793e-01 -1.44425368e+00 -6.23686552e-01
7.63559192e-02 -3.74261260e-01 -5.78446865e-01 1.03269708e+00
4.00656164e-01 3.16951126e-01 4.39092010e-01 -1.30393386e+00
-8.54562044e-01 7.85197079e-01 2.77328968e-01 1.92351546e-02
7.62514532e-01 -1.54797867e-01 1.38985634e+00 -6.55767143e-01
5.63623250e-01 1.35648406e+00 3.67436498e-01 3.08803707e-01
-9.34564888e-01 -5.95393896e-01 -1.54816672e-01 7.05213606e-01
-1.48905575e+00 -8.96620229e-02 7.40658939e-01 9.11020034e-04
1.59046435e+00 2.49318808e-01 7.49100387e-01 7.47500837e-01
-5.73139638e-03 1.04439032e+00 4.03209299e-01 -4.16912436e-01
1.97086498e-01 -1.07234590e-01 2.15798199e-01 1.12306511e+00
1.67092159e-01 -3.94476913e-02 -4.61218476e-01 -2.81779677e-01
3.65907192e-01 -1.12060457e-01 -1.60732776e-01 -1.83742329e-01
-9.80126560e-01 1.02081740e+00 6.99245989e-01 1.42560020e-01
-4.00872111e-01 5.08628666e-01 7.92523503e-01 5.86955369e-01
-1.01839237e-01 6.45919740e-01 -4.34870124e-01 -1.20348871e-01
-4.71443743e-01 3.89011443e-01 1.17567015e+00 1.03223372e+00
9.10869479e-01 -3.66752297e-02 -7.08871782e-01 6.89736128e-01
2.82003880e-01 5.05343556e-01 5.02332389e-01 -7.95107782e-01
5.56683779e-01 9.53109980e-01 -2.18956187e-01 -1.21838999e+00
-1.04130529e-01 -3.46035004e-01 -6.89947188e-01 -5.93977153e-01
1.66968144e-02 2.90550768e-01 -6.80352807e-01 1.62234247e+00
2.16468185e-01 2.48885509e-02 3.47445130e-01 6.65370524e-01
9.97980118e-01 6.61151111e-01 1.65167242e-01 5.23461029e-02
1.62573564e+00 -1.01867080e+00 -6.28442407e-01 -1.36147365e-01
1.08112037e+00 -2.95394838e-01 1.35486221e+00 9.56360325e-02
-9.73865867e-01 -4.75045264e-01 -1.14428592e+00 -4.58158791e-01
-8.21134031e-01 -4.65473570e-02 8.78252268e-01 3.66522014e-01
-9.62825000e-01 3.48239481e-01 -5.85228324e-01 -1.63294226e-01
3.87613416e-01 3.05559605e-01 -1.80213973e-01 -5.16520262e-01
-1.74286389e+00 6.88571036e-01 1.06542015e+00 9.07958373e-02
-9.12929058e-01 -6.38520360e-01 -1.14049470e+00 5.79743385e-01
8.41697812e-01 -1.08660686e+00 1.25445545e+00 -2.06649601e-01
-1.44340575e+00 4.98403966e-01 -3.25863898e-01 -5.33833861e-01
-1.24524914e-01 -9.21216458e-02 -5.52582860e-01 2.75568366e-01
-7.62524977e-02 4.55666721e-01 6.44501209e-01 -8.48541677e-01
-4.40369368e-01 -1.94221467e-01 7.82665133e-01 1.03435114e-01
-3.50748211e-01 3.62433828e-02 -7.65812516e-01 -4.01088864e-01
1.00926898e-01 -5.52868426e-01 1.25426978e-01 -3.39522779e-01
-5.26139319e-01 -8.68199289e-01 6.35968089e-01 -6.15116417e-01
1.72189057e+00 -1.93381631e+00 8.33363235e-02 3.82518470e-01
2.47870520e-01 3.54790896e-01 -1.68872371e-01 7.61061251e-01
1.86126962e-01 1.65628672e-01 -1.65569276e-01 2.36022528e-02
4.66169417e-01 5.96508861e-01 -5.17964423e-01 -2.21288219e-01
2.50263840e-01 1.54095924e+00 -1.19455612e+00 -6.54105842e-01
-1.66300163e-01 1.86196908e-01 -7.73367047e-01 4.34511065e-01
-7.96557248e-01 -4.51723188e-01 -6.82072282e-01 7.64752507e-01
2.56977379e-01 -5.09334207e-01 2.53516525e-01 -5.11183560e-01
5.97391725e-01 4.89909649e-01 -9.07999516e-01 1.86604905e+00
-5.65875769e-01 2.30994374e-01 -5.05338848e-01 -1.27944613e+00
8.43788445e-01 1.63047418e-01 6.90615326e-02 -1.17071545e+00
-1.58098325e-01 2.18405023e-01 -2.19644010e-01 -7.80819952e-01
7.70898819e-01 5.86790442e-02 -3.51809978e-01 3.76609713e-01
1.17367081e-01 -1.81127310e-01 3.94719422e-01 5.09414852e-01
1.24795377e+00 -1.66414976e-02 1.25386417e-01 3.08339208e-01
7.66831517e-01 -7.08287675e-03 2.93624580e-01 8.34981740e-01
1.74996004e-01 -1.99234277e-01 6.81248426e-01 -4.02992010e-01
-5.04295230e-01 -1.01147878e+00 4.43375170e-01 1.23687708e+00
8.64510238e-02 -8.52374375e-01 -5.40185571e-01 -9.17860270e-01
1.80576161e-01 8.80854964e-01 -2.69701660e-01 -7.47141242e-01
-5.72440028e-01 -2.53344864e-01 1.07499373e+00 6.76037610e-01
7.84417331e-01 -1.35829067e+00 -5.67124844e-01 3.89894009e-01
-2.98823833e-01 -1.14657211e+00 -3.72038633e-01 4.55382317e-02
-5.10219395e-01 -1.20189953e+00 -1.73948646e-01 -7.94069767e-01
5.33931971e-01 2.31818929e-02 1.38773668e+00 3.30558121e-01
-7.22709447e-02 5.17902315e-01 -5.23176134e-01 -1.99320033e-01
-4.07707244e-01 2.30281472e-01 -3.47980797e-01 -2.16401041e-01
4.79181498e-01 -4.37386423e-01 -5.27911901e-01 3.85292321e-02
-1.29475379e+00 -1.07624203e-01 5.36630392e-01 1.03210044e+00
7.01845109e-01 3.95793974e-01 5.61136425e-01 -9.30328548e-01
1.07669377e+00 -4.28223163e-01 -7.63103306e-01 8.85720849e-01
-6.99979365e-01 5.24143755e-01 7.93968916e-01 -1.85161695e-01
-8.90456676e-01 -3.20825338e-01 -1.52934164e-01 -6.23367608e-01
3.28035593e-01 1.24142230e+00 -1.40518904e-01 3.87342051e-02
4.96816903e-01 5.91497838e-01 -2.53614992e-01 -1.62828445e-01
5.64398289e-01 3.73154134e-01 5.94862640e-01 -1.00582230e+00
6.84634626e-01 3.11914105e-02 1.43386960e-01 -4.79770660e-01
-7.24249542e-01 -3.37621927e-01 -1.54887140e-01 1.93141356e-01
8.06148112e-01 -4.88580674e-01 -1.25580120e+00 2.16228798e-01
-1.35526848e+00 -3.06868792e-01 -1.88292399e-01 2.84513891e-01
-3.77128869e-01 3.22230697e-01 -6.78089857e-01 -6.27532780e-01
-6.39352620e-01 -1.12573481e+00 9.99334991e-01 6.38509495e-03
-3.88698578e-02 -1.00335550e+00 -3.00250769e-01 3.10012490e-01
5.52191257e-01 -9.62822586e-02 1.79994678e+00 -7.78328896e-01
-8.16347837e-01 -3.57697248e-01 -4.34354901e-01 4.15849924e-01
-1.27406418e-01 -3.06490749e-01 -7.00773299e-01 -3.43222857e-01
-3.95179868e-01 -6.69808567e-01 6.15218699e-01 -2.36832842e-01
1.54564226e+00 -4.74602371e-01 -2.26802886e-01 6.65947497e-01
1.54172707e+00 3.98760349e-01 6.88617349e-01 2.09805787e-01
7.16133416e-01 3.39705467e-01 3.84211898e-01 1.49654657e-01
8.44355166e-01 5.11601508e-01 4.70216393e-01 3.64467144e-01
7.79918358e-02 -7.98444629e-01 2.68943667e-01 1.01364779e+00
1.42717659e-01 -3.14128548e-01 -9.57567155e-01 3.58665377e-01
-1.55166686e+00 -7.84295738e-01 2.88024515e-01 1.98122180e+00
8.53302777e-01 6.75966367e-02 -2.89823800e-01 2.34108090e-01
1.17665023e-01 1.75235629e-01 -5.79887629e-01 -4.91612375e-01
8.51106644e-02 5.78781366e-01 3.93970191e-01 3.85385185e-01
-5.87367654e-01 1.26551652e+00 5.82983017e+00 1.07353318e+00
-1.02620852e+00 1.16075808e-02 2.57861428e-02 2.99850017e-01
-6.41615570e-01 1.84339821e-01 -7.24428594e-01 3.02425504e-01
9.59600508e-01 -3.90409082e-01 8.16813111e-01 7.07565129e-01
-5.20904720e-01 1.25676155e-01 -1.26250207e+00 9.77886975e-01
-3.48088853e-02 -1.48903131e+00 5.87880254e-01 -3.55293453e-01
1.72858521e-01 -1.19390033e-01 -2.87543893e-01 1.11372268e+00
2.77461261e-01 -1.12945962e+00 4.09909546e-01 6.11400545e-01
7.93209076e-01 -7.42116809e-01 7.65849173e-01 3.11060429e-01
-1.46182847e+00 -1.74658522e-01 -5.17162800e-01 1.70254141e-01
6.78139850e-02 1.27959505e-01 -9.26102459e-01 1.12553108e+00
7.21934378e-01 3.42732757e-01 -6.45134389e-01 5.80978811e-01
-5.22583485e-01 4.67735916e-01 -1.57966375e-01 -3.07894260e-01
2.84444690e-01 -1.38706267e-02 1.07172504e-01 1.02780402e+00
2.66934335e-01 1.57796919e-01 1.70416415e-01 1.03421259e+00
-3.54774088e-01 9.60413590e-02 -5.53959191e-01 -6.24937654e-01
7.97557771e-01 7.66191602e-01 -3.49203140e-01 -7.10705400e-01
-4.85633612e-01 1.08876407e+00 7.43190587e-01 7.44460285e-01
-7.72108555e-01 -9.46922600e-01 2.60097414e-01 -1.79107919e-01
4.48933482e-01 -9.07891840e-02 4.42111045e-01 -1.04378247e+00
3.82351458e-01 -8.66396070e-01 8.96597624e-01 -1.01486111e+00
-9.61348116e-01 4.93637949e-01 3.84192646e-01 -6.82986796e-01
-5.13910294e-01 -3.76704931e-01 -4.28719491e-01 8.75286579e-01
-1.76538002e+00 -1.06288159e+00 -2.87489593e-01 7.86623061e-01
1.20959125e-01 -1.53848723e-01 1.14289534e+00 4.87028122e-01
-4.70770866e-01 8.93180668e-01 -3.59156787e-01 2.60852516e-01
1.47778437e-01 -1.17032635e+00 3.32697541e-01 4.07139361e-01
2.30868250e-01 8.63241434e-01 1.26487985e-01 -4.05325681e-01
-2.04548454e+00 -1.08989596e+00 1.15331864e+00 -6.61936402e-02
6.71673536e-01 -2.20949247e-01 -1.22392905e+00 8.01620841e-01
1.56588722e-02 1.39822215e-01 4.63153124e-01 4.15556412e-03
-6.60340071e-01 -2.98498750e-01 -8.17314565e-01 4.29857999e-01
1.16026056e+00 -1.14516270e+00 -7.94485450e-01 2.80987829e-01
1.44634938e+00 -4.99003112e-01 -1.15717828e+00 4.34709549e-01
4.32436496e-01 -6.75832629e-01 1.07853401e+00 -8.86875212e-01
2.47909844e-01 -2.67748982e-01 -4.16175514e-01 -1.08068144e+00
-1.39181212e-01 -4.18217152e-01 -5.56477010e-01 1.03301609e+00
3.64371568e-01 -8.89210582e-01 8.07073057e-01 3.15648884e-01
-2.81916797e-01 -1.27950740e+00 -6.77885532e-01 -8.47821951e-01
-3.14721882e-01 -6.34436309e-01 1.15023577e+00 8.79896998e-01
-7.88682997e-02 4.58910018e-01 1.97864100e-01 2.00697750e-01
2.45647550e-01 3.04252744e-01 6.54037952e-01 -1.05168438e+00
-4.52748895e-01 -4.08462852e-01 -2.64691174e-01 -1.31426656e+00
5.00146508e-01 -1.45046210e+00 -2.16654420e-01 -1.84821653e+00
-1.99091598e-01 -3.51139694e-01 -4.71565425e-01 7.22916126e-01
-1.42050041e-02 -5.44230103e-01 -6.04583472e-02 -3.60463671e-02
-5.89110613e-01 6.56248271e-01 1.34006226e+00 -5.43079078e-01
3.81514579e-02 -2.80991077e-01 -5.25811791e-01 1.98940590e-01
4.85574841e-01 -4.14579123e-01 -1.01769280e+00 -6.33970678e-01
7.82787204e-01 4.01464254e-01 4.43359584e-01 -7.32546210e-01
6.28782094e-01 -9.04290453e-02 -2.14443162e-01 -7.17639327e-01
3.21562290e-01 -7.13275433e-01 -1.22259092e-02 6.12094820e-01
-2.88316220e-01 4.15205270e-01 1.22936510e-01 5.45148611e-01
-7.04871118e-01 -3.73751968e-01 -9.93115176e-03 -9.94819030e-02
-1.03831089e+00 4.42537189e-01 1.18814312e-01 3.56514096e-01
5.75458527e-01 -8.86313617e-02 -3.94867510e-01 -2.29339093e-01
-3.54321867e-01 5.08372426e-01 -8.93655047e-02 2.95997292e-01
8.81220758e-01 -1.46947169e+00 -3.10354114e-01 2.73196250e-01
4.89569455e-01 3.85305971e-01 1.81813642e-01 5.80668449e-01
-6.79917753e-01 8.44152808e-01 2.23121166e-01 -3.13510984e-01
-8.99715662e-01 7.83851385e-01 3.14036965e-01 -6.73311651e-01
-3.34087342e-01 9.08001602e-01 -1.45240977e-01 -5.11256158e-01
4.53440279e-01 -6.91136360e-01 -9.73793119e-02 -4.73515540e-02
2.44832531e-01 2.23608002e-01 1.89095542e-01 -1.35913163e-01
-3.36037368e-01 2.48523802e-01 -3.40083539e-02 1.98046431e-01
8.44224751e-01 3.89934003e-01 -3.90058935e-01 -3.81532051e-02
1.61019242e+00 -4.11268741e-01 -2.69542605e-01 -6.79525077e-01
3.23853880e-01 -1.50713339e-01 -1.09141663e-01 -5.83310664e-01
-9.63308990e-01 7.94267237e-01 1.31129045e-02 2.30627298e-01
1.18186760e+00 4.15505581e-02 1.13374186e+00 1.24722135e+00
4.16028172e-01 -8.77698362e-01 2.69772977e-01 7.75811076e-01
8.88605058e-01 -8.40261996e-01 -2.33473361e-01 -4.25807178e-01
-2.66560078e-01 9.80979145e-01 5.54603934e-01 2.49811187e-01
4.43838894e-01 -2.35210955e-01 -4.68389779e-01 -6.44782722e-01
-8.51907670e-01 -1.97529048e-01 3.78007323e-01 2.57376909e-01
1.08174860e-01 -6.71986444e-03 -1.51351273e-01 6.96270704e-01
-4.15894210e-01 2.30707660e-01 -1.34054005e-01 1.02133644e+00
-1.97492182e-01 -1.06791234e+00 2.64827609e-01 5.57655990e-01
-1.77847013e-01 -2.97333777e-01 -1.89557970e-02 8.40363264e-01
-1.32986575e-01 7.31065273e-01 -1.44821912e-01 -5.83779752e-01
6.79493248e-01 4.30606365e-01 5.00511169e-01 -5.64759314e-01
-5.04487872e-01 -7.01184511e-01 3.63224924e-01 -8.26982439e-01
-3.87750268e-02 6.40474632e-02 -1.53784335e+00 -1.73746005e-01
-2.46116281e-01 4.46766287e-01 4.44445938e-01 9.26928520e-01
5.20916164e-01 7.47394860e-01 2.68429309e-01 1.59316644e-01
-7.67640054e-01 -6.82267249e-01 -2.70053446e-01 5.51777244e-01
1.23657454e-02 -5.08983254e-01 -8.70230645e-02 -2.47118562e-01] | [9.279297828674316, 7.7220025062561035] |
aa5b485d-a5cb-44a4-a775-5cf036be6810 | transforming-the-interactive-segmentation-for | 2208.09592 | null | https://arxiv.org/abs/2208.09592v2 | https://arxiv.org/pdf/2208.09592v2.pdf | Transforming the Interactive Segmentation for Medical Imaging | The goal of this paper is to interactively refine the automatic segmentation on challenging structures that fall behind human performance, either due to the scarcity of available annotations or the difficulty nature of the problem itself, for example, on segmenting cancer or small organs. Specifically, we propose a novel Transformer-based architecture for Interactive Segmentation (TIS), that treats the refinement task as a procedure for grouping pixels with similar features to those clicks given by the end users. Our proposed architecture is composed of Transformer Decoder variants, which naturally fulfills feature comparison with the attention mechanisms. In contrast to existing approaches, our proposed TIS is not limited to binary segmentations, and allows the user to edit masks for arbitrary number of categories. To validate the proposed approach, we conduct extensive experiments on three challenging datasets and demonstrate superior performance over the existing state-of-the-art methods. The project page is: https://wtliu7.github.io/tis/. | ['Ya zhang', 'Weidi Xie', 'Yuhuan Yang', 'Chaofan Ma', 'Wentao Liu'] | 2022-08-20 | null | null | null | null | ['interactive-segmentation'] | ['computer-vision'] | [ 5.24813592e-01 5.35776556e-01 -4.13087346e-02 -4.06183034e-01
-7.44077384e-01 -5.34865499e-01 3.83818179e-01 5.10755833e-03
-4.64146137e-01 5.12426078e-01 2.88583376e-02 -4.80513901e-01
2.99717486e-01 -3.67248535e-01 -5.99429607e-01 -5.05123317e-01
2.22846359e-01 5.35560489e-01 7.15164483e-01 -5.06251678e-02
2.04002827e-01 9.32537988e-02 -1.21791780e+00 2.55674660e-01
1.07656181e+00 9.53681529e-01 4.57160026e-01 5.37467301e-01
-1.59733042e-01 3.19870591e-01 -3.80886376e-01 -6.27396226e-01
1.54294893e-01 -4.38843280e-01 -1.04734910e+00 2.20926523e-01
3.99624407e-01 -2.47492418e-01 -9.90948007e-02 1.23917389e+00
2.07618162e-01 -1.71912894e-01 6.38024807e-01 -1.03743196e+00
-5.98098040e-01 7.31710613e-01 -6.25557363e-01 2.52829075e-01
1.98625177e-01 8.23928714e-02 1.09090793e+00 -8.26570988e-01
5.85058331e-01 1.05727422e+00 5.17268598e-01 7.88377464e-01
-1.11163747e+00 -5.00142097e-01 4.72482681e-01 7.98926875e-02
-1.39384925e+00 -3.47366869e-01 4.16076571e-01 -5.83536327e-01
4.51164007e-01 4.27750260e-01 6.19651020e-01 1.01048315e+00
-3.01918457e-03 1.10414028e+00 9.21879232e-01 -1.52622178e-01
7.42807761e-02 1.11379549e-01 2.74643064e-01 7.72382915e-01
1.12305351e-01 -3.38898003e-01 -2.50259697e-01 1.38870955e-01
7.68005908e-01 -1.02169842e-01 -4.65051532e-01 -5.02948225e-01
-1.41315722e+00 5.60124815e-01 6.36955261e-01 2.89848089e-01
-1.95992872e-01 7.94391781e-02 2.83651263e-01 -1.25398889e-01
3.87214631e-01 3.71470094e-01 -3.96074712e-01 9.12473649e-02
-1.18155634e+00 -1.09176785e-02 5.89706361e-01 1.22710538e+00
5.34336507e-01 -3.29414725e-01 -5.72431982e-01 5.22661865e-01
4.72852737e-01 -6.98752776e-02 3.86410683e-01 -6.30940795e-01
3.59247714e-01 7.21768916e-01 2.40677670e-02 -4.42686737e-01
-2.78459519e-01 -7.14741051e-01 -8.64571869e-01 -5.06504998e-02
5.48530340e-01 -8.83328095e-02 -1.33719337e+00 1.57233465e+00
4.90347445e-01 1.19645335e-01 -2.93508768e-01 9.99857545e-01
1.05630302e+00 1.91732302e-01 2.53574342e-01 7.60434568e-02
1.66460502e+00 -1.29272473e+00 -6.96274042e-01 -2.18205273e-01
4.77528870e-01 -7.38073826e-01 1.25891757e+00 3.20244282e-01
-1.03800261e+00 -5.28999627e-01 -8.06662560e-01 -2.81361341e-01
-2.24113271e-01 3.72519135e-01 5.86037636e-01 5.88186979e-01
-1.09381628e+00 4.20246482e-01 -1.08063972e+00 -5.19978881e-01
7.49394298e-01 3.87434006e-01 1.01659775e-01 2.81598061e-01
-8.68738532e-01 5.36469221e-01 4.25995171e-01 2.92116582e-01
-9.68503475e-01 -6.55982018e-01 -6.59554303e-01 2.36493185e-01
5.43948591e-01 -7.59724140e-01 1.56014097e+00 -1.08702946e+00
-1.34835041e+00 1.06952250e+00 -2.35462636e-01 -3.56709778e-01
8.83294761e-01 -4.05340075e-01 1.43358305e-01 4.56548706e-02
1.85827523e-01 1.00712931e+00 6.59111500e-01 -1.14369166e+00
-7.66884387e-01 -2.67826855e-01 2.37734109e-01 2.65220940e-01
-3.82966340e-01 -1.03305250e-01 -1.17457902e+00 -7.27426827e-01
2.24187896e-01 -9.99676287e-01 -6.02236271e-01 1.39023155e-01
-9.16429698e-01 -2.45479718e-01 6.73566520e-01 -5.88889301e-01
1.26509082e+00 -2.18883038e+00 3.43594760e-01 3.07458658e-02
3.87349188e-01 1.36134923e-01 1.69763580e-01 7.97811225e-02
6.70404211e-02 4.15894747e-01 -6.80895448e-01 -5.03000796e-01
3.12194098e-02 -2.71749552e-02 6.73258910e-03 3.16245496e-01
1.17864348e-01 9.58892226e-01 -8.20524514e-01 -9.14074183e-01
-5.62170371e-02 2.18636185e-01 -4.31085080e-01 1.96886256e-01
-4.28944886e-01 8.53810370e-01 -7.67780364e-01 8.29041600e-01
6.61091924e-01 -6.04965746e-01 1.13528751e-01 -3.35657507e-01
-1.63442373e-01 3.16619754e-01 -9.73251283e-01 2.02284360e+00
4.45056297e-02 3.07395250e-01 2.97910452e-01 -8.39398444e-01
3.86187911e-01 3.21816772e-01 3.91658694e-01 -4.45623130e-01
2.70989656e-01 1.64490059e-01 9.49223191e-02 -3.17969799e-01
5.26106238e-01 2.45786265e-01 -1.07432336e-01 2.13631272e-01
8.46048594e-02 5.31817460e-03 4.91168052e-01 3.34941149e-01
1.09460843e+00 3.54738861e-01 3.95854652e-01 -3.72386038e-01
6.37459993e-01 9.58215296e-02 7.55005479e-01 6.82433844e-01
-3.69962662e-01 7.39415526e-01 5.22251248e-01 -6.67822883e-02
-8.63876641e-01 -1.01496947e+00 -2.95775980e-01 1.06035256e+00
5.59942245e-01 -5.16832173e-01 -1.10313010e+00 -1.10715806e+00
-3.97641838e-01 4.33497846e-01 -7.32858300e-01 2.88992971e-01
-4.48834062e-01 -7.30483055e-01 5.21572053e-01 5.32558262e-01
6.39783919e-01 -1.07290637e+00 -7.19759405e-01 1.12004010e-02
-3.73391598e-01 -1.07885480e+00 -7.65039384e-01 1.47397950e-01
-8.52502942e-01 -1.02207744e+00 -9.03806388e-01 -9.79025185e-01
9.58427429e-01 8.82463530e-02 1.10640121e+00 2.79556274e-01
-3.57657909e-01 1.06142573e-01 -4.29688781e-01 -2.75498420e-01
-1.37678027e-01 6.00677192e-01 -5.77997386e-01 3.14486325e-02
2.82506775e-02 -2.08030581e-01 -9.26433563e-01 5.54754555e-01
-9.20053124e-01 6.12513006e-01 8.26727390e-01 6.64082348e-01
9.03145015e-01 -2.68468827e-01 3.63010496e-01 -1.50213122e+00
3.50692451e-01 -3.40923727e-01 -5.25833845e-01 2.49731138e-01
-4.05453265e-01 -1.33559220e-02 3.94022286e-01 -3.18977416e-01
-1.08506680e+00 4.40271497e-01 -3.53284419e-01 -1.63913280e-01
-3.59255254e-01 3.31278324e-01 -1.70427263e-01 -2.15464272e-02
3.38115841e-01 2.04793468e-01 -3.27331543e-01 -6.30867839e-01
2.81534404e-01 5.84466100e-01 5.87670982e-01 -3.32125753e-01
7.66520262e-01 3.57530177e-01 -3.99537802e-01 -4.91903156e-01
-8.28954577e-01 -5.63469887e-01 -8.24360609e-01 -1.84810698e-01
1.13820338e+00 -7.36240804e-01 -5.09667277e-01 3.53686035e-01
-1.01319873e+00 -4.68231469e-01 -1.49432123e-01 1.54728115e-01
-5.02782404e-01 4.17899340e-01 -7.59512067e-01 -4.90394443e-01
-4.60278809e-01 -1.40467072e+00 1.20918512e+00 4.70987290e-01
-2.42884517e-01 -7.14620948e-01 -2.84325510e-01 5.47963917e-01
3.25670242e-01 1.90542072e-01 8.65361571e-01 -7.02610612e-01
-9.04472172e-01 3.75971906e-02 -3.27082396e-01 -8.61268258e-04
6.63984865e-02 -7.98181146e-02 -9.66945231e-01 -3.66518766e-01
-2.13498518e-01 -1.93360537e-01 9.61590052e-01 3.23513955e-01
1.64898968e+00 -2.40832251e-02 -7.63037860e-01 7.27199376e-01
1.15700412e+00 1.96276531e-02 6.16523921e-01 2.53132582e-01
7.46259749e-01 4.28352237e-01 8.18709731e-01 2.01315194e-01
3.16210717e-01 6.62189722e-01 6.44498944e-01 -5.54627776e-01
-1.21056631e-01 -1.04363561e-01 -1.04594037e-01 6.81733131e-01
2.19506342e-02 -3.49505901e-01 -8.49208832e-01 7.37138629e-01
-1.85428977e+00 -5.55346370e-01 -3.72696519e-01 2.08973885e+00
8.48390162e-01 3.63284856e-01 9.79517251e-02 -1.16136476e-01
8.14629853e-01 6.47759661e-02 -6.11790657e-01 -1.39681756e-01
3.88383210e-01 1.47191092e-01 6.00900352e-01 2.80468345e-01
-1.37050307e+00 1.16119647e+00 5.78371811e+00 1.06067038e+00
-9.54948723e-01 2.53741413e-01 8.81054640e-01 1.74476057e-01
-2.64229804e-01 -8.38100836e-02 -7.75652289e-01 3.79816860e-01
3.77922982e-01 5.00577018e-02 5.13805076e-02 5.78145862e-01
1.38297066e-01 -3.21553648e-02 -1.32605231e+00 6.41719997e-01
-6.01351112e-02 -1.09919477e+00 -7.93562457e-02 -6.88281506e-02
5.08463025e-01 1.50245065e-02 2.02682853e-01 2.19143599e-01
3.19998384e-01 -9.72883165e-01 8.84495795e-01 2.76553214e-01
9.04404283e-01 -1.66537687e-01 5.38465023e-01 3.66333634e-01
-1.28359306e+00 1.30045637e-01 -3.00499275e-02 4.09773737e-01
2.82872677e-01 3.91635656e-01 -1.01644289e+00 4.56653297e-01
8.46830308e-01 6.16443336e-01 -7.75349855e-01 1.41830659e+00
-3.39848846e-01 7.49079466e-01 -2.94223756e-01 1.01430118e-01
2.60461062e-01 -3.27100046e-02 4.48127478e-01 1.50544429e+00
2.12650806e-01 4.97508980e-02 1.72752589e-01 9.97933030e-01
-2.08456293e-01 2.71780312e-01 -2.37105742e-01 1.08615056e-01
2.62862265e-01 1.60123968e+00 -1.26544297e+00 -4.05123025e-01
-3.83888155e-01 1.25517559e+00 3.55904818e-01 3.29165250e-01
-1.07073522e+00 -3.25009614e-01 2.90046781e-01 1.42001420e-01
5.79929173e-01 6.33553863e-02 -4.79539365e-01 -9.84628320e-01
3.22842412e-02 -8.20928633e-01 4.71607476e-01 -5.35259426e-01
-9.51994300e-01 7.71595061e-01 -2.31126398e-01 -1.26851547e+00
2.35116854e-01 -3.39026541e-01 -6.34542584e-01 5.99267364e-01
-1.40070534e+00 -1.22706282e+00 -4.86810565e-01 4.12369311e-01
9.46033895e-01 3.93028647e-01 5.33474982e-01 5.31719387e-01
-6.21035993e-01 7.09378123e-01 -2.10267782e-01 2.00208843e-01
5.23695171e-01 -1.48994279e+00 6.42850637e-01 9.39239502e-01
8.80396888e-02 3.86479735e-01 6.19348824e-01 -6.67320549e-01
-9.50226426e-01 -1.16016936e+00 7.93378413e-01 -2.62406409e-01
4.31610972e-01 -6.28213942e-01 -7.99646974e-01 7.48123348e-01
3.75222445e-01 -4.48946953e-02 5.95263958e-01 -6.43171370e-02
5.20284697e-02 1.81416616e-01 -1.10699046e+00 7.61340082e-01
1.23189747e+00 -7.64615163e-02 -3.54383349e-01 3.60769868e-01
7.69672155e-01 -7.88180470e-01 -6.65421784e-01 4.88854378e-01
3.91583025e-01 -7.84788787e-01 8.11874926e-01 -3.99353623e-01
4.43018556e-01 -5.73027551e-01 2.28057086e-01 -1.00665140e+00
-4.01785403e-01 -5.80235779e-01 2.44301409e-01 1.28492939e+00
7.25486755e-01 -2.36769035e-01 9.83592749e-01 5.09131968e-01
-5.59783757e-01 -9.08534288e-01 -9.07032847e-01 -2.24002466e-01
-1.71813563e-01 -1.74541116e-01 4.09774512e-01 6.44748092e-01
-1.44600406e-01 3.89609188e-01 -1.57446429e-01 3.23706388e-01
5.95769465e-01 1.64223671e-01 5.90777516e-01 -1.09180653e+00
-4.42662001e-01 -5.16138732e-01 -1.25412956e-01 -1.52709436e+00
-2.47104019e-01 -1.02009010e+00 3.46437693e-01 -1.68947232e+00
3.12373519e-01 -5.87859333e-01 -2.68120170e-01 6.49516761e-01
-3.45334500e-01 5.29831111e-01 2.19491810e-01 2.00390965e-01
-1.01719105e+00 5.05875587e-01 1.60972965e+00 -2.61588067e-01
-1.32864177e-01 2.66831785e-01 -8.97478580e-01 8.23010802e-01
6.36072993e-01 -4.52930868e-01 -3.09201598e-01 -6.32438242e-01
-2.90626548e-02 8.46325159e-02 3.21527690e-01 -1.00664663e+00
4.07677501e-01 9.41602979e-03 1.39584735e-01 -7.53529131e-01
1.51924983e-01 -7.64242828e-01 7.65288621e-02 5.27093709e-01
-5.10817170e-01 -1.34192273e-01 3.13651934e-02 5.22131801e-01
-1.58687025e-01 -2.05148309e-01 7.92478740e-01 -2.40771934e-01
-6.40053093e-01 4.27425146e-01 -3.97694796e-01 1.63567021e-01
1.03634143e+00 -2.84224302e-01 -2.23451599e-01 4.99223685e-03
-1.01152539e+00 6.18321002e-01 4.08746451e-01 1.96216330e-01
4.52313066e-01 -9.95027363e-01 -5.78133881e-01 -3.14809531e-02
1.70740753e-01 4.34510767e-01 1.00845456e-01 1.19042754e+00
-6.05082870e-01 3.31849933e-01 -5.32930903e-02 -7.00485051e-01
-1.42113197e+00 4.08831626e-01 4.50862378e-01 -3.76451582e-01
-8.55944395e-01 1.11458588e+00 7.57754803e-01 -2.81913817e-01
4.48386431e-01 -5.87435961e-01 -3.99530649e-01 -7.51525089e-02
1.54696852e-01 5.96526917e-03 4.36086990e-02 -4.33941424e-01
-3.57485712e-01 4.18962985e-01 -4.69158649e-01 9.91851538e-02
1.05430758e+00 -3.34150672e-01 1.19361028e-01 2.74751693e-01
6.68169498e-01 -2.71409214e-01 -1.27824593e+00 -2.44917378e-01
1.02584429e-01 -3.02352160e-01 -1.39231682e-01 -8.97053003e-01
-1.26804090e+00 7.81953096e-01 6.75382972e-01 5.74785173e-02
1.12834203e+00 1.59200087e-01 8.46726775e-01 5.91832958e-02
2.13351458e-01 -9.50740397e-01 -1.23760998e-01 3.32721412e-01
5.97558022e-01 -1.12714767e+00 -2.31364816e-01 -9.20406878e-01
-6.75886631e-01 7.65925527e-01 1.01155567e+00 6.55549318e-02
5.41222334e-01 2.44893476e-01 9.10818800e-02 -2.90124685e-01
-6.80884242e-01 -4.62924957e-01 3.25588763e-01 3.58671159e-01
7.69354045e-01 9.75310430e-03 -7.56551445e-01 6.97769761e-01
-1.10838950e-01 7.81927537e-03 3.81786764e-01 7.21617877e-01
-3.82723451e-01 -1.07411182e+00 -1.84809119e-01 6.28293633e-01
-6.78991795e-01 -1.51438832e-01 -4.11626548e-01 6.49260461e-01
2.05505669e-01 6.82230294e-01 4.28783558e-02 -1.68029889e-01
2.32287169e-01 -1.20851696e-01 1.84704453e-01 -8.68695796e-01
-7.57375777e-01 3.94217819e-01 4.65827510e-02 -5.52940786e-01
-2.42136434e-01 -6.71982110e-01 -1.35782218e+00 2.58061677e-01
-2.52072752e-01 1.81805670e-01 3.58803928e-01 8.39186132e-01
1.36390239e-01 9.64414954e-01 2.33789697e-01 -7.69854903e-01
-2.65137345e-01 -1.04528296e+00 -3.71181339e-01 4.60085064e-01
1.47924513e-01 -6.49868548e-01 -3.88240837e-03 2.88601011e-01] | [14.391654014587402, -1.9858976602554321] |
c44eea4d-8f4f-42a2-9300-ebd0adc5d3b8 | towards-multi-spatiotemporal-scale | 2209.15616 | null | https://arxiv.org/abs/2209.15616v2 | https://arxiv.org/pdf/2209.15616v2.pdf | Towards Multi-spatiotemporal-scale Generalized PDE Modeling | Partial differential equations (PDEs) are central to describing complex physical system simulations. Their expensive solution techniques have led to an increased interest in deep neural network based surrogates. However, the practical utility of training such surrogates is contingent on their ability to model complex multi-scale spatio-temporal phenomena. Various neural network architectures have been proposed to target such phenomena, most notably Fourier Neural Operators (FNOs), which give a natural handle over local & global spatial information via parameterization of different Fourier modes, and U-Nets which treat local and global information via downsampling and upsampling paths. However, generalizing across different equation parameters or time-scales still remains a challenge. In this work, we make a comprehensive comparison between various FNO, ResNet, and U-Net like approaches to fluid mechanics problems in both vorticity-stream and velocity function form. For U-Nets, we transfer recent architectural improvements from computer vision, most notably from object segmentation and generative modeling. We further analyze the design considerations for using FNO layers to improve performance of U-Net architectures without major degradation of computational cost. Finally, we show promising results on generalization to different PDE parameters and time-scales with a single surrogate model. Source code for our PyTorch benchmark framework is available at https://github.com/microsoft/pdearena. | ['Johannes Brandstetter', 'Jayesh K. Gupta'] | 2022-09-30 | null | null | null | null | ['pde-surrogate-modeling'] | ['miscellaneous'] | [-2.13711441e-01 -4.15966421e-01 3.07118177e-01 -5.50543293e-02
-2.93362349e-01 -3.73359472e-01 8.14270020e-01 -3.19343656e-01
-2.30601326e-01 9.19789910e-01 7.33693466e-02 -3.73701125e-01
-1.74575418e-01 -1.05742812e+00 -6.71511114e-01 -7.57273197e-01
-3.94316405e-01 2.88917691e-01 2.15101957e-01 -3.80224198e-01
1.02440931e-01 8.80101562e-01 -1.19087029e+00 -1.30040413e-02
8.65240932e-01 9.99368548e-01 -5.94759174e-02 1.05639577e+00
-2.61730939e-01 6.72708094e-01 -3.72987509e-01 1.71962664e-01
3.56470913e-01 -5.01925230e-01 -8.25350404e-01 -2.98332363e-01
3.69598180e-01 -5.10679066e-01 -6.05514228e-01 7.28392303e-01
6.29191399e-01 6.49514318e-01 8.33288968e-01 -1.17515278e+00
-8.15796018e-01 3.56440425e-01 -3.08933020e-01 7.16321886e-01
-1.58356279e-01 5.29357612e-01 6.93895876e-01 -8.44781578e-01
6.25121653e-01 1.37043834e+00 1.15565670e+00 7.17466474e-01
-1.19743454e+00 -2.87413329e-01 -1.56906515e-01 -2.22232968e-01
-1.05543303e+00 -2.36665234e-01 6.83125436e-01 -5.28945744e-01
1.19613516e+00 4.23722446e-01 7.29895592e-01 9.75826323e-01
4.19655293e-01 6.86534822e-01 1.02106380e+00 8.64108130e-02
2.57688522e-01 -2.11878881e-01 1.14461519e-01 4.18406039e-01
9.10502672e-02 2.49237120e-01 -3.02145660e-01 -2.40296915e-01
1.49296117e+00 -6.69767633e-02 -4.10344452e-01 -1.32651731e-01
-9.12859082e-01 9.26235199e-01 7.34808803e-01 4.23827738e-01
-3.66950452e-01 6.31329238e-01 5.27925491e-01 5.11660948e-02
8.01401317e-01 6.92018747e-01 -3.79586250e-01 -2.97085971e-01
-1.14775479e+00 7.74052620e-01 9.66375828e-01 7.29534507e-01
7.34252274e-01 7.26220369e-01 4.45253551e-02 7.68701673e-01
1.44580051e-01 4.31263626e-01 4.74553674e-01 -1.34720707e+00
-4.51622419e-02 2.13040203e-01 4.72978614e-02 -8.58172596e-01
-5.49738646e-01 -4.88757223e-01 -1.33936906e+00 2.99500734e-01
2.62486994e-01 -5.13744771e-01 -1.10260749e+00 1.42028701e+00
3.63597304e-01 5.78267515e-01 -2.57015023e-02 1.03611898e+00
1.00150311e+00 9.83080804e-01 3.50337289e-03 1.43057570e-01
8.85649502e-01 -9.20298874e-01 -4.31377888e-01 8.70857537e-02
6.60469294e-01 -5.94407916e-01 7.65378356e-01 -1.19764790e-01
-1.16723812e+00 -5.62814832e-01 -7.96919346e-01 -3.14111203e-01
-5.24723768e-01 -3.16453844e-01 7.92479098e-01 2.69721448e-01
-1.51297247e+00 1.43455625e+00 -1.25856686e+00 -2.27391526e-01
4.30348247e-01 3.41466218e-01 2.22042963e-01 4.77947384e-01
-1.35868216e+00 5.46132922e-01 9.20536965e-02 3.66479158e-01
-6.51743174e-01 -1.31788015e+00 -7.89217174e-01 7.63892159e-02
-3.33568245e-01 -1.08162379e+00 1.21195126e+00 -5.66820264e-01
-1.60226440e+00 4.09523129e-01 -1.03894249e-01 -7.48677850e-01
8.02928984e-01 -2.48101339e-01 -2.14704573e-01 1.83677062e-01
5.31779192e-02 7.63381243e-01 6.09091103e-01 -1.17311347e+00
-3.74791384e-01 2.09063023e-01 3.76233459e-02 1.42027587e-01
-6.99230190e-03 -1.19522929e-01 5.64078726e-02 -7.84882188e-01
-1.31910652e-01 -8.05306494e-01 -5.37973881e-01 1.60512999e-01
-1.85077175e-01 -1.07585587e-01 1.03726411e+00 -5.67103982e-01
1.04230952e+00 -1.74618435e+00 9.75666195e-02 -1.61959812e-01
3.54583949e-01 5.02015352e-01 1.21842399e-01 3.91808867e-01
-3.91904414e-02 3.57581228e-01 -6.18393421e-01 -4.36984926e-01
-2.33738571e-02 3.51770848e-01 -6.51585400e-01 5.44099271e-01
4.12268341e-01 1.17896235e+00 -7.96342432e-01 -9.40379500e-02
4.99555767e-01 8.98424327e-01 -7.50846446e-01 4.18240130e-02
-3.27849895e-01 7.74963021e-01 -4.36885089e-01 2.61110872e-01
9.70235705e-01 -4.69872057e-01 -4.12583441e-01 1.15078665e-01
-4.50570077e-01 2.63185918e-01 -1.13196802e+00 1.28447735e+00
-6.22365475e-01 8.72315645e-01 4.91737992e-01 -9.89187360e-01
6.31200492e-01 3.84826988e-01 7.97861218e-01 -3.80252481e-01
2.44743645e-01 5.28940797e-01 2.29997952e-02 -3.53585094e-01
7.33382940e-01 -3.02215099e-01 2.82764167e-01 3.48186880e-01
6.97064623e-02 -3.57514024e-01 3.42437029e-01 1.07873373e-01
9.37442720e-01 4.66093630e-01 -4.19406630e-02 -7.57412434e-01
3.12491953e-01 2.24024817e-01 2.57737428e-01 7.18981266e-01
-1.68283597e-01 8.08946908e-01 3.39657813e-01 -8.41914237e-01
-1.22186840e+00 -8.84416640e-01 -5.32464683e-01 7.52224147e-01
3.05326632e-03 -1.69132590e-01 -7.15037107e-01 9.26506072e-02
2.68388957e-01 4.17278349e-01 -7.87575901e-01 1.23531306e-02
-1.21878469e+00 -1.01420343e+00 7.49079585e-01 7.55729079e-01
8.04374754e-01 -1.34377217e+00 -7.86313951e-01 4.70216393e-01
3.48164111e-01 -9.46007967e-01 -2.14000583e-01 1.35446951e-01
-1.12415266e+00 -7.50716925e-01 -1.13814175e+00 -5.31960189e-01
2.89438874e-01 5.83735391e-06 1.08340991e+00 5.03010757e-04
-2.47912571e-01 3.79787445e-01 1.97697729e-02 -2.00352035e-02
-4.60552394e-01 2.48478368e-01 1.11162275e-01 -2.72012264e-01
-1.47253528e-01 -1.01131952e+00 -8.76013458e-01 6.33289665e-02
-1.01109445e+00 -7.32645486e-03 -5.30411443e-03 6.39619291e-01
3.09384286e-01 -3.66056234e-01 2.93308109e-01 -8.39852333e-01
8.82434845e-01 -7.01993942e-01 -4.06383514e-01 -3.89005363e-01
-2.10478187e-01 -1.89700071e-02 9.40999091e-01 -4.18574661e-01
-1.11521220e+00 -5.12182176e-01 -4.17879671e-01 -8.11365604e-01
-2.20479965e-01 4.74193990e-01 5.99408209e-01 -1.38809502e-01
7.67924130e-01 2.18800932e-01 -2.21278705e-02 -4.65772539e-01
2.28969961e-01 2.32874453e-01 4.90095109e-01 -9.56056118e-01
6.34686410e-01 7.36131191e-01 2.60936648e-01 -1.25168729e+00
-4.78056520e-01 -3.78934860e-01 -6.63702846e-01 -8.52637589e-02
8.23974192e-01 -6.29142404e-01 -5.93613565e-01 8.05783093e-01
-1.27234852e+00 -8.32142055e-01 -5.92350245e-01 3.04726154e-01
-6.21237338e-01 5.84872346e-03 -1.18153715e+00 -6.74622715e-01
-4.32016820e-01 -1.10435808e+00 9.87285793e-01 4.97605085e-01
-3.48860979e-01 -1.69584537e+00 1.55766413e-01 -3.44563514e-01
1.01797891e+00 7.39640534e-01 6.01747513e-01 -1.79628417e-01
-6.23393655e-01 3.06026787e-01 -3.47827584e-01 1.90406680e-01
5.97610734e-02 3.32325220e-01 -1.02559578e+00 -1.57179520e-01
2.87544042e-01 -3.24502923e-02 1.13619673e+00 1.05275440e+00
1.02570629e+00 -2.41039589e-01 -2.93323517e-01 1.24352944e+00
1.34491587e+00 3.52324694e-02 3.10345739e-01 3.47873271e-02
1.06467533e+00 3.26766104e-01 -1.30552068e-01 3.65934819e-01
4.60562371e-02 2.01675832e-01 1.57992005e-01 -3.89129251e-01
-2.95662582e-01 1.63490847e-01 1.06323972e-01 8.65194261e-01
-4.44806933e-01 -2.89741457e-01 -1.07321131e+00 6.62731349e-01
-1.81533849e+00 -8.08014691e-01 -5.70484102e-01 1.48895597e+00
6.69223189e-01 -4.18278389e-02 1.18559904e-01 -1.97093189e-01
5.91746271e-01 4.53238904e-01 -7.59723425e-01 -5.32310784e-01
-4.39261608e-02 4.50784594e-01 5.87304235e-01 6.82063401e-01
-1.22961235e+00 8.86999309e-01 6.38073874e+00 7.73623168e-01
-1.64530456e+00 1.93042517e-01 8.17375660e-01 -1.19902179e-01
-1.83607951e-01 -2.45316267e-01 -7.06571043e-01 3.11810195e-01
1.17181373e+00 -1.34636804e-01 3.81331831e-01 5.56451440e-01
5.05169094e-01 2.25556805e-03 -8.89474869e-01 6.76066756e-01
-5.04351735e-01 -1.90025210e+00 5.77212609e-02 1.63110718e-01
1.03963268e+00 5.22579134e-01 3.06070223e-02 2.19864473e-01
4.38999981e-01 -1.17884684e+00 5.78792214e-01 4.33920324e-01
7.86199093e-01 -4.27019775e-01 3.69367689e-01 1.93241984e-01
-1.44120646e+00 4.19913858e-01 -3.95307571e-01 -3.31693560e-01
5.17169595e-01 4.77709293e-01 -4.07305628e-01 4.06107426e-01
8.31307709e-01 9.21878338e-01 5.12792282e-02 9.42679584e-01
3.30416024e-01 6.98436022e-01 -7.11961329e-01 -5.90862706e-02
6.50769591e-01 -4.10959691e-01 7.75578976e-01 1.31653690e+00
4.97116297e-01 1.36994347e-01 3.51576097e-02 1.38023877e+00
8.63389447e-02 -2.54909605e-01 -6.00620687e-01 2.41236147e-02
6.67584464e-02 1.13125205e+00 -1.02669704e+00 -3.66162330e-01
-1.58714622e-01 6.23267055e-01 1.58676207e-01 7.03946710e-01
-9.14659023e-01 -1.70176774e-01 1.13292313e+00 4.14168984e-01
1.96136117e-01 -5.80910861e-01 -4.80128676e-01 -1.08228743e+00
-2.63257891e-01 -1.62883073e-01 1.08920440e-01 -5.89988470e-01
-1.24744689e+00 6.91384196e-01 2.17409059e-01 -1.02970767e+00
-2.72954822e-01 -9.43385124e-01 -9.55869377e-01 1.23232913e+00
-1.52596986e+00 -8.94364178e-01 -2.18944207e-01 2.99943626e-01
5.02143800e-01 1.58185422e-01 5.32891750e-01 2.46867225e-01
-4.50184882e-01 -1.47467449e-01 2.89195180e-01 2.31777847e-01
1.87278941e-01 -1.43375361e+00 1.01462424e+00 8.32865775e-01
-4.18099821e-01 8.17030489e-01 9.49531019e-01 -7.44221270e-01
-1.12978315e+00 -1.19945943e+00 2.88088024e-01 -2.80975938e-01
8.55183125e-01 -2.45379642e-01 -1.28465426e+00 5.42619884e-01
2.92039514e-01 5.27142942e-01 8.27866271e-02 -3.11253399e-01
2.23215312e-01 3.86517376e-01 -1.04496622e+00 6.14466906e-01
1.01547849e+00 -3.32308859e-01 3.77809890e-02 2.38270536e-01
6.27269030e-01 -8.17724884e-01 -1.02051008e+00 4.73968446e-01
2.55481243e-01 -1.11765158e+00 1.01751602e+00 -5.30504525e-01
9.17727053e-01 -5.88620976e-02 2.33216807e-01 -1.46330392e+00
-3.12626719e-01 -9.02436495e-01 -4.06935722e-01 8.86529684e-01
7.97455534e-02 -9.48273063e-01 7.71341443e-01 6.93863750e-01
-4.54462439e-01 -1.19354117e+00 -9.84429181e-01 -5.92581332e-01
7.52863407e-01 -4.46165890e-01 4.68875289e-01 1.10293174e+00
-4.83316332e-01 -1.02004800e-02 -3.15150470e-01 -5.53186461e-02
4.11653399e-01 1.28108934e-01 5.42130768e-01 -1.21299064e+00
-2.37504244e-01 -8.72777402e-01 -1.63637530e-02 -1.27675986e+00
9.13538709e-02 -8.22430134e-01 -6.21839501e-02 -1.69027245e+00
-4.32886839e-01 -4.88328844e-01 1.47914544e-01 -5.43673225e-02
3.97860520e-02 3.91918272e-01 -1.03317477e-01 3.77645493e-01
1.06897697e-01 6.33308411e-01 1.65223229e+00 2.22094998e-01
-4.52316076e-01 -5.74271977e-02 -1.74965039e-01 8.16493034e-01
1.11760557e+00 -1.40785828e-01 -2.67977983e-01 -5.32010913e-01
-7.28270635e-02 1.98086187e-01 6.97533429e-01 -1.19349694e+00
1.65354639e-01 -2.04798922e-01 3.76575142e-01 -3.32903475e-01
4.70237821e-01 -2.78106511e-01 5.72736152e-02 5.46520889e-01
-1.45754278e-01 1.81829646e-01 5.60288548e-01 2.49454379e-01
-1.44463316e-01 1.68923419e-02 9.89437819e-01 -4.75744903e-01
-5.97883046e-01 6.53261721e-01 -4.97063160e-01 2.27533564e-01
6.97885096e-01 -2.81586140e-01 -3.63569230e-01 -3.41032267e-01
-7.47460246e-01 1.03804208e-01 3.81320685e-01 1.73800111e-01
3.85131359e-01 -1.18546653e+00 -6.04749143e-01 2.03520760e-01
-7.53882170e-01 4.92304415e-01 5.18204868e-01 9.19340968e-01
-1.31756592e+00 4.18947041e-01 -3.02343726e-01 -6.92733347e-01
-5.59948146e-01 9.06096995e-02 9.41596687e-01 -3.17821234e-01
-8.76506567e-01 9.80305374e-01 4.49747384e-01 -6.06983662e-01
-3.19238842e-01 -7.33380198e-01 -3.02231009e-03 -1.34792373e-01
1.73503697e-01 7.67639577e-01 -1.00558400e-01 -7.23524570e-01
-1.20639175e-01 5.14357924e-01 3.84698421e-01 -1.01361774e-01
1.54896867e+00 5.11994809e-02 -1.59611315e-01 5.26489317e-01
1.35833073e+00 -3.24567020e-01 -1.63695025e+00 -7.48604089e-02
-3.66048694e-01 3.65559496e-02 2.22320974e-01 -2.25376502e-01
-1.30050826e+00 1.20675230e+00 1.13588102e-01 6.30426228e-01
6.79783165e-01 -2.73062319e-01 1.19316697e+00 -2.06624735e-02
1.49978967e-02 -7.85157859e-01 -3.27557534e-01 9.98470783e-01
8.70475292e-01 -8.37077439e-01 -7.49376416e-02 -4.15726662e-01
-1.58027977e-01 1.29962802e+00 6.59580350e-01 -8.59267354e-01
1.08552670e+00 6.58366263e-01 9.71248969e-02 -3.02792132e-01
-5.19735634e-01 8.66089538e-02 2.35383302e-01 2.16500133e-01
7.36368954e-01 -5.22794873e-02 -8.38461425e-03 1.26215741e-01
-3.52434069e-01 -2.68242136e-02 5.34205794e-01 8.93599868e-01
-1.74869597e-01 -7.22571135e-01 -2.79690266e-01 5.57411015e-01
-4.48226959e-01 -1.28629923e-01 4.48449813e-02 9.03300047e-01
-7.40129575e-02 2.97443151e-01 3.65110874e-01 1.09221883e-01
-6.05044961e-02 -9.99038480e-03 3.60341370e-01 -3.95144314e-01
-7.13329196e-01 1.51574463e-01 -1.74983919e-01 -4.08810407e-01
-5.89589596e-01 -7.07120597e-01 -1.54726112e+00 -7.05815971e-01
2.46061534e-01 2.41769794e-02 2.82933891e-01 9.36319232e-01
2.88942039e-01 1.00604916e+00 9.60198492e-02 -1.69902551e+00
-3.08366418e-01 -1.11432624e+00 -6.36629939e-01 2.60284036e-01
5.99021971e-01 -5.84941268e-01 -4.84953433e-01 6.65484145e-02] | [6.524408340454102, 3.3893496990203857] |
d45f35fd-d1eb-4fa6-988c-b6bcaa19590c | meta-evaluation-of-conversational-search | 2104.13453 | null | https://arxiv.org/abs/2104.13453v1 | https://arxiv.org/pdf/2104.13453v1.pdf | Meta-evaluation of Conversational Search Evaluation Metrics | Conversational search systems, such as Google Assistant and Microsoft Cortana, enable users to interact with search systems in multiple rounds through natural language dialogues. Evaluating such systems is very challenging given that any natural language responses could be generated, and users commonly interact for multiple semantically coherent rounds to accomplish a search task. Although prior studies proposed many evaluation metrics, the extent of how those measures effectively capture user preference remains to be investigated. In this paper, we systematically meta-evaluate a variety of conversational search metrics. We specifically study three perspectives on those metrics: (1) reliability: the ability to detect "actual" performance differences as opposed to those observed by chance; (2) fidelity: the ability to agree with ultimate user preference; and (3) intuitiveness: the ability to capture any property deemed important: adequacy, informativeness, and fluency in the context of conversational search. By conducting experiments on two test collections, we find that the performance of different metrics varies significantly across different scenarios whereas consistent with prior studies, existing metrics only achieve a weak correlation with ultimate user preference and satisfaction. METEOR is, comparatively speaking, the best existing single-turn metric considering all three perspectives. We also demonstrate that adapted session-based evaluation metrics can be used to measure multi-turn conversational search, achieving moderate concordance with user satisfaction. To our knowledge, our work establishes the most comprehensive meta-evaluation for conversational search to date. | ['Max L. Wilson', 'Ke Zhou', 'Zeyang Liu'] | 2021-04-27 | null | null | null | null | ['conversational-search'] | ['natural-language-processing'] | [-2.25900203e-01 1.25045300e-01 -4.48600322e-01 -3.42931777e-01
-9.10645783e-01 -1.01263976e+00 1.05376053e+00 2.86314070e-01
-6.99113011e-01 5.66896737e-01 6.39205515e-01 -5.54096699e-01
-3.72786045e-01 -2.74093211e-01 1.90518290e-01 1.85584590e-01
2.76826888e-01 7.33784080e-01 2.46922821e-02 -6.30937397e-01
6.93262637e-01 2.25896940e-01 -1.58530951e+00 3.57618809e-01
1.15378177e+00 7.11391687e-01 1.84693336e-01 7.70312548e-01
-2.50777453e-01 6.52099907e-01 -6.04108334e-01 -5.18794775e-01
-7.92055950e-02 -4.47774947e-01 -1.46158886e+00 -2.29434341e-01
5.91927581e-02 -3.54894221e-01 -7.26981610e-02 6.23064339e-01
5.84043920e-01 3.77543867e-01 4.65654016e-01 -1.11354446e+00
-7.10066319e-01 3.99852842e-01 2.26778731e-01 4.37428981e-01
1.29853475e+00 4.64583755e-01 1.38244891e+00 -7.23255515e-01
5.44043660e-01 1.19644690e+00 4.23055708e-01 4.27771389e-01
-1.03394687e+00 -2.86798239e-01 -6.60743341e-02 -3.95100042e-02
-1.12266707e+00 -6.23852432e-01 2.50401050e-01 -4.66737002e-01
1.09552324e+00 7.86535680e-01 5.84353805e-01 1.02170050e+00
-1.58301070e-01 5.98297358e-01 1.35276294e+00 -4.72593933e-01
1.27891049e-01 7.44208157e-01 4.16845918e-01 4.15094167e-01
-7.90019408e-02 -6.91986382e-02 -7.09898651e-01 -6.17953300e-01
1.73577607e-01 -2.60257930e-01 -4.62774038e-01 1.12478361e-02
-1.10952175e+00 7.53101707e-01 1.80166259e-01 5.84362507e-01
-3.85537267e-01 -3.86587322e-01 2.94338733e-01 5.63607991e-01
4.18107420e-01 1.14783645e+00 -4.18581843e-01 -1.05704224e+00
-6.99831426e-01 4.17129457e-01 1.64451516e+00 9.27705884e-01
5.08346915e-01 -7.05277503e-01 -5.43091238e-01 1.25343287e+00
1.89535752e-01 3.11169714e-01 8.74461651e-01 -9.81078982e-01
1.62439376e-01 7.91042447e-01 4.85549599e-01 -9.24804688e-01
-4.07260805e-01 -3.38530093e-01 -7.11319922e-03 -3.73945653e-01
4.64091271e-01 -3.45219672e-02 -2.04135060e-01 1.67473924e+00
-1.88117608e-01 -8.12554419e-01 -2.49044504e-02 1.04952466e+00
1.11396134e+00 1.93963870e-01 1.19131185e-01 -2.77099758e-01
1.61710179e+00 -1.00679564e+00 -6.51508152e-01 -4.37656641e-01
6.83841944e-01 -1.09222233e+00 1.85402095e+00 3.61379758e-02
-1.19317937e+00 -2.49906540e-01 -6.90156281e-01 -6.36962950e-02
-4.50375557e-01 -2.58005351e-01 7.18754768e-01 9.66595292e-01
-1.29415846e+00 1.13930210e-01 -1.85884640e-01 -8.71852815e-01
-5.10014236e-01 1.49888307e-01 1.69115774e-02 2.33804286e-01
-1.27869439e+00 1.44444883e+00 -2.03703597e-01 -2.61561334e-01
-3.53560776e-01 -3.60377669e-01 -3.07246566e-01 2.44260415e-01
2.84839302e-01 -8.24932694e-01 1.94763887e+00 -6.36724353e-01
-1.59340394e+00 9.36790109e-01 -5.02228498e-01 -3.38577852e-02
3.78526121e-01 7.85697624e-03 -3.89458537e-01 -1.44044265e-01
2.73758769e-01 4.05852705e-01 -1.09579407e-01 -1.09707510e+00
-6.15761161e-01 -2.23101974e-01 5.57989538e-01 9.31853294e-01
-3.43733758e-01 2.32511148e-01 -5.65113604e-01 -1.83202282e-01
-1.20288886e-01 -9.79457021e-01 2.07011253e-02 -4.80804324e-01
-9.77518037e-03 -6.83323383e-01 3.89620423e-01 -5.34837782e-01
1.63983810e+00 -1.66072834e+00 -1.59217164e-01 4.66393344e-02
2.41469905e-01 1.32868931e-01 -8.09703022e-02 9.04044986e-01
4.56763357e-01 5.73352158e-01 1.75377324e-01 -2.01592341e-01
2.80965745e-01 -1.78350568e-01 3.22661139e-02 1.70780141e-02
-4.40936297e-01 1.17139697e+00 -1.07966590e+00 -5.17589390e-01
6.13612942e-02 1.41518697e-01 -5.79815626e-01 1.75204396e-01
1.53067783e-02 7.42725283e-02 -3.70575249e-01 5.40756762e-01
-2.39446741e-02 -4.65910167e-01 1.14040039e-01 1.50551885e-01
-1.88152865e-01 9.36141193e-01 -6.06231630e-01 1.44131315e+00
-9.02215302e-01 7.45265603e-01 7.23178312e-02 -4.05054986e-01
6.74734294e-01 4.14161056e-01 2.58696556e-01 -1.01536977e+00
-1.08658947e-01 3.26483577e-01 4.82940152e-02 -5.26294231e-01
1.00109994e+00 1.60537377e-01 -2.52395868e-01 8.85714173e-01
-2.16334775e-01 -4.81578022e-01 1.06698066e-01 2.88677633e-01
1.15632272e+00 -5.21318257e-01 4.33717668e-01 -5.11414111e-01
4.52585250e-01 1.54962197e-01 -1.57249421e-01 1.25027931e+00
-3.57868552e-01 2.40734830e-01 2.69628823e-01 -7.90489651e-03
-6.18293762e-01 -8.40748727e-01 -6.05636137e-03 1.44893646e+00
4.60896134e-01 -5.90678751e-01 -7.05376327e-01 -6.09933674e-01
-2.73597986e-01 9.72356617e-01 -1.34153441e-01 -7.85012692e-02
6.70848191e-02 -3.56954873e-01 6.62312448e-01 -4.92232777e-02
4.61404622e-01 -1.17100132e+00 -6.70846820e-01 -1.96978939e-03
-7.66534865e-01 -1.03853047e+00 -8.63069296e-01 -2.00978160e-01
-5.32172441e-01 -1.01904345e+00 -4.02990907e-01 -5.82997680e-01
6.67917505e-02 6.73162580e-01 1.59322321e+00 4.16611761e-01
-3.30663659e-02 1.10011506e+00 -5.59022367e-01 -1.02119200e-01
-2.85896927e-01 2.26906672e-01 -6.36475310e-02 -7.35545933e-01
7.96681285e-01 -3.57281208e-01 -8.31719220e-01 7.37188697e-01
-6.48577690e-01 -2.82268763e-01 5.20883977e-01 7.70695388e-01
-3.11265826e-01 -3.67723644e-01 4.79734421e-01 -5.62715828e-01
1.97093916e+00 -6.16951346e-01 7.60418624e-02 4.36365426e-01
-1.29621768e+00 -1.45622000e-01 8.34826678e-02 -1.71298519e-01
-1.00522590e+00 -6.38312280e-01 8.12438503e-02 4.48788524e-01
-8.13691393e-02 8.07076812e-01 3.87527764e-01 -1.85809419e-01
9.13901269e-01 1.88667312e-01 1.43567309e-01 -1.63107917e-01
4.23714101e-01 1.14097154e+00 2.88940758e-01 -7.50936389e-01
2.56364167e-01 -6.07687868e-02 -8.40683401e-01 -9.76807475e-01
-4.71447021e-01 -9.75080669e-01 -4.67106141e-02 -4.03809488e-01
5.23641050e-01 -6.90570474e-01 -1.27449024e+00 -9.19740647e-02
-1.01469445e+00 -3.25399041e-01 5.73194697e-02 5.04658461e-01
-5.42634666e-01 6.00876510e-01 -5.13896227e-01 -1.13207936e+00
-5.17029643e-01 -1.38875389e+00 9.21778440e-01 2.84870893e-01
-1.10320306e+00 -1.06432152e+00 1.07063241e-01 8.76112878e-01
8.06053460e-01 -5.41501522e-01 6.19706035e-01 -9.80102122e-01
-1.25588194e-01 -3.37995827e-01 -1.83864698e-01 -6.63837343e-02
2.93172538e-01 -2.97597945e-01 -6.91976428e-01 -2.03153983e-01
2.07523286e-01 -4.60540771e-01 2.80180812e-01 8.22862610e-02
6.96871579e-01 -5.07408679e-01 -2.95353234e-01 -2.05261290e-01
8.86703610e-01 3.14423174e-01 4.29160446e-01 5.61416388e-01
-1.17522061e-01 8.05259347e-01 5.50777853e-01 3.04789454e-01
7.80128479e-01 1.16630900e+00 -1.12503596e-01 3.04047644e-01
8.05367604e-02 -2.24045277e-01 1.94746763e-01 6.00214422e-01
-1.02053583e-02 -5.84207892e-01 -9.56383884e-01 4.06434327e-01
-1.77386677e+00 -1.04116714e+00 2.67027199e-01 2.18221807e+00
7.26918101e-01 8.26494843e-02 3.39993745e-01 -3.25948685e-01
5.05025804e-01 1.48942932e-01 -2.54640996e-01 -6.81164205e-01
1.31475553e-01 5.98540939e-02 1.63920999e-01 9.96545434e-01
-3.44211340e-01 7.65713036e-01 7.42188025e+00 4.12723184e-01
-8.35936546e-01 1.56107888e-01 5.95522106e-01 -1.16613954e-01
-6.97221160e-01 2.81821728e-01 -4.49002802e-01 4.93324190e-01
7.81824172e-01 -6.56275511e-01 8.05438161e-01 8.39102924e-01
2.95636326e-01 -1.99107930e-01 -1.17402089e+00 9.32447910e-01
9.95243192e-02 -8.71898830e-01 -1.91351831e-01 1.16215669e-01
3.17000508e-01 -7.73235783e-02 3.94509397e-02 6.86588168e-01
5.07328391e-01 -1.11081135e+00 4.02460814e-01 4.51076329e-01
5.42326033e-01 -2.72626758e-01 6.25410318e-01 6.00370467e-01
-7.68189907e-01 8.31205398e-02 2.41155878e-01 -2.86321402e-01
2.44044468e-01 2.30882034e-01 -1.16812778e+00 1.45639420e-01
5.69171131e-01 -1.06950223e-01 -5.60038686e-01 8.81113529e-01
1.19809352e-01 3.65921259e-01 -3.14848602e-01 -7.75660157e-01
4.07231510e-01 -1.50092050e-01 7.43111551e-01 1.45627093e+00
2.16759384e-01 1.85709253e-01 1.01358570e-01 7.71280408e-01
3.76914404e-02 3.71245056e-01 -6.35603964e-01 -2.55749673e-01
1.02438939e+00 1.20475864e+00 -5.76984704e-01 -2.02380672e-01
-4.57666814e-01 1.03124511e+00 2.20173612e-01 2.79290020e-01
-3.26474398e-01 -1.69499815e-01 6.59681022e-01 5.59459301e-03
-4.27482933e-01 -2.22654954e-01 -5.21525502e-01 -9.99945760e-01
3.10206085e-01 -1.30025256e+00 4.14869606e-01 -7.44353652e-01
-1.16949975e+00 7.54751086e-01 -3.29948589e-02 -8.54792476e-01
-6.64819062e-01 -1.73893929e-01 -5.73652506e-01 1.15715981e+00
-9.52443838e-01 -5.94924688e-01 -6.51046276e-01 2.65807569e-01
7.93556869e-01 -1.32334501e-01 9.83744681e-01 1.02419436e-01
-1.09543890e-01 6.38411164e-01 -2.76147038e-01 -4.75803375e-01
9.33596313e-01 -1.21380711e+00 2.66566247e-01 2.42923066e-01
2.85722483e-02 1.20950711e+00 8.61498654e-01 -5.21879196e-01
-1.26275849e+00 9.09409970e-02 1.21296191e+00 -8.28592837e-01
4.71285611e-01 -8.51070955e-02 -5.20236373e-01 3.66735578e-01
4.67039317e-01 -8.21813941e-01 8.16901445e-01 8.46942365e-01
-1.06988639e-01 2.68579781e-01 -1.01128900e+00 8.24248552e-01
1.09836674e+00 -1.03222120e+00 -7.26959527e-01 4.56365228e-01
5.74710727e-01 -4.37281728e-01 -8.73583555e-01 2.39900514e-01
8.85991156e-01 -1.24715734e+00 9.75106597e-01 -4.15900886e-01
1.88666254e-01 1.61010846e-01 -1.44467235e-01 -1.47397149e+00
-2.35175908e-01 -7.86347389e-01 2.74206042e-01 1.08254564e+00
8.04437518e-01 -8.23209167e-01 5.69160461e-01 1.34351838e+00
-1.42342553e-01 -7.97546327e-01 -6.33030355e-01 -4.63191658e-01
-1.09522484e-01 -3.13778132e-01 6.87814951e-01 9.76898491e-01
8.24663758e-01 6.57303572e-01 -7.78257772e-02 -3.31649452e-01
-2.21503098e-02 -1.06076390e-01 8.55290711e-01 -1.18198383e+00
-2.28098691e-01 -1.20957899e+00 2.98504513e-02 -1.53121209e+00
3.33051048e-02 -7.15892076e-01 2.67717578e-02 -1.59201872e+00
3.61548036e-01 -4.90241379e-01 8.01547393e-02 5.56932390e-02
-3.32677066e-01 -2.12185368e-01 -3.04182824e-02 5.41425169e-01
-7.22601354e-01 3.58093888e-01 1.06171036e+00 1.52230293e-01
-5.75108409e-01 1.57135084e-01 -1.40543056e+00 3.22891295e-01
6.36466503e-01 1.78878084e-01 -7.45661378e-01 -2.66057342e-01
2.76714742e-01 4.29233462e-01 1.93780899e-01 -5.90527892e-01
5.98862052e-01 -3.02672625e-01 -2.27313578e-01 -4.47931141e-02
3.54825675e-01 -4.86615270e-01 -9.43856612e-02 2.02016875e-01
-7.31703579e-01 4.74987715e-01 7.95658678e-02 3.05986047e-01
-3.65073621e-01 -2.71994233e-01 1.57035798e-01 -2.77287513e-01
-5.93707323e-01 -1.67472437e-01 -6.85897887e-01 2.24220008e-01
5.40168762e-01 -3.21475744e-01 -4.86302286e-01 -1.18802762e+00
-3.19380492e-01 4.39214289e-01 5.29055297e-01 7.95975804e-01
2.26174444e-01 -1.08125710e+00 -4.77398068e-01 -3.30671310e-01
4.05840576e-01 -6.19536400e-01 -1.68524399e-01 8.41365874e-01
-2.29117498e-01 9.80654538e-01 2.56098688e-01 -4.44317400e-01
-1.16300154e+00 1.49926826e-01 3.35228503e-01 -2.73121446e-01
2.52196547e-02 5.91782570e-01 -6.52174726e-02 -6.63369715e-01
3.58233720e-01 6.34652898e-02 -2.27216974e-01 1.11425430e-01
3.32171142e-01 4.14105415e-01 2.58407414e-01 -4.55244541e-01
-3.76580864e-01 7.85696378e-04 -1.12809546e-01 -6.39720142e-01
7.06398726e-01 -5.27877688e-01 -1.49126798e-01 3.85084450e-01
1.04707217e+00 1.67746082e-01 -2.11242616e-01 -2.40393743e-01
2.60390431e-01 -5.08506119e-01 -7.40244761e-02 -1.18990934e+00
-3.08451623e-01 1.51984110e-01 4.33628082e-01 7.64298081e-01
7.86955535e-01 1.30592033e-01 4.61756736e-01 8.09165180e-01
3.08963150e-01 -1.19977939e+00 1.79864615e-01 5.63219786e-01
9.62125897e-01 -1.37329149e+00 -2.88248211e-01 -1.53410330e-01
-9.29022729e-01 9.10957456e-01 7.36275494e-01 6.91262782e-01
4.81307924e-01 -2.87845224e-01 1.92840397e-01 -6.02359891e-01
-9.12501454e-01 -2.21634209e-01 5.15340209e-01 2.07922399e-01
1.15374160e+00 1.39387459e-01 -1.06592453e+00 4.58690315e-01
-8.08166504e-01 -1.50795996e-01 2.17369437e-01 8.12862992e-01
-4.02965009e-01 -9.77658808e-01 1.17633482e-02 6.17868841e-01
-1.72167733e-01 -2.60738671e-01 -1.08681059e+00 6.19225025e-01
-8.10916781e-01 1.71696317e+00 -1.39658555e-01 -5.98465681e-01
6.25435889e-01 2.82081425e-01 9.52225924e-02 -6.32655323e-01
-9.61968720e-01 -5.42908967e-01 7.17818499e-01 -6.76119685e-01
-3.43249083e-01 -6.05128706e-01 -4.54136759e-01 -5.31234980e-01
-6.03279054e-01 8.03664863e-01 6.41627133e-01 9.25160289e-01
5.08888185e-01 -1.51273966e-01 5.22459805e-01 -3.79993469e-01
-7.42990851e-01 -1.30287874e+00 -3.84471603e-02 5.95484793e-01
1.13303497e-01 -5.86578488e-01 -6.13853455e-01 -5.97535908e-01] | [12.20368766784668, 7.763775825500488] |
b36aae9e-05a7-405c-8d09-d15c6163253f | deep-learning-based-detection-of-motion | 2303.10987 | null | https://arxiv.org/abs/2303.10987v1 | https://arxiv.org/pdf/2303.10987v1.pdf | Deep Learning-Based Detection of Motion-Affected k-Space Lines for T2*-Weighted MRI | T2*-weighted gradient echo MR imaging is strongly impacted by subject head motion due to motion-related changes in B0 inhomogeneities. Within the oxygenation-sensitive mqBOLD protocol, even mild motion during the acquisition of the T2*-weighted data propagates into errors in derived quantitative parameter maps. In order to correct these images without the need of repeated measurements, we propose to learn a classification of motion-affected k-space lines. To test this, we perform realistic motion simulations including motion-induced field inhomogeneity changes for supervised training. To detect the presence of motion in each phase encoding line, we train a convolutional neural network, leveraging the multi-echo information of the T2*-weighted images. The proposed network accurately detects motion-affected k-space lines for simulated displacements of $\geq$ 0.5mm (accuracy on test set: 92.5%). Finally, we show example reconstructions where we include these classification labels as weights in the data consistency term of an iterative reconstruction procedure, opening up exciting opportunities of k-space line detection in combination with more powerful reconstruction methods. | ['Julia A. Schnabel', 'Christine Preibisch', 'Daniel Rueckert', 'Samira M. Epp', 'Veronika Spieker', 'Kerstin Hammernik', 'Hannah Eichhorn'] | 2023-03-20 | null | null | null | null | ['line-detection'] | ['computer-vision'] | [ 4.30081397e-01 -1.10399775e-01 1.56989038e-01 -5.15832365e-01
-7.36797929e-01 -3.29019248e-01 1.65742144e-01 8.08992609e-02
-7.44678319e-01 7.87709653e-01 1.76972240e-01 -6.11155666e-02
-3.62830818e-01 -3.54847580e-01 -5.68684578e-01 -9.53634918e-01
-7.70593584e-01 3.36776316e-01 5.37955403e-01 1.25218317e-01
4.18454438e-01 6.46239936e-01 -7.63880551e-01 2.56787539e-01
7.75659621e-01 7.39848495e-01 3.90211344e-01 4.22733128e-01
2.65459836e-01 9.64440823e-01 -2.90828675e-01 2.95137793e-01
9.07381531e-03 -5.52073061e-01 -9.86192584e-01 -1.36579782e-01
5.49588621e-01 -5.34453273e-01 -4.60128188e-01 1.05034173e+00
1.04033983e+00 2.86060363e-01 4.47334230e-01 -6.62463069e-01
-6.52820766e-02 7.86427379e-01 -4.65761065e-01 1.08185697e+00
-1.32382974e-01 3.33388299e-01 3.50989774e-02 -9.65408504e-01
9.20162559e-01 3.24650437e-01 7.68124878e-01 3.72254312e-01
-1.20966029e+00 -4.87876803e-01 -5.54595053e-01 4.67232853e-01
-1.17527413e+00 -6.84398592e-01 9.47772980e-01 -6.51096106e-01
9.94157493e-01 2.20767349e-01 4.81544226e-01 4.55309570e-01
7.48317361e-01 4.17158931e-01 1.35401142e+00 -2.78266311e-01
1.28335699e-01 -3.75448287e-01 3.53140622e-01 6.56978130e-01
9.96643007e-02 3.36150937e-02 -3.32115948e-01 -3.38900136e-03
7.10525095e-01 -4.26024675e-01 -9.50269163e-01 -4.22011167e-01
-1.56683838e+00 4.14073855e-01 6.94775581e-01 7.99143910e-01
-3.99647951e-01 1.92725211e-01 3.77067924e-01 7.94551894e-02
2.02748060e-01 4.91985351e-01 4.37635370e-02 2.16011479e-01
-1.28412044e+00 -1.22399233e-01 9.13269669e-02 3.12326699e-01
4.43748176e-01 1.80834636e-01 -4.10065293e-01 6.69839859e-01
1.83927327e-01 3.44264090e-01 1.08100688e+00 -1.01837301e+00
5.81564866e-02 -6.99717253e-02 1.16929062e-01 -7.19549894e-01
-1.09221685e+00 -7.29686439e-01 -9.17261600e-01 2.79335678e-02
5.97357810e-01 -1.97267309e-01 -9.47250247e-01 1.47337389e+00
4.40957010e-01 2.18795121e-01 -2.78248250e-01 1.48074758e+00
6.27038240e-01 1.60134017e-01 -9.72585306e-02 -6.02066576e-01
8.93603206e-01 -7.19862461e-01 -1.05408478e+00 -2.76224650e-02
1.08565736e+00 -3.80742669e-01 4.99679714e-01 -9.63719636e-02
-1.47383070e+00 -2.72599876e-01 -1.21886456e+00 3.46721351e-01
2.04833239e-01 -3.09649289e-01 3.96542490e-01 4.21418846e-01
-1.15004623e+00 1.05126846e+00 -9.99098301e-01 1.83671117e-01
3.73586446e-01 5.40903628e-01 -4.65572804e-01 -3.19386452e-01
-1.51415479e+00 1.46547341e+00 3.74802917e-01 5.79654992e-01
-7.66711116e-01 -1.10249257e+00 -6.51391387e-01 -4.60861921e-01
6.47185594e-02 -2.80162662e-01 8.68652344e-01 -3.79168481e-01
-1.05338120e+00 8.85088027e-01 -3.91879044e-02 -3.70276421e-01
7.71393418e-01 2.60592967e-01 -6.58128262e-01 5.95167935e-01
7.97476172e-02 4.46600884e-01 6.59040034e-01 -9.79678333e-01
3.49171683e-02 -6.20141268e-01 -5.19413173e-01 1.03226677e-01
3.59042466e-01 1.47691429e-01 2.18574390e-01 -2.30042636e-01
6.73996031e-01 -9.02140617e-01 -2.36234874e-01 -6.87997863e-02
-2.62809992e-01 4.63822693e-01 3.95615906e-01 -9.86915350e-01
9.86718416e-01 -1.92536592e+00 -2.24207625e-01 3.59766275e-01
5.82247972e-01 1.30070686e-01 -2.48603877e-02 -3.44494939e-01
-5.30440867e-01 -3.37836117e-01 -7.76273191e-01 4.67866212e-01
-4.16994125e-01 -2.58170843e-01 2.25669935e-01 1.14126360e+00
-8.87481943e-02 1.12764442e+00 -1.06867373e+00 -2.06380606e-01
1.56589955e-01 5.62511027e-01 -1.91266418e-01 -9.95044596e-03
5.37793577e-01 9.56794918e-01 -4.88905199e-02 7.26251155e-02
8.76577020e-01 -5.44426069e-02 3.94516587e-01 -6.22080803e-01
-3.32379520e-01 2.70396948e-01 -8.40988576e-01 1.74340880e+00
-6.08283542e-02 7.02471435e-01 4.67578582e-02 -1.26476824e+00
6.59710407e-01 5.22672415e-01 9.40290928e-01 -1.11072063e+00
7.73603916e-02 6.15515530e-01 6.19876504e-01 -7.26259053e-01
9.37942788e-02 -5.48958778e-01 5.52801788e-01 6.20369196e-01
6.86919168e-02 -2.95764863e-01 -4.13562916e-02 -1.74905077e-01
1.11393225e+00 -1.90924704e-01 -3.30023944e-01 -8.37099314e-01
4.95448321e-01 -2.48119205e-01 3.86439532e-01 7.89945602e-01
-7.70269930e-01 7.28150606e-01 1.54358029e-01 -7.11715043e-01
-1.20014167e+00 -7.33925462e-01 -7.77275026e-01 1.89132616e-01
5.32770231e-02 3.01799208e-01 -7.52494574e-01 -3.84922892e-01
-2.89232969e-01 5.52378714e-01 -6.00680292e-01 -3.23711038e-01
-1.21325290e+00 -1.51725280e+00 4.43529427e-01 3.39130253e-01
5.34222186e-01 -1.08350635e+00 -9.88632262e-01 5.99454105e-01
-5.35832286e-01 -1.16555929e+00 -3.60420525e-01 5.09108007e-01
-1.11315262e+00 -9.18260038e-01 -1.25461364e+00 -6.38035536e-01
7.13058770e-01 7.03955591e-02 9.88912582e-01 -2.54673474e-02
-4.56424743e-01 1.48219362e-01 1.32830918e-01 1.86686188e-01
-3.17950398e-01 -1.93152279e-01 -2.38692667e-02 -2.35161453e-01
1.06877545e-02 -5.71272075e-01 -1.00878501e+00 2.80752957e-01
-8.81240129e-01 -1.22017935e-01 2.16812611e-01 7.24151611e-01
5.31037450e-01 -5.95789813e-02 4.12940741e-01 -4.56236362e-01
4.71379519e-01 -4.22215253e-01 -3.88194710e-01 2.14196131e-01
-5.26381433e-01 3.24522108e-01 6.00201860e-02 -4.77126986e-01
-5.80622137e-01 -1.98554084e-01 -2.11884215e-01 1.41198980e-02
-1.45731807e-01 5.23993015e-01 3.56567383e-01 -5.00836730e-01
7.60314107e-01 6.76575959e-01 1.55134574e-01 6.09248467e-02
-1.44588381e-01 1.96219459e-01 7.39550591e-01 -2.14233771e-01
3.06596458e-01 5.38197219e-01 3.16283762e-01 -8.36246014e-01
-4.76993233e-01 -2.87573010e-01 -9.64587867e-01 -6.32284403e-01
9.28547859e-01 -4.65666562e-01 -4.48887438e-01 6.73580945e-01
-9.13576782e-01 -7.55265534e-01 -9.34568197e-02 1.00244606e+00
-4.58309889e-01 5.56014359e-01 -7.67069161e-01 -3.16365659e-01
-3.98403317e-01 -1.63163483e+00 3.03570509e-01 -2.73753535e-02
-8.95150974e-02 -1.16096985e+00 5.19202538e-02 2.72407740e-01
9.25788462e-01 3.68273467e-01 9.46824551e-01 -4.15636837e-01
-2.88715959e-01 -5.01245670e-02 7.74202198e-02 1.44834043e-02
-4.68682460e-02 -7.68416107e-01 -7.81991839e-01 -2.25742444e-01
4.03580695e-01 -1.05081044e-01 8.36244166e-01 1.02168357e+00
1.12998903e+00 2.68839180e-01 1.89144537e-02 6.93239570e-01
1.10277772e+00 2.94257879e-01 7.80325055e-01 3.47759992e-01
5.93989730e-01 6.41257107e-01 1.96894985e-02 1.70210525e-01
2.95892477e-01 7.11348593e-01 1.74784958e-01 -2.36443177e-01
-4.27700698e-01 3.13561380e-01 3.19050290e-02 1.00003815e+00
-1.47342235e-01 2.95573503e-01 -1.16221917e+00 5.30632317e-01
-1.19089472e+00 -9.12393510e-01 -8.02710652e-01 2.41134953e+00
9.61713731e-01 -4.78020571e-02 -1.64919108e-01 1.82889998e-01
8.23246896e-01 1.50332645e-01 -6.28248334e-01 1.91640273e-01
-1.70347974e-01 2.10596666e-01 6.90267444e-01 9.65252638e-01
-8.91496420e-01 2.92862207e-01 6.53950071e+00 1.09613359e-01
-1.70326078e+00 4.49686199e-01 6.39490366e-01 -1.08561553e-01
-4.07295763e-01 -2.81769603e-01 -2.05798194e-01 5.94331861e-01
1.15297294e+00 1.25440925e-01 4.24478531e-01 1.46116980e-03
4.70672667e-01 -3.47776771e-01 -6.78857505e-01 9.54829812e-01
6.06502928e-02 -1.30639434e+00 -7.32234955e-01 -1.89269453e-01
4.99157399e-01 4.13916796e-01 -5.14966529e-03 -1.72509179e-01
-4.54632938e-01 -1.02757347e+00 4.81627613e-01 6.42307580e-01
1.04246533e+00 -4.23387885e-01 6.85840249e-01 1.70685858e-01
-5.91848791e-01 3.82844031e-01 -2.44595334e-01 4.23480868e-01
3.50988179e-01 8.74750733e-01 -7.80684054e-01 3.05330306e-01
4.74167079e-01 3.12174499e-01 -3.50763679e-01 1.33779418e+00
-1.26020759e-01 4.75166500e-01 -2.95136571e-01 5.80528021e-01
2.42679179e-01 -4.74244356e-02 3.84813726e-01 1.09750843e+00
3.27757895e-01 4.25439775e-01 -2.54215389e-01 9.31766748e-01
7.42693245e-02 -1.92344591e-01 -2.06623212e-01 4.90769833e-01
3.18069868e-02 1.15559745e+00 -8.62969875e-01 -4.21617150e-01
-7.32564852e-02 1.02018785e+00 1.28123626e-01 4.97477144e-01
-5.55244207e-01 -1.56877831e-01 -2.02168271e-01 5.46273708e-01
-1.33594468e-01 -4.29925919e-01 -5.86074851e-02 -1.15699172e+00
2.99144238e-02 -3.43529254e-01 -9.24160555e-02 -8.06498468e-01
-7.19415724e-01 5.88147819e-01 -2.69328933e-02 -5.08137941e-01
-1.23105690e-01 -2.65355438e-01 -5.55018246e-01 1.07115877e+00
-1.95532358e+00 -2.62039214e-01 -4.59325850e-01 4.85272110e-01
-1.52352527e-01 4.03113216e-01 5.44339776e-01 5.71648896e-01
-3.63432497e-01 3.31661671e-01 1.42400190e-01 1.46828189e-01
9.04765904e-01 -1.19962585e+00 5.92153259e-02 7.79941499e-01
-4.92896587e-01 6.17828071e-01 7.06660986e-01 -6.90699279e-01
-1.16806746e+00 -9.27638113e-01 8.21274281e-01 -7.39923120e-02
6.35955095e-01 4.37350161e-02 -1.36304951e+00 4.77558345e-01
5.23084663e-02 8.99957001e-01 5.50526083e-01 -7.86126971e-01
1.65865690e-01 8.74318928e-02 -1.48667014e+00 -8.22328124e-03
6.82082415e-01 -3.98673803e-01 -4.36083615e-01 4.11659747e-01
2.00053722e-01 -7.04154551e-01 -1.35261047e+00 5.33874750e-01
5.20138025e-01 -9.53628898e-01 9.44976985e-01 -3.76284331e-01
2.57647216e-01 -3.25350642e-01 3.16173404e-01 -1.44288933e+00
-4.86324608e-01 -4.05502886e-01 4.58478965e-02 2.46533304e-01
1.73385382e-01 -5.90300441e-01 6.52116537e-01 8.17213416e-01
-5.10421455e-01 -2.84764498e-01 -1.27087128e+00 -3.10315102e-01
4.49139386e-01 -1.85873374e-01 1.16889365e-01 1.46027005e+00
1.92402467e-01 -1.38109326e-01 -2.04090461e-01 1.73205242e-01
1.03196704e+00 -2.59012043e-01 -2.57424742e-01 -7.82696843e-01
-1.57278664e-02 -3.56275529e-01 -2.01094612e-01 -7.98006713e-01
3.84444445e-02 -1.26136315e+00 2.18269885e-01 -1.31890869e+00
2.72763252e-01 -6.58761442e-01 -6.84921861e-01 2.04308927e-01
-2.29828134e-01 5.17901361e-01 -8.16259980e-02 3.47102433e-01
-2.64221758e-01 2.93024153e-01 1.68812203e+00 -1.86889216e-01
1.57942493e-02 -4.62900668e-01 1.62066966e-01 3.70291829e-01
6.71453774e-01 -4.70588684e-01 -1.77390024e-01 -3.87824804e-01
-3.47495191e-02 7.42013812e-01 3.37257475e-01 -1.27275145e+00
2.08808094e-01 1.87344342e-01 7.13200092e-01 -3.49590153e-01
-1.25967860e-01 -7.31587231e-01 1.09055772e-01 8.24301839e-01
-5.65685809e-01 -9.26990360e-02 9.75267887e-02 -2.48387307e-01
-9.03318971e-02 -4.06839997e-01 1.21124649e+00 -2.91659623e-01
-4.57131803e-01 5.03124595e-01 -7.41773844e-01 6.25885651e-02
5.58400214e-01 -1.40281558e-01 -1.13412261e-01 -6.26724884e-02
-1.09779477e+00 -9.08728763e-02 8.56818333e-02 -8.37605149e-02
7.42406249e-01 -1.20123351e+00 -6.91227257e-01 2.03153536e-01
-1.85840026e-01 -3.35050285e-01 8.64442945e-01 1.79396605e+00
-8.40161026e-01 3.84675473e-01 -5.20297587e-01 -6.84810579e-01
-4.94110584e-01 2.35553503e-01 1.32774961e+00 -2.02052549e-01
-8.47210169e-01 7.28904665e-01 -2.51093239e-01 -2.74978071e-01
-2.10297406e-01 -4.35062319e-01 -2.78913885e-01 -2.41504252e-01
8.13933432e-01 2.61894643e-01 5.08557975e-01 -7.00409472e-01
-6.07649326e-01 4.46077645e-01 9.94142890e-02 -1.65179595e-01
1.35009015e+00 -2.26652831e-01 -1.33078650e-01 4.16579664e-01
1.23145986e+00 -3.30603927e-01 -1.40536952e+00 -2.14418963e-01
1.20222248e-01 -1.71067968e-01 6.98856950e-01 -1.06197178e+00
-1.33117735e+00 1.03851008e+00 1.30805683e+00 -1.68011054e-01
7.43644297e-01 -4.14704144e-01 8.33840549e-01 -8.21840838e-02
3.46005350e-01 -7.27130830e-01 -2.29053050e-01 2.68528044e-01
7.95560837e-01 -1.33994913e+00 -1.28134474e-01 2.34704018e-02
-5.77098429e-01 1.23154593e+00 2.09201053e-01 -8.43338817e-02
4.98859555e-01 4.66286927e-01 1.82324320e-01 -4.52158928e-01
-1.71818167e-01 3.46177191e-01 2.72048086e-01 6.31219327e-01
8.68832171e-01 9.12011974e-03 -5.22975564e-01 -9.65375006e-02
9.97181386e-02 1.69353917e-01 7.17588365e-01 8.76583278e-01
-4.76512820e-01 -6.27276123e-01 -4.50803339e-01 4.61252153e-01
-5.38156390e-01 -9.32895318e-02 4.37784076e-01 4.59974021e-01
-8.35201591e-02 6.11228764e-01 2.73473769e-01 1.46259815e-01
1.55595034e-01 3.74587506e-01 1.02532351e+00 -5.92059754e-02
-5.57342410e-01 1.72266483e-01 -1.46607727e-01 -5.89541674e-01
-6.35571063e-01 -4.74547386e-01 -1.70614612e+00 -1.05579376e-01
-3.02303713e-02 2.57880270e-01 8.72786880e-01 9.87191439e-01
9.01391357e-02 7.30959833e-01 4.97339875e-01 -9.94982541e-01
-3.59163851e-01 -1.11512673e+00 -7.65830219e-01 5.34084141e-01
5.05642772e-01 -6.31613195e-01 -3.07917804e-01 -2.68513620e-01] | [13.632637023925781, -2.3905084133148193] |
e581cbc4-6001-4f70-af87-fbc15cba5090 | accu-help-a-machine-learning-based-smart | 2212.02346 | null | https://arxiv.org/abs/2212.02346v1 | https://arxiv.org/pdf/2212.02346v1.pdf | Accu-Help: A Machine Learning based Smart Healthcare Framework for Accurate Detection of Obsessive Compulsive Disorder | In recent years the importance of Smart Healthcare cannot be overstated. The current work proposed to expand the state-of-art of smart healthcare in integrating solutions for Obsessive Compulsive Disorder (OCD). Identification of OCD from oxidative stress biomarkers (OSBs) using machine learning is an important development in the study of OCD. However, this process involves the collection of OCD class labels from hospitals, collection of corresponding OSBs from biochemical laboratories, integrated and labeled dataset creation, use of suitable machine learning algorithm for designing OCD prediction model, and making these prediction models available for different biochemical laboratories for OCD prediction for unlabeled OSBs. Further, from time to time, with significant growth in the volume of the dataset with labeled samples, redesigning the prediction model is required for further use. The whole process requires distributed data collection, data integration, coordination between the hospital and biochemical laboratory, dynamic machine learning OCD prediction mode design using a suitable machine learning algorithm, and making the machine learning model available for the biochemical laboratories. Keeping all these things in mind, Accu-Help a fully automated, smart, and accurate OCD detection conceptual model is proposed to help the biochemical laboratories for efficient detection of OCD from OSBs. OSBs are classified into three classes: Healthy Individual (HI), OCD Affected Individual (OAI), and Genetically Affected Individual (GAI). The main component of this proposed framework is the machine learning OCD prediction model design. In this Accu-Help, a neural network-based approach is presented with an OCD prediction accuracy of 86 percent. | ['Saraju Prasad Mohanty', 'Susanta Kumar Padhy', 'Sujita Kumar Kar', 'Laxmi Narayan Padhy', 'Ajaya Kumar Tripathy', 'Kabita Patel'] | 2022-12-05 | null | null | null | null | ['data-integration'] | ['knowledge-base'] | [-2.35959142e-01 -1.05327688e-01 -1.36368215e-01 -2.35240325e-01
-1.60684362e-01 -1.15966447e-01 -1.98218942e-01 5.38810015e-01
-1.43267019e-02 5.35758078e-01 -9.36086625e-02 7.84679577e-02
-4.75935906e-01 -6.97418809e-01 1.33058578e-01 -6.67304158e-01
-1.83734283e-01 1.17336798e+00 4.34698723e-02 1.28566548e-01
6.27328753e-01 7.91393816e-01 -1.50215352e+00 5.62216401e-01
9.89283979e-01 8.94934773e-01 8.67341012e-02 4.31943476e-01
-2.32220978e-01 5.41984022e-01 -2.96396315e-01 -2.03775987e-02
1.03487328e-01 -2.12850973e-01 -5.84259152e-01 1.45933911e-01
-5.20896196e-01 -2.58384377e-01 3.49958032e-01 6.42547488e-01
6.05229855e-01 -2.54026026e-01 8.58209312e-01 -1.27849102e+00
-6.55825675e-01 3.12645584e-01 -3.64564769e-02 -1.37652919e-01
2.87532449e-01 4.23719168e-01 2.14423865e-01 -7.19648004e-01
6.58715427e-01 1.10864592e+00 6.63506866e-01 5.55590928e-01
-1.03694761e+00 -8.42443943e-01 -4.20607209e-01 8.05223107e-01
-1.33495307e+00 -1.36769488e-02 4.18664932e-01 -8.41656327e-01
1.12524283e+00 1.13973945e-01 1.09108508e+00 6.79457963e-01
3.72892141e-01 2.55887747e-01 1.31054771e+00 -4.43620414e-01
7.76706278e-01 3.77486557e-01 4.72365707e-01 3.50530684e-01
4.67489541e-01 -3.42696428e-01 -3.38411450e-01 -1.61686718e-01
3.68637383e-01 4.18561399e-01 3.20595324e-01 -2.39963219e-01
-2.28837714e-01 6.95908904e-01 -1.36065349e-01 6.59556568e-01
-3.97777766e-01 -3.53445023e-01 4.81704384e-01 6.51343465e-02
7.84579292e-02 2.63305664e-01 -3.68614048e-01 2.31816415e-02
-6.67182088e-01 -3.46247740e-02 5.59613407e-01 4.72132385e-01
5.77361107e-01 1.17742501e-01 2.55326152e-01 1.05529606e+00
5.80619574e-01 3.75721574e-01 1.13072741e+00 -7.77609587e-01
-3.04519147e-01 1.24634933e+00 -5.66140935e-02 -1.10151720e+00
-7.20257580e-01 -2.63036817e-01 -6.35840535e-01 5.10754548e-02
1.49685860e-01 -1.05174787e-01 -6.06578112e-01 1.07721257e+00
5.74241221e-01 1.14804953e-01 -1.69207733e-02 7.57449985e-01
2.47942001e-01 3.18466961e-01 3.63463223e-01 -2.68838644e-01
1.44237721e+00 -4.73525703e-01 -7.10690141e-01 7.17134327e-02
1.08556366e+00 -4.73070711e-01 7.25543737e-01 1.01217687e+00
-6.55545831e-01 -3.00314695e-01 -1.01279449e+00 1.68663338e-01
-7.34498739e-01 1.50448069e-01 8.34122181e-01 8.68921340e-01
-8.06839943e-01 4.50560510e-01 -9.02056277e-01 -7.69509077e-01
8.70560706e-02 8.52911294e-01 -4.85906571e-01 -4.02170271e-01
-8.98697972e-01 8.76073539e-01 3.65463316e-01 5.85918799e-02
-7.67254412e-01 -6.05068982e-01 -3.55793446e-01 4.44360524e-02
-1.78625152e-01 -6.08615279e-01 6.08689606e-01 -6.76224768e-01
-1.31185389e+00 7.91461229e-01 2.07383081e-01 -3.09844524e-01
1.88797891e-01 1.41479243e-02 -7.21472144e-01 4.21872765e-01
1.36339560e-01 1.85422331e-01 4.77767944e-01 -1.01511860e+00
-5.92718601e-01 -1.10715067e+00 -6.80556118e-01 -4.89655137e-01
-3.05452377e-01 1.28029332e-01 5.97024187e-02 -4.41549793e-02
3.69457513e-01 -8.39544296e-01 -2.18397632e-01 -3.15513372e-01
-1.56551987e-01 -5.38496852e-01 7.76526511e-01 -6.53714597e-01
1.18001831e+00 -2.02077985e+00 -3.02719653e-01 2.41061702e-01
1.68226033e-01 4.10835057e-01 1.30271345e-01 6.57716513e-01
-4.64094281e-01 1.45443203e-02 1.67712644e-01 9.70868468e-02
2.16687873e-01 -4.29405943e-02 2.90602416e-01 4.29579169e-01
4.06612277e-01 2.08185837e-01 -5.59926152e-01 -4.17384148e-01
3.77779037e-01 3.89740616e-01 -8.06158483e-01 1.76007599e-01
-5.99802211e-02 4.57055897e-01 -5.47736585e-01 9.46553707e-01
8.17218184e-01 -1.44826457e-01 5.82844138e-01 -1.07509464e-01
-9.28947031e-02 -3.26995552e-01 -1.54432225e+00 9.92115021e-01
-1.10877626e-01 -3.09554726e-01 -6.64869696e-02 -1.17246139e+00
1.29313898e+00 4.17618394e-01 1.01104212e+00 -7.48552442e-01
2.39911079e-01 5.70868254e-01 5.93396388e-02 -1.12929595e+00
-2.13523865e-01 -2.43911505e-01 3.79173845e-01 3.22963655e-01
5.67009561e-02 3.71501863e-01 3.00034285e-01 -6.65841252e-02
1.16415346e+00 -2.00148314e-01 2.20501706e-01 1.47418063e-02
8.15359116e-01 8.88837427e-02 7.87442982e-01 -1.22644026e-02
-4.68474567e-01 -4.76419963e-02 5.68058550e-01 -4.71150696e-01
-1.03998387e+00 -7.40277410e-01 -5.01423120e-01 4.25004125e-01
-2.98620850e-01 -1.80704162e-01 -7.84984589e-01 -1.19985931e-01
5.12634031e-02 3.97108287e-01 -3.08754653e-01 -2.44370744e-01
-6.20470718e-02 -1.03433883e+00 3.17710549e-01 5.83916567e-02
2.76192665e-01 -8.02419543e-01 -3.70042473e-01 5.51866174e-01
2.24563166e-01 -6.60322785e-01 4.64991957e-01 3.06268364e-01
-1.04569769e+00 -1.43521583e+00 -2.89145827e-01 -5.68850279e-01
3.99847567e-01 -1.39103457e-01 2.43320823e-01 2.34832183e-01
-8.41693163e-01 3.16955775e-01 -5.30551612e-01 -5.72730482e-01
-2.52252758e-01 -3.28796357e-01 2.02166036e-01 2.22276166e-01
9.47499454e-01 -6.84754014e-01 -7.67917097e-01 9.92986858e-02
-8.53549361e-01 -3.96562874e-01 5.64923227e-01 4.96300608e-01
5.55988073e-01 1.82427213e-01 1.10617936e+00 -5.95866978e-01
3.49475652e-01 -9.79821086e-01 -2.81873554e-01 7.06743076e-02
-9.87996221e-01 -2.24906623e-01 4.51979339e-01 -2.86861569e-01
-7.74230957e-01 1.83344334e-01 -2.29765967e-01 -8.33174810e-02
-6.72046363e-01 3.56114388e-01 -1.56687349e-01 4.04646754e-01
4.52061653e-01 5.59600592e-02 2.66695738e-01 -6.29434288e-01
-1.99216142e-01 1.34793890e+00 -1.15807585e-01 -2.17823371e-01
-1.64946616e-01 3.81328613e-01 1.01984434e-01 -7.55501986e-01
-1.11312710e-01 -7.76119351e-01 -6.72701418e-01 -2.53393531e-01
1.15524578e+00 -5.82253218e-01 -1.13964784e+00 6.84338510e-01
-7.56423473e-01 -7.93513842e-03 3.44083101e-01 7.99395323e-01
-4.38822985e-01 2.71142215e-01 -6.57213569e-01 -1.19747090e+00
-6.51568294e-01 -1.31950760e+00 6.36856556e-01 1.34342954e-01
-3.69255930e-01 -7.64988184e-01 2.57683247e-01 8.85433912e-01
2.11817473e-01 2.56793410e-01 1.39421988e+00 -1.12778580e+00
-4.52588022e-01 -6.94847524e-01 -8.36495832e-02 3.18206578e-01
-4.69475333e-03 3.45052093e-01 -6.47973537e-01 1.81354716e-01
3.55545282e-01 -2.36021832e-01 8.00655112e-02 1.43672362e-01
8.27994168e-01 -3.00920140e-02 -3.49978566e-01 8.11620802e-02
1.87659705e+00 9.85388815e-01 7.35537171e-01 4.19130355e-01
1.23429328e-01 8.25145543e-01 7.36595571e-01 6.24380887e-01
3.91008377e-01 4.18999046e-01 4.33890522e-01 2.42740065e-01
4.58282866e-02 4.91637975e-01 2.92970836e-01 9.85671878e-01
5.26065342e-02 4.35463041e-01 -1.20532250e+00 4.92104501e-01
-1.72494102e+00 -8.45171690e-01 -7.87613571e-01 1.96070719e+00
6.53821409e-01 -2.79952407e-01 2.80102402e-01 4.13731903e-01
5.91223240e-01 -1.07766163e+00 -5.08150816e-01 -8.18641603e-01
3.23761553e-01 2.99433976e-01 -1.16452940e-01 1.30559236e-01
-5.20026505e-01 5.03576159e-01 5.87947750e+00 5.97253799e-01
-1.49207473e+00 3.91829193e-01 4.37085569e-01 -8.29482377e-02
2.36847848e-01 -1.28225461e-01 -7.51255572e-01 9.27048326e-01
1.17899942e+00 1.57433018e-01 3.84104431e-01 1.07944262e+00
7.81376362e-01 -4.30086613e-01 -7.68635333e-01 9.69572604e-01
3.52080837e-02 -1.01815856e+00 -2.59678274e-01 1.87927574e-01
3.55182290e-01 -7.35699013e-02 -3.75981897e-01 1.80700973e-01
1.44751547e-02 -6.99116170e-01 1.19752869e-01 1.07984352e+00
2.05742955e-01 -6.32397234e-01 1.02595282e+00 1.95299029e-01
-6.51964307e-01 -6.43399775e-01 -3.46188962e-01 -2.21787440e-03
-3.50455999e-01 9.29893911e-01 -1.01685798e+00 4.56784010e-01
7.74461031e-01 3.81120801e-01 -6.30750060e-01 1.20714831e+00
2.72559196e-01 3.40243191e-01 1.29077673e-01 -8.31766576e-02
-1.07414976e-01 -5.07514238e-01 -2.21693680e-01 8.78076911e-01
5.06388545e-01 3.26319754e-01 5.75960167e-02 8.44888508e-01
7.89451718e-01 6.70390725e-01 -4.16045070e-01 -2.96005845e-01
2.97130138e-01 1.27054691e+00 -7.95128882e-01 -2.73357660e-01
-2.31187612e-01 4.50709164e-01 2.20321074e-01 -1.87635705e-01
-4.48815644e-01 -2.88433522e-01 4.40307945e-01 4.37191904e-01
-2.36724541e-01 4.88612289e-03 -7.46951342e-01 -7.15825438e-01
-5.22224665e-01 -8.96818876e-01 5.14628410e-01 -9.58611190e-01
-1.38918889e+00 1.17480710e-01 -3.92309725e-01 -1.10444868e+00
-5.91501854e-02 -7.75080025e-01 -4.38233793e-01 6.82669938e-01
-8.07142735e-01 -1.12310517e+00 -2.27113754e-01 4.84521568e-01
2.95099348e-01 -2.56912082e-01 8.90282691e-01 5.86755276e-01
-1.09916568e+00 1.96839478e-02 3.21034580e-01 -2.35412702e-01
9.28263009e-01 -8.17312837e-01 -1.10694599e+00 2.99549341e-01
-8.61718237e-01 8.38086963e-01 3.22570890e-01 -1.12339401e+00
-1.42176259e+00 -9.14587796e-01 1.04029775e+00 -1.19048633e-01
8.24944854e-01 -5.71207069e-02 -8.13390136e-01 2.55260855e-01
-1.25754029e-01 -6.30021632e-01 1.44276083e+00 4.80890349e-02
5.34145832e-02 -3.33433479e-01 -1.60166955e+00 4.81050819e-01
1.62713006e-01 -2.25714415e-01 -4.35289264e-01 7.07451403e-01
2.65209913e-01 3.31419736e-01 -1.38319004e+00 3.94706428e-02
5.56107223e-01 -1.15953422e+00 7.15847731e-01 -7.18085289e-01
4.41047668e-01 -2.36146033e-01 -8.82782042e-02 -7.34215975e-01
-2.55949795e-01 2.27320820e-01 2.02591240e-01 1.08506489e+00
-2.33680569e-02 -6.97322190e-01 7.50594497e-01 1.18807971e+00
-2.77693570e-01 -8.76039743e-01 -7.93712735e-01 -6.23800993e-01
-1.25764474e-01 -5.50871909e-01 5.57117522e-01 1.16308320e+00
5.59612811e-01 -1.72677916e-03 3.19145061e-02 2.53627121e-01
4.90194321e-01 -2.95920342e-01 6.14834249e-01 -1.52252698e+00
-1.28567621e-01 -1.09145850e-01 -8.58230889e-01 6.12556040e-01
-3.63351583e-01 -8.31250608e-01 -7.99935043e-01 -1.67178059e+00
1.98120743e-01 -6.58536971e-01 -1.73843667e-01 6.03661478e-01
2.74162769e-01 1.25663400e-01 1.48589477e-01 3.29064667e-01
-1.93724439e-01 -5.68449311e-02 9.17172790e-01 3.27669866e-02
-2.53423899e-01 -3.01269203e-01 -4.22708601e-01 6.35495245e-01
8.74260604e-01 -5.96170545e-01 -4.07656580e-01 2.17560515e-01
-1.17882388e-02 9.13900323e-03 2.45341018e-01 -1.30076826e+00
1.36260673e-01 -4.29300308e-01 6.07862234e-01 -6.65859938e-01
3.40368867e-01 -1.01743531e+00 7.25541234e-01 1.20045340e+00
1.84719652e-01 -5.16033590e-01 -1.38910100e-01 1.02362975e-01
1.80075690e-01 -7.08079755e-01 9.71294999e-01 -3.08078170e-01
-4.76832241e-01 -1.81201119e-02 -6.93989336e-01 -6.89195216e-01
1.60109639e+00 -6.63820088e-01 -3.94559890e-01 2.38884702e-01
-1.46899319e+00 -1.24852926e-01 1.50269568e-01 6.60601854e-02
2.82278746e-01 -9.61283863e-01 -1.48190111e-01 2.47694880e-01
6.76668286e-02 -4.84867960e-01 8.96184444e-01 1.38948107e+00
-7.23038137e-01 8.36166799e-01 -7.05673337e-01 -3.71802032e-01
-1.06409347e+00 9.87046480e-01 2.22970665e-01 -1.81677952e-01
-9.34534147e-02 8.05999711e-02 -4.20677036e-01 -3.60689431e-01
7.70230740e-02 -1.04957655e-01 -5.38512290e-01 2.81626225e-01
6.05541945e-01 8.92613530e-01 5.00690222e-01 -3.39816630e-01
-2.79054523e-01 4.06742305e-01 2.33292788e-01 2.70086288e-01
1.64440632e+00 -1.83053032e-01 -6.78475261e-01 4.67437059e-01
8.99085641e-01 -1.74711198e-01 -2.48788625e-01 4.38229412e-01
2.18643919e-01 -2.48524323e-01 -2.40460098e-01 -1.17145586e+00
-7.68766344e-01 8.61203909e-01 1.31319225e+00 1.81084484e-01
1.01076031e+00 -2.67969608e-01 4.41899985e-01 1.92258388e-01
5.85861981e-01 -1.48043692e+00 2.44306847e-01 1.12583779e-01
7.16559172e-01 -9.58773255e-01 -3.41813982e-01 -4.37876791e-01
-5.80323219e-01 1.24194574e+00 7.58890927e-01 -1.87706232e-01
6.90075397e-01 1.84241787e-01 -6.13468438e-02 -3.95174384e-01
-8.00173044e-01 1.16070511e-03 -2.76530266e-01 1.15458453e+00
3.24393094e-01 1.66862905e-01 -9.99678493e-01 1.33951664e+00
1.65578514e-01 5.75037301e-01 5.49800336e-01 8.22986960e-01
-9.01749849e-01 -1.38844895e+00 -7.45378554e-01 6.94600403e-01
-3.85957778e-01 1.17792457e-01 -5.61412573e-01 6.54794872e-01
7.11694956e-01 1.17721653e+00 -8.84244777e-03 -3.17128360e-01
1.65512696e-01 6.38095140e-01 -2.88773403e-02 -7.45335937e-01
-5.51352918e-01 3.76014471e-01 1.16583947e-02 -4.33218837e-01
-5.67146242e-02 -6.57037735e-01 -1.76631463e+00 -2.74726391e-01
-2.25785524e-01 9.80889350e-02 1.19201577e+00 1.08453155e+00
7.24522948e-01 3.75409663e-01 7.17753112e-01 -2.96620935e-01
-1.40239015e-01 -9.95344818e-01 -1.11892354e+00 9.85611379e-02
-4.43842202e-01 -7.01808333e-01 -1.27487451e-01 1.14309214e-01] | [8.409306526184082, 4.884030818939209] |
35b6ef58-7ff2-42b9-a70f-1d87da6c3961 | positional-diffusion-ordering-unordered-sets | 2303.11120 | null | https://arxiv.org/abs/2303.11120v1 | https://arxiv.org/pdf/2303.11120v1.pdf | Positional Diffusion: Ordering Unordered Sets with Diffusion Probabilistic Models | Positional reasoning is the process of ordering unsorted parts contained in a set into a consistent structure. We present Positional Diffusion, a plug-and-play graph formulation with Diffusion Probabilistic Models to address positional reasoning. We use the forward process to map elements' positions in a set to random positions in a continuous space. Positional Diffusion learns to reverse the noising process and recover the original positions through an Attention-based Graph Neural Network. We conduct extensive experiments with benchmark datasets including two puzzle datasets, three sentence ordering datasets, and one visual storytelling dataset, demonstrating that our method outperforms long-lasting research on puzzle solving with up to +18% compared to the second-best deep learning method, and performs on par against the state-of-the-art methods on sentence ordering and visual storytelling. Our work highlights the suitability of diffusion models for ordering problems and proposes a novel formulation and method for solving various ordering tasks. Project website at https://iit-pavis.github.io/Positional_Diffusion/ | ['Alessio Del Bue', 'Yiming Wang', 'Stuart James', 'Gianluca Scarpellini', 'Francesco Giuliari'] | 2023-03-20 | null | null | null | null | ['visual-storytelling', 'sentence-ordering'] | ['natural-language-processing', 'natural-language-processing'] | [ 5.61554804e-02 5.05163491e-01 -2.39362568e-01 -2.11168509e-02
-8.30557227e-01 -7.69262433e-01 5.70494115e-01 3.04329604e-01
8.63819420e-02 5.04310668e-01 9.11221981e-01 -5.55728376e-01
-7.38051653e-01 -9.87435699e-01 -1.02034605e+00 -5.45654416e-01
-2.39482388e-01 1.31798518e+00 2.22773373e-01 -3.40436310e-01
5.69164038e-01 -7.17424750e-02 -8.64331841e-01 9.53934371e-01
6.56206727e-01 4.21052098e-01 3.81678462e-01 8.04516733e-01
-3.92969638e-01 1.61919177e+00 -7.00253546e-01 -5.84697485e-01
1.42591998e-01 -5.61827540e-01 -1.36436987e+00 -1.07579075e-01
2.47235686e-01 -3.22401136e-01 -8.21223974e-01 6.77269936e-01
3.50336820e-01 1.08029976e-01 4.94877517e-01 -1.19922173e+00
-1.57435465e+00 1.29170692e+00 -8.42550039e-01 4.89609122e-01
9.90913868e-01 2.48281971e-01 1.45846486e+00 -2.20513731e-01
1.14755714e+00 9.66897666e-01 8.90432179e-01 2.90806592e-01
-1.21398437e+00 -7.12919533e-02 2.48034000e-01 6.17079556e-01
-1.06360459e+00 3.29461172e-02 1.01664615e+00 -4.89574462e-01
1.28384304e+00 3.39898914e-01 9.57282484e-01 1.33116949e+00
2.25850478e-01 1.28753281e+00 8.54938805e-01 -1.22460604e-01
2.73458034e-01 -8.45208347e-01 3.32084328e-01 9.80777085e-01
6.69354498e-02 -4.29089159e-01 -1.05954289e+00 -1.40673652e-01
8.21070969e-01 -5.22781432e-01 -4.09803152e-01 -7.93767273e-01
-1.13468361e+00 7.83558011e-01 6.29582286e-01 2.87304580e-01
-4.82827574e-01 5.39442301e-01 3.17461193e-01 3.67215991e-01
5.48810899e-01 8.09993804e-01 -1.83334127e-01 -5.15822470e-01
-7.55230665e-01 6.23616755e-01 9.71934676e-01 9.04231429e-01
2.74518877e-01 -4.02677208e-01 -5.25837243e-01 4.40176934e-01
2.12315306e-01 -2.46004924e-01 1.41436711e-01 -1.20936263e+00
8.75873744e-01 6.05980396e-01 -3.68798561e-02 -1.20199609e+00
-4.66867745e-01 -5.54753542e-01 -5.94820976e-01 -3.82532552e-02
5.10123730e-01 2.66444445e-01 -6.65764749e-01 1.59931958e+00
1.63121954e-01 9.73154753e-02 -1.76122084e-01 9.60489511e-01
1.07420754e+00 7.83915997e-01 -3.67095232e-01 2.44258508e-01
1.13435984e+00 -1.47033429e+00 -5.00296772e-01 -4.45033044e-01
6.14548206e-01 -3.24276388e-01 1.34182453e+00 4.90433723e-01
-1.48892796e+00 -5.46508841e-02 -1.06060219e+00 -5.77507973e-01
3.89417671e-02 -3.50737512e-01 8.94363701e-01 3.78856003e-01
-1.52519917e+00 8.79180014e-01 -5.69744766e-01 -2.67569095e-01
9.11510229e-01 1.04189396e-01 -3.89277756e-01 -3.00379306e-01
-8.28192115e-01 7.08660960e-01 3.04333251e-02 -2.82071736e-02
-9.74955797e-01 -9.67187166e-01 -8.04573894e-01 2.25207418e-01
4.35433984e-01 -1.22528338e+00 1.29189456e+00 -9.79679450e-02
-1.24214327e+00 7.73216665e-01 -1.79785527e-02 -5.76413333e-01
8.16381514e-01 -2.72040784e-01 3.29640836e-01 8.08689669e-02
4.73303199e-01 3.81986171e-01 2.85443842e-01 -1.28935623e+00
-1.09940074e-01 -2.52090961e-01 5.33994257e-01 3.88562649e-01
1.40064001e-01 -2.85440773e-01 -6.55174851e-01 -4.19193208e-01
3.78519207e-01 -5.95035195e-01 -1.41157538e-01 -2.32082814e-01
-6.62593365e-01 -4.54876065e-01 3.36456656e-01 -9.57002521e-01
1.14722586e+00 -1.57386148e+00 8.54468346e-01 -1.98049501e-01
6.54225111e-01 -4.07238930e-01 -3.74671996e-01 9.83358443e-01
-1.53029636e-02 2.55267799e-01 -3.37347478e-01 -7.22854972e-01
4.22905684e-01 2.55180717e-01 -3.14878345e-01 4.73333091e-01
-3.35027874e-01 1.47167587e+00 -1.13091552e+00 -1.06360622e-01
-1.67905003e-01 2.79912591e-01 -6.14964426e-01 6.53234199e-02
-6.58785403e-01 3.51086229e-01 -7.09950998e-02 4.19925243e-01
5.75530112e-01 -6.34029448e-01 5.13772488e-01 2.07732499e-01
2.79771328e-01 6.91967428e-01 -7.31981277e-01 2.36381054e+00
-4.04602475e-02 9.51181114e-01 -2.52628535e-01 -6.98382616e-01
6.90712750e-01 -1.08962327e-01 3.48439932e-01 -7.35978544e-01
-5.26614860e-02 -3.23563784e-01 -6.21365719e-02 -5.88173628e-01
7.61020899e-01 3.33850235e-01 -2.05671743e-01 1.10342908e+00
-1.31005898e-01 -2.59101987e-01 4.69118953e-01 8.35079849e-01
1.80186427e+00 2.29614317e-01 -1.00328602e-01 -1.31913051e-01
-2.84974456e-01 2.24477112e-01 8.85450095e-02 1.13909304e+00
9.30217355e-02 8.33288968e-01 1.42907810e+00 -7.11951613e-01
-1.08426535e+00 -1.18553126e+00 6.25281870e-01 1.25648654e+00
2.53895402e-01 -7.45212376e-01 -7.52325773e-01 -7.77178049e-01
-2.13492252e-02 9.33627546e-01 -1.16417003e+00 1.32022962e-01
-7.87451565e-01 -6.39120519e-01 4.57067132e-01 7.08181143e-01
3.46453726e-01 -1.19419289e+00 -3.50854516e-01 1.50286242e-01
-3.68118227e-01 -5.86121738e-01 -4.70047295e-01 -6.71335906e-02
-4.27931786e-01 -1.27532804e+00 -5.89591384e-01 -8.81363690e-01
5.41326046e-01 2.93764681e-01 1.56375515e+00 1.88697308e-01
-4.14647125e-02 3.50775838e-01 -4.84320462e-01 1.84939817e-01
-2.46554300e-01 3.66473436e-01 -7.00159848e-01 -4.47726071e-01
2.07643816e-03 -8.09094667e-01 -7.73147523e-01 -3.20837796e-02
-5.77430844e-01 4.45558816e-01 1.49877220e-01 6.95930898e-01
3.01598608e-01 -4.59077246e-02 1.22355528e-01 -1.13963580e+00
1.40925717e+00 -7.09845901e-01 -3.23771983e-01 2.70980328e-01
-4.05383080e-01 7.82689080e-02 1.97091833e-01 -1.71680093e-01
-7.02939510e-01 -4.91052777e-01 -4.16978784e-02 -1.38443187e-01
2.88282543e-01 8.01081359e-01 1.04305893e-01 3.34554046e-01
6.93690062e-01 2.16441989e-01 -2.12898940e-01 -3.47758502e-01
7.24900067e-01 -1.55308053e-01 7.50530660e-01 -7.01044142e-01
3.91688555e-01 6.26848698e-01 -2.37789646e-01 -7.75913745e-02
-8.52532268e-01 -2.30052933e-01 -6.61424220e-01 -1.45819575e-01
9.08781946e-01 -4.19154137e-01 -8.07293355e-01 4.37248349e-01
-1.57983148e+00 -9.70688939e-01 -4.95793998e-01 -2.96478152e-01
-8.92039657e-01 4.84749496e-01 -1.12077248e+00 -4.69976664e-01
-2.70680279e-01 -9.53071177e-01 8.85481894e-01 7.43831918e-02
-6.63396299e-01 -1.05696940e+00 4.32906091e-01 8.26836467e-01
3.96184266e-01 2.72368163e-01 1.33204424e+00 -4.04087275e-01
-1.14799690e+00 1.34318054e-01 -3.04445446e-01 -6.14807725e-01
-2.85967767e-01 -2.90276587e-01 -3.84287745e-01 -4.24679145e-02
-1.44275203e-01 -2.70493776e-01 1.08198571e+00 6.38857901e-01
1.05767882e+00 -2.91871727e-01 -3.62697512e-01 6.57497764e-01
1.11407197e+00 4.28831577e-02 9.01188433e-01 7.96854317e-01
9.13058221e-01 4.70499367e-01 6.55641593e-03 5.33787727e-01
1.04835653e+00 3.73040885e-01 7.61851788e-01 2.66912162e-01
-5.07788301e-01 -7.28519976e-01 -2.01553121e-01 6.70263708e-01
-6.02803566e-02 -8.33706975e-01 -1.26549113e+00 8.27014089e-01
-2.29607701e+00 -1.17791033e+00 -5.73038697e-01 1.48012435e+00
6.15471482e-01 1.79940134e-01 4.65216264e-02 2.64974069e-02
5.59306264e-01 8.06616604e-01 -3.82192016e-01 -3.68642867e-01
-1.76103920e-01 -1.97196528e-02 2.41138682e-01 8.13305318e-01
-9.09457505e-01 9.73998547e-01 6.68565083e+00 6.83485985e-01
-3.73939335e-01 3.35195571e-01 8.17303121e-01 -3.53975952e-01
-8.33330393e-01 -4.75814417e-02 -1.15077972e-01 2.96421826e-01
4.34209764e-01 -2.23108009e-01 9.55244184e-01 5.40767789e-01
-1.46858573e-01 -1.64066806e-01 -1.05195594e+00 8.16911221e-01
3.06162745e-01 -2.16658401e+00 -1.62326053e-01 -2.64831837e-02
9.72158432e-01 8.97274092e-02 1.96957588e-01 1.63488448e-01
9.64114368e-01 -1.25927818e+00 1.14247251e+00 5.45801282e-01
3.20495307e-01 -5.74948847e-01 5.55727243e-01 4.02889639e-01
-9.24805105e-01 -1.80462763e-01 -4.03938264e-01 -4.13226455e-01
5.94367445e-01 2.28559062e-01 -8.55801761e-01 6.21518970e-01
7.61743963e-01 8.08098257e-01 -3.86544943e-01 1.23396194e+00
-8.26938152e-01 4.56460685e-01 1.59279376e-01 -2.36425459e-01
4.09600556e-01 -1.75058231e-01 6.57544315e-01 8.42338622e-01
1.82544127e-01 1.08706310e-01 -1.28291339e-01 1.03856969e+00
-2.60839492e-01 -2.70526975e-01 -5.85122585e-01 -1.10458165e-01
3.73704761e-01 8.07596207e-01 -1.02840567e+00 -1.94488585e-01
1.60528377e-01 1.23396146e+00 9.73785162e-01 3.03735793e-01
-1.09424782e+00 7.74233341e-02 3.17172021e-01 1.90268934e-01
5.80692410e-01 -5.71350098e-01 -9.17359531e-01 -7.34436750e-01
-1.94582697e-02 -8.47585797e-01 5.51893473e-01 -1.27823675e+00
-1.25673902e+00 6.55289590e-01 -6.93171322e-02 -4.83636856e-01
7.75423944e-02 -2.10159808e-01 -7.81434834e-01 6.87627256e-01
-8.76540780e-01 -1.15998709e+00 -3.72152597e-01 7.33579099e-02
8.79390180e-01 7.89836794e-02 5.67614794e-01 -2.02785090e-01
-3.57126176e-01 2.70145357e-01 5.50326109e-02 9.70966294e-02
5.23461066e-02 -1.63335490e+00 1.34734643e+00 5.28872430e-01
5.75639546e-01 4.61065859e-01 8.01923633e-01 -9.52016652e-01
-1.54502976e+00 -4.61254388e-01 9.13154185e-01 -9.25065339e-01
1.00207675e+00 -6.76453531e-01 -9.04978693e-01 9.63485956e-01
9.05713320e-01 -5.97677052e-01 4.72381532e-01 3.29739541e-01
-5.09387016e-01 5.34306526e-01 -8.02721858e-01 7.99657881e-01
1.87060142e+00 -5.48713684e-01 -8.49086940e-01 7.99169183e-01
8.80793154e-01 -1.00958300e+00 -4.58792001e-01 -1.78383023e-01
4.80416596e-01 -1.23060894e+00 9.44164038e-01 -6.26509309e-01
1.03895688e+00 -1.03033110e-01 5.31288534e-02 -1.58927214e+00
-1.00688291e+00 -1.00472462e+00 -2.61715978e-01 9.21931207e-01
6.02727354e-01 -3.39648932e-01 1.41620660e+00 4.87426102e-01
-2.80048281e-01 -1.04716396e+00 -1.00306892e+00 -5.65596581e-01
3.85472625e-02 -4.56619203e-01 9.55144286e-01 9.39462662e-01
2.87863165e-01 4.77144361e-01 -3.78591031e-01 1.62431985e-01
4.70721275e-01 3.34181398e-01 7.95566499e-01 -8.18898559e-01
-7.61391878e-01 -7.23773658e-01 -5.89295439e-02 -1.48439038e+00
6.99584708e-02 -1.16685164e+00 -1.10765092e-01 -2.92374897e+00
3.01115543e-01 -2.59526849e-01 1.18713364e-01 4.05197352e-01
1.45842195e-01 8.39883685e-02 2.72065282e-01 2.52691656e-01
-1.01987016e+00 6.22337580e-01 1.46470058e+00 -6.58676803e-01
-1.62636161e-01 -3.46618652e-01 -1.18771529e+00 3.85433823e-01
9.17023659e-01 -6.25086010e-01 -8.28901112e-01 -1.36385345e+00
7.91333675e-01 3.00063848e-01 2.21769184e-01 -6.38531744e-01
8.19126368e-01 3.04136313e-02 1.26914322e-01 -6.79229319e-01
4.54074740e-01 -1.14289105e-01 2.87582576e-01 3.82702261e-01
-5.60691953e-01 5.68453193e-01 9.93300527e-02 6.99666917e-01
3.31956625e-01 -2.06207424e-01 -7.05588162e-02 -3.38256150e-01
-7.33518004e-01 5.16667329e-02 -3.71061772e-01 1.93265304e-01
9.63468194e-01 -1.76188141e-01 -1.22499204e+00 -9.01443124e-01
-7.69654632e-01 3.07726145e-01 4.09908563e-01 4.15326476e-01
6.90207958e-01 -1.10937560e+00 -8.35000634e-01 -4.52108346e-02
-3.14816475e-01 1.38389081e-01 5.47437310e-01 7.56954968e-01
-1.17196119e+00 2.38691866e-01 -1.65459201e-01 -3.82608771e-01
-1.00942922e+00 5.93660414e-01 5.47010601e-02 -6.08348906e-01
-9.25318718e-01 1.49428046e+00 -6.69548288e-03 -3.39261502e-01
1.00376457e-01 -3.51976067e-01 -4.86616530e-02 -8.26058239e-02
3.81882697e-01 4.70736295e-01 -1.84087962e-01 -1.54293805e-01
-2.35025942e-01 2.06414998e-01 -1.67893216e-01 -1.85332388e-01
1.70492816e+00 -7.97690079e-03 -4.70770240e-01 2.25422323e-01
7.29267597e-01 4.82361615e-02 -1.35484910e+00 1.15609735e-01
1.91640675e-01 -6.40684903e-01 -2.23100930e-01 -1.02210176e+00
-8.37256849e-01 4.23883229e-01 -3.75882328e-01 6.94558322e-01
5.48115253e-01 6.46239579e-01 6.79063380e-01 1.01263650e-01
4.96377736e-01 -6.11144304e-01 4.49574322e-01 6.63294613e-01
1.51794779e+00 -6.97748959e-01 5.75877987e-02 -4.95523810e-01
-8.25476766e-01 7.49703228e-01 5.15952170e-01 -4.25615728e-01
2.14224532e-01 4.22613174e-01 -1.96522459e-01 -7.75940359e-01
-1.09574771e+00 8.95879120e-02 1.47535250e-01 8.46428871e-01
1.51672706e-01 -1.21514488e-03 -2.51064390e-01 7.84990489e-01
-6.37761116e-01 -1.79440871e-01 7.31982052e-01 8.87486219e-01
-2.27154806e-01 -8.99895847e-01 -1.73598886e-01 3.79326999e-01
1.41396999e-01 -2.77048916e-01 -7.75026202e-01 7.83434272e-01
-2.54265875e-01 9.50142503e-01 2.31685400e-01 -4.97566104e-01
1.81235999e-01 -1.54036537e-01 7.98623919e-01 -5.40340066e-01
-5.85362792e-01 -3.80504668e-01 4.67420250e-01 -7.16398239e-01
1.78257078e-01 -7.76244700e-01 -1.13247383e+00 -7.81372070e-01
2.49768063e-01 3.47921103e-02 2.15271324e-01 7.01950490e-01
5.05150437e-01 9.55185056e-01 -1.46365836e-01 -7.41567552e-01
-4.87385802e-02 -7.37142563e-01 -4.78692681e-01 3.00722986e-01
-6.35543317e-02 -4.73667175e-01 -7.10568056e-02 -3.54256123e-01] | [10.747859954833984, 8.050457000732422] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.