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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
449e9e5c-b556-44e7-b7c4-34cb1773be86 | decentralized-motor-skill-learning-for | 2306.17411 | null | https://arxiv.org/abs/2306.17411v1 | https://arxiv.org/pdf/2306.17411v1.pdf | Decentralized Motor Skill Learning for Complex Robotic Systems | Reinforcement learning (RL) has achieved remarkable success in complex robotic systems (eg. quadruped locomotion). In previous works, the RL-based controller was typically implemented as a single neural network with concatenated observation input. However, the corresponding learned policy is highly task-specific. Since all motors are controlled in a centralized way, out-of-distribution local observations can impact global motors through the single coupled neural network policy. In contrast, animals and humans can control their limbs separately. Inspired by this biological phenomenon, we propose a Decentralized motor skill (DEMOS) learning algorithm to automatically discover motor groups that can be decoupled from each other while preserving essential connections and then learn a decentralized motor control policy. Our method improves the robustness and generalization of the policy without sacrificing performance. Experiments on quadruped and humanoid robots demonstrate that the learned policy is robust against local motor malfunctions and can be transferred to new tasks. | ['Jianyu Chen', 'Jingyue Gao', 'Yen-Jen Wang', 'Zheyuan Jiang', 'Yanjiang Guo'] | 2023-06-30 | null | null | null | null | ['reinforcement-learning-1'] | ['methodology'] | [ 1.91659499e-02 5.10648131e-01 -3.30180019e-01 2.85466373e-01
1.77031830e-01 -5.03058553e-01 4.53046471e-01 -4.53911394e-01
-4.91570264e-01 1.26005900e+00 -2.05199718e-01 2.19148353e-01
-2.94791996e-01 -5.49663186e-01 -1.01256204e+00 -1.27758420e+00
-2.12193757e-01 3.27657282e-01 4.07387614e-01 -4.64987248e-01
1.04801320e-01 4.80634779e-01 -1.49637389e+00 -3.23130518e-01
9.35258687e-01 2.79286712e-01 8.19202244e-01 6.08800530e-01
6.54757380e-01 1.17903996e+00 -5.10882139e-01 6.97408736e-01
3.17157596e-01 -8.00332606e-01 -7.22848952e-01 1.81369826e-01
-2.85585910e-01 -1.20199785e-01 -6.55335248e-01 9.02496994e-01
6.35231614e-01 2.29777709e-01 6.78300261e-01 -1.44256759e+00
-4.37272102e-01 8.34668219e-01 -3.70144635e-01 -3.61780912e-01
-2.56719083e-01 5.63329279e-01 7.61936545e-01 -5.47853857e-02
7.90204644e-01 1.06342614e+00 3.33804995e-01 9.00194347e-01
-1.41704082e+00 -6.89069331e-01 2.29354188e-01 1.23128414e-01
-1.06742942e+00 -8.02044123e-02 6.76321685e-01 -2.59963870e-01
9.45627034e-01 -5.52189350e-01 8.69129837e-01 1.30771875e+00
5.89320004e-01 7.87130773e-01 8.98829162e-01 -1.42879441e-01
5.13693929e-01 -6.27526164e-01 -7.14492440e-01 7.73358941e-01
3.65731120e-01 3.81276578e-01 -6.36294544e-01 1.85370177e-01
1.22163939e+00 -6.58036098e-02 -1.69300869e-01 -9.94379282e-01
-1.70350254e+00 6.13834977e-01 7.07829535e-01 2.98968315e-01
-6.41101837e-01 6.94020092e-01 3.16063583e-01 7.56284297e-01
-3.77493411e-01 7.16591537e-01 -7.48010874e-01 2.91386303e-02
-2.46607617e-01 5.37645936e-01 7.21025527e-01 8.23743403e-01
7.58758724e-01 5.13546705e-01 2.11234301e-01 7.37517953e-01
-1.25653014e-01 6.73929870e-01 6.61212683e-01 -1.46520197e+00
5.37439361e-02 4.86659437e-01 4.59901504e-02 -9.01369035e-01
-7.84434199e-01 -2.72490889e-01 -9.69431758e-01 7.90034473e-01
3.32261175e-01 -5.75367808e-01 -6.62961900e-01 2.09733558e+00
1.78422347e-01 -2.02368736e-01 1.96910158e-01 7.93276072e-01
-1.70283020e-01 4.14963543e-01 -2.66600735e-02 -3.90434153e-02
6.35575652e-01 -1.00428104e+00 -4.32188779e-01 -3.91436756e-01
7.13468790e-01 5.68829738e-02 7.56019771e-01 4.80830312e-01
-8.89985204e-01 -2.93560982e-01 -1.20014179e+00 4.38250512e-01
-1.66887060e-01 1.87203940e-02 2.52922982e-01 -3.61682504e-01
-9.32610691e-01 9.97229934e-01 -1.16274142e+00 -7.04705358e-01
1.71060547e-01 7.72451341e-01 -4.52392489e-01 3.36745948e-01
-9.87136781e-01 1.18115532e+00 8.32184434e-01 -2.28012621e-01
-1.36256695e+00 -1.23989336e-01 -4.20465380e-01 -1.95061788e-01
6.88281775e-01 -1.01481700e+00 1.27560854e+00 -8.65578830e-01
-2.18236899e+00 4.24166501e-01 3.82465154e-01 -5.61571658e-01
1.98251784e-01 -8.57693180e-02 2.42265284e-01 3.54843110e-01
3.00843149e-01 9.66337204e-01 1.12544370e+00 -1.32996500e+00
-7.84438074e-01 -1.49712443e-01 -8.14964250e-02 3.89656842e-01
-3.75072986e-01 -4.94541585e-01 1.45394206e-01 -6.98559582e-01
8.25924277e-02 -1.33853996e+00 -3.78202140e-01 8.49135891e-02
-1.55741185e-01 -2.51637995e-01 7.77538359e-01 -1.28046155e-01
7.38314271e-01 -1.81186891e+00 1.15438688e+00 -6.19509816e-03
1.24670818e-01 3.79530005e-02 -2.16455370e-01 7.82665908e-01
2.16038108e-01 -3.05827439e-01 -2.56429583e-01 3.86739969e-01
-7.22820014e-02 8.73074353e-01 -1.49957895e-01 5.35720885e-01
5.31580299e-02 9.29094553e-01 -1.23784697e+00 -3.75162959e-01
-1.56615786e-02 1.44039303e-01 -6.85509980e-01 3.48941147e-01
-3.02576482e-01 1.10141802e+00 -6.22678816e-01 5.11628866e-01
-1.47818819e-01 -1.75174549e-01 6.37081861e-01 2.58183777e-01
-1.56065762e-01 -1.88208908e-01 -8.72425556e-01 1.70253766e+00
-4.44271743e-01 2.99135238e-01 4.37277347e-01 -1.64586866e+00
8.58651698e-01 4.39526260e-01 9.10783708e-01 -5.11789322e-01
2.50886679e-01 3.36512774e-01 4.73629534e-01 -5.31741917e-01
-4.39430848e-02 6.62077814e-02 -3.20045918e-01 4.89383459e-01
5.06003082e-01 -3.55933636e-01 1.66035190e-01 -1.40801057e-01
1.42357647e+00 6.57700062e-01 5.75743854e-01 -3.92598599e-01
2.36517683e-01 3.13018858e-01 8.37563872e-01 8.42548788e-01
-3.06207418e-01 1.35854632e-01 3.50694627e-01 -1.20292082e-01
-1.10214269e+00 -1.21120036e+00 5.04759371e-01 1.32534397e+00
2.77295321e-01 -3.33200321e-02 -7.36905158e-01 -3.30449194e-01
2.35458702e-01 2.02292979e-01 -5.06225169e-01 -5.81095159e-01
-1.06870532e+00 -1.41229719e-01 5.65028310e-01 5.22383511e-01
5.61819077e-01 -1.82592344e+00 -1.16154325e+00 6.56158507e-01
-1.03497297e-01 -8.59486938e-01 -1.68978259e-01 6.80962741e-01
-9.27851796e-01 -1.15986848e+00 -7.65209198e-01 -1.10751843e+00
7.48825729e-01 1.57317117e-01 6.09478593e-01 9.09864381e-02
-1.07908435e-01 3.06400597e-01 -3.21621448e-01 -1.87067345e-01
-3.99196863e-01 2.67577261e-01 6.73113227e-01 -4.72542465e-01
-3.09400082e-01 -1.24565279e+00 -3.94766778e-01 6.04469538e-01
-7.03076959e-01 3.58214206e-03 8.99524748e-01 1.38333547e+00
4.03640032e-01 1.51194289e-01 8.98875952e-01 -4.77936044e-02
6.48964107e-01 -4.13518727e-01 -5.18580556e-01 -4.58259555e-03
-4.02386129e-01 3.57937843e-01 1.12367272e+00 -8.02650809e-01
-7.81561255e-01 4.69501406e-01 5.44215679e-01 -2.60415375e-01
-3.72502834e-01 1.75107971e-01 1.15785867e-01 -9.50568318e-02
6.58345044e-01 5.96946895e-01 4.31255639e-01 -2.88898021e-01
5.02594292e-01 4.47901070e-01 7.38907099e-01 -8.83755803e-01
8.47802222e-01 3.06105584e-01 2.34710589e-01 -8.44399989e-01
-1.44508824e-01 -5.88710308e-02 -8.98423791e-01 -4.60699528e-01
7.41124213e-01 -8.19599748e-01 -1.11276317e+00 7.30704904e-01
-8.60220194e-01 -9.30800974e-01 -3.26386213e-01 4.39991504e-01
-1.49275088e+00 1.29719585e-01 -6.32052064e-01 -5.25624752e-01
7.79804885e-02 -9.09332275e-01 7.28551626e-01 2.12154433e-01
-4.21060890e-01 -5.97640932e-01 2.78973341e-01 -3.97395104e-01
2.53529042e-01 2.44028509e-01 8.39626312e-01 -1.87396370e-02
-3.70458841e-01 2.71308601e-01 4.22901899e-01 2.54158258e-01
3.46533597e-01 -1.88402772e-01 -3.19577783e-01 -7.44556069e-01
-2.36619666e-01 -8.28574359e-01 4.76393670e-01 3.11640561e-01
7.52317190e-01 -3.25672418e-01 -6.08830571e-01 2.25100160e-01
1.17963743e+00 3.21238965e-01 2.83478230e-01 6.15311027e-01
4.86232102e-01 8.09805572e-01 4.92803007e-01 2.29318142e-01
1.17210206e-02 4.78567868e-01 7.51845658e-01 3.24811637e-01
3.66630778e-02 -4.77721512e-01 7.62721658e-01 9.52619970e-01
-3.91944349e-01 2.95939092e-02 -6.09381139e-01 6.67507350e-01
-2.44063401e+00 -8.33128214e-01 4.90052551e-01 1.75233293e+00
8.40188205e-01 -4.60803211e-02 4.07052487e-01 7.07383901e-02
6.55749917e-01 -2.83247232e-01 -1.09100544e+00 -1.43798485e-01
-1.78339422e-01 -1.18860245e-01 6.94313228e-01 -4.59444299e-02
-7.84331143e-01 1.17453885e+00 6.84846449e+00 6.09819889e-01
-1.14427531e+00 -2.65814573e-01 -3.27647597e-01 -1.41343236e-01
4.23369795e-01 -9.57227275e-02 -4.17233676e-01 5.30903459e-01
4.41206902e-01 -2.16955096e-01 8.76451552e-01 1.00237763e+00
2.90208727e-01 -3.70070413e-02 -8.94104838e-01 6.07448459e-01
-4.55458432e-01 -9.37910736e-01 -1.96663111e-01 2.32127503e-01
9.44543004e-01 3.62148076e-01 -1.67984679e-01 2.47081175e-01
1.04038405e+00 -8.54831874e-01 6.71225667e-01 4.89862323e-01
3.45318347e-01 -7.65883386e-01 1.60316557e-01 9.41129267e-01
-1.01307154e+00 -7.41699398e-01 -4.13210660e-01 -4.02442634e-01
1.18966073e-01 -8.14641789e-02 -4.44459945e-01 3.75193357e-01
7.40490437e-01 1.00114179e+00 -1.40037417e-01 6.59451485e-01
-7.31017888e-01 4.86650795e-01 -4.25927937e-01 -4.72769439e-01
3.18597823e-01 -2.59755254e-01 8.40476632e-01 5.09609163e-01
1.46100044e-01 -1.03840508e-01 2.58793443e-01 7.70306110e-01
2.28224427e-01 -2.41759509e-01 -1.08553469e+00 -7.30653629e-02
2.79977620e-01 1.05640996e+00 -6.81266904e-01 -5.81038035e-02
-5.34351170e-02 9.74003971e-01 7.90813804e-01 4.33806807e-01
-6.10052764e-01 -4.23750669e-01 9.40867543e-01 -3.32333535e-01
8.54818106e-01 -4.21545595e-01 2.64354069e-02 -1.03764033e+00
-1.28747076e-01 -8.70017529e-01 -1.38630599e-01 -6.37362659e-01
-1.30328143e+00 1.06550589e-01 -9.20263156e-02 -1.67186892e+00
-8.14529002e-01 -5.98224461e-01 -4.78443652e-01 1.11277185e-01
-1.07250822e+00 -8.77795100e-01 1.11556746e-01 7.99036205e-01
1.97315544e-01 -4.49335337e-01 8.47395539e-01 -2.44944200e-01
-4.77430373e-01 2.21493885e-01 4.98876750e-01 1.15154264e-02
6.79924905e-01 -1.23986864e+00 -2.44353384e-01 5.81994414e-01
-4.26243663e-01 4.63584661e-01 7.94847429e-01 -6.70866609e-01
-1.27523160e+00 -1.06025088e+00 3.06010395e-01 1.65111616e-01
9.52995300e-01 -1.60234809e-01 -5.23614526e-01 6.37578547e-01
4.25702304e-01 -2.56306469e-01 -1.11075260e-01 -2.65837044e-01
6.25729710e-02 -1.28808483e-01 -9.32228684e-01 1.15230107e+00
1.36868811e+00 -6.66909143e-02 -7.96795726e-01 1.47418752e-01
6.28770471e-01 4.92366068e-02 -8.30187440e-01 4.78272110e-01
8.50870371e-01 -5.45046210e-01 8.52227986e-01 -9.64757681e-01
4.69043285e-01 -6.05053604e-01 1.07205532e-01 -1.91222835e+00
-6.39221191e-01 -7.55779624e-01 -2.63758421e-01 5.62665045e-01
5.34981117e-02 -7.98349738e-01 6.31109595e-01 -3.72890681e-01
-1.00779183e-01 -5.43679535e-01 -9.35570598e-01 -1.33606410e+00
5.98547876e-01 4.88056749e-01 1.49281368e-01 8.08532953e-01
5.87352812e-01 3.84096891e-01 -7.99321651e-01 1.16589837e-01
6.68322623e-01 1.15633443e-01 1.01132250e+00 -9.42973197e-01
-5.14032304e-01 -5.34406602e-01 -3.78675014e-01 -1.12763679e+00
4.11875278e-01 -7.87416816e-01 6.31838202e-01 -1.34409976e+00
-1.64626986e-01 -4.88344282e-02 -3.12562495e-01 8.90332401e-01
2.71638930e-01 -6.87271804e-02 2.17440516e-01 3.39120746e-01
-5.19645333e-01 7.53397167e-01 1.58914256e+00 -1.11620478e-01
-3.18938196e-01 -1.31879404e-01 -4.19347107e-01 8.44371557e-01
1.40517139e+00 -7.39327610e-01 -4.42678124e-01 -3.34656984e-01
9.29532796e-02 6.43382445e-02 3.05263340e-01 -1.31586111e+00
5.17831385e-01 -4.53617603e-01 -1.72031764e-02 -2.29358345e-01
-1.32738380e-03 -8.40281606e-01 -1.01035431e-01 1.23692584e+00
-2.90601611e-01 -1.06286943e-01 -2.40439206e-01 9.15640116e-01
-1.58526748e-01 1.49251461e-01 8.55569661e-01 -3.28334004e-01
-6.80117965e-01 -6.35352209e-02 -1.22163081e+00 -1.31900221e-01
1.27153826e+00 -2.13326328e-02 -1.92572951e-01 -2.78334498e-01
-8.24842691e-01 5.91656625e-01 5.33690512e-01 4.10971761e-01
2.82703131e-01 -1.33200991e+00 -4.19861376e-01 -4.09066118e-02
1.30457997e-01 6.49136528e-02 -1.45119159e-02 7.87769258e-01
-4.21994984e-01 3.56964558e-01 -1.05226982e+00 -7.22005606e-01
-8.94436538e-01 5.56704998e-01 4.95678812e-01 -1.92163840e-01
-7.94710934e-01 2.80102104e-01 1.87289819e-01 -9.56367731e-01
1.09951206e-01 -2.26178840e-01 -1.72315210e-01 -5.56202948e-01
-7.94324949e-02 3.83242279e-01 -6.15425110e-01 -3.81622881e-01
-1.00544721e-01 6.84457421e-01 3.23415637e-01 -1.29937217e-01
1.54182684e+00 -1.20313346e-01 -1.91418156e-01 3.71199042e-01
6.89597487e-01 -5.50687015e-01 -1.71536183e+00 -2.24864811e-01
-5.35246171e-02 2.23635882e-01 -4.39953357e-01 -5.37970781e-01
-1.06003273e+00 5.11980832e-01 1.17667735e-01 6.16640262e-02
1.01925945e+00 -1.08333919e-02 5.40940762e-01 1.17806852e+00
9.20213282e-01 -1.69931924e+00 4.82257336e-01 7.40667105e-01
9.68997478e-01 -9.29842174e-01 -1.94837853e-01 1.16996862e-01
-6.34931207e-01 1.04669809e+00 9.02604580e-01 -7.64653087e-01
5.24466038e-01 4.56164807e-01 -9.75489542e-02 5.35221323e-02
-9.76491153e-01 -3.01858187e-01 -4.04635936e-01 1.02139246e+00
-2.90970862e-01 -1.33410757e-02 -3.03089261e-01 2.88701832e-01
-1.16514098e-02 1.97979346e-01 4.42452550e-01 1.51371419e+00
-8.39026570e-01 -1.07714081e+00 -1.35652348e-01 1.06267929e-01
6.58754706e-02 5.28459013e-01 -4.47735995e-01 1.05831182e+00
1.24370597e-01 7.82184422e-01 -2.32831091e-01 -5.84273219e-01
3.14195842e-01 -1.37692645e-01 7.24689960e-01 -5.43683112e-01
-3.07172716e-01 2.34366767e-02 -3.49830300e-01 -7.83959270e-01
-6.39638603e-01 -5.85672140e-01 -1.64818704e+00 -1.11636080e-01
1.48197624e-03 -7.35018449e-03 2.22779647e-01 9.16731954e-01
3.98896992e-01 7.27240443e-01 7.11870432e-01 -1.09810936e+00
-9.37846363e-01 -1.05254877e+00 -8.61792207e-01 2.94633836e-01
3.78633410e-01 -1.16040289e+00 -3.44993979e-01 2.85463810e-01] | [4.380160808563232, 1.2929694652557373] |
9d80e25c-2597-4d84-86f7-2c3a7681a781 | task-robust-pre-training-for-worst-case | 2306.12070 | null | https://arxiv.org/abs/2306.12070v2 | https://arxiv.org/pdf/2306.12070v2.pdf | Task-Robust Pre-Training for Worst-Case Downstream Adaptation | Pre-training has achieved remarkable success when transferred to downstream tasks. In machine learning, we care about not only the good performance of a model but also its behavior under reasonable shifts of condition. The same philosophy holds when pre-training a foundation model. However, the foundation model may not uniformly behave well for a series of related downstream tasks. This happens, for example, when conducting mask recovery regression where the recovery ability or the training instances diverge like pattern features are extracted dominantly on pre-training, but semantic features are also required on a downstream task. This paper considers pre-training a model that guarantees a uniformly good performance over the downstream tasks. We call this goal as $\textit{downstream-task robustness}$. Our method first separates the upstream task into several representative ones and applies a simple minimax loss for pre-training. We then design an efficient algorithm to solve the minimax loss and prove its convergence in the convex setting. In the experiments, we show both on large-scale natural language processing and computer vision datasets our method increases the metrics on worse-case downstream tasks. Additionally, some theoretical explanations for why our loss is beneficial are provided. Specifically, we show fewer samples are inherently required for the most challenging downstream task in some cases. | ['Yang Chen', 'Zhouchen Lin', 'Cong Fang', 'Xingyu Xie', 'Jianghui Wang'] | 2023-06-21 | null | null | null | null | ['philosophy'] | ['miscellaneous'] | [ 4.85668659e-01 4.14833218e-01 2.76540760e-02 -4.29931641e-01
-1.14222968e+00 -4.98100907e-01 4.83632296e-01 1.67565614e-01
-5.54186285e-01 5.28598368e-01 1.92765549e-01 -3.55706871e-01
-3.27030927e-01 -4.23811406e-01 -8.26896012e-01 -1.04946828e+00
-1.30059302e-01 2.23925769e-01 9.76272747e-02 -2.41913199e-01
2.86546707e-01 3.53350401e-01 -1.32128954e+00 3.91329497e-01
7.66410530e-01 1.00307167e+00 3.36062312e-01 5.39853454e-01
3.23692501e-01 4.84990448e-01 -4.49905336e-01 -4.44824666e-01
7.70340502e-01 -2.58179754e-01 -1.00425112e+00 2.00624108e-01
3.08933198e-01 -1.99302705e-03 -1.01300813e-01 1.13664389e+00
4.90874767e-01 2.33018652e-01 8.90554905e-01 -1.32272184e+00
-4.93926466e-01 8.14736128e-01 -5.81847072e-01 3.09813291e-01
-2.81435877e-01 2.69449234e-01 1.07618117e+00 -1.02116799e+00
3.22581232e-01 1.13354111e+00 9.68620360e-01 4.24773574e-01
-1.23296905e+00 -3.48573506e-01 5.07534921e-01 2.46563852e-02
-1.02829683e+00 -5.50378382e-01 3.74484986e-01 -5.14395177e-01
7.72005439e-01 8.40239972e-02 -1.44048352e-02 1.21542847e+00
9.71050411e-02 9.28899765e-01 1.04211295e+00 -2.78683245e-01
1.17735185e-01 1.71818540e-01 3.46809328e-01 5.10312796e-01
1.47739187e-01 2.09128916e-01 -3.54771763e-01 -4.30443250e-02
5.01167953e-01 -3.38375717e-02 -4.79484379e-01 -6.36503696e-02
-1.04854858e+00 7.86455154e-01 4.32292372e-01 2.43659914e-01
-1.15977749e-01 2.37631220e-02 5.21479011e-01 6.43912554e-01
6.47249341e-01 4.04437363e-01 -7.71757007e-01 4.84363921e-02
-8.71022046e-01 6.84592798e-02 6.32565200e-01 1.19902110e+00
5.69745719e-01 -3.66426706e-01 -4.08358991e-01 7.83295512e-01
1.61045521e-01 1.07370444e-01 5.95151544e-01 -9.74252760e-01
8.56133938e-01 2.23984540e-01 -2.51175258e-02 -6.82457924e-01
-3.48592907e-01 -6.68075085e-01 -9.17746007e-01 1.31907091e-01
1.03508055e+00 -9.23476368e-02 -8.34993184e-01 2.19528389e+00
1.05144009e-01 -1.13994651e-01 8.47479850e-02 8.15164149e-01
1.72315434e-01 6.41693413e-01 6.64424896e-02 -3.93685073e-01
1.12769270e+00 -1.07347727e+00 -3.60760689e-01 -4.33866322e-01
1.00477028e+00 -7.28028059e-01 1.26919758e+00 4.97301579e-01
-9.09937263e-01 -5.03556430e-01 -8.92628491e-01 -2.35208720e-01
-2.32171923e-01 2.23409623e-01 3.43994796e-01 2.54206419e-01
-9.94697213e-01 1.08250892e+00 -6.12900317e-01 -3.40050012e-01
4.07388121e-01 3.63460571e-01 -4.56702828e-01 -1.84465408e-01
-9.69574690e-01 8.73270690e-01 2.89996684e-01 1.66948915e-01
-9.65235770e-01 -8.55344892e-01 -7.03011930e-01 2.72077888e-01
3.69772881e-01 -6.76640689e-01 1.31845009e+00 -1.11835098e+00
-1.24419975e+00 9.52617943e-01 -1.70412973e-01 -4.63006139e-01
8.95693004e-01 -4.65932220e-01 1.56776056e-01 -7.60187507e-02
2.83412158e-01 3.64953876e-01 1.22428191e+00 -1.22284544e+00
-7.22032607e-01 -5.28009653e-01 1.35610476e-01 1.49764687e-01
-3.38229984e-01 -2.05862492e-01 -4.07060325e-01 -7.27226079e-01
-6.03408515e-02 -7.67847300e-01 -4.69150424e-01 1.37507766e-01
-4.17499542e-01 -4.17649299e-01 6.01653218e-01 -2.67125279e-01
1.05437279e+00 -2.43062806e+00 -6.52536675e-02 1.37171432e-01
1.29936278e-01 2.05906987e-01 -3.39289546e-01 3.57786089e-01
-2.99722701e-01 6.81053475e-02 -6.29106700e-01 -6.81740284e-01
-3.87182981e-02 7.35141411e-02 -6.67250156e-01 9.04051423e-01
2.41260633e-01 6.93251312e-01 -8.15377772e-01 -3.51835012e-01
-1.17326982e-01 7.15917572e-02 -6.22858763e-01 9.90831107e-02
1.45732149e-01 3.49488258e-01 -4.95160043e-01 1.97595388e-01
5.55722117e-01 -3.72643381e-01 -2.03645080e-01 2.17949245e-02
5.64762726e-02 1.52011991e-01 -9.58203495e-01 1.70451975e+00
-6.42179251e-01 6.02218211e-01 4.14557010e-01 -1.59923255e+00
6.34990335e-01 1.25867978e-01 3.41342181e-01 -3.28735650e-01
2.21294835e-01 2.26502791e-01 -2.03820355e-02 -6.06420040e-01
7.96813518e-02 -6.24845862e-01 -2.10698664e-01 4.50741708e-01
-6.22416101e-02 3.41539708e-04 6.70933872e-02 7.37704709e-02
1.22927558e+00 5.42162359e-02 3.36495966e-01 -6.24273479e-01
5.29427648e-01 1.56277299e-01 6.48588777e-01 1.01609647e+00
-4.39012706e-01 7.44217396e-01 5.83054841e-01 5.39138168e-03
-8.55529547e-01 -1.01466250e+00 -2.07021400e-01 1.24316585e+00
1.18098959e-01 -2.57589787e-01 -6.46569252e-01 -8.96942854e-01
8.46272707e-02 5.54016471e-01 -6.32565022e-01 -6.41074002e-01
-5.41494071e-01 -9.29047048e-01 5.35226643e-01 5.92880905e-01
3.00430894e-01 -8.92019451e-01 -3.48523319e-01 -3.02783698e-02
-1.14899650e-02 -9.95803952e-01 -6.38003707e-01 5.57055295e-01
-1.09093046e+00 -1.11100364e+00 -7.35374689e-01 -9.30235147e-01
7.57278383e-01 4.02028173e-01 1.04398811e+00 6.58721765e-05
-1.30171984e-01 4.35363613e-02 -3.00507188e-01 -3.18175942e-01
-4.23525035e-01 7.17055947e-02 7.12214038e-02 1.49524570e-01
2.62190759e-01 -6.63250268e-01 -5.08535206e-01 2.81976104e-01
-7.94458926e-01 -2.25166440e-01 9.12440240e-01 9.24616516e-01
5.51553786e-01 5.81665263e-02 6.81178331e-01 -1.03253031e+00
5.80669224e-01 -5.98645985e-01 -3.38719875e-01 1.90977722e-01
-7.41628230e-01 2.47303307e-01 1.17828906e+00 -3.72845173e-01
-9.29430485e-01 1.35145441e-01 8.35350074e-04 -6.02729201e-01
-1.67233452e-01 2.96139032e-01 -2.42926553e-01 4.31871057e-01
8.11033845e-01 8.10125768e-02 9.63461846e-02 -6.01180494e-01
4.17797208e-01 5.69723427e-01 3.66929829e-01 -7.66385436e-01
1.05053914e+00 7.77226686e-01 4.57766503e-02 -8.67782414e-01
-1.34634197e+00 -5.93956113e-01 -5.57231724e-01 1.58387154e-01
5.79896867e-01 -7.42977977e-01 -6.84003651e-01 3.37878019e-01
-1.07195973e+00 -6.81384087e-01 -6.28282964e-01 2.96128660e-01
-7.94559240e-01 4.38038260e-01 -5.68081498e-01 -7.60377467e-01
-8.42717364e-02 -1.04642642e+00 1.24557269e+00 -2.56318867e-01
-8.21247604e-03 -1.06008756e+00 -1.94824755e-01 1.98272660e-01
1.22106947e-01 -4.41088751e-02 9.13782299e-01 -7.80910969e-01
-1.43567637e-01 2.92999372e-02 -3.32701534e-01 6.54975653e-01
1.15773968e-01 -4.14447010e-01 -1.16531265e+00 -3.09437037e-01
4.87023741e-01 -3.22577089e-01 1.40451503e+00 5.52255809e-01
1.46196866e+00 -4.44627672e-01 -4.19072539e-01 7.83916771e-01
1.38009751e+00 -2.33357862e-01 2.67621696e-01 1.75821096e-01
5.35993099e-01 1.24332201e+00 8.90288949e-01 2.93236464e-01
-7.09946230e-02 5.06980240e-01 2.39847645e-01 4.59831432e-02
9.07134712e-02 -2.21239641e-01 5.95985591e-01 4.62180346e-01
2.58736134e-01 -1.48015991e-01 -7.18982458e-01 6.36157036e-01
-2.01968884e+00 -8.35790932e-01 -2.51200825e-01 2.17563629e+00
9.23845470e-01 1.29479691e-01 5.71671613e-02 3.42783898e-01
8.59310389e-01 -1.53475538e-01 -5.86852133e-01 -2.93703198e-01
-6.18551150e-02 1.02542102e-01 6.16453528e-01 5.50182521e-01
-1.29074991e+00 9.61929798e-01 6.13527679e+00 1.10324109e+00
-9.99044240e-01 3.74756932e-01 8.84238303e-01 -1.94647014e-01
-7.17964619e-02 5.54544702e-02 -7.76353121e-01 3.35347325e-01
6.88737392e-01 -3.01978350e-01 1.49169207e-01 1.10756803e+00
5.15683413e-01 1.05246693e-01 -1.55509555e+00 8.88518751e-01
-2.57733047e-01 -1.03421533e+00 -1.52747899e-01 9.00486335e-02
5.33014894e-01 -7.09829945e-03 2.46478632e-01 4.72818285e-01
1.09260425e-01 -1.10604072e+00 9.81186688e-01 -1.94139369e-02
8.85432839e-01 -6.60002708e-01 5.17244935e-01 7.50195086e-01
-8.62628400e-01 -3.73673975e-01 -5.26976943e-01 -2.33220801e-01
5.27694039e-02 9.46251035e-01 -5.18578649e-01 3.77273470e-01
5.61310887e-01 7.31166244e-01 -3.17908615e-01 8.74029040e-01
-2.78910726e-01 5.05039692e-01 -3.51846427e-01 3.16105813e-01
3.73460561e-01 -2.99106479e-01 5.80724001e-01 1.48489392e+00
1.76857024e-01 2.20238436e-02 8.76040384e-02 6.59886420e-01
-9.06368196e-02 1.03460424e-01 -8.83129895e-01 1.91653207e-01
2.23264888e-01 1.09291422e+00 -6.12122715e-01 -1.97808594e-01
-2.60084450e-01 1.05729902e+00 6.17591500e-01 3.86789709e-01
-7.81049907e-01 -1.84718832e-01 9.59466159e-01 5.70400394e-02
2.08277121e-01 -3.76529805e-02 -6.73054814e-01 -1.16680014e+00
2.28531525e-01 -6.48419559e-01 6.01515174e-01 -2.00785086e-01
-1.73486674e+00 4.22575444e-01 -4.33466658e-02 -1.27800250e+00
-1.04750551e-01 -8.43346536e-01 -5.90791643e-01 7.72709608e-01
-1.67646968e+00 -7.69886792e-01 1.59151331e-02 6.87205493e-01
9.04479623e-01 1.74669176e-01 3.85939240e-01 2.21468613e-01
-5.59856772e-01 8.15449774e-01 1.08595528e-01 6.62993193e-02
8.53552997e-01 -1.31413317e+00 4.98952568e-02 1.18419683e+00
3.65764159e-03 6.37228906e-01 8.10973644e-01 -3.14860314e-01
-1.17666376e+00 -1.38518357e+00 8.44362319e-01 -4.02836412e-01
7.70238101e-01 -2.77146459e-01 -7.69527376e-01 7.94691682e-01
-3.09145808e-01 1.37531847e-01 3.49236727e-01 1.16432279e-01
-3.63627732e-01 -1.42535076e-01 -1.10346818e+00 6.86194599e-01
1.50968456e+00 -3.38398904e-01 -7.55851328e-01 6.07112229e-01
6.63534820e-01 -6.60815835e-02 -4.88036573e-01 4.66672570e-01
1.04325943e-01 -1.12327695e+00 8.77929807e-01 -9.01772678e-01
7.17883050e-01 -4.44157124e-02 -1.05811924e-01 -1.26774776e+00
-2.83533245e-01 -8.64348471e-01 9.49750021e-02 1.16729498e+00
6.59078777e-01 -6.61244094e-01 6.88215733e-01 5.08653045e-01
-4.46168244e-01 -8.54630828e-01 -1.00551581e+00 -1.19106162e+00
5.11517048e-01 -8.09363246e-01 -5.73023297e-02 8.59605491e-01
-3.63358632e-02 4.47591305e-01 -3.35013196e-02 1.70026988e-01
6.58353925e-01 1.59805864e-01 5.60626328e-01 -9.84697700e-01
-5.58748484e-01 -7.62791872e-01 -9.34993997e-02 -1.49802423e+00
1.92713484e-01 -1.26484787e+00 4.82447326e-01 -1.16914427e+00
2.02634200e-01 -6.54879093e-01 -2.00001702e-01 4.58079487e-01
-4.59366769e-01 -4.85936254e-02 2.21265435e-01 3.40277225e-01
-2.52447784e-01 5.67720890e-01 1.26114309e+00 -3.30457613e-02
-1.92656636e-01 4.16372597e-01 -1.11056328e+00 8.44175756e-01
7.64845192e-01 -6.29181087e-01 -5.81232190e-01 -5.31809866e-01
1.14636667e-01 -2.82107055e-01 2.79209882e-01 -5.68589151e-01
2.18706746e-02 -4.04137634e-02 5.66387698e-02 3.33484896e-02
2.63109446e-01 -9.28617358e-01 -5.07471085e-01 5.82143426e-01
-5.59897959e-01 -2.19645143e-01 -2.94067524e-02 8.51246059e-01
-1.02303080e-01 -3.95369709e-01 1.08808899e+00 -1.04686320e-01
-4.24820662e-01 4.02963281e-01 -3.08193058e-01 3.60817939e-01
1.07780468e+00 -2.88533300e-01 -2.11484015e-01 -2.63124883e-01
-9.58933771e-01 3.27105165e-01 3.08225930e-01 3.46058249e-01
4.29509729e-01 -9.54445839e-01 -8.50542247e-01 1.05617329e-01
9.54771191e-02 2.44687542e-01 -2.03093551e-02 1.32947218e+00
-1.02002569e-01 2.48925149e-01 2.53049493e-01 -6.19223177e-01
-1.08733439e+00 7.33449876e-01 1.85760051e-01 -5.07428348e-01
-7.62468100e-01 1.22846460e+00 8.81177008e-01 -3.26862633e-02
3.81313086e-01 -5.27841210e-01 -4.78544692e-03 1.60375252e-01
4.18433994e-01 3.00827503e-01 5.83090372e-02 -3.72771323e-01
-4.02503431e-01 4.19295758e-01 6.17496995e-03 -1.21970423e-01
1.50633538e+00 -2.35402197e-01 7.79862329e-02 4.58768487e-01
1.58755517e+00 -8.51183161e-02 -1.53205550e+00 -2.57386267e-01
3.15752745e-01 -2.60939389e-01 -2.18489960e-01 -2.94177771e-01
-9.92354333e-01 1.06723344e+00 1.41043812e-01 3.05538028e-01
1.34985566e+00 1.83662057e-01 6.52874529e-01 4.29643840e-01
2.81137675e-01 -1.13951492e+00 -4.42989133e-02 7.03783453e-01
1.03393531e+00 -1.32588172e+00 -2.10657611e-01 -5.10605991e-01
-5.90908766e-01 1.07623053e+00 3.87695163e-01 -5.00975192e-01
8.69298220e-01 3.08775067e-01 -1.48392424e-01 -1.72420904e-01
-8.30155849e-01 -2.70341635e-01 9.51640084e-02 4.42268670e-01
1.90166339e-01 -2.15793148e-01 -4.82291073e-01 8.22193265e-01
-2.79056996e-01 -3.52656573e-01 3.28179330e-01 7.52335966e-01
-4.81955707e-01 -8.38579714e-01 -1.81691870e-01 3.21294844e-01
-6.55349672e-01 -2.75854498e-01 -4.65567440e-01 8.17317009e-01
7.29896575e-02 1.04284978e+00 -9.85610634e-02 -2.86225051e-01
2.16445953e-01 9.58872661e-02 4.62795615e-01 -8.48334432e-01
-7.01166749e-01 -6.61429316e-02 4.88444604e-02 -7.61668861e-01
-3.06348383e-01 -6.25362039e-01 -1.10957491e+00 -2.76942879e-01
-1.70613393e-01 1.33559510e-01 3.54212433e-01 1.31910455e+00
2.96933591e-01 2.63070434e-01 7.30228603e-01 -7.77019560e-01
-1.16993916e+00 -9.81947362e-01 -6.26453102e-01 5.90432227e-01
4.13392961e-01 -4.61362869e-01 -9.27461982e-01 1.85606897e-01] | [9.326844215393066, 3.4979746341705322] |
05fb7e8d-b9be-4015-b327-636e81f9ea30 | interactive-data-synthesis-for-systematic | 2305.12799 | null | https://arxiv.org/abs/2305.12799v1 | https://arxiv.org/pdf/2305.12799v1.pdf | Interactive Data Synthesis for Systematic Vision Adaptation via LLMs-AIGCs Collaboration | Recent text-to-image generation models have shown promising results in generating high-fidelity photo-realistic images. In parallel, the problem of data scarcity has brought a growing interest in employing AIGC technology for high-quality data expansion. However, this paradigm requires well-designed prompt engineering that cost-less data expansion and labeling remain under-explored. Inspired by LLM's powerful capability in task guidance, we propose a new paradigm of annotated data expansion named as ChatGenImage. The core idea behind it is to leverage the complementary strengths of diverse models to establish a highly effective and user-friendly pipeline for interactive data augmentation. In this work, we extensively study how LLMs communicate with AIGC model to achieve more controllable image generation and make the first attempt to collaborate them for automatic data augmentation for a variety of downstream tasks. Finally, we present fascinating results obtained from our ChatGenImage framework and demonstrate the powerful potential of our synthetic data for systematic vision adaptation. Our codes are available at https://github.com/Yuqifan1117/Labal-Anything-Pipeline. | ['Yueting Zhuang', 'Siliang Tang', 'Wentao Ye', 'Juncheng Li', 'Qifan Yu'] | 2023-05-22 | null | null | null | null | ['prompt-engineering'] | ['natural-language-processing'] | [ 2.50808299e-01 2.59452552e-01 2.17288062e-01 -3.41698378e-01
-9.12916243e-01 -3.81299645e-01 7.33621597e-01 -2.69837409e-01
-2.23266751e-01 5.50932884e-01 3.30976665e-01 -3.32541257e-01
3.02446902e-01 -5.18211842e-01 -8.85311782e-01 -5.26442230e-01
3.26479673e-01 4.29753006e-01 -4.14285921e-02 -3.75727713e-01
1.91204950e-01 3.79005909e-01 -1.65767109e+00 2.87204295e-01
1.08657384e+00 5.74696839e-01 5.65202236e-01 6.86679184e-01
1.34247899e-01 5.33926427e-01 -2.95911402e-01 -4.06175286e-01
4.76791054e-01 -5.74880898e-01 -8.73559415e-01 2.31846556e-01
5.92527390e-01 -3.38974535e-01 -2.82748163e-01 6.93350255e-01
8.23202550e-01 5.87421432e-02 3.42768461e-01 -1.15585160e+00
-9.35967505e-01 3.42089862e-01 -6.26359880e-01 -3.59150693e-02
2.90855765e-02 7.32459486e-01 7.23025918e-01 -1.09145617e+00
8.56540978e-01 9.88335192e-01 3.93568933e-01 8.13949764e-01
-1.30808556e+00 -5.73594749e-01 8.65065232e-02 -2.69436445e-02
-9.59191620e-01 -7.77474821e-01 6.09137237e-01 -4.05855447e-01
8.48163068e-01 3.49830091e-01 9.69758689e-01 1.17096543e+00
-3.89146268e-01 8.50837588e-01 1.13946629e+00 -6.30850673e-01
-1.47046819e-01 5.01293242e-02 -3.19306076e-01 7.11855292e-01
-7.26328939e-02 2.18055367e-01 -6.56206012e-01 2.10425168e-01
8.82890403e-01 -2.36183211e-01 -4.13140386e-01 -2.44705647e-01
-1.37236381e+00 5.11101186e-01 5.93861043e-01 1.43493786e-01
-2.37385958e-01 4.44109917e-01 2.33619213e-01 -2.25440282e-02
5.44851184e-01 9.72162127e-01 -4.92678493e-01 -8.99660587e-02
-6.71799481e-01 4.31561053e-01 2.52960265e-01 1.11028123e+00
6.70305252e-01 2.37721130e-01 -3.10025483e-01 8.69291186e-01
1.29956469e-01 4.31131989e-01 3.57671052e-01 -1.28857934e+00
3.29915673e-01 5.31227529e-01 4.43607084e-02 -5.51299751e-01
-3.96940142e-01 -5.29187024e-01 -8.02768171e-01 3.35352600e-01
4.71871436e-01 -1.06916437e-02 -1.14588273e+00 1.73265100e+00
3.47953588e-01 9.39008817e-02 -7.67208487e-02 1.03252888e+00
7.61466026e-01 4.73577797e-01 1.74026981e-01 2.21081093e-01
1.24053466e+00 -1.21723568e+00 -3.84846210e-01 -1.83803573e-01
6.34564400e-01 -7.95088887e-01 1.68668342e+00 3.45101565e-01
-1.22323322e+00 -6.80101931e-01 -6.64867103e-01 -3.99155349e-01
-1.14173390e-01 1.26114190e-01 9.56428230e-01 2.03220576e-01
-1.24965310e+00 3.76869291e-01 -7.82640457e-01 -5.06137252e-01
7.22879589e-01 1.12070814e-01 -4.27345753e-01 -1.71798944e-01
-8.57131422e-01 6.69905365e-01 3.67195845e-01 1.50799364e-01
-9.35609698e-01 -9.02226567e-01 -6.54010117e-01 -3.78720820e-01
3.79694730e-01 -1.23482955e+00 1.31519067e+00 -6.73244774e-01
-1.43274760e+00 9.43519235e-01 1.80544127e-02 -3.59808445e-01
5.85773766e-01 -2.93515444e-01 4.56440859e-02 -3.98463272e-02
-8.22168961e-02 1.61821198e+00 6.95535958e-01 -1.42685449e+00
-3.73819411e-01 -2.33555660e-01 -7.83976987e-02 3.88443142e-01
-4.99014348e-01 -1.21144660e-01 -8.86943519e-01 -5.96405447e-01
-1.69161528e-01 -1.03787935e+00 -6.02228701e-01 -4.08968218e-02
-5.46273530e-01 7.46154934e-02 6.54325426e-01 -5.55237412e-01
9.81552541e-01 -1.97769761e+00 1.22291975e-01 -2.05032870e-01
3.62386256e-01 5.25472999e-01 -4.94659960e-01 6.26437306e-01
-5.06743118e-02 3.72536331e-01 -4.23214763e-01 -6.42618656e-01
-3.09671402e-01 2.25467980e-03 -3.55073512e-01 2.61948965e-02
5.21584213e-01 1.41457283e+00 -8.62130523e-01 -3.65895778e-01
2.95801818e-01 4.87117231e-01 -8.25624526e-01 3.48856717e-01
-6.52997315e-01 1.09984505e+00 -3.56796175e-01 6.10175550e-01
5.49277961e-01 -5.46236038e-01 -1.58171505e-01 -2.70918310e-01
-2.95143276e-01 9.17479396e-02 -6.22633338e-01 1.97726858e+00
-3.97206247e-01 6.98612213e-01 -1.39012381e-01 -6.08783662e-01
8.25311244e-01 -2.50131767e-02 2.93760031e-01 -7.09473014e-01
3.79174054e-02 3.70988026e-02 -5.79375215e-02 -6.44511700e-01
6.55027270e-01 7.44495392e-02 1.35548040e-01 3.72323036e-01
-4.71556559e-02 -6.58123195e-01 3.44727248e-01 4.29302216e-01
8.27400863e-01 6.15224719e-01 -4.51045260e-02 -6.47033975e-02
1.16313554e-01 4.18236554e-01 3.62467498e-01 6.08065486e-01
-1.71407312e-02 1.16838419e+00 1.49481758e-01 -3.37373227e-01
-1.44288337e+00 -6.56354427e-01 -7.05012605e-02 9.91792679e-01
-1.99821159e-01 -4.25383151e-01 -6.87546849e-01 -5.01023829e-01
-1.09561846e-01 6.77257359e-01 -5.45517862e-01 -3.38232145e-02
-4.37028497e-01 -7.99961388e-01 5.74988186e-01 4.90161657e-01
6.70646727e-01 -1.24095154e+00 -3.33759636e-01 -3.71609293e-02
-2.56211311e-01 -1.14065230e+00 -5.09170949e-01 -2.36717090e-01
-7.65167177e-01 -8.39059293e-01 -8.13005805e-01 -7.01146245e-01
8.49139512e-01 4.49632794e-01 1.25202179e+00 3.89157265e-01
-4.88345087e-01 3.72944593e-01 -4.37373161e-01 -6.46239579e-01
-5.58548868e-01 1.58190981e-01 -1.44696563e-01 -2.93020278e-01
-1.27680883e-01 -5.01364350e-01 -9.50972378e-01 3.09092790e-01
-1.02448690e+00 8.22811067e-01 6.73658371e-01 7.02788770e-01
7.21231282e-01 -6.43476129e-01 6.64542496e-01 -9.97232437e-01
6.26888752e-01 -1.57366484e-01 -7.30387032e-01 1.64873615e-01
-6.89126432e-01 4.14445475e-02 5.45836389e-01 -3.05450112e-01
-1.19437897e+00 3.31235141e-01 -3.15028518e-01 -3.97910178e-01
-1.67306021e-01 3.53140056e-01 -1.12588175e-01 1.17961066e-02
9.11277115e-01 2.36613512e-01 1.03111230e-01 -4.51498419e-01
1.05037916e+00 4.72573608e-01 8.09149921e-01 -6.29491270e-01
7.49556363e-01 5.05688310e-01 -2.23865509e-01 -8.12249303e-01
-6.99603140e-01 -2.62852281e-01 -5.21795869e-01 -3.18617582e-01
1.00369632e+00 -1.00551498e+00 -5.51775634e-01 7.42327392e-01
-1.04962575e+00 -8.78172338e-01 -4.51359749e-01 1.92995340e-01
-6.47636890e-01 2.53123373e-01 -3.87188017e-01 -5.90059876e-01
-5.25033593e-01 -1.27377701e+00 1.28969300e+00 3.49333465e-01
-4.22487147e-02 -7.38153160e-01 2.04145946e-02 8.13431263e-01
5.05651355e-01 5.83501346e-02 5.44147670e-01 -1.79999962e-01
-1.05569685e+00 -4.95010614e-02 -4.40662056e-01 3.19257319e-01
-1.58271909e-01 1.74718350e-01 -1.10402739e+00 -2.42014915e-01
-4.94036347e-01 -7.79256165e-01 9.76713836e-01 2.57577866e-01
1.54819489e+00 3.75650264e-02 -1.87242985e-01 1.09761727e+00
1.10458255e+00 -1.04075611e-01 8.06168377e-01 2.68841803e-01
1.02415335e+00 5.17093301e-01 6.04527533e-01 3.29672843e-01
5.57523131e-01 7.35669553e-01 4.78703916e-01 -5.89847147e-01
-6.00970268e-01 -4.44883019e-01 1.21088093e-02 6.37940407e-01
-2.18969196e-01 -4.59849417e-01 -1.01301098e+00 4.17495072e-01
-1.76395857e+00 -6.87048316e-01 -2.77200162e-01 2.01764989e+00
9.77391720e-01 -1.51419595e-01 1.34836227e-01 -4.71862167e-01
3.82081121e-01 5.63191175e-02 -7.05815315e-01 -5.88599518e-02
-2.70780385e-01 -7.25586191e-02 2.83137888e-01 3.87684077e-01
-8.96927536e-01 1.39921069e+00 6.13590479e+00 6.51029527e-01
-8.86819899e-01 -6.40121624e-02 1.11683357e+00 -1.69393852e-01
-5.81787944e-01 1.54780611e-01 -8.21915329e-01 2.60218799e-01
6.50198936e-01 5.86928055e-02 4.27738935e-01 6.25306785e-01
4.74665612e-01 -2.03600734e-01 -7.95510411e-01 8.79961848e-01
-7.90581703e-02 -1.72081184e+00 1.44892350e-01 1.83516100e-01
9.80167031e-01 3.23083431e-01 2.08755270e-01 5.64207211e-02
5.49097478e-01 -8.95978391e-01 5.34328759e-01 6.87550485e-01
1.22009790e+00 -2.97810555e-01 1.82996735e-01 2.59715289e-01
-8.54725838e-01 1.42226398e-01 -3.53567392e-01 -2.59705950e-02
3.12203348e-01 5.77448308e-01 -1.06049919e+00 3.95286143e-01
6.10651851e-01 6.25996053e-01 -8.02648425e-01 1.03038013e+00
-2.48310700e-01 4.96503681e-01 -1.93087712e-01 4.14521694e-01
-1.65191200e-02 -2.79751629e-01 4.11208928e-01 1.05451286e+00
3.48056346e-01 1.70332462e-01 6.88019320e-02 1.04253256e+00
-2.14081496e-01 -1.02383289e-02 -7.87549973e-01 -3.01807940e-01
4.61785346e-01 1.42924905e+00 -5.37926435e-01 -2.59454936e-01
-3.20546359e-01 1.03780627e+00 5.90580344e-01 3.38002056e-01
-6.38428509e-01 1.12425670e-01 5.39723158e-01 1.88298151e-01
-1.36812463e-01 -4.12562609e-01 -4.80619490e-01 -1.10580039e+00
-9.46033150e-02 -9.71602917e-01 1.04407556e-01 -1.21953583e+00
-1.10080373e+00 7.80862451e-01 -5.76089658e-02 -1.12866676e+00
-2.33954743e-01 -3.46800596e-01 -5.22948325e-01 1.02764499e+00
-1.30998814e+00 -1.74704075e+00 -7.44155467e-01 4.33881968e-01
6.22241795e-01 -1.41862765e-01 6.97447896e-01 1.07602544e-01
-7.71873415e-01 5.11027336e-01 -1.75088167e-01 -5.28484918e-02
8.71242821e-01 -1.13824010e+00 1.07707572e+00 9.76606250e-01
2.39865512e-01 4.74588245e-01 5.97992063e-01 -7.71683574e-01
-1.30784380e+00 -1.17465711e+00 4.56308424e-01 -8.02224994e-01
3.97458494e-01 -4.56905305e-01 -8.72442722e-01 5.40746570e-01
3.55163604e-01 7.41245300e-02 4.10087377e-01 -4.30257954e-02
-2.32781038e-01 2.22440530e-02 -7.74094522e-01 8.28424335e-01
1.47391760e+00 -2.02075347e-01 -4.92706411e-02 5.25805414e-01
1.00298119e+00 -5.32810986e-01 -6.06413364e-01 4.05127048e-01
3.00136417e-01 -1.03246713e+00 9.65946198e-01 -6.99454963e-01
7.46023476e-01 -3.50266457e-01 1.54823631e-01 -1.39087594e+00
-3.01304132e-01 -7.74516940e-01 2.51523972e-01 1.28938782e+00
5.87383628e-01 -4.69953865e-01 8.67441177e-01 7.54595459e-01
-5.93570590e-01 -7.85031259e-01 -3.63978654e-01 -4.79363292e-01
2.28578113e-02 -5.84609866e-01 5.57913601e-01 8.40006411e-01
-1.81358650e-01 3.28665763e-01 -5.97134292e-01 -1.63628250e-01
3.41921866e-01 1.37474481e-02 1.29976678e+00 -9.12387967e-01
-4.10219908e-01 -3.34452212e-01 2.51840074e-02 -1.18817222e+00
-1.87033996e-01 -1.00634933e+00 1.06972851e-01 -1.69906974e+00
3.43522340e-01 -8.04692805e-01 1.45160884e-01 6.48437619e-01
-4.83608246e-01 6.11573994e-01 3.91432375e-01 3.72936696e-01
-5.33576787e-01 7.56095469e-01 1.90262902e+00 2.41335243e-01
-3.30748469e-01 -1.63964108e-01 -1.03507698e+00 5.01092315e-01
1.01621699e+00 3.10245901e-03 -6.14298165e-01 -7.84542203e-01
3.40267628e-01 -8.91616717e-02 4.53406006e-01 -9.54898775e-01
-9.22032967e-02 -1.61511794e-01 3.29981536e-01 -4.33194965e-01
4.29233968e-01 -4.13958281e-01 1.36451811e-01 8.41810480e-02
-3.68042141e-01 9.14079174e-02 3.96352261e-01 3.37952763e-01
4.58419928e-03 5.94970509e-02 6.56585991e-01 -3.00684035e-01
-8.12235177e-01 5.39181650e-01 1.53676225e-02 1.37455985e-01
8.16191018e-01 -4.04559448e-02 -6.72993660e-01 -4.66351658e-01
-6.20206356e-01 4.18741643e-01 8.32159042e-01 4.77825433e-01
6.39594853e-01 -1.07114375e+00 -8.97659838e-01 3.12240660e-01
4.16072905e-01 3.70223850e-01 3.66584271e-01 7.46249616e-01
-5.02380013e-01 2.04852447e-01 -1.78173646e-01 -6.43369377e-01
-1.12734830e+00 3.97298366e-01 2.40595683e-01 -2.45358143e-02
-7.58488715e-01 1.14478087e+00 4.50185448e-01 -4.82948124e-01
-1.53109387e-01 -8.22569728e-02 8.66273120e-02 -2.28003591e-01
4.35054868e-01 1.86377928e-01 7.81261921e-02 -3.65194261e-01
8.12902972e-02 2.48224258e-01 -1.69133186e-01 -2.95581400e-01
1.40680134e+00 -1.39057979e-01 1.17071576e-01 2.24381150e-03
7.79872537e-01 -1.66256651e-01 -1.67159975e+00 -8.48359838e-02
-2.69720763e-01 -5.98526776e-01 -6.13095425e-02 -1.07756197e+00
-1.07759202e+00 8.36180508e-01 4.42536622e-01 -4.51370403e-02
1.27176392e+00 2.00850561e-01 7.49403775e-01 1.36570856e-01
2.93622017e-01 -9.43360865e-01 3.38404715e-01 3.79437029e-01
1.33727300e+00 -1.52895236e+00 -1.29329816e-01 -5.00597119e-01
-1.01415229e+00 6.95219576e-01 1.00485551e+00 2.66658008e-01
8.59404802e-02 1.52203828e-01 3.01538467e-01 -2.01268688e-01
-7.44341969e-01 -5.02462506e-01 3.53230149e-01 9.52103078e-01
5.61470926e-01 -1.17602043e-01 -6.73599169e-02 2.88147330e-02
-3.93755853e-01 2.00660124e-01 5.61192155e-01 5.98695099e-01
-2.32227892e-01 -1.40662110e+00 -2.29725733e-01 4.99322146e-01
-2.48252470e-02 -3.52408081e-01 -2.95327246e-01 7.86912799e-01
-1.71610098e-02 7.26936638e-01 -7.87472948e-02 -3.82397622e-01
2.30489701e-01 4.86418828e-02 5.29180467e-01 -8.23072970e-01
-2.87138253e-01 2.25893661e-01 1.82258740e-01 -5.79724729e-01
-2.79807359e-01 -6.71446145e-01 -1.13292050e+00 -8.07497948e-02
-9.74252820e-02 -2.17974648e-01 7.98906624e-01 6.22583270e-01
8.21596980e-01 5.51404476e-01 3.32896680e-01 -1.13783395e+00
-1.42760471e-01 -1.03223670e+00 6.28551543e-02 6.22294307e-01
1.26862880e-02 -3.40662509e-01 7.89856389e-02 3.48714828e-01] | [11.346965789794922, -0.26973533630371094] |
dab9b13b-13f3-4211-bf65-e0f804f98209 | monocular-3d-human-pose-estimation-by | null | null | http://openaccess.thecvf.com/content_iccv_2017/html/Nie_Monocular_3D_Human_ICCV_2017_paper.html | http://openaccess.thecvf.com/content_ICCV_2017/papers/Nie_Monocular_3D_Human_ICCV_2017_paper.pdf | Monocular 3D Human Pose Estimation by Predicting Depth on Joints | This paper aims at estimating full-body 3D human poses from monocular images of which the biggest challenge is the inherent ambiguity introduced by lifting the 2D pose into 3D space. We propose a novel framework focusing on reducing this ambiguity by predicting the depth of human joints based on 2D human joint locations and body part images. Our approach is built on a two-level hierarchy of Long Short-Term Memory (LSTM) Networks which can be trained end-to-end. The first level consists of two components: 1) a skeleton-LSTM which learns the depth information from global human skeleton features; 2) a patch-LSTM which utilizes the local image evidence around joint locations. The both networks have tree structure defined on the kinematic relation of human skeleton, thus the information at different joints is broadcast through the whole skeleton in a top-down fashion. The two networks are first pre-trained separately on different data sources and then aggregated in the second layer for final depth prediction. The empirical evaluation on Human3.6M and HHOI dataset demonstrates the advantage of combining global 2D skeleton and local image patches for depth prediction, and our superior quantitative and qualitative performance relative to state-of-the-art methods.
| ['Song-Chun Zhu', 'Bruce Xiaohan Nie', 'Ping Wei'] | 2017-10-01 | null | null | null | iccv-2017-10 | ['monocular-3d-human-pose-estimation'] | ['computer-vision'] | [ 5.13888747e-02 4.23042297e-01 -3.13174069e-01 -2.60993659e-01
-6.26366317e-01 3.49885434e-01 4.23066646e-01 -3.76986504e-01
-6.07317448e-01 4.38426554e-01 4.29980040e-01 3.99927318e-01
1.46698698e-01 -6.98032796e-01 -8.93167138e-01 -5.63296795e-01
-2.25502387e-01 7.41651237e-01 4.89682704e-01 -9.66835693e-02
-1.30780637e-01 2.91523993e-01 -1.40062785e+00 3.36437732e-01
2.69145727e-01 1.38036478e+00 5.08393683e-02 6.43522263e-01
1.13459699e-01 8.68317664e-01 -2.97956169e-01 -1.61481395e-01
3.50820839e-01 -3.32885623e-01 -8.09932649e-01 3.09891909e-01
4.14407879e-01 -5.93725264e-01 -5.62610149e-01 6.40219033e-01
8.28916609e-01 -5.67372702e-02 5.30246258e-01 -7.02338278e-01
-9.03843567e-02 1.79299787e-01 -7.13729978e-01 -1.20192148e-01
5.40214837e-01 1.86507292e-02 8.32370877e-01 -8.65779877e-01
5.94759464e-01 1.48734903e+00 7.72021055e-01 5.28776467e-01
-7.29812384e-01 -2.22404674e-01 1.43863901e-01 2.59623975e-01
-1.14599931e+00 -1.01026468e-01 7.67232478e-01 -6.17979288e-01
1.02670026e+00 -3.99114639e-01 1.01044893e+00 1.05274320e+00
6.99026227e-01 9.44153428e-01 7.23542988e-01 -4.69230354e-01
-2.18103394e-01 -5.89367747e-01 -1.41590312e-01 1.25195742e+00
-1.87696010e-01 3.18283558e-01 -8.92725945e-01 9.45110321e-02
1.26673865e+00 -2.22787312e-05 -7.26813276e-04 -6.70775175e-01
-1.38135731e+00 6.88158214e-01 8.58492792e-01 5.99579923e-02
-6.33421481e-01 5.13019919e-01 4.17197913e-01 7.30934888e-02
4.94106919e-01 -2.14342698e-01 -6.20757461e-01 2.13549450e-01
-8.06166053e-01 3.52086365e-01 4.57265884e-01 5.01933813e-01
6.76433444e-01 -2.26482511e-01 -2.76201218e-03 8.79221857e-01
7.20915616e-01 3.59699070e-01 7.00560153e-01 -8.52185369e-01
6.30957425e-01 6.31961763e-01 -1.46927744e-01 -8.85616601e-01
-7.60857821e-01 -5.96341074e-01 -8.09470654e-01 2.93370634e-01
3.88930589e-01 -2.19135031e-01 -1.19766343e+00 1.65697062e+00
6.38871908e-01 -2.47914001e-01 -3.66001785e-01 1.26307476e+00
8.30441415e-01 4.51253980e-01 2.16972008e-02 3.93382370e-01
1.39171171e+00 -1.33938754e+00 -3.86730313e-01 -4.94109333e-01
4.33540553e-01 -3.41824502e-01 5.14704823e-01 3.18324238e-01
-1.33723247e+00 -1.01029968e+00 -1.15010858e+00 -5.05229354e-01
-1.75698549e-01 2.53470033e-01 2.73200899e-01 1.95048451e-01
-9.60667491e-01 6.93636715e-01 -1.12519729e+00 -3.60236406e-01
1.19711623e-01 5.80043256e-01 -5.18875301e-01 1.18131816e-01
-1.32242978e+00 9.53056395e-01 2.46036828e-01 5.39349675e-01
-8.59605908e-01 -1.46251723e-01 -1.33887053e+00 -2.65086591e-01
2.78988451e-01 -1.47265160e+00 1.17216790e+00 -6.50708795e-01
-1.65752482e+00 1.10432398e+00 -1.25931464e-02 -5.87750494e-01
7.35946357e-01 -7.05253124e-01 2.72858113e-01 4.23040956e-01
2.66589195e-01 1.10107374e+00 9.73218262e-01 -1.10967648e+00
-7.26697385e-01 -9.30962980e-01 -1.83966368e-01 6.12538815e-01
5.87320402e-02 -3.03648800e-01 -8.33882034e-01 -7.46597588e-01
6.16997123e-01 -9.41636026e-01 -3.62747848e-01 2.88852334e-01
-4.89267766e-01 -1.62741050e-01 3.03036928e-01 -1.03522420e+00
8.99731636e-01 -1.62588644e+00 7.92392075e-01 1.76336974e-01
1.40114322e-01 -9.86617059e-02 5.71057983e-02 3.04486156e-01
1.13169171e-01 -5.58444619e-01 -9.71977338e-02 -6.65829837e-01
-7.69054191e-03 2.77531117e-01 1.85639262e-01 6.88482523e-01
4.50127050e-02 1.06112707e+00 -6.41276479e-01 -8.78310502e-01
4.19985354e-01 5.88741422e-01 -3.50934893e-01 2.98536897e-01
-2.13067144e-01 5.66171050e-01 -4.01112765e-01 8.07780921e-01
2.49856919e-01 -2.50322461e-01 1.16455220e-01 -2.35460296e-01
9.71494988e-02 2.44843155e-01 -1.07762837e+00 2.50026155e+00
-3.20505977e-01 -8.27311166e-03 3.64273638e-02 -9.42806482e-01
6.44011319e-01 4.14355934e-01 6.74152672e-01 -7.90372550e-01
2.15725079e-01 1.03548653e-01 -2.20224470e-01 -5.01075625e-01
2.73824967e-02 -2.50998527e-01 -2.02621028e-01 2.72031575e-01
4.07605678e-01 2.48863846e-01 -1.36643216e-01 -2.71287352e-01
7.93498695e-01 9.09856796e-01 2.93314427e-01 2.45155152e-02
5.82056820e-01 -3.52344334e-01 5.11223018e-01 2.99474150e-01
-1.54814392e-01 6.84640765e-01 3.34915966e-01 -7.20941365e-01
-8.26668739e-01 -1.30620968e+00 2.62406796e-01 1.04998064e+00
5.99270500e-02 -1.99883744e-01 -7.89481401e-01 -5.57238460e-01
1.07595965e-01 -2.17668459e-01 -8.67211282e-01 -8.15373510e-02
-8.69805276e-01 -3.51479620e-01 4.09906834e-01 8.22550595e-01
8.53404641e-01 -1.01876283e+00 -1.12483108e+00 1.59983918e-01
-3.14674407e-01 -1.11986446e+00 -3.74395311e-01 2.56029874e-01
-1.18693328e+00 -9.79111850e-01 -1.06064141e+00 -8.58970046e-01
4.75873619e-01 -2.06782266e-01 1.02918863e+00 -1.21064194e-01
-3.77773315e-01 3.46849203e-01 -4.90849838e-02 -5.92421181e-02
1.35996208e-01 2.86851555e-01 -3.09003461e-02 -2.35828422e-02
1.28551677e-01 -6.51124954e-01 -8.23851407e-01 3.55332106e-01
-3.59388977e-01 2.08740383e-01 1.05294883e+00 8.75603616e-01
7.38946021e-01 -2.84257203e-01 1.56892180e-01 -3.78239572e-01
-1.90209240e-01 -2.29448676e-01 -1.13570891e-01 6.16647266e-02
-7.54945874e-02 1.95241645e-01 1.95990026e-01 2.16402505e-02
-1.15335369e+00 6.16445780e-01 -4.01428223e-01 -3.29169154e-01
-1.36065245e-01 4.20109689e-01 -2.26406470e-01 5.26660234e-02
3.78549159e-01 2.58685082e-01 2.55975246e-01 -8.10967088e-01
2.20323652e-01 2.83238620e-01 7.67765284e-01 -5.86639047e-01
6.04296148e-01 6.17737651e-01 1.75137714e-01 -8.70188653e-01
-1.05250514e+00 -4.57685292e-01 -1.41981888e+00 -4.41547930e-01
1.24367738e+00 -1.21531594e+00 -4.30527061e-01 9.30650055e-01
-1.14835656e+00 -4.70756561e-01 -1.48847461e-01 5.97369313e-01
-7.26327181e-01 6.28263593e-01 -1.11391616e+00 -4.96791929e-01
-4.91221607e-01 -8.97131979e-01 1.65984321e+00 -1.34765014e-01
-1.88286379e-01 -9.87765789e-01 2.78955817e-01 6.48181081e-01
-2.68361598e-01 5.91279685e-01 8.61488581e-01 -5.30884191e-02
-4.33433086e-01 -3.27576429e-01 -6.69894088e-03 3.79871488e-01
-1.51589140e-01 -4.64665085e-01 -8.62499297e-01 -1.87203780e-01
5.93637228e-02 -5.29664218e-01 1.08120704e+00 8.33377719e-01
6.46065116e-01 -4.91448492e-02 -3.93117636e-01 5.63907325e-01
1.08534050e+00 -3.39437187e-01 5.52444756e-01 3.44084293e-01
1.03786039e+00 8.61481071e-01 6.56400621e-01 3.82863045e-01
6.44890249e-01 8.70861888e-01 5.44382513e-01 -3.71729881e-01
-4.04987782e-01 -5.48452616e-01 5.23052275e-01 7.54389286e-01
-3.86207312e-01 3.09262872e-01 -7.80554712e-01 4.79246914e-01
-2.05822659e+00 -6.20772779e-01 1.11769356e-01 2.20318246e+00
7.02397645e-01 4.08338100e-01 5.66311538e-01 2.39124447e-01
4.78084981e-01 2.91915625e-01 -4.34128881e-01 -4.98205461e-02
2.34673947e-01 1.64032713e-01 2.76196450e-01 4.50875342e-01
-1.21809101e+00 8.13052237e-01 5.98861408e+00 5.26610553e-01
-9.74601328e-01 1.20160893e-01 3.06541353e-01 -2.57810771e-01
3.22872758e-01 -3.20484072e-01 -7.46321619e-01 1.98854595e-01
5.49669743e-01 5.29280424e-01 -2.95501292e-01 7.42748201e-01
-1.92056410e-02 -2.20086277e-01 -1.15183389e+00 9.16625977e-01
1.14995167e-01 -7.87635982e-01 1.56299546e-01 2.46341780e-01
4.50460970e-01 2.43604615e-01 -1.90327629e-01 8.60447064e-02
-6.70364127e-02 -9.81015265e-01 9.72825468e-01 7.75823772e-01
6.18331730e-01 -7.10069716e-01 7.71239400e-01 6.50346637e-01
-1.39657569e+00 -7.57914931e-02 -3.62062991e-01 -4.38215226e-01
3.97225946e-01 6.00718677e-01 -4.86386955e-01 6.38849020e-01
8.31641257e-01 8.79526675e-01 -4.41035867e-01 7.93704748e-01
-5.56055486e-01 -8.98510441e-02 -3.62433165e-01 3.51366729e-01
1.90805241e-01 1.36106402e-01 3.03170949e-01 9.68977392e-01
5.84605634e-02 -8.14637840e-02 3.93067032e-01 5.45015633e-01
1.50454491e-01 -1.91905662e-01 -4.73982751e-01 4.59349066e-01
-2.23576859e-01 1.08189237e+00 -4.29956019e-01 -3.28876793e-01
-5.44931412e-01 1.21457803e+00 3.40297520e-01 9.98966843e-02
-9.27367628e-01 -7.73556083e-02 2.93249011e-01 2.90031403e-01
3.06947380e-01 -4.04895157e-01 -2.76640385e-01 -1.12351131e+00
2.03303218e-01 -5.48523962e-01 7.32052326e-01 -6.81169689e-01
-1.08525670e+00 4.95829016e-01 -5.87247014e-02 -1.08052289e+00
-6.18484914e-01 -7.62429893e-01 -1.61305100e-01 6.62476301e-01
-1.26908624e+00 -1.57662034e+00 -3.49315941e-01 6.67678893e-01
5.89307845e-01 2.29834780e-01 7.00061738e-01 3.32953453e-01
-3.83157313e-01 4.11687225e-01 -5.70749104e-01 3.56290460e-01
5.07204711e-01 -1.26086354e+00 6.61950886e-01 6.78750575e-01
1.30421802e-01 3.25627834e-01 3.07349414e-01 -8.52622747e-01
-1.07766080e+00 -7.92603314e-01 1.12469971e+00 -4.72582251e-01
1.69101536e-01 -3.13112587e-01 -3.76236379e-01 8.27853918e-01
-1.00052737e-01 -1.48489252e-02 4.44684029e-01 2.04030629e-02
-1.90780625e-01 -4.73721065e-02 -8.57656002e-01 2.64134079e-01
1.26462734e+00 -4.29153204e-01 -9.25430000e-01 1.20054618e-01
5.28602183e-01 -6.70174241e-01 -1.17312670e+00 6.72540843e-01
9.86178398e-01 -1.21105790e+00 1.22875810e+00 -3.79084200e-01
6.70407236e-01 -1.06201783e-01 -1.80709407e-01 -8.89200330e-01
-1.33356735e-01 -1.76461279e-01 -3.68819743e-01 3.90173674e-01
1.08713739e-01 -1.44086957e-01 1.14788389e+00 2.09260449e-01
-2.25257009e-01 -1.14531636e+00 -1.07080221e+00 -5.00532568e-01
2.41170786e-02 -2.41043136e-01 -3.46310027e-02 2.39874467e-01
-3.47418971e-02 6.38058484e-01 -7.45416343e-01 -7.57580856e-03
9.35581803e-01 1.76884934e-01 9.50785637e-01 -1.18233740e+00
-5.39326608e-01 -8.25618580e-02 -6.14193380e-01 -1.70792329e+00
-9.48832557e-02 -5.38034976e-01 2.11207241e-01 -1.74287117e+00
2.09227473e-01 2.83294678e-01 -1.06654406e-01 5.35039961e-01
9.99399424e-02 4.00803000e-01 -2.06666086e-02 2.16565698e-01
-6.24641955e-01 8.03249836e-01 1.41531873e+00 8.88474435e-02
-3.00950408e-02 2.02100992e-01 8.88023451e-02 1.17783320e+00
2.93931514e-01 -3.04266274e-01 -1.59904823e-01 -6.93819404e-01
1.61657985e-02 5.11631787e-01 8.70982289e-01 -1.52528119e+00
3.07198495e-01 3.21001589e-01 8.63598943e-01 -1.09189367e+00
7.11643815e-01 -6.05605543e-01 -1.63991868e-01 1.03734791e+00
-1.57739326e-01 -7.29477927e-02 -2.74906427e-01 6.64928854e-01
-3.12612981e-01 2.45366022e-01 8.24389219e-01 -6.51696146e-01
-7.50652134e-01 5.63121915e-01 -1.48723766e-01 -2.11633548e-01
6.10967815e-01 -4.33593810e-01 3.01353246e-01 -3.50849897e-01
-1.16780031e+00 1.74319103e-01 3.49553794e-01 3.77746820e-01
6.92104280e-01 -1.45310783e+00 -5.25946796e-01 3.28827083e-01
-6.87071085e-02 3.48528773e-01 3.96456480e-01 8.71345878e-01
-6.15739942e-01 6.64913177e-01 -5.15792608e-01 -1.00733483e+00
-1.08447313e+00 3.37514877e-01 6.72698259e-01 -5.90988636e-01
-7.85157442e-01 1.04359579e+00 3.77265215e-01 -5.78807533e-01
5.36005020e-01 -4.04031575e-01 -5.21701314e-02 -1.00639708e-01
7.17897862e-02 4.53151166e-01 -4.92849089e-02 -9.82104182e-01
-4.03622448e-01 1.20576048e+00 2.77236700e-01 -3.42825830e-01
1.17468846e+00 -2.68665105e-01 -1.46087080e-01 4.85118657e-01
1.28585136e+00 -4.52737391e-01 -1.50329912e+00 -5.14109910e-01
-6.40043989e-02 -4.03465539e-01 -2.01772600e-01 -7.28388011e-01
-1.17739093e+00 1.19689965e+00 5.51421404e-01 -4.13470775e-01
9.92287874e-01 8.16766769e-02 1.33172095e+00 2.43471801e-01
5.68559647e-01 -1.26337802e+00 5.14905930e-01 4.42787826e-01
8.75573754e-01 -1.06521404e+00 1.18919663e-01 -2.10328400e-01
-3.30073357e-01 1.11938334e+00 6.99522793e-01 -2.52020717e-01
6.92102075e-01 -1.32599339e-01 6.55943304e-02 -4.60163176e-01
-5.13971567e-01 -3.75072241e-01 6.69551075e-01 4.92973119e-01
4.01371837e-01 -1.57532066e-01 -2.86932200e-01 6.43012941e-01
-1.24719605e-01 1.45283595e-01 -2.05953553e-01 9.07219112e-01
-5.84658086e-01 -1.02050078e+00 -4.20523614e-01 4.50018756e-02
-4.97493565e-01 3.69426131e-01 -4.06010836e-01 9.26447392e-01
5.17781675e-01 4.93784696e-01 -5.72546721e-02 -5.04845381e-01
4.47318792e-01 7.18460605e-02 9.08817291e-01 -7.12785244e-01
-3.52190465e-01 4.12688017e-01 1.08386770e-01 -9.67456043e-01
-5.94137311e-01 -5.46041489e-01 -1.40819180e+00 1.08742312e-01
-4.58264432e-04 -2.34068662e-01 5.84845304e-01 1.25040293e+00
1.99157596e-01 4.83873278e-01 2.05907494e-01 -1.52563167e+00
-6.33075118e-01 -9.99573886e-01 -6.38034940e-01 2.75400996e-01
3.02002311e-01 -9.97127116e-01 1.49403671e-02 6.75499365e-02] | [7.053112030029297, -0.8292876482009888] |
faa9368d-838d-44a0-99ec-35a1cce5ec07 | logical-entity-representation-in-knowledge | 2305.12738 | null | https://arxiv.org/abs/2305.12738v1 | https://arxiv.org/pdf/2305.12738v1.pdf | Logical Entity Representation in Knowledge-Graphs for Differentiable Rule Learning | Probabilistic logical rule learning has shown great strength in logical rule mining and knowledge graph completion. It learns logical rules to predict missing edges by reasoning on existing edges in the knowledge graph. However, previous efforts have largely been limited to only modeling chain-like Horn clauses such as $R_1(x,z)\land R_2(z,y)\Rightarrow H(x,y)$. This formulation overlooks additional contextual information from neighboring sub-graphs of entity variables $x$, $y$ and $z$. Intuitively, there is a large gap here, as local sub-graphs have been found to provide important information for knowledge graph completion. Inspired by these observations, we propose Logical Entity RePresentation (LERP) to encode contextual information of entities in the knowledge graph. A LERP is designed as a vector of probabilistic logical functions on the entity's neighboring sub-graph. It is an interpretable representation while allowing for differentiable optimization. We can then incorporate LERP into probabilistic logical rule learning to learn more expressive rules. Empirical results demonstrate that with LERP, our model outperforms other rule learning methods in knowledge graph completion and is comparable or even superior to state-of-the-art black-box methods. Moreover, we find that our model can discover a more expressive family of logical rules. LERP can also be further combined with embedding learning methods like TransE to make it more interpretable. | ['Heng Ji', 'Hanghang Tong', 'Xinya Du', 'Charles Yu', 'Qizheng He', 'Chi Han'] | 2023-05-22 | null | null | null | null | ['knowledge-graph-completion'] | ['knowledge-base'] | [-8.25767815e-02 8.47780824e-01 -7.92355895e-01 -3.96170944e-01
-3.27576578e-01 -3.64927053e-01 3.58157009e-01 2.67721355e-01
1.84410721e-01 8.66777360e-01 2.34369844e-01 -6.94906712e-01
-7.58352876e-01 -1.48433185e+00 -1.05450141e+00 -2.68550158e-01
-4.82541800e-01 5.40476143e-01 1.75076723e-01 -2.45020851e-01
-3.90960127e-01 2.92194277e-01 -1.34247983e+00 3.30207705e-01
7.70166039e-01 7.90536284e-01 -2.23049060e-01 9.41576436e-03
-3.21778357e-01 1.47865069e+00 -7.88117647e-02 -7.05558836e-01
5.20325557e-04 7.27776736e-02 -1.07888091e+00 -4.63057995e-01
1.73233315e-01 -1.69153474e-02 -7.07223773e-01 9.86532092e-01
-1.95416883e-01 2.19371960e-01 6.27382815e-01 -1.44547629e+00
-9.86054122e-01 1.45381749e+00 -4.78744715e-01 -3.20476353e-01
6.94037735e-01 -1.51884481e-01 1.74302721e+00 -7.81866133e-01
9.64365661e-01 1.50031316e+00 6.63944244e-01 2.78122067e-01
-1.19239354e+00 -6.32742941e-01 4.40016448e-01 7.19759047e-01
-1.69777226e+00 5.71428351e-02 9.61277723e-01 -1.20831743e-01
1.11892188e+00 2.21951947e-01 5.06304324e-01 7.88522422e-01
2.13747546e-01 9.03490782e-01 6.67056203e-01 -5.67872167e-01
1.65193364e-01 4.19020653e-02 3.76191199e-01 1.49495232e+00
4.87797588e-01 1.60319477e-01 -7.70437181e-01 -1.90912291e-01
5.00653744e-01 1.90691743e-02 -1.64492622e-01 -6.57168448e-01
-9.36372340e-01 1.01601303e+00 6.04515731e-01 3.81164178e-02
-2.45405391e-01 5.07333338e-01 1.46192342e-01 2.23288804e-01
-8.66559669e-02 4.61260468e-01 -8.21725488e-01 1.83546513e-01
-5.25286078e-01 2.99101144e-01 9.87438142e-01 1.10337019e+00
1.20606375e+00 1.56931882e-03 -5.87278046e-02 4.27990556e-01
5.68875432e-01 2.15641081e-01 -2.61600047e-01 -1.01407444e+00
5.06198764e-01 1.19152975e+00 -6.96536079e-02 -1.26309335e+00
-3.05349827e-01 -3.06943525e-02 -6.31609082e-01 1.50634900e-01
2.14150533e-01 -3.52662131e-02 -8.27777863e-01 1.66918528e+00
2.72493184e-01 4.43408936e-01 1.77870527e-01 3.60231191e-01
8.34080338e-01 5.76968610e-01 9.51642543e-02 7.31544346e-02
1.30546165e+00 -4.62161362e-01 -6.89431131e-01 -1.18128493e-01
7.35609889e-01 8.29396595e-04 9.45896983e-01 3.62104386e-01
-7.15583920e-01 -1.77053228e-01 -9.39367235e-01 -1.18761592e-01
-8.60758662e-01 -3.96729037e-02 1.42429280e+00 6.00336611e-01
-7.89776027e-01 5.08804679e-01 -8.07564676e-01 -3.90178226e-02
6.34558678e-01 4.33849663e-01 -6.30370855e-01 -6.36285722e-01
-1.70936930e+00 7.46174753e-01 8.83481026e-01 1.17512429e-02
-6.09317064e-01 -8.93644035e-01 -1.62921929e+00 2.56957650e-01
1.19783413e+00 -8.43694031e-01 5.50464153e-01 -2.15922788e-01
-1.03710222e+00 5.08369684e-01 -3.81329983e-01 -5.82579374e-01
2.98863668e-02 6.71753436e-02 -6.89365029e-01 -7.91554227e-02
1.35523647e-01 6.11663580e-01 5.74731052e-01 -1.38852835e+00
-6.06347978e-01 -4.80948031e-01 6.02792263e-01 -1.89301372e-01
-8.19729343e-02 -2.37552583e-01 -5.83958089e-01 -4.62158948e-01
1.20814674e-01 -6.27848446e-01 -2.79005766e-02 -2.43092269e-01
-9.47010756e-01 -5.97493887e-01 9.11399901e-01 -4.28684890e-01
1.46753705e+00 -1.81600654e+00 4.07085270e-02 8.13343883e-01
4.68454927e-01 4.29762341e-02 1.75498933e-01 4.00088519e-01
-3.74948293e-01 4.74722296e-01 -1.56036064e-01 2.44133890e-01
4.34779376e-01 6.74136758e-01 -4.88153279e-01 6.59240559e-02
3.04539919e-01 1.32024384e+00 -9.87639844e-01 -5.23216307e-01
2.27828428e-01 3.91474456e-01 -5.84606946e-01 -2.72667348e-01
-6.92774594e-01 -3.13184738e-01 -6.62600875e-01 9.65212047e-01
4.92346823e-01 -4.70380098e-01 7.63458192e-01 -3.31250459e-01
2.90125579e-01 2.70664513e-01 -1.52843332e+00 1.39599264e+00
-2.57711858e-01 3.14312547e-01 -3.05997610e-01 -1.21592295e+00
7.41171479e-01 1.08954608e-01 3.60745907e-01 -2.75176018e-01
-9.86683294e-02 -3.24469805e-01 -2.09687680e-01 -5.09899139e-01
2.30968446e-01 -1.90212935e-01 -1.68238178e-01 1.74590141e-01
-3.64170969e-02 -6.63931519e-02 3.59330803e-01 5.11773169e-01
1.49241042e+00 2.66195387e-01 3.38650852e-01 1.16003394e-01
5.14580250e-01 3.47314849e-02 1.00280523e+00 7.74852037e-01
1.45765945e-01 -1.32742539e-01 9.84563172e-01 -5.48486471e-01
-3.18332762e-01 -1.29544067e+00 -6.06484246e-04 1.01083207e+00
2.39046261e-01 -1.06372464e+00 -1.06996298e-01 -1.14263618e+00
5.71211517e-01 1.03448069e+00 -7.83443689e-01 -2.46497422e-01
-4.65653211e-01 -4.44688112e-01 5.93113482e-01 8.12207699e-01
3.11891854e-01 -9.63931382e-01 3.66218504e-03 1.30465776e-01
-5.59620559e-02 -1.02366793e+00 1.73163507e-02 3.08221072e-01
-6.26425087e-01 -1.67160046e+00 5.41165352e-01 -6.64859772e-01
7.31510103e-01 -3.24946910e-01 1.12292099e+00 1.48628294e-01
-3.75711650e-01 4.03868824e-01 -3.96200448e-01 -3.39967817e-01
2.45172959e-02 -2.14228243e-01 1.11600786e-01 -1.39585480e-01
5.86696923e-01 -5.84017396e-01 -5.40952235e-02 1.18016198e-01
-9.90322113e-01 -2.51530081e-01 3.81175965e-01 9.03000832e-01
8.31438363e-01 1.04039717e+00 3.62473339e-01 -1.40531504e+00
4.90941286e-01 -4.50092971e-01 -5.24353683e-01 7.35733449e-01
-8.62799406e-01 6.26654506e-01 6.67056322e-01 -4.52612005e-02
-1.16901541e+00 -1.30903106e-02 1.27276808e-01 -4.90164250e-01
1.31067252e-02 1.18721867e+00 -4.72920179e-01 1.75324440e-01
4.93810445e-01 -1.53860658e-01 -4.12217081e-01 -2.22490817e-01
7.82934248e-01 -1.78775899e-02 4.28313673e-01 -8.95172536e-01
1.06277478e+00 5.85635483e-01 2.96405107e-01 -3.92335027e-01
-8.09802115e-01 -3.69576477e-02 -3.95504117e-01 3.01361889e-01
5.99798143e-01 -8.17833304e-01 -1.22172523e+00 -3.74436587e-01
-8.47169340e-01 -2.25839123e-01 -4.65580702e-01 4.52661663e-01
-2.82193691e-01 3.34136218e-01 -4.44720149e-01 -7.59674370e-01
1.30761817e-01 -7.86896110e-01 6.87503815e-01 -4.13706675e-02
-4.46172893e-01 -1.27683043e+00 -1.28845766e-01 4.01766062e-01
-2.45575577e-01 3.20994675e-01 1.72912490e+00 -5.40247858e-01
-8.46270740e-01 -2.99569339e-01 -2.28662163e-01 5.65595627e-02
3.44262347e-02 2.29882836e-01 -4.22052383e-01 1.84797317e-01
-6.40777051e-01 -3.96313280e-01 1.06895888e+00 2.75645345e-01
1.35051823e+00 -5.81657052e-01 -7.88634479e-01 5.61868608e-01
1.37720978e+00 -5.29833622e-02 7.71033883e-01 6.26321584e-02
8.99183154e-01 4.63556290e-01 6.21461093e-01 3.82795393e-01
8.60327482e-01 4.58053112e-01 4.37682271e-01 2.08711043e-01
-1.38705401e-02 -8.01549554e-01 4.77460831e-01 5.88602647e-02
-2.68272132e-01 -1.06244631e-01 -1.10059583e+00 4.78819430e-01
-2.02685761e+00 -1.20174944e+00 -4.67538796e-02 1.78800893e+00
9.29847896e-01 2.79808640e-01 -2.76806384e-01 1.84916049e-01
4.84865814e-01 1.61784962e-01 -4.51238185e-01 -1.33580983e-01
-2.26896510e-01 5.91465831e-01 4.76423174e-01 8.83008838e-01
-8.55389237e-01 1.27337241e+00 6.07860088e+00 9.10316408e-01
-5.90300858e-01 -8.77032056e-02 2.02396289e-01 1.40138134e-01
-1.01098800e+00 5.31171203e-01 -8.93820941e-01 2.00616166e-01
2.90337294e-01 -8.18075761e-02 6.62636459e-01 8.88030708e-01
-1.71351776e-01 -4.09045294e-02 -1.39170623e+00 8.20239186e-01
-2.95299925e-02 -1.63344467e+00 2.45424554e-01 1.74196512e-02
7.64915407e-01 -2.96813428e-01 -2.30649009e-01 8.36727917e-01
9.98198330e-01 -1.43670595e+00 3.08277875e-01 6.43286526e-01
7.13862181e-01 -9.50762331e-01 7.34725654e-01 1.31101385e-01
-1.40799892e+00 -2.88497180e-01 -2.90191054e-01 -5.41141964e-02
6.66092485e-02 9.81182039e-01 -9.32190895e-01 1.05413353e+00
6.36305630e-01 7.63518631e-01 -3.78479213e-01 2.29747802e-01
-1.07039773e+00 5.53804636e-01 -2.95805037e-01 1.01229645e-01
3.07693221e-02 -2.09118605e-01 3.11709911e-01 9.18721795e-01
-6.23519756e-02 2.31047764e-01 1.66558012e-01 1.27211058e+00
-4.61574882e-01 -2.92812854e-01 -1.01348257e+00 -1.86833322e-01
6.02113843e-01 9.78885531e-01 -3.81270677e-01 -2.57047594e-01
-4.89066035e-01 4.58818853e-01 6.91452146e-01 6.77285135e-01
-9.34673846e-01 -5.66253543e-01 7.76652277e-01 1.46211505e-01
4.84058052e-01 -7.49676973e-02 -2.13341489e-01 -1.14721549e+00
1.56431705e-01 -6.07238531e-01 8.70449543e-01 -7.48688877e-01
-1.13270164e+00 -1.71777625e-02 1.42136067e-01 -4.21511024e-01
-2.78612137e-01 -9.70527172e-01 -5.87957442e-01 5.95400572e-01
-1.67203009e+00 -1.34833705e+00 2.14875266e-01 8.83073449e-01
-1.13884106e-01 -1.94894820e-01 9.20103073e-01 -1.05134375e-01
-6.77990735e-01 7.64201105e-01 -3.13512594e-01 3.52740526e-01
2.29051471e-01 -1.38853323e+00 -1.95899785e-01 8.10837150e-01
4.06086653e-01 1.08571172e+00 4.33660477e-01 -9.97491241e-01
-1.91685247e+00 -1.19340825e+00 1.05449367e+00 -7.31010139e-01
7.73584008e-01 -1.97353303e-01 -8.60344410e-01 1.34547222e+00
-1.77089766e-01 2.15248629e-01 8.16731334e-01 9.34016764e-01
-7.86470056e-01 -2.96832711e-01 -1.10725188e+00 6.92283571e-01
1.19833028e+00 -6.98643804e-01 -9.08676267e-01 1.78762898e-01
7.83209145e-01 -9.41483080e-02 -1.08573079e+00 7.72495866e-01
4.97172087e-01 -6.06050253e-01 1.25574982e+00 -9.63518262e-01
3.78622353e-01 -5.51346064e-01 -1.55707866e-01 -1.05629158e+00
-4.29137856e-01 -5.64476788e-01 -8.15849543e-01 1.12933958e+00
8.26577067e-01 -5.32943368e-01 1.04133177e+00 9.05698180e-01
5.20283431e-02 -9.43119884e-01 -8.96493316e-01 -7.80585647e-01
-1.80056915e-01 -9.25444484e-01 7.90371776e-01 1.30486631e+00
7.68601537e-01 3.14863890e-01 -2.12783262e-01 4.76130098e-01
5.26730478e-01 4.87943828e-01 6.09789073e-01 -1.35962069e+00
-3.25028807e-01 -3.63648683e-01 -4.82796729e-01 -7.40549326e-01
7.66113102e-01 -1.37662661e+00 -3.58734071e-01 -1.89576435e+00
7.59425461e-02 -6.52922213e-01 -4.24723357e-01 1.31855571e+00
-2.15408742e-01 -2.82866567e-01 -1.29565641e-01 -3.70105296e-01
-6.34713948e-01 6.35314822e-01 1.12891734e+00 -5.44182479e-01
-1.45837516e-01 -4.20875341e-01 -1.04872584e+00 9.01929557e-01
4.84611154e-01 -4.06979471e-01 -5.41417778e-01 -2.26227522e-01
8.93218577e-01 7.54524991e-02 4.04803485e-01 -4.18126434e-01
4.01360661e-01 -4.43308651e-01 2.18234777e-01 -3.16979349e-01
2.50906318e-01 -9.14522171e-01 2.12632060e-01 1.07006647e-01
-4.17792127e-02 -2.92289317e-01 -1.47667807e-02 8.38945568e-01
-3.82199049e-01 -1.00805901e-01 -2.20124591e-02 -5.31629426e-03
-1.04169130e+00 3.55737358e-01 4.37839665e-02 -5.32150045e-02
9.81435716e-01 -1.68867856e-01 -2.22701028e-01 -2.19467491e-01
-8.71041894e-01 5.59195280e-01 1.29433632e-01 2.37010568e-01
8.39882314e-01 -1.39858699e+00 -4.40676183e-01 7.85191450e-03
3.41957569e-01 3.49685252e-01 7.03051239e-02 6.27269149e-01
-2.83872575e-01 3.86425346e-01 2.20657334e-01 -6.60565915e-03
-8.67077231e-01 7.26970196e-01 6.23780824e-02 -6.58350766e-01
-5.60799778e-01 9.04726207e-01 -1.12321869e-01 -7.13857889e-01
3.17287534e-01 -7.16796637e-01 -8.42828602e-02 -1.37255356e-01
2.70575225e-01 3.92144084e-01 -2.36665025e-01 -1.73031092e-01
-6.24093771e-01 3.27761143e-01 -9.41786021e-02 2.91912526e-01
1.38591850e+00 3.70134264e-01 -3.68575722e-01 1.52863100e-01
8.17489922e-01 3.88105959e-01 -7.58665204e-01 -3.10664982e-01
1.58777997e-01 -4.26210105e-01 -1.48148993e-02 -8.54985178e-01
-1.05026317e+00 4.58449364e-01 -1.73904359e-01 1.20452404e-01
8.45841646e-01 5.56261778e-01 4.12887365e-01 8.45345914e-01
6.09996378e-01 -1.16483605e+00 -2.81198442e-01 7.51845062e-01
6.07357562e-01 -1.32366204e+00 2.91824549e-01 -9.52609539e-01
-5.87173223e-01 9.88130689e-01 3.75323802e-01 1.55044839e-01
7.90480316e-01 2.02004746e-01 -4.36028540e-01 -6.59872115e-01
-7.90642679e-01 -3.15663487e-01 2.31252626e-01 6.18861675e-01
2.86774039e-01 3.69187981e-01 -4.81577106e-02 9.76556301e-01
-2.03084052e-01 1.14174098e-01 1.30822390e-01 8.86273623e-01
-2.25923419e-01 -1.41120660e+00 -2.21501812e-01 5.64918339e-01
-2.04200640e-01 -2.88686991e-01 -3.38833511e-01 1.13346553e+00
3.79614145e-01 9.28381562e-01 -3.82642448e-01 -4.80940729e-01
3.72273117e-01 4.13985670e-01 6.14372253e-01 -7.60452390e-01
8.10823776e-03 -6.28544748e-01 2.89839476e-01 -6.40123904e-01
-1.92260966e-01 -5.51292479e-01 -1.86190057e+00 -4.43329871e-01
-3.09456974e-01 2.12991223e-01 3.93952653e-02 1.03680122e+00
4.28416610e-01 6.07617497e-01 1.51798412e-01 7.72984475e-02
-2.05787495e-01 -4.13449347e-01 -1.04387343e+00 4.52434212e-01
-1.37884961e-02 -9.36959743e-01 -1.43193722e-01 9.48298648e-02] | [8.843635559082031, 7.702932357788086] |
3d184742-6c48-4c9a-98cd-a350ee605fad | digitizing-handwriting-with-a-sensor-pen-a | 2107.03704 | null | https://arxiv.org/abs/2107.03704v1 | https://arxiv.org/pdf/2107.03704v1.pdf | Digitizing Handwriting with a Sensor Pen: A Writer-Independent Recognizer | Online handwriting recognition has been studied for a long time with only few practicable results when writing on normal paper. Previous approaches using sensor-based devices encountered problems that limited the usage of the developed systems in real-world applications. This paper presents a writer-independent system that recognizes characters written on plain paper with the use of a sensor-equipped pen. This system is applicable in real-world applications and requires no user-specific training for recognition. The pen provides linear acceleration, angular velocity, magnetic field, and force applied by the user, and acts as a digitizer that transforms the analogue signals of the sensors into timeseries data while writing on regular paper. The dataset we collected with this pen consists of Latin lower-case and upper-case alphabets. We present the results of a convolutional neural network model for letter classification and show that this approach is practical and achieves promising results for writer-independent character recognition. This work aims at providing a realtime handwriting recognition system to be used for writing on normal paper. | ['Bjoern Eskofier', 'Jens Barth', 'Tim Hamann', 'Mohamad Wehbi'] | 2021-07-08 | null | null | null | null | ['handwriting-recognition'] | ['computer-vision'] | [ 4.05074269e-01 -6.32908583e-01 -1.55009523e-01 -3.65002126e-01
1.43926069e-01 -7.73122489e-01 6.03799582e-01 -1.08931221e-01
-5.64943314e-01 4.24364775e-01 -4.17711765e-01 -4.52907056e-01
-2.93260962e-01 -8.51236343e-01 -6.63107038e-01 -5.71617186e-01
3.69173259e-01 5.94967723e-01 2.50018060e-01 -4.99236465e-01
7.11433053e-01 1.23950052e+00 -1.47516382e+00 2.69491136e-01
3.61828119e-01 6.99609101e-01 4.95451570e-01 1.28882241e+00
6.32309243e-02 7.91307569e-01 -1.09948456e+00 9.99774039e-03
1.36105701e-01 -3.28777462e-01 -4.31938946e-01 2.58188456e-01
3.33369702e-01 -6.36006355e-01 -6.95735335e-01 5.21570623e-01
6.32921815e-01 1.14727933e-02 6.23895347e-01 -5.54272711e-01
-8.95049453e-01 4.28020746e-01 -4.52974811e-02 1.08247787e-01
4.91037786e-01 -1.07522167e-01 3.36123735e-01 -3.78315419e-01
7.35485792e-01 7.58077323e-01 6.19286776e-01 3.25798631e-01
-6.84570730e-01 -1.61890581e-01 -3.77016902e-01 3.15716386e-01
-9.14987803e-01 -1.64726913e-01 7.87123799e-01 -3.23029727e-01
1.32077670e+00 5.72135329e-01 4.38355803e-01 1.15890610e+00
4.21271950e-01 7.99731910e-01 1.13950145e+00 -8.65351319e-01
2.55358964e-01 1.16303995e-01 6.29146934e-01 4.06479478e-01
1.45876445e-02 -2.02714339e-01 -5.32792628e-01 4.11378652e-01
1.16912913e+00 9.43029076e-02 -2.50938296e-01 2.09429637e-01
-9.70489204e-01 8.01644847e-02 -5.13189584e-02 7.50267088e-01
-3.54249805e-01 -3.49173881e-02 3.98728371e-01 6.13864779e-01
-9.63616669e-02 3.11527103e-01 -5.21441512e-02 -8.54668975e-01
-1.07148790e+00 1.02921098e-01 1.15053940e+00 1.02943802e+00
-2.05489665e-01 8.94921198e-02 -4.45908830e-02 8.15914750e-01
-7.20256045e-02 8.56543005e-01 7.33494222e-01 -1.78171381e-01
5.31008601e-01 4.96473253e-01 3.86325687e-01 -9.78313148e-01
-4.59679544e-01 -5.16745914e-03 -6.44147217e-01 4.07494992e-01
6.91935122e-01 -1.58010006e-01 -8.28897774e-01 6.15661025e-01
-4.98905301e-01 -3.86204779e-01 2.25270586e-03 1.16555095e+00
5.29860377e-01 5.54956019e-01 -6.48185849e-01 -2.13378537e-02
1.01798916e+00 -7.22927034e-01 -1.21560407e+00 1.39503062e-01
6.58994988e-02 -8.10620666e-01 1.11326861e+00 9.81922448e-01
-8.84143233e-01 -7.26792276e-01 -1.49535310e+00 -1.11426324e-01
-6.03334546e-01 9.06256020e-01 4.44315910e-01 1.02398920e+00
-6.58632636e-01 9.62921739e-01 -1.07568085e+00 -4.82230246e-01
1.21154860e-01 5.11649966e-01 -3.20126235e-01 4.49956894e-01
-7.23718524e-01 1.04527259e+00 1.33434653e-01 4.83871728e-01
-1.89864084e-01 1.54092059e-01 -2.04125300e-01 -8.21613744e-02
-1.38917923e-01 5.08070827e-01 1.15637600e+00 -7.16025949e-01
-2.34717155e+00 6.76767707e-01 -4.76981811e-02 -3.50102752e-01
9.86053884e-01 -1.15929089e-01 -8.35210800e-01 -2.33476590e-02
-5.41391671e-01 -4.45789039e-01 9.02390778e-01 -4.13296431e-01
-2.82107800e-01 -2.76194930e-01 -4.86130655e-01 -8.93279240e-02
-6.52518332e-01 1.41517192e-01 -5.86775720e-01 -5.11678636e-01
1.30161628e-01 -7.14066923e-01 6.27721548e-01 -6.55778795e-02
-3.54186267e-01 -1.85893387e-01 1.11562431e+00 -9.14651930e-01
8.85255754e-01 -1.97205412e+00 5.71211129e-02 3.65744889e-01
-4.67946857e-01 8.37644815e-01 1.34416088e-01 7.20686674e-01
2.42436498e-01 -5.66969454e-01 1.18301995e-01 2.85786428e-02
-2.75300220e-02 2.43954107e-01 -5.78527510e-01 4.53571349e-01
5.43603450e-02 6.80042207e-01 -5.70750177e-01 1.69557124e-01
5.42893171e-01 4.18180823e-01 3.49070698e-01 1.95484713e-01
-2.57033631e-02 -1.00263238e-01 -2.15640157e-01 6.61116481e-01
6.75267935e-01 1.15417562e-01 3.37097526e-01 1.78762916e-02
-3.92808467e-01 8.99562389e-02 -1.32968605e+00 1.03522074e+00
-4.66389596e-01 1.43837428e+00 -1.00024752e-01 -7.78346896e-01
1.57496440e+00 1.37439460e-01 -2.52011031e-01 -8.49521279e-01
3.96649629e-01 6.12360477e-01 -1.23583883e-01 -9.35161352e-01
6.60734594e-01 3.83004785e-01 4.47730273e-01 6.44527733e-01
-1.22456208e-01 -2.28619367e-01 3.07722241e-01 -3.75904590e-01
9.39300835e-01 3.38191867e-01 -1.95519939e-01 1.25696948e-02
4.82746184e-01 -2.28785813e-01 -3.80343914e-01 7.95721054e-01
1.57318383e-01 6.34941697e-01 2.40285367e-01 -5.80094635e-01
-1.22996736e+00 -5.37161529e-01 -3.12880009e-01 8.02023947e-01
1.21609122e-01 1.73793390e-01 -5.13301015e-01 -3.27623039e-02
3.25175434e-01 5.48673987e-01 -3.26914728e-01 3.69089425e-01
-7.97416091e-01 -4.67792928e-01 8.98637772e-01 9.36813712e-01
4.91686612e-01 -1.29808366e+00 -9.70236421e-01 5.53874552e-01
5.11642694e-01 -9.64714706e-01 7.97633082e-03 4.69690382e-01
-7.15470850e-01 -8.82243454e-01 -9.08991754e-01 -6.86615825e-01
6.87362611e-01 -1.41813412e-01 3.48507375e-01 -2.47645125e-01
-3.67412120e-01 1.95742458e-01 -4.25548673e-01 -6.76346123e-01
-3.76677662e-01 1.71203390e-01 2.44375154e-01 8.00536424e-02
7.17965782e-01 -2.12162435e-01 4.61470634e-02 3.54012638e-01
-6.49004877e-01 -2.69406825e-01 5.22440553e-01 5.68170309e-01
5.54345250e-02 -2.14795992e-01 2.08942324e-01 -5.12050867e-01
1.22662711e+00 2.55868733e-01 -8.17565858e-01 3.76620412e-01
-3.29108179e-01 4.21867557e-02 1.17812943e+00 -6.78392828e-01
-8.24805617e-01 -1.16755152e-02 -1.94863021e-01 -3.53829935e-02
-3.12944591e-01 2.26689830e-01 4.31033745e-02 -3.77558708e-01
9.32041168e-01 6.80605471e-01 1.85442224e-01 -7.77689338e-01
-2.85931051e-01 1.76812923e+00 1.00994658e+00 -2.87537307e-01
2.04315215e-01 4.18630391e-01 -5.63616958e-03 -1.37536502e+00
1.24524266e-01 -2.08867893e-01 -8.78298819e-01 -4.19558227e-01
1.74120128e-01 -2.56814927e-01 -1.14520621e+00 1.41664064e+00
-1.20708978e+00 -3.80732715e-01 1.45568848e-02 5.10617673e-01
-4.75630105e-01 3.86947423e-01 -7.80587614e-01 -1.05855739e+00
-3.44507664e-01 -8.30212831e-01 7.72367358e-01 4.86842513e-01
-2.17308700e-01 -8.29622805e-01 -1.10207118e-01 9.90930498e-02
6.16920173e-01 -1.24662472e-02 3.48851681e-01 -5.01651049e-01
-3.03236336e-01 -9.26493227e-01 -1.58958986e-01 5.95213532e-01
3.26076150e-01 3.23061287e-01 -9.66711342e-01 -2.77363569e-01
-2.34088898e-02 -1.95592329e-01 6.12910032e-01 -4.19517681e-02
1.27458358e+00 9.60291252e-02 -1.59103066e-01 4.90665555e-01
1.26032114e+00 6.68495059e-01 8.26593399e-01 5.01434505e-01
6.34006262e-01 -2.03733817e-02 3.07023466e-01 5.27395308e-01
-2.86411107e-01 6.60142183e-01 -2.63739288e-01 1.82826728e-01
9.61910412e-02 1.16510340e-03 2.96599895e-01 6.87664747e-01
-7.43983150e-01 -6.18695617e-01 -1.03726149e+00 2.75881141e-01
-1.55129707e+00 -8.84908676e-01 -4.36531663e-01 2.06024146e+00
8.84059250e-01 2.18886673e-01 -2.35759304e-03 7.65623093e-01
6.53914809e-01 -2.55521297e-01 -5.58937311e-01 -1.13641429e+00
-3.57850969e-01 4.50953066e-01 7.77917743e-01 3.78049999e-01
-1.05818844e+00 5.79363883e-01 6.68702459e+00 2.99444199e-01
-1.93424249e+00 -5.08807063e-01 1.49647789e-02 5.95467649e-02
5.04312813e-01 -6.37452960e-01 -8.51953387e-01 6.85771942e-01
9.28850710e-01 1.50439695e-01 5.86509287e-01 6.74870372e-01
1.24259457e-01 -1.64719671e-01 -1.29044342e+00 1.17784262e+00
2.19609171e-01 -1.37248468e+00 -1.54315293e-01 -2.52465516e-01
3.37416142e-01 -1.96801037e-01 -4.69009094e-02 -2.69801557e-01
-4.65205222e-01 -1.06948113e+00 7.31437385e-01 1.12121820e+00
1.03009939e+00 -5.24551332e-01 9.00836825e-01 5.11863172e-01
-4.34902042e-01 -2.14558560e-02 -4.70174432e-01 -7.21044838e-01
-8.35401863e-02 3.05377543e-01 -9.21655238e-01 1.38242275e-01
1.71193823e-01 5.99530458e-01 -3.02712262e-01 7.41477072e-01
-7.96268433e-02 7.09971368e-01 -5.42019546e-01 -8.98952186e-01
4.99123484e-02 -3.33782852e-01 3.33420396e-01 1.69425452e+00
2.51466215e-01 -3.90992090e-02 -6.06039226e-01 6.39352202e-01
9.40861702e-02 -6.08626828e-02 -4.38582540e-01 -5.54912150e-01
1.81080803e-01 8.81163359e-01 -8.52231979e-01 -2.86659449e-01
-6.16384931e-02 1.37014174e+00 3.40592302e-02 1.58195898e-01
-3.24789643e-01 -1.37981737e+00 -2.38267973e-01 7.37006739e-02
4.11723882e-01 -6.84836447e-01 -5.80954790e-01 -8.91930521e-01
6.22651696e-01 -7.69899666e-01 -3.95164549e-01 -7.39092290e-01
-9.88706470e-01 3.66221219e-01 -5.46485484e-01 -1.07824373e+00
-4.02247638e-01 -1.51473379e+00 -5.53949654e-01 1.34748280e+00
-8.96194458e-01 -8.45370710e-01 -6.07842863e-01 2.73571908e-01
4.07905072e-01 -8.98460448e-01 1.00360167e+00 8.40244293e-02
-4.80825126e-01 8.39560926e-01 9.46554899e-01 5.03689647e-01
5.97431839e-01 -1.18892395e+00 5.21791101e-01 7.16031611e-01
-6.87713996e-02 5.78318655e-01 6.38983071e-01 -6.94896579e-01
-2.11067486e+00 -4.92545485e-01 1.04706466e+00 -4.18633908e-01
3.95815790e-01 -5.55016279e-01 -6.86914861e-01 4.64911759e-01
-1.68033764e-02 -2.26579562e-01 1.89110249e-01 -2.85130233e-01
-6.70795068e-02 -2.59777427e-01 -1.14190948e+00 3.31296474e-01
5.54246187e-01 -7.82861114e-01 -5.63189030e-01 4.53421772e-01
-4.90311414e-01 -8.65932226e-01 -6.17445290e-01 -3.09635520e-01
1.07246327e+00 -8.02579284e-01 4.05849755e-01 -3.87362778e-01
4.35003012e-01 -2.14597404e-01 2.11241931e-01 -1.00476289e+00
-1.20805010e-01 -6.54095292e-01 -3.02010864e-01 8.09189260e-01
5.46223707e-02 -7.38424361e-01 6.47600770e-01 5.45235217e-01
2.19824731e-01 -4.81870651e-01 -6.66731298e-01 -9.71982896e-01
-2.60535657e-01 -2.60554940e-01 4.01494920e-01 7.13614047e-01
3.99521708e-01 -2.59909660e-01 -2.99201936e-01 -1.47545915e-02
1.81784302e-01 1.39408454e-01 6.51659608e-01 -9.74228799e-01
-3.59848768e-01 -4.35015827e-01 -9.19425368e-01 -1.11487031e+00
-2.07085162e-01 -5.78780055e-01 -1.18205875e-01 -1.32762539e+00
-3.40322345e-01 -1.01079941e-01 1.60915047e-01 3.83124471e-01
4.30560261e-01 1.21529743e-01 1.46491200e-01 1.50502726e-01
4.57922649e-03 -1.46340519e-01 1.02183020e+00 -2.25149259e-01
-1.95192009e-01 2.16449410e-01 2.12615713e-01 2.12890640e-01
8.41132224e-01 1.79558590e-01 -4.48052697e-02 -5.06024659e-01
1.30878702e-01 -3.06082647e-02 2.68565983e-01 -1.26106620e+00
5.33187747e-01 7.99827576e-02 9.20281410e-01 -6.80387437e-01
2.78968699e-02 -7.36104012e-01 -3.17408256e-02 4.71125305e-01
-3.08239788e-01 -1.07526235e-01 2.30767980e-01 1.94433145e-02
-1.60377055e-01 -4.55216050e-01 5.78837991e-01 2.25245506e-01
-6.68611228e-01 -5.13777435e-01 -6.98878169e-01 -7.88837314e-01
7.62688696e-01 -6.20826423e-01 -3.38986427e-01 -2.29574025e-01
-4.78664994e-01 -3.69645089e-01 3.60269547e-01 6.12376153e-01
6.21102095e-01 -9.08207238e-01 -6.27856374e-01 8.66972566e-01
-2.74054199e-01 -5.44028521e-01 -9.91625041e-02 2.62542576e-01
-1.47741938e+00 8.49743187e-01 -8.37267041e-01 -4.32334244e-01
-1.33756614e+00 1.54884040e-01 5.55809021e-01 4.46833819e-01
-7.13383675e-01 6.11881673e-01 -1.43991745e+00 -2.28874579e-01
5.41209519e-01 -5.42500615e-01 -2.31189832e-01 -8.51817057e-02
7.85456300e-01 8.24784040e-01 6.65945470e-01 -1.69876352e-01
-2.42278248e-01 6.12462342e-01 -1.21546283e-01 -2.68881470e-01
1.45124710e+00 5.42990327e-01 -4.15771678e-02 7.73551822e-01
8.68259013e-01 -1.24292083e-01 -9.89262283e-01 1.62215158e-01
-1.32562533e-01 -5.12137473e-01 -5.56080528e-02 -1.07260334e+00
-5.70240080e-01 9.98498499e-01 8.36879730e-01 3.32225651e-01
8.24010432e-01 -7.93025613e-01 4.44974720e-01 1.29079080e+00
4.24862534e-01 -1.81648505e+00 -3.88121933e-01 8.68152916e-01
1.06637549e+00 -8.11916232e-01 -1.49393916e-01 1.65878475e-01
-3.75398159e-01 2.05094242e+00 2.09640056e-01 -3.03400576e-01
3.14688534e-01 7.79667795e-01 3.18455338e-01 9.27906558e-02
-2.26531804e-01 4.57627416e-01 1.15735732e-01 6.16028428e-01
8.18014264e-01 4.22167242e-01 -6.04703844e-01 4.71546710e-01
-3.14704448e-01 5.77798426e-01 9.24502611e-01 1.39807248e+00
-2.25400835e-01 -1.11501169e+00 -6.62483573e-01 5.87488592e-01
-1.30266234e-01 5.01319766e-01 -6.82155073e-01 4.09621090e-01
-1.32533565e-01 6.44071698e-01 3.92775744e-01 -3.47911358e-01
6.30206347e-01 2.27759659e-01 8.24725986e-01 1.69390544e-01
-5.22604644e-01 -3.64859402e-01 -1.00615904e-01 -1.28870934e-01
2.37593167e-02 -6.15756154e-01 -1.17112410e+00 -4.38503206e-01
-1.67897403e-01 -2.40482062e-01 1.33009982e+00 8.59213829e-01
2.78418273e-01 5.02218783e-01 5.44062197e-01 -9.33542550e-01
-8.53411794e-01 -1.24399841e+00 -9.62456346e-01 3.56539428e-01
4.18806136e-01 -1.71793476e-01 7.33445361e-02 6.84956238e-02] | [11.882561683654785, 2.5682590007781982] |
10eb52c1-b863-4c0f-9c19-333d13d2736a | muscaps-generating-captions-for-music-audio | 2104.11984 | null | https://arxiv.org/abs/2104.11984v1 | https://arxiv.org/pdf/2104.11984v1.pdf | MusCaps: Generating Captions for Music Audio | Content-based music information retrieval has seen rapid progress with the adoption of deep learning. Current approaches to high-level music description typically make use of classification models, such as in auto-tagging or genre and mood classification. In this work, we propose to address music description via audio captioning, defined as the task of generating a natural language description of music audio content in a human-like manner. To this end, we present the first music audio captioning model, MusCaps, consisting of an encoder-decoder with temporal attention. Our method combines convolutional and recurrent neural network architectures to jointly process audio-text inputs through a multimodal encoder and leverages pre-training on audio data to obtain representations that effectively capture and summarise musical features in the input. Evaluation of the generated captions through automatic metrics shows that our method outperforms a baseline designed for non-music audio captioning. Through an ablation study, we unveil that this performance boost can be mainly attributed to pre-training of the audio encoder, while other design choices - modality fusion, decoding strategy and the use of attention - contribute only marginally. Our model represents a shift away from classification-based music description and combines tasks requiring both auditory and linguistic understanding to bridge the semantic gap in music information retrieval. | ['Gyorgy Fazekas', 'Elio Quinton', 'Emmanouil Benetos', 'Ilaria Manco'] | 2021-04-24 | null | null | null | null | ['audio-captioning', 'music-information-retrieval'] | ['audio', 'music'] | [ 5.34304142e-01 1.64394200e-01 -8.58461764e-03 -1.40618607e-01
-1.47737181e+00 -7.57748008e-01 7.14221358e-01 1.61701709e-01
-3.53954062e-02 2.06928685e-01 1.04453623e+00 2.74439812e-01
-2.98092276e-01 -2.40151078e-01 -6.49647713e-01 -3.15581381e-01
-4.54304926e-03 5.65003633e-01 -4.01184350e-01 -4.86818939e-01
2.64315039e-01 -1.54151574e-01 -1.82493722e+00 8.77164066e-01
6.34955913e-02 1.08328915e+00 8.36936086e-02 1.04534268e+00
-2.05471009e-01 1.00976765e+00 -6.83396459e-01 -3.98173481e-01
-1.76546648e-01 -7.64994621e-01 -1.02597523e+00 -1.74754709e-01
4.06510890e-01 -9.34912339e-02 -1.68472320e-01 4.07872409e-01
8.61778677e-01 1.44930094e-01 6.40936673e-01 -1.11498511e+00
-6.12554610e-01 1.23979104e+00 -1.21501880e-02 -2.48751089e-01
8.88240159e-01 -1.01660080e-01 1.60404325e+00 -6.08792543e-01
4.39249158e-01 1.07968044e+00 1.03034699e+00 7.29041159e-01
-1.41857588e+00 -6.05253637e-01 -2.79743880e-01 3.11296791e-01
-1.39336872e+00 -7.94853628e-01 9.45122778e-01 -5.56748211e-01
1.02185166e+00 4.36120063e-01 8.16329241e-01 1.38343143e+00
-2.96120882e-01 8.36033046e-01 6.00756407e-01 -5.60028613e-01
4.51516738e-04 -1.95746347e-01 -2.76586235e-01 5.79520091e-02
-5.35266876e-01 -4.55549434e-02 -9.94092822e-01 -2.28401259e-01
5.80994248e-01 -4.48862702e-01 -5.31454310e-02 -6.38103932e-02
-1.47454071e+00 6.31617188e-01 2.76985884e-01 4.94499534e-01
-5.24598718e-01 7.15837896e-01 9.65201497e-01 2.30929911e-01
2.12778136e-01 8.61583829e-01 -2.22283721e-01 -4.83137429e-01
-1.36700702e+00 5.02871513e-01 7.98771203e-01 7.27040708e-01
1.90768272e-01 1.45643041e-01 -5.47721207e-01 1.09095252e+00
2.87018627e-01 3.09559077e-01 8.22024345e-01 -1.10230529e+00
2.19155446e-01 1.31886721e-01 -6.92291707e-02 -5.54342330e-01
-3.50418985e-01 -5.60235441e-01 -4.80263591e-01 -2.36986771e-01
1.16079114e-02 2.10805923e-01 -7.28272676e-01 1.97591150e+00
-3.28236520e-01 2.67200679e-01 1.92984581e-01 9.64148819e-01
1.06453621e+00 6.53547108e-01 2.81989336e-01 -1.55338924e-02
1.49085236e+00 -8.31876397e-01 -7.34012663e-01 1.28100872e-01
3.83848280e-01 -9.97867644e-01 1.06939960e+00 3.76361161e-01
-1.45942175e+00 -6.89514518e-01 -1.08232880e+00 -2.64662445e-01
-1.57077700e-01 1.53290033e-01 3.71658921e-01 2.16896445e-01
-1.11458886e+00 6.87486053e-01 -3.65805328e-01 -3.41349334e-01
1.94228485e-01 4.31104660e-01 -1.66088328e-01 5.87799907e-01
-1.21566081e+00 5.14375329e-01 6.57340407e-01 -2.35987663e-01
-9.73710895e-01 -8.14500451e-01 -8.50490570e-01 2.50791460e-01
-4.32930291e-02 -1.09138203e+00 1.79843020e+00 -1.42799151e+00
-1.73680890e+00 7.99409509e-01 5.11838570e-02 -6.39059544e-01
-2.34747261e-01 -2.68709868e-01 -3.33960950e-01 2.85647124e-01
8.67067799e-02 1.19371879e+00 8.64718676e-01 -1.11855829e+00
-3.21100026e-01 1.05987743e-01 -5.24162985e-02 3.53720248e-01
-3.19938242e-01 2.13121757e-01 -2.77765661e-01 -9.71851826e-01
-1.98068023e-01 -1.12187183e+00 2.12113202e-01 -5.81293166e-01
-4.58122879e-01 -1.56227291e-01 4.05381143e-01 -6.87936842e-01
1.46464217e+00 -2.32204318e+00 3.31115097e-01 -1.38596958e-02
-2.04521820e-01 2.70547476e-02 -4.32087332e-01 8.55420291e-01
-4.34190303e-01 5.56214713e-02 -2.10977599e-01 -5.41379690e-01
5.01007378e-01 -1.34767950e-01 -7.95921326e-01 -1.36416301e-01
3.46762359e-01 1.16625535e+00 -8.38841617e-01 -3.31755280e-01
-1.35600548e-02 9.06704485e-01 -7.73167193e-01 1.77030906e-01
-5.40866137e-01 5.18659055e-01 -9.08489153e-03 4.89627749e-01
-2.31518298e-01 -6.69087749e-03 -1.17207952e-01 -2.93895155e-01
-3.55079137e-02 1.00595760e+00 -9.67313409e-01 2.66315198e+00
-6.91145539e-01 7.38840163e-01 -7.51642510e-02 -7.06437349e-01
7.44185984e-01 1.05594170e+00 6.58811927e-01 -6.80239737e-01
1.45125434e-01 4.16608810e-01 -1.29439384e-01 -4.38564628e-01
6.15447521e-01 -4.95785028e-01 -6.24228656e-01 6.44036949e-01
3.84053975e-01 -2.46561289e-01 -2.78414134e-02 -2.08727326e-02
1.11200023e+00 5.55119038e-01 1.57680064e-01 2.84518868e-01
3.48302633e-01 5.51667586e-02 3.23867351e-02 7.42648363e-01
1.90064028e-01 1.14064860e+00 9.95898321e-02 -5.92515841e-02
-1.21611154e+00 -9.32950497e-01 2.84925811e-02 1.73551512e+00
-4.17597622e-01 -9.88076925e-01 -6.68300867e-01 -1.80611074e-01
-2.00414717e-01 5.25247872e-01 -5.23378074e-01 -3.14096034e-01
-5.28518140e-01 -2.43944243e-01 1.11010993e+00 4.34529811e-01
-2.76540462e-02 -1.60856950e+00 -4.70777780e-01 5.36414623e-01
-4.51791465e-01 -8.37870061e-01 -5.52036107e-01 2.67161131e-01
-6.39679313e-01 -5.58055639e-01 -9.96871114e-01 -8.64581764e-01
-3.92812639e-01 -2.57031024e-01 1.33582151e+00 -3.62268269e-01
-2.96519548e-01 7.63350546e-01 -6.29968584e-01 -5.95155656e-01
-5.64529300e-01 5.90315700e-01 -5.13860881e-02 4.77438569e-02
1.94047764e-01 -1.01835382e+00 -5.40837109e-01 -1.53203115e-01
-9.96869445e-01 3.27357709e-01 8.10519636e-01 6.60450041e-01
4.76905704e-01 -5.58657110e-01 1.03178239e+00 -3.79578173e-01
8.29694271e-01 -4.93169904e-01 3.82696152e-01 -7.75384232e-02
-2.24641889e-01 2.30458051e-01 2.98191637e-01 -7.29905903e-01
-5.85790932e-01 3.20550680e-01 -3.63767177e-01 -6.74472094e-01
-2.75431663e-01 6.77449048e-01 4.93089557e-02 4.56757218e-01
7.28406072e-01 2.64389336e-01 -1.27202958e-01 -6.44791484e-01
6.38353348e-01 1.00801170e+00 1.05097592e+00 -5.92923403e-01
5.84944367e-01 1.60805315e-01 -1.13493413e-01 -5.09057164e-01
-1.04542470e+00 -5.67271769e-01 -3.25488567e-01 -3.28651845e-01
9.00973678e-01 -1.12888336e+00 -7.44416833e-01 -1.79645211e-01
-1.20897818e+00 -9.67450216e-02 -6.88055575e-01 4.51759517e-01
-1.16252089e+00 -7.11972564e-02 -7.44886398e-01 -1.04906332e+00
-7.73015618e-01 -6.46960616e-01 1.70723474e+00 -1.98373422e-01
-1.01677334e+00 -5.82364500e-01 4.73385245e-01 4.79465336e-01
6.19425356e-01 2.15607211e-01 8.08076203e-01 -8.61022472e-01
-2.74191916e-01 -8.65398645e-02 1.05844297e-01 4.82247360e-02
-2.31547982e-01 -5.00458121e-01 -1.55290306e+00 5.02674654e-03
-4.56127495e-01 -7.98162282e-01 9.16345417e-01 2.10075408e-01
1.09784591e+00 -3.53856891e-01 2.95955896e-01 4.33872908e-01
1.11986089e+00 5.41764125e-02 6.40771866e-01 3.98534566e-01
5.88603437e-01 6.47003055e-01 3.16520542e-01 5.06785274e-01
3.08478385e-01 1.08900416e+00 3.53029191e-01 4.56690527e-02
-6.36809051e-01 -6.21268690e-01 6.24946773e-01 1.13098145e+00
6.56623989e-02 8.16028491e-02 -7.54166722e-01 7.06269979e-01
-1.89773083e+00 -1.23673999e+00 2.88870662e-01 1.96402180e+00
1.15092707e+00 -2.07804039e-01 5.73101342e-01 6.81496382e-01
5.29220164e-01 -4.81826365e-02 -3.53420764e-01 -6.02670670e-01
-6.66472390e-02 4.80468839e-01 -2.24432111e-01 2.14954272e-01
-1.09133792e+00 7.84456193e-01 6.42327118e+00 8.26709390e-01
-1.09706879e+00 1.98413879e-01 -1.49211893e-02 -5.66753805e-01
-3.08086187e-01 -1.43093377e-01 -2.30490610e-01 2.23883852e-01
1.48745692e+00 3.84297594e-02 6.34810150e-01 3.27295244e-01
2.47600466e-01 6.09074891e-01 -1.36883330e+00 1.30608320e+00
3.89767647e-01 -1.18204606e+00 4.04310048e-01 -1.60773337e-01
4.12634730e-01 -5.82763851e-02 1.58678457e-01 5.47970355e-01
-1.99237645e-01 -1.20882273e+00 1.28810704e+00 8.16077173e-01
9.57252204e-01 -6.13140345e-01 5.88658273e-01 -1.47848427e-01
-1.36488593e+00 -2.02421740e-01 3.55465949e-01 -2.96668410e-01
2.70813197e-01 -4.64909896e-02 -9.98656750e-01 4.91286337e-01
4.96227950e-01 8.86733770e-01 -3.43925267e-01 1.20613825e+00
1.38416886e-01 8.12403500e-01 -6.07578009e-02 2.45331362e-01
5.78669235e-02 3.37401390e-01 7.57262707e-01 1.64026308e+00
5.64679503e-01 -2.86080450e-01 3.73016968e-02 8.63480031e-01
-1.05275400e-01 4.39223439e-01 -5.17343938e-01 -5.97937644e-01
2.89112598e-01 9.94525433e-01 -2.44695216e-01 -1.51987746e-01
-1.54183870e-02 1.03433239e+00 -3.10281273e-02 1.62681788e-01
-7.04321325e-01 -4.79826927e-01 5.28322041e-01 2.35559210e-01
2.92181641e-01 6.87772483e-02 -2.20215991e-01 -9.07432973e-01
-1.00277774e-01 -8.88400555e-01 4.08573240e-01 -1.31897068e+00
-1.14766943e+00 6.20797515e-01 -1.21281542e-01 -1.38918936e+00
-8.82294834e-01 -3.67022790e-02 -5.09461880e-01 7.18595922e-01
-1.24245930e+00 -1.52237833e+00 1.43180471e-02 3.17566961e-01
6.46059752e-01 -2.47433588e-01 1.26673675e+00 5.76000154e-01
-9.08216536e-02 4.98994827e-01 -3.47453743e-01 5.51457927e-02
9.34154510e-01 -1.25775576e+00 1.24869145e-01 6.83410764e-02
8.11941206e-01 4.07862633e-01 8.04456115e-01 -1.31085679e-01
-1.07579792e+00 -1.00182724e+00 1.24063730e+00 -6.02669358e-01
6.61765158e-01 -4.82381135e-01 -6.63419724e-01 3.88041377e-01
4.57770526e-01 -7.07514584e-01 1.24651027e+00 2.60688365e-01
-6.07453704e-01 -5.65027781e-02 -3.34508896e-01 3.77746791e-01
9.73332822e-01 -1.19047928e+00 -9.42247152e-01 3.32662798e-02
8.83508861e-01 -1.36828929e-01 -9.77949142e-01 3.82524580e-01
8.72789562e-01 -4.93687153e-01 1.11374533e+00 -6.11270785e-01
6.15425289e-01 -4.05197263e-01 -2.24489227e-01 -1.05899000e+00
-3.05312753e-01 -1.02039635e+00 -2.53164023e-03 1.56666815e+00
5.06367087e-01 4.51641291e-01 4.42205191e-01 -8.87848437e-02
-4.30174530e-01 -2.83744872e-01 -8.38903785e-01 -4.97988313e-01
-2.10057840e-01 -8.52144897e-01 5.25402009e-01 8.16495478e-01
4.24182743e-01 9.55202520e-01 -5.42282701e-01 -2.81210244e-01
1.63390428e-01 2.24691421e-01 6.75525546e-01 -1.35073638e+00
-6.91165805e-01 -7.65509486e-01 -5.76856971e-01 -6.06226742e-01
3.97211790e-01 -1.41403079e+00 1.72273770e-01 -1.58099997e+00
3.42625350e-01 -2.29186080e-02 -4.87341493e-01 8.16977322e-01
4.47629720e-01 8.91255319e-01 3.99940193e-01 4.42926645e-01
-1.04924536e+00 6.41423047e-01 6.88490152e-01 -2.10256949e-01
-4.31680977e-01 -2.44010627e-01 -9.14075255e-01 3.89633834e-01
7.16557443e-01 -4.90186751e-01 -2.59564728e-01 -2.55761147e-01
8.02488208e-01 3.32019150e-01 5.25058150e-01 -1.35461259e+00
1.78970337e-01 4.51290637e-01 1.22663565e-01 -3.34762096e-01
8.47967625e-01 -4.77345765e-01 4.61903781e-01 6.79954365e-02
-1.07496130e+00 -1.31206006e-01 4.96921092e-01 3.96096081e-01
-5.26353121e-01 -1.76866099e-01 2.31157467e-01 3.78205664e-02
-2.35510111e-01 -2.19703466e-01 -5.29137194e-01 3.40044498e-02
2.35174879e-01 3.72547321e-02 5.78668118e-02 -8.52670491e-01
-1.22067332e+00 -2.53128648e-01 3.95504944e-02 6.41054392e-01
3.80777150e-01 -1.80931652e+00 -1.05118859e+00 -9.21504945e-02
4.38823253e-01 -5.59534252e-01 -3.00026834e-02 6.47124350e-01
1.10689932e-02 7.38529086e-01 -7.88432136e-02 -5.79363346e-01
-1.22104573e+00 2.90314198e-01 2.23965615e-01 -6.06385469e-02
-6.25042617e-01 5.44378936e-01 -1.17939666e-01 -1.24060281e-01
5.87270796e-01 -1.49320468e-01 -3.72144073e-01 3.46814483e-01
5.82801282e-01 -6.68490529e-02 -4.89169173e-02 -9.03371394e-01
-1.50373533e-01 5.85649431e-01 2.78364033e-01 -8.71240079e-01
1.28820312e+00 -7.61120617e-02 7.19357729e-02 1.02925897e+00
1.16907191e+00 -7.01900572e-02 -7.13131428e-01 3.34307109e-03
2.29556695e-01 2.19227657e-01 2.04585612e-01 -1.22591650e+00
-5.31309307e-01 8.74001801e-01 5.65842152e-01 2.77396977e-01
1.14444745e+00 3.14764798e-01 1.09596765e+00 4.14834231e-01
-1.08542539e-01 -9.96012628e-01 2.52637088e-01 5.80274701e-01
1.19296885e+00 -7.89509952e-01 -5.06286025e-01 2.32401252e-01
-6.41306102e-01 1.12389219e+00 -1.46078467e-02 -2.88677029e-02
3.66953909e-01 4.68450151e-02 1.20924897e-01 -1.81901008e-01
-9.77007389e-01 -6.23563766e-01 8.11569333e-01 3.09726954e-01
9.22475517e-01 -3.02492231e-02 8.18018466e-02 1.27176118e+00
-8.73336136e-01 9.52378064e-02 2.15933755e-01 6.25128210e-01
-2.68954009e-01 -1.29537368e+00 -3.77736330e-01 9.78869051e-02
-7.20091224e-01 -3.21629077e-01 -8.86657119e-01 2.56518155e-01
1.99367344e-01 9.68948483e-01 1.54151246e-01 -5.03239691e-01
2.85031408e-01 6.88354075e-01 5.16530097e-01 -7.93571949e-01
-1.18037999e+00 4.99188930e-01 2.60929376e-01 -4.46368635e-01
-6.95637822e-01 -5.76540649e-01 -1.14631355e+00 3.65275115e-01
-5.68753146e-02 3.93255204e-01 9.33627248e-01 8.32826495e-01
5.39691925e-01 9.77756262e-01 3.15143049e-01 -1.30631316e+00
-2.44260877e-01 -1.16452980e+00 -4.42130923e-01 5.78125596e-01
3.96666795e-01 -3.35905433e-01 -3.09631467e-01 3.31583023e-01] | [15.627087593078613, 5.170064926147461] |
ae7d114c-74d0-4471-8e95-7cfcdb0d3d14 | does-multimodality-help-human-and-machine-for | 1605.09186 | null | http://arxiv.org/abs/1605.09186v4 | http://arxiv.org/pdf/1605.09186v4.pdf | Does Multimodality Help Human and Machine for Translation and Image Captioning? | This paper presents the systems developed by LIUM and CVC for the WMT16
Multimodal Machine Translation challenge. We explored various comparative
methods, namely phrase-based systems and attentional recurrent neural networks
models trained using monomodal or multimodal data. We also performed a human
evaluation in order to estimate the usefulness of multimodal data for human
machine translation and image description generation. Our systems obtained the
best results for both tasks according to the automatic evaluation metrics BLEU
and METEOR. | ['Joost Van de Weijer', 'Loïc Barrault', 'Mercedes García-Martínez', 'Marc Masana', 'Walid Aransa', 'Ozan Caglayan', 'Yaxing Wang', 'Fethi Bougares'] | 2016-05-30 | does-multimodality-help-human-and-machine-for-1 | https://aclanthology.org/W16-2358 | https://aclanthology.org/W16-2358.pdf | ws-2016-8 | ['multimodal-machine-translation'] | ['natural-language-processing'] | [ 3.17872882e-01 -1.00051969e-01 -3.72556061e-01 -5.06184064e-02
-1.36640942e+00 -4.87661123e-01 1.35865271e+00 -9.21796039e-02
-7.25441217e-01 9.13771570e-01 3.46138567e-01 -5.13445973e-01
2.88724065e-01 -1.61760580e-02 -4.99145836e-01 -4.32491273e-01
5.05144119e-01 1.03802061e+00 -1.71094805e-01 -4.03364658e-01
4.91613984e-01 1.13617927e-01 -1.14222121e+00 7.56714225e-01
7.65052736e-01 7.14972079e-01 3.05873960e-01 7.99188197e-01
-1.59793377e-01 4.65470433e-01 -5.22858381e-01 -8.39503467e-01
-1.90874740e-01 -6.48960948e-01 -9.54542696e-01 -2.16573328e-01
3.70202065e-01 1.46275312e-01 -2.69688547e-01 8.78260970e-01
9.59590614e-01 1.06161505e-01 8.76191497e-01 -1.16244233e+00
-1.11011684e+00 7.25499451e-01 -3.16655159e-01 2.78716177e-01
7.34925151e-01 7.16654863e-03 6.61036134e-01 -1.51656282e+00
1.01134264e+00 1.37480712e+00 3.99862677e-02 8.86477232e-01
-1.11765754e+00 -3.91349167e-01 -5.88445365e-01 5.01093984e-01
-1.29353321e+00 -8.24107945e-01 3.37847918e-01 -1.96257912e-04
1.31589854e+00 4.29066002e-01 1.17235318e-01 1.64272666e+00
3.44121456e-01 8.69064391e-01 1.27856040e+00 -9.67496216e-01
-1.89924404e-01 4.44448948e-01 -1.81017891e-01 4.29924250e-01
-3.06721717e-01 8.89474228e-02 -7.62742400e-01 -1.46736160e-01
6.04923368e-01 -5.33822954e-01 -1.80226400e-01 2.95815822e-02
-2.13691854e+00 9.56812620e-01 1.54068649e-01 5.16447127e-01
-5.15778601e-01 2.35425867e-02 6.41852677e-01 5.01577795e-01
4.11769181e-01 7.03414500e-01 -1.78085119e-01 -2.61892498e-01
-7.85378158e-01 -9.67036560e-02 5.16530573e-01 1.13155472e+00
9.08314884e-02 -2.64154136e-01 -5.64041197e-01 1.08575583e+00
4.09040809e-01 1.03395593e+00 8.50098670e-01 -9.12626803e-01
1.06922531e+00 1.50896445e-01 2.79578149e-01 -5.48320293e-01
-3.11093628e-01 3.03437840e-02 -6.35641038e-01 -4.30486858e-01
1.09245880e-02 -1.36760816e-01 -1.29193771e+00 1.26893079e+00
-4.93640214e-01 -7.64811933e-01 4.55883592e-01 1.10658193e+00
1.22860646e+00 1.05381846e+00 3.55050772e-01 -3.33333373e-01
1.21951067e+00 -1.36023426e+00 -1.19026268e+00 1.27710342e-01
6.07005358e-01 -1.44018173e+00 8.87422442e-01 -9.22172740e-02
-1.45075905e+00 -4.75631505e-01 -6.80778027e-01 -5.24100177e-02
-7.42175281e-01 5.28825700e-01 1.68037295e-01 5.23166835e-01
-1.45265067e+00 1.22616820e-01 -4.03711796e-01 -9.28755581e-01
-1.00545928e-01 5.04674852e-01 -4.38873380e-01 9.50747579e-02
-1.46805668e+00 1.65025985e+00 3.40655923e-01 2.31516823e-01
-7.78914571e-01 2.92780101e-01 -8.15232038e-01 -3.28472167e-01
-2.90710479e-01 -8.06059778e-01 1.28710985e+00 -1.01227450e+00
-1.58917630e+00 1.11245322e+00 -5.04878521e-01 -6.64738297e-01
1.59967050e-01 9.29779261e-02 -5.68162024e-01 4.05318558e-01
-1.11373305e-01 1.49427092e+00 7.64140308e-01 -1.09140444e+00
-5.09971797e-01 8.46033618e-02 -3.50101650e-01 7.13968039e-01
-1.65835574e-01 5.98813236e-01 -3.33540857e-01 -5.13152480e-01
-1.67105913e-01 -1.24962187e+00 -8.40038620e-03 -8.43336761e-01
-2.52043128e-01 -4.59095359e-01 6.34649754e-01 -9.47046936e-01
9.44022954e-01 -1.78835976e+00 7.78258920e-01 2.53883377e-02
-3.21494520e-01 1.34330168e-01 -7.43176043e-01 6.22210741e-01
1.86166298e-02 3.24984848e-01 1.16591096e-01 -6.24223948e-01
7.16854706e-02 1.04626901e-01 -4.99951601e-01 -1.79731116e-01
2.27427080e-01 1.47350979e+00 -5.52640676e-01 -6.99527204e-01
2.14088529e-01 5.90297639e-01 4.67714220e-01 2.61715174e-01
-1.50221244e-01 3.93432200e-01 -1.08505249e-01 8.60463977e-01
1.42629789e-02 -1.59932077e-02 -5.27935140e-02 -8.48765224e-02
-2.61350553e-02 1.70732990e-01 2.38040853e-02 2.00474763e+00
-4.84246463e-01 1.07735872e+00 -4.23746079e-01 -1.86654344e-01
7.50802636e-01 1.07688653e+00 -1.31334960e-01 -1.10492587e+00
3.46020311e-01 4.95735675e-01 -1.23570286e-01 -5.64266264e-01
1.07594252e+00 -1.80161018e-02 -8.74042287e-02 4.73435998e-01
3.11636597e-01 5.87710319e-03 2.77326941e-01 1.18116051e-01
5.15608609e-01 3.18172514e-01 3.97662725e-03 -1.06622741e-01
4.63360280e-01 3.41535926e-01 -2.86227167e-01 7.20932901e-01
-1.20617971e-01 9.12512839e-01 1.15902610e-01 -2.30290011e-01
-1.26698101e+00 -9.08727169e-01 1.86323479e-01 1.35427690e+00
1.62890971e-01 -1.67232454e-01 -7.26412475e-01 -6.42663121e-01
-8.81435335e-01 9.66317892e-01 -6.12789392e-01 -5.49233444e-02
-4.11628395e-01 -7.69540191e-01 5.82338750e-01 3.69351596e-01
1.56828344e-01 -1.55551612e+00 -3.94379556e-01 -6.05852790e-02
-9.32739437e-01 -1.14060903e+00 -6.08515620e-01 -1.67648792e-02
-9.21701550e-01 -2.78221309e-01 -1.39498508e+00 -1.04476237e+00
5.57846665e-01 1.53837591e-01 1.20587969e+00 -3.37256372e-01
4.50684950e-02 3.35503966e-01 -6.72912538e-01 -2.37339437e-01
-6.02002382e-01 5.56786895e-01 -1.02635540e-01 -4.25816119e-01
4.37019110e-01 5.43674491e-02 -2.19192192e-01 5.04056454e-01
-7.62454152e-01 2.60649323e-01 9.94423568e-01 8.88077199e-01
2.58924097e-01 -8.43970895e-01 3.61536860e-01 -1.31526649e-01
1.13142145e+00 -2.44788185e-01 -1.73809916e-01 7.17393696e-01
-4.92685497e-01 2.06554532e-01 -1.13043532e-01 -6.88240111e-01
-9.31738496e-01 -1.56639963e-01 9.88058373e-02 -4.59156573e-01
-7.33272657e-02 6.36040986e-01 2.42242754e-01 -1.42032966e-01
6.63443208e-01 3.80430639e-01 -9.25240368e-02 -2.14931816e-01
5.00415683e-01 8.69279087e-01 4.19627964e-01 -3.92580748e-01
4.32566524e-01 -1.88774988e-01 -2.26789802e-01 -5.12132406e-01
-4.17089947e-02 -2.81916142e-01 -6.26151383e-01 -3.64979982e-01
1.22725666e+00 -7.17968285e-01 -3.11578572e-01 9.75084230e-02
-1.64767861e+00 -1.13115031e-02 2.38501683e-01 6.63129032e-01
-6.82909131e-01 2.86179874e-02 -8.06323528e-01 -7.87900269e-01
-7.47807562e-01 -1.38692153e+00 1.33359301e+00 1.22864015e-01
-4.89914507e-01 -1.06999040e+00 2.67581671e-01 6.97486818e-01
6.98897243e-01 -1.63058072e-01 1.00363433e+00 -6.65485978e-01
-2.87370592e-01 -2.06067771e-01 -3.41824800e-01 -3.17377418e-01
-3.18870008e-01 -2.26231769e-01 -6.87806308e-01 -1.23167694e-01
-5.20219803e-01 -7.77907431e-01 8.93503308e-01 1.76805601e-01
2.63201505e-01 -8.67296942e-03 -2.08714634e-01 -1.86081260e-01
1.01842391e+00 3.21353137e-01 9.30533111e-01 3.63082469e-01
4.96222585e-01 7.97886729e-01 5.32259881e-01 -2.77723551e-01
2.64091581e-01 9.80779827e-01 1.18894249e-01 -4.85913381e-02
-2.19924599e-01 -2.06681550e-01 6.02573574e-01 1.01581168e+00
-3.81715924e-01 -7.32547462e-01 -1.03842664e+00 5.13420463e-01
-2.07376242e+00 -8.21507335e-01 -3.43825780e-02 1.90282011e+00
6.45533383e-01 -2.17574075e-01 1.75935298e-01 -5.00873625e-01
1.04707825e+00 -1.61208287e-01 2.52677858e-01 -9.29117620e-01
-5.65179825e-01 1.41256765e-01 3.75023872e-01 3.94020259e-01
-8.42449248e-01 1.20083618e+00 7.84082508e+00 9.94935215e-01
-9.31644320e-01 4.80740219e-01 5.34894049e-01 -8.39406252e-02
-1.90815955e-01 -2.72790074e-01 -5.53024530e-01 1.80216923e-01
1.75110340e+00 1.56076640e-01 5.79009533e-01 1.50921151e-01
-2.68158652e-02 -1.57004073e-01 -9.52311754e-01 1.10391831e+00
6.06491923e-01 -1.22754920e+00 5.90122342e-01 1.50653884e-01
7.66729295e-01 4.03844684e-01 5.15961409e-01 4.58293617e-01
-9.00264457e-02 -1.23016262e+00 5.81778467e-01 8.05596948e-01
7.53994524e-01 -6.47512138e-01 1.16743457e+00 1.25694871e-01
-4.32498723e-01 4.88446027e-01 -2.97839552e-01 3.15792382e-01
1.61239862e-01 -4.45537031e-01 -9.33392584e-01 6.27256572e-01
2.29549199e-01 3.90140653e-01 -7.92886198e-01 7.75706291e-01
-1.73875704e-01 2.51946568e-01 5.66491224e-02 -4.52381134e-01
4.98661339e-01 -1.44508407e-02 6.53320730e-01 1.39828086e+00
4.73592281e-01 -5.59078753e-01 -2.90774643e-01 5.41572392e-01
-2.70903677e-01 4.58254993e-01 -6.86457396e-01 -5.23613334e-01
8.82757455e-02 1.27939379e+00 -5.56619525e-01 -4.22437370e-01
-3.07485998e-01 1.22788167e+00 -2.21479740e-02 6.80932343e-01
-8.11055660e-01 -2.33866274e-01 -1.90080225e-01 -2.98994601e-01
4.92206663e-02 -2.35740945e-01 -2.05896586e-01 -1.30432487e+00
-1.87514156e-01 -1.00926113e+00 2.66568542e-01 -1.43300092e+00
-1.11475670e+00 1.22918212e+00 1.52439520e-01 -1.35815144e+00
-7.88925529e-01 -5.99895239e-01 -2.05028698e-01 1.11747384e+00
-1.05779386e+00 -1.44462574e+00 4.05063093e-01 3.69002789e-01
7.57611692e-01 -7.80870974e-01 1.26787639e+00 1.75638244e-01
-3.43709707e-01 6.06995881e-01 -6.12503150e-03 -2.36975309e-02
1.10643446e+00 -8.65103960e-01 2.88741887e-01 4.96165991e-01
3.73033911e-01 6.12299681e-01 7.69444466e-01 -4.40708727e-01
-1.43734670e+00 -7.18658090e-01 1.63986897e+00 -8.57387662e-01
5.53363979e-01 -4.17156890e-02 -4.81737494e-01 5.05146623e-01
1.42036009e+00 -8.15698266e-01 5.54966807e-01 5.00133783e-02
-2.54868954e-01 4.68866915e-01 -1.01512063e+00 7.65685022e-01
5.17747462e-01 -7.21760929e-01 -8.49404037e-01 5.02866685e-01
7.43226647e-01 -2.93285578e-01 -8.20445538e-01 5.01580536e-01
4.73379970e-01 -2.15053901e-01 6.40055060e-01 -5.77773571e-01
7.77222514e-01 -1.64005190e-01 -3.47426832e-01 -1.33165443e+00
-6.51155412e-02 -5.99008322e-01 3.50266248e-02 9.29115295e-01
1.32475579e+00 -3.94275635e-01 1.74050212e-01 5.38458288e-01
5.49101718e-02 -3.64980221e-01 -1.11030579e+00 -4.27759111e-01
9.26318392e-02 -5.01304604e-02 1.42223477e-01 7.59659886e-01
2.59852141e-01 1.03875268e+00 -6.09338045e-01 -4.66537148e-01
3.00161272e-01 1.68121718e-02 3.16835582e-01 -3.41134131e-01
2.46380568e-01 -7.98938155e-01 -4.22530562e-01 -5.07487655e-01
2.88125068e-01 -1.02036262e+00 9.80153456e-02 -1.60478687e+00
7.17349470e-01 5.26533663e-01 -2.87876129e-01 4.48484033e-01
1.16016418e-01 9.68075275e-01 2.72186667e-01 6.21423602e-01
-9.90296423e-01 4.76148367e-01 1.21780431e+00 -3.63191307e-01
2.85398960e-02 -2.97306627e-01 -2.58466184e-01 1.76351350e-02
7.43879974e-01 -2.25662783e-01 -7.71220401e-02 -8.43098938e-01
3.76247615e-01 3.58106226e-01 3.69454443e-01 -6.33562863e-01
2.12240279e-01 1.31656393e-01 5.20870924e-01 -8.61277401e-01
6.26016736e-01 -6.37296796e-01 -5.34214266e-02 2.87438393e-01
-8.11199546e-01 7.48492718e-01 2.55340755e-01 7.99162239e-02
-5.52932680e-01 -2.60277271e-01 4.09384727e-01 -5.89469895e-02
-5.31462848e-01 -2.18944699e-01 -7.07047343e-01 -4.79191452e-01
5.46257079e-01 1.43365994e-01 -6.10249460e-01 -7.96208024e-01
-9.48178530e-01 3.09104025e-01 3.82569194e-01 9.19278800e-01
9.28163588e-01 -1.78004372e+00 -9.69633520e-01 -3.66874129e-01
3.98747176e-01 -1.19060218e+00 -2.27205977e-01 1.29856277e+00
-3.32696885e-01 1.13042915e+00 -5.80919027e-01 -6.28276527e-01
-1.35835648e+00 4.91129875e-01 8.78087729e-02 -2.07783401e-01
1.31646097e-01 3.25193316e-01 -3.93513829e-01 -4.76384550e-01
1.33020446e-01 2.75324047e-01 -5.17080247e-01 7.60541409e-02
4.41609710e-01 3.03449661e-01 8.92822295e-02 -1.03356826e+00
-3.15601885e-01 4.19183642e-01 -1.73581570e-01 -1.15306187e+00
6.53451979e-01 -4.16994840e-01 -3.85988265e-01 7.35578120e-01
1.13555682e+00 -5.20589352e-01 -1.12758711e-01 -2.69500047e-01
1.77993685e-01 -8.11089352e-02 -1.35382311e-02 -1.41813445e+00
-4.03422296e-01 9.81107950e-01 8.80995154e-01 6.26788214e-02
1.09536183e+00 1.26254186e-01 8.86288881e-01 6.21486604e-01
3.94143820e-01 -1.10578954e+00 -1.64628029e-01 8.93578291e-01
1.15714622e+00 -1.48780918e+00 -4.71428663e-01 2.62543738e-01
-1.11879086e+00 1.33550704e+00 2.25635573e-01 4.14667606e-01
-2.08369002e-01 -3.80754948e-01 5.82779169e-01 7.68149421e-02
-1.12085474e+00 -2.15013355e-01 9.89406407e-01 4.63022530e-01
8.84568095e-01 1.18167229e-01 -7.49777377e-01 -7.12083951e-02
7.32877627e-02 -2.13641748e-01 1.70636445e-01 7.11825371e-01
-3.29087824e-01 -1.17493582e+00 -5.15270114e-01 -7.64079317e-02
-2.91527420e-01 -3.93462241e-01 -9.96053815e-01 7.05986381e-01
-4.00708407e-01 1.04046071e+00 -1.68916062e-01 -5.79206884e-01
3.71075459e-02 5.82777858e-01 8.12882483e-01 -4.41853255e-01
-7.15418160e-01 4.18170035e-01 3.96799237e-01 -2.37475693e-01
-8.62928092e-01 -7.76842535e-01 -7.05610812e-01 -1.92416627e-02
-3.38642478e-01 4.10200059e-01 1.20544672e+00 9.09654617e-01
2.78747737e-01 7.17568770e-02 4.34006304e-01 -9.13525641e-01
-1.12919882e-01 -1.73819005e+00 3.18497062e-01 1.45244583e-01
-5.48943691e-02 -1.67502224e-01 -1.70066521e-01 2.33488098e-01] | [11.513670921325684, 1.5288325548171997] |
102af934-8458-4433-9bb7-740713886f20 | align-rudder-learning-from-few-demonstrations | 2009.14108 | null | https://arxiv.org/abs/2009.14108v2 | https://arxiv.org/pdf/2009.14108v2.pdf | Align-RUDDER: Learning From Few Demonstrations by Reward Redistribution | Reinforcement learning algorithms require many samples when solving complex hierarchical tasks with sparse and delayed rewards. For such complex tasks, the recently proposed RUDDER uses reward redistribution to leverage steps in the Q-function that are associated with accomplishing sub-tasks. However, often only few episodes with high rewards are available as demonstrations since current exploration strategies cannot discover them in reasonable time. In this work, we introduce Align-RUDDER, which utilizes a profile model for reward redistribution that is obtained from multiple sequence alignment of demonstrations. Consequently, Align-RUDDER employs reward redistribution effectively and, thereby, drastically improves learning on few demonstrations. Align-RUDDER outperforms competitors on complex artificial tasks with delayed rewards and few demonstrations. On the Minecraft ObtainDiamond task, Align-RUDDER is able to mine a diamond, though not frequently. Code is available at https://github.com/ml-jku/align-rudder. YouTube: https://youtu.be/HO-_8ZUl-UY | ['Jose A. Arjona-Medina', 'Johannes Brandstetter', 'Patrick M. Blies', 'Matthias Dorfer', 'Marius-Constantin Dinu', 'Sepp Hochreiter', 'Markus Hofmarcher', 'Vihang P. Patil'] | 2020-09-29 | align-rudder-learning-from-few-demonstrations-1 | https://openreview.net/forum?id=8bZC3CyF-f7 | https://openreview.net/pdf?id=8bZC3CyF-f7 | null | ['multiple-sequence-alignment', 'safe-exploration'] | ['medical', 'robots'] | [-2.63792872e-01 3.78595898e-03 -3.63152981e-01 -1.78904384e-01
-9.59919095e-01 -6.04187965e-01 2.40128115e-01 8.45657215e-02
-6.13541067e-01 1.34060669e+00 -7.07364902e-02 -9.02110860e-02
-1.60089403e-01 -3.59398454e-01 -7.22187936e-01 -5.18950939e-01
-3.43765378e-01 4.93084788e-01 6.10918412e-03 -1.63172781e-01
3.37174863e-01 -4.19540480e-02 -1.50625217e+00 -1.52609676e-01
1.02263176e+00 5.78786612e-01 8.72057855e-01 8.27523232e-01
3.75938773e-01 7.83147216e-01 -6.11333907e-01 7.86226541e-02
5.15825629e-01 -6.20361090e-01 -7.09228098e-01 -3.24958026e-01
5.76621965e-02 -7.83919454e-01 -3.76473963e-01 9.82893169e-01
5.49480259e-01 6.89294815e-01 4.58618701e-01 -1.58585787e+00
-1.49307117e-01 8.26846063e-01 -6.48085356e-01 4.59139913e-01
4.21253532e-01 7.21849620e-01 1.15403903e+00 -6.30786657e-01
6.69226885e-01 1.19127285e+00 2.29781389e-01 6.76716030e-01
-1.19560599e+00 -9.04985785e-01 2.43716180e-01 3.76879752e-01
-8.60131800e-01 -2.89091855e-01 5.84764242e-01 -3.31301898e-01
9.86193478e-01 -6.34750649e-02 9.16576147e-01 1.31704426e+00
-1.76251262e-01 1.48963368e+00 1.15433848e+00 1.08910166e-01
4.77898061e-01 -4.01262671e-01 5.60371904e-03 7.32624531e-01
-1.93140302e-02 4.92761195e-01 -7.54740894e-01 -1.89943891e-02
7.88944006e-01 2.07357243e-01 -2.86976725e-01 -5.44734359e-01
-1.36651635e+00 9.39629257e-01 4.15947020e-01 -1.03265889e-01
-6.39194965e-01 5.65103292e-01 4.44651723e-01 7.31274962e-01
-4.07547466e-02 8.74651492e-01 -3.94305527e-01 -9.11142647e-01
-6.97929382e-01 7.43894041e-01 6.39698565e-01 1.12557459e+00
8.22911501e-01 2.85835654e-01 -2.23853797e-01 6.21336758e-01
-2.20451191e-01 2.85629690e-01 5.19662201e-01 -1.47658241e+00
6.28558278e-01 1.93753779e-01 5.08433819e-01 -1.38853177e-01
-4.36567426e-01 -2.37218648e-01 -2.44003758e-01 6.75901890e-01
5.82316339e-01 -6.25618696e-01 -6.47412539e-01 1.73652005e+00
6.33958220e-01 3.55717838e-02 8.81045163e-02 1.18626213e+00
3.57096314e-01 4.85937059e-01 -4.36979420e-02 -3.93729433e-02
8.21827233e-01 -1.45211160e+00 -4.60141480e-01 -2.53390878e-01
6.27902865e-01 -3.76565963e-01 1.46015859e+00 6.35705411e-01
-1.29490423e+00 -5.64128935e-01 -8.70888114e-01 5.87761328e-02
1.38080969e-01 1.11209311e-01 5.94276905e-01 -8.49900953e-03
-6.47933900e-01 1.19527662e+00 -1.12714159e+00 -7.50202835e-02
5.43578386e-01 3.16123605e-01 -1.28047243e-01 -6.75001219e-02
-9.75675285e-01 8.38034272e-01 3.13960850e-01 -1.34807125e-01
-1.72745883e+00 -5.77055156e-01 -9.00660694e-01 7.81458989e-02
7.95070469e-01 -5.06239414e-01 1.82276058e+00 -5.29993534e-01
-1.83594942e+00 1.26661941e-01 1.27256542e-01 -6.90987289e-01
7.37675011e-01 -6.83130860e-01 3.85610551e-01 1.80290163e-01
2.60973006e-01 9.27318513e-01 1.03757310e+00 -8.73660505e-01
-8.48961532e-01 -6.83342740e-02 2.93113887e-01 7.24696696e-01
5.27086444e-02 -3.68956387e-01 2.69493219e-02 -3.52029175e-01
-4.78433937e-01 -1.09886670e+00 -5.28202116e-01 -1.59045607e-01
-3.67432863e-01 -4.77048069e-01 4.24497932e-01 -2.42733538e-01
8.00704837e-01 -2.01338601e+00 5.48404336e-01 -1.83273494e-01
2.93649465e-01 9.75819901e-02 -4.54410523e-01 5.94088197e-01
1.90234199e-01 -3.13296139e-01 -1.13386385e-01 -3.72425556e-01
3.08705837e-01 1.85613289e-01 -1.79060325e-01 3.90190899e-01
2.81067155e-02 9.31873500e-01 -1.35893416e+00 -1.31530121e-01
2.21321642e-01 -5.60797416e-02 -6.62028670e-01 5.64185321e-01
-5.75189710e-01 7.83687532e-01 -7.26802707e-01 6.77057028e-01
1.73642308e-01 -1.51672512e-01 1.34025007e-01 5.51826894e-01
-1.30540788e-01 4.84747410e-01 -9.84464109e-01 2.03611565e+00
-3.50673795e-01 5.17387271e-01 1.50978640e-01 -1.08364058e+00
7.32784510e-01 3.23291197e-02 4.07023698e-01 -7.28861272e-01
9.82296169e-02 2.24105299e-01 1.55680522e-01 -3.76208961e-01
6.31795943e-01 2.69714147e-01 -3.92922685e-02 6.99409485e-01
1.05752081e-01 -3.78506541e-01 7.44121313e-01 3.00150871e-01
1.49077177e+00 7.63168454e-01 3.43143791e-01 1.37722954e-01
-8.76459554e-02 1.96547464e-01 6.42836571e-01 9.16941285e-01
-5.23266256e-01 8.53681043e-02 6.63523495e-01 -1.81714848e-01
-9.90730643e-01 -9.95657682e-01 3.67893875e-01 1.28009236e+00
1.42805293e-01 -5.68996012e-01 -5.87232053e-01 -7.46292830e-01
3.08717102e-01 8.89568567e-01 -4.10138994e-01 1.32813342e-02
-4.61845279e-01 -5.95652871e-03 3.02097380e-01 3.47730845e-01
1.12159483e-01 -1.64636600e+00 -1.46692383e+00 3.07545304e-01
1.87552515e-02 -6.85580790e-01 -4.98326600e-01 7.14983702e-01
-1.01102495e+00 -1.22283256e+00 -1.04116464e+00 -4.57356960e-01
5.84239185e-01 2.33083740e-01 9.55747306e-01 1.04086936e-01
-5.19577980e-01 3.98366481e-01 -4.94075269e-01 -2.88379878e-01
-9.33689773e-02 1.14993460e-01 2.60855198e-01 -9.50231314e-01
1.58059523e-01 -8.32483172e-01 -7.64640510e-01 3.23028058e-01
-3.93266350e-01 9.55324173e-02 6.59580529e-01 1.24571383e+00
4.16265696e-01 -2.69315422e-01 8.68269026e-01 -5.32624483e-01
9.03313637e-01 -5.65003514e-01 -1.00233614e+00 -3.04947317e-01
-6.48396313e-01 1.49122953e-01 7.21575141e-01 -8.06654751e-01
-6.37067437e-01 9.81542170e-02 3.45780849e-02 -9.19592857e-01
-1.26874581e-01 3.40468079e-01 5.08361220e-01 3.45350921e-01
7.59948313e-01 3.29334795e-01 1.68437794e-01 -4.39848989e-01
3.18378389e-01 2.19814271e-01 1.30814314e-01 -8.57233822e-01
7.02801883e-01 -3.75987627e-02 -2.54115969e-01 -3.76350522e-01
-6.30515814e-01 -4.18380499e-01 -1.72953993e-01 -2.34476000e-01
3.84636283e-01 -9.67463613e-01 -1.21365225e+00 2.69999295e-01
-6.26066804e-01 -1.33821630e+00 -7.39145279e-01 7.57040858e-01
-1.16799700e+00 2.53283203e-01 -8.01647007e-01 -9.38516974e-01
-2.03542292e-01 -1.26524341e+00 7.84221768e-01 4.71045852e-01
-4.30677354e-01 -3.94537091e-01 1.14515267e-01 2.98984110e-01
1.69743732e-01 9.95564982e-02 4.20271397e-01 -4.92275268e-01
-7.23585963e-01 2.96101958e-01 2.42766976e-01 1.73074245e-01
1.29517943e-01 -3.92830312e-01 -5.17439783e-01 -6.22768939e-01
-3.36550951e-01 -1.22487581e+00 7.12274313e-01 4.05224681e-01
1.00237858e+00 -3.04553598e-01 -2.43005157e-02 4.19613451e-01
8.47188950e-01 3.96531105e-01 1.83283657e-01 5.64799726e-01
3.43051165e-01 5.22953689e-01 1.44980860e+00 9.70435798e-01
2.52510637e-01 5.79222083e-01 8.18201423e-01 2.94024765e-01
1.38276055e-01 -5.30467093e-01 7.14301705e-01 4.12199497e-01
2.47425269e-02 1.37621894e-01 -5.12958586e-01 6.17231309e-01
-2.12904716e+00 -1.10903513e+00 2.99248844e-01 2.00209141e+00
1.20790517e+00 2.17590898e-01 5.03544271e-01 -2.44451508e-01
4.24377859e-01 1.71648785e-01 -1.36205840e+00 -2.74536073e-01
3.68545711e-01 1.34477586e-01 3.44877183e-01 6.22978866e-01
-7.00315416e-01 1.06396246e+00 5.19636440e+00 9.04960930e-01
-7.11319685e-01 -4.19814996e-02 2.38507524e-01 -8.13707173e-01
3.34498589e-03 1.56348884e-01 -8.13697815e-01 5.44031262e-01
7.46394813e-01 -4.58269000e-01 9.40953135e-01 1.30912423e+00
3.28291088e-01 -5.83447814e-01 -1.28692055e+00 9.25537109e-01
-5.18464088e-01 -9.67967808e-01 -5.83258986e-01 -5.01763225e-02
6.99186683e-01 2.94731975e-01 2.42946237e-01 7.63798773e-01
9.39944088e-01 -9.43912625e-01 6.67793930e-01 1.31838366e-01
5.80563545e-01 -7.70024538e-01 1.69632941e-01 7.16655493e-01
-8.50374162e-01 -3.13306183e-01 -5.13481557e-01 -3.08446079e-01
1.10542268e-01 2.14476734e-01 -1.11248338e+00 2.69284666e-01
6.52720809e-01 9.01045561e-01 5.51142246e-02 9.76714611e-01
-6.96840167e-01 4.12871957e-01 -3.10591221e-01 -5.20410478e-01
5.30608654e-01 -3.35273266e-01 6.85177267e-01 7.10453928e-01
4.31646913e-01 8.16368237e-02 4.72698808e-01 9.57916319e-01
1.56270191e-01 -1.29633233e-01 -5.99279761e-01 -2.25528970e-01
6.72902048e-01 1.35214972e+00 -2.87698895e-01 -3.01926553e-01
1.19336940e-01 1.00404084e+00 7.09637702e-01 3.58818591e-01
-8.84326339e-01 -3.24721515e-01 8.99854600e-01 -1.37904644e-01
4.97953892e-01 -4.96696651e-01 3.53738010e-01 -9.43877399e-01
-6.43534139e-02 -1.15604162e+00 3.50071937e-01 -8.85133624e-01
-9.04158771e-01 4.00010198e-01 -1.13542691e-01 -1.36455500e+00
-7.34647214e-01 -2.96343178e-01 -5.76101243e-01 6.83197856e-01
-1.60926402e+00 -4.64946002e-01 -1.44749746e-01 7.42123783e-01
1.13158047e+00 -1.64096132e-01 5.68077147e-01 -1.52038768e-01
-5.68774283e-01 3.89895797e-01 2.09512264e-01 -2.30359867e-01
7.72776783e-01 -1.51420522e+00 4.34041321e-01 3.84446472e-01
-4.46444526e-02 4.95633692e-01 8.83199394e-01 -7.17535436e-01
-1.33646894e+00 -5.61769068e-01 -3.45459394e-02 -1.18068799e-01
8.40554774e-01 -2.25052625e-01 -5.91525793e-01 6.35973632e-01
2.05116019e-01 -1.13164872e-01 4.30901408e-01 1.65160879e-01
-1.06642485e-01 2.72409290e-01 -9.98998404e-01 7.98780203e-01
9.83571589e-01 -2.98087627e-01 -5.70191443e-01 4.26503330e-01
6.20897353e-01 -7.74329185e-01 -7.60869026e-01 -2.02657841e-02
5.10744691e-01 -9.95836198e-01 7.88511813e-01 -7.70299315e-01
6.53652132e-01 -2.45449454e-01 1.86968476e-01 -1.84895849e+00
-1.15258142e-01 -1.04115570e+00 -4.75368321e-01 5.23673773e-01
3.16788644e-01 -6.38895452e-01 9.78022337e-01 2.91498303e-01
-1.68254241e-01 -1.02884102e+00 -8.46015930e-01 -1.16790056e+00
-9.52648968e-02 -4.10632789e-03 3.44656616e-01 6.42333150e-01
4.04957592e-01 2.54070818e-01 -7.04578996e-01 -2.64185488e-01
8.15401137e-01 3.74208033e-01 1.05744874e+00 -7.74930179e-01
-7.37285435e-01 -3.53949040e-01 2.75376350e-01 -1.45626259e+00
2.46395126e-01 -6.42311156e-01 2.54042298e-01 -1.32350385e+00
6.04282096e-02 -5.78556478e-01 -3.03700238e-01 6.70340836e-01
-2.84456462e-01 -4.27013278e-01 5.16231298e-01 2.96926469e-01
-9.21446204e-01 9.97730434e-01 1.61181760e+00 5.85800335e-02
-4.76479530e-01 2.39774972e-01 -4.67765749e-01 5.41496575e-01
1.29195154e+00 -6.43976092e-01 -5.80339074e-01 -2.13298798e-01
-1.10992603e-02 5.99657774e-01 1.71800554e-01 -8.81957173e-01
1.41664043e-01 -3.46310228e-01 3.10359359e-01 -5.53906977e-01
7.14983284e-01 -3.89768481e-01 -1.29789904e-01 8.46537054e-01
-7.80485332e-01 3.77292275e-01 4.04902026e-02 6.42861903e-01
3.96029800e-02 -5.55256844e-01 5.30842841e-01 -3.99065644e-01
-7.35451460e-01 2.94510305e-01 -4.70609516e-01 2.62177199e-01
1.14521539e+00 -6.23392127e-02 -4.35206264e-01 -5.41171789e-01
-7.83536553e-01 1.00469470e+00 4.74682927e-01 3.68480891e-01
7.23388374e-01 -1.02980423e+00 -4.83679742e-01 -1.57349661e-01
-1.47702739e-01 3.97446752e-01 4.40267980e-01 8.71906877e-01
-7.96630010e-02 1.24485396e-01 -6.98043108e-01 -3.93167406e-01
-1.38425398e+00 5.66669881e-01 7.27972165e-02 -5.14665782e-01
-7.63473809e-01 8.90107989e-01 1.43709816e-02 -4.26836520e-01
4.20643419e-01 -3.48019212e-01 -5.49530573e-02 4.02278919e-03
3.38195592e-01 5.64518988e-01 -5.89793384e-01 1.59567043e-01
-4.76344712e-02 3.09937093e-02 -2.15173453e-01 -3.74215066e-01
1.42261958e+00 1.57469492e-02 5.02957761e-01 4.58607465e-01
7.50755668e-01 -2.84589499e-01 -2.13949561e+00 -9.69825462e-02
-6.62149191e-02 -7.46444464e-01 -3.14535528e-01 -7.12926626e-01
-7.58331299e-01 8.10264349e-01 3.63219470e-01 -1.66072041e-01
8.54135573e-01 -2.83650607e-01 8.07259440e-01 9.46674049e-01
7.01755583e-01 -1.25949121e+00 6.57881916e-01 5.02636015e-01
8.79098058e-01 -1.24581718e+00 1.04740679e-01 2.92408913e-01
-1.08420038e+00 9.18794572e-01 9.22410071e-01 -3.07722211e-01
1.53223658e-02 2.07886800e-01 -2.20377624e-01 -1.06425472e-01
-1.11499631e+00 -3.63754123e-01 -3.61316562e-01 5.41851163e-01
-1.72107462e-02 2.06949025e-01 -2.42700875e-01 2.17476219e-01
-3.13549191e-01 2.21850779e-02 6.53889179e-01 1.27187085e+00
-6.89348578e-01 -1.12499106e+00 -8.39694291e-02 5.50044060e-01
-2.50998318e-01 -3.96894738e-02 -2.10294407e-02 6.58102870e-01
-4.74750757e-01 7.81277716e-01 -2.08710879e-01 -2.50661165e-01
1.26492411e-01 -2.84813672e-01 6.10631227e-01 -7.47286499e-01
-6.13603234e-01 6.93389922e-02 1.99704751e-01 -1.01663554e+00
-1.09051161e-01 -7.91089475e-01 -1.53389215e+00 -2.24386945e-01
1.50347333e-02 4.81254458e-01 5.60957789e-01 4.38279659e-01
2.33276471e-01 5.51889181e-01 7.55635798e-01 -1.19605517e+00
-1.27634013e+00 -9.77065623e-01 -7.51261353e-01 3.24973315e-02
6.29259467e-01 -9.50115740e-01 -3.88237208e-01 -3.78449053e-01] | [4.122997760772705, 1.6287682056427002] |
5aebde59-1208-4a54-b63c-f2545adf8f7d | dtw-at-quran-qa-2022-utilising-transfer | null | null | https://aclanthology.org/2022.osact-1.10 | https://aclanthology.org/2022.osact-1.10.pdf | DTW at Qur’an QA 2022: Utilising Transfer Learning with Transformers for Question Answering in a Low-resource Domain | The task of machine reading comprehension (MRC) is a useful benchmark to evaluate the natural language understanding of machines. It has gained popularity in the natural language processing (NLP) field mainly due to the large number of datasets released for many languages. However, the research in MRC has been understudied in several domains, including religious texts. The goal of the Qur’an QA 2022 shared task is to fill this gap by producing state-of-the-art question answering and reading comprehension research on Qur’an. This paper describes the DTW entry to the Quran QA 2022 shared task. Our methodology uses transfer learning to take advantage of available Arabic MRC data. We further improve the results using various ensemble learning strategies. Our approach provided a partial Reciprocal Rank (pRR) score of 0.49 on the test set, proving its strong performance on the task. | ['Ruslan Mitkov', 'Wajdi Zaghouani', 'Tharindu Ranasinghe', 'Damith Premasiri'] | null | null | null | null | osact-lrec-2022-6 | ['machine-reading-comprehension'] | ['natural-language-processing'] | [ 4.42231447e-01 3.96149099e-01 1.48915440e-01 -3.41275990e-01
-1.30532968e+00 -6.64958537e-01 7.34711289e-01 3.97485107e-01
-4.65021938e-01 9.62730587e-01 6.48382902e-01 -6.41685426e-01
-1.07811555e-01 -8.64701867e-01 -5.55956602e-01 -4.38878119e-01
1.02327749e-01 7.31609166e-01 2.45679095e-01 -9.17954266e-01
7.53518760e-01 -1.92783564e-01 -1.38898945e+00 7.92229533e-01
1.46725130e+00 8.53432715e-01 2.92232394e-01 7.53987968e-01
5.40409461e-02 1.43174338e+00 -7.18376577e-01 -5.57324469e-01
-2.33754054e-01 -9.42116022e-01 -1.64143157e+00 -6.82459712e-01
4.49269116e-01 -6.97941557e-02 9.91862938e-02 6.50475144e-01
5.10824323e-01 1.52800620e-01 7.00556874e-01 -8.25565934e-01
-9.43469584e-01 6.60766423e-01 -8.78571644e-02 3.01129073e-01
9.02673483e-01 -3.32527280e-01 1.49331474e+00 -8.74246359e-01
5.02193511e-01 1.07780612e+00 2.26734295e-01 4.54287052e-01
-6.46427572e-01 -2.56408691e-01 -6.03179097e-01 9.60036874e-01
-9.51651275e-01 -2.76348442e-01 5.86691439e-01 -2.63642550e-01
9.96239066e-01 2.03275725e-01 6.14162087e-02 8.44322562e-01
2.61420995e-01 9.37407792e-01 1.55837584e+00 -9.26771700e-01
2.16545820e-01 -2.80628979e-01 2.63122827e-01 5.60759842e-01
-3.80026937e-01 -3.95129651e-01 -8.10580015e-01 1.71108961e-01
1.35676950e-01 -8.64241719e-01 -3.48842919e-01 2.62462676e-01
-1.16803610e+00 1.24516630e+00 3.24808717e-01 5.31573594e-01
-2.81396836e-01 -2.17539132e-01 3.45475018e-01 9.67104256e-01
5.34271896e-01 8.29503179e-01 -8.83786917e-01 -5.60223281e-01
-6.50038660e-01 3.51736099e-01 1.14711070e+00 2.01440066e-01
4.25130576e-01 -5.52288115e-01 -1.99990928e-01 1.15010011e+00
2.11649105e-01 6.02548957e-01 5.07237494e-01 -8.96714449e-01
8.74489784e-01 6.64031029e-01 9.43037216e-03 -9.33628142e-01
-3.31214428e-01 -6.23789579e-02 -6.80652618e-01 -1.70418546e-01
8.63584399e-01 -1.62848964e-01 -5.64895988e-01 1.55535817e+00
3.35204750e-02 -4.43386883e-01 4.83178973e-01 6.84211075e-01
1.10411668e+00 9.97465372e-01 9.97973140e-03 -2.36241948e-02
1.58850408e+00 -9.87880290e-01 -6.76834702e-01 -1.39262468e-01
8.42336893e-01 -8.32994759e-01 1.11821401e+00 5.64004064e-01
-8.90278816e-01 -5.17108977e-01 -9.63412642e-01 -4.27988023e-01
-3.80150229e-01 5.73187135e-02 3.53169620e-01 7.10460901e-01
-8.45813692e-01 2.61927933e-01 -4.35733467e-01 -1.57590926e-01
2.96673685e-01 5.77894002e-02 -3.61013025e-01 -5.33672690e-01
-1.71027899e+00 1.54348314e+00 5.30426085e-01 -5.82928434e-02
-6.94095016e-01 -5.19946516e-01 -8.51666689e-01 -2.22457629e-02
4.70876008e-01 -3.75444889e-01 1.41498733e+00 -1.05411077e+00
-1.69355094e+00 1.07081270e+00 -8.03026184e-02 -6.39417708e-01
3.65726382e-01 -5.28429389e-01 -2.92930573e-01 2.92082012e-01
4.89874408e-02 3.68388504e-01 4.72756952e-01 -8.42648506e-01
-5.63046277e-01 -4.76601362e-01 2.70394951e-01 4.34462965e-01
1.31128013e-01 2.17878208e-01 2.53609478e-01 -3.92139703e-01
-1.49955735e-01 -6.01807594e-01 3.01969051e-01 -6.56847239e-01
1.30515248e-01 -9.12555277e-01 4.20200616e-01 -1.38739526e+00
1.21375167e+00 -1.69287145e+00 4.46393073e-01 -1.94862023e-01
-5.74510135e-02 2.93517798e-01 -3.38752478e-01 9.17067826e-01
1.01358719e-01 -1.28933892e-01 -5.74381173e-01 1.50367290e-01
-1.57817051e-01 2.14388996e-01 -4.40513879e-01 2.10671633e-01
3.19489628e-01 1.13741553e+00 -1.02649689e+00 -3.08215499e-01
-8.89344662e-02 -6.48652436e-03 -2.40333214e-01 3.62741292e-01
-5.09089351e-01 4.66273487e-01 -4.55650061e-01 5.09835422e-01
2.70124048e-01 -1.58768848e-01 -4.13158126e-02 2.30581030e-01
-3.18055823e-02 7.73051620e-01 -5.08145690e-01 1.78878951e+00
-4.95550871e-01 6.69980288e-01 -3.04998666e-01 -1.21450377e+00
1.02848327e+00 3.61039847e-01 1.58254787e-01 -1.24308991e+00
8.42307657e-02 4.63687360e-01 4.72981095e-01 -7.55311131e-01
6.05759025e-01 -2.81248927e-01 -2.51110911e-01 6.80487990e-01
3.53835188e-02 -4.54068333e-01 1.52239889e-01 2.29220346e-01
1.05682039e+00 2.81708151e-01 6.83038473e-01 -5.45261443e-01
9.73337293e-01 2.34081432e-01 -5.62961847e-02 5.78338563e-01
-1.10488296e-01 5.11451125e-01 4.87738073e-01 -1.58322573e-01
-8.51544380e-01 -8.04203808e-01 -2.32419476e-01 1.44694448e+00
-1.24518842e-01 -1.94990218e-01 -8.37568223e-01 -7.19334602e-01
-4.42227006e-01 1.16129577e+00 -7.62553215e-01 1.28138348e-01
-8.03208411e-01 -7.44306862e-01 7.92108893e-01 1.61495849e-01
7.43363082e-01 -1.45376956e+00 -7.92146146e-01 3.30335289e-01
-8.78166854e-01 -8.36924136e-01 1.61348388e-01 -1.69531807e-01
-5.88092804e-01 -1.44726086e+00 -8.31756949e-01 -9.51475441e-01
-2.92305965e-02 -2.02027723e-01 1.52625322e+00 1.34245932e-01
2.55352557e-01 3.32537532e-01 -1.16414869e+00 -6.20083869e-01
-9.67353761e-01 4.75361973e-01 -4.73787904e-01 -1.55047357e-01
5.37402689e-01 1.80176031e-02 -1.52255803e-01 2.69174248e-01
-8.62442195e-01 2.20085829e-01 3.81738186e-01 9.85092759e-01
-7.74940178e-02 -2.62721688e-01 1.26712215e+00 -6.62959814e-01
9.97907996e-01 -5.73120117e-01 -2.66206741e-01 5.88610530e-01
-2.83516049e-01 7.90778697e-02 4.22898144e-01 1.74296767e-01
-1.18739879e+00 -6.45414054e-01 -5.19159973e-01 8.66716683e-01
-1.20103791e-01 1.06524456e+00 -1.00105412e-01 2.58474946e-01
8.40658128e-01 2.29460016e-01 9.16462392e-02 -4.19831902e-01
3.63171041e-01 1.08069217e+00 4.55229640e-01 -6.97789133e-01
3.84919137e-01 -1.41222820e-01 -9.68070552e-02 -1.08442974e+00
-1.38613546e+00 -3.95361036e-01 -6.47867858e-01 -1.28019527e-01
1.02266860e+00 -7.98694611e-01 -5.82826674e-01 5.66248298e-01
-1.16757250e+00 -4.22268212e-01 8.47118571e-02 2.57634103e-01
-6.07294202e-01 4.18266892e-01 -5.06577075e-01 -7.40718246e-01
-5.12948275e-01 -8.66340458e-01 7.45985925e-01 1.38195366e-01
-3.53012085e-01 -1.06491709e+00 5.76194644e-01 1.17454028e+00
5.83070278e-01 1.93246156e-01 1.31897330e+00 -9.93101776e-01
-1.62993878e-01 1.91619486e-01 -1.14787482e-01 6.00198627e-01
9.49713811e-02 -5.02977133e-01 -9.25715744e-01 -1.02984481e-01
2.21414298e-01 -9.84803796e-01 8.47627699e-01 -8.06741491e-02
6.93661094e-01 -6.87039942e-02 5.40187776e-01 -4.90652621e-01
1.20254886e+00 9.26257893e-02 1.03325045e+00 6.01937771e-01
2.77632147e-01 9.95123208e-01 8.71103406e-01 3.26574184e-02
7.54090011e-01 5.09407043e-01 4.00014073e-01 3.36417884e-01
-1.09176919e-01 -2.04270348e-01 5.17982423e-01 1.18157566e+00
-3.11001956e-01 -3.07960898e-01 -1.50982904e+00 7.59514689e-01
-1.72511160e+00 -8.30567420e-01 -3.80135685e-01 1.88701081e+00
1.08025110e+00 -2.12067470e-01 -1.81946903e-01 3.20886612e-01
1.70424163e-01 8.19898844e-02 -1.88008733e-02 -6.31376207e-01
-3.61758590e-01 6.28439844e-01 -2.46431485e-01 7.17981637e-01
-8.72070372e-01 9.98241603e-01 5.69666815e+00 9.13047135e-01
-4.46408957e-01 1.24334864e-01 5.80404997e-01 8.23683023e-01
-2.37034529e-01 -1.14972308e-01 -3.07261288e-01 1.18380040e-01
1.13026166e+00 -4.17283811e-02 3.24134052e-01 2.71124363e-01
-1.23983119e-02 -6.30551696e-01 -1.06698203e+00 6.14697337e-01
6.50596321e-01 -1.13542247e+00 1.54266238e-01 -1.85762450e-01
7.27672815e-01 4.18573260e-01 -2.44622275e-01 6.50229394e-01
1.54619008e-01 -1.64483762e+00 3.31252426e-01 3.92196923e-01
5.73439121e-01 -7.24519610e-01 1.30610800e+00 8.05314124e-01
-4.53349382e-01 -2.84207962e-03 -2.68438905e-01 -6.14731669e-01
2.03027830e-01 2.20398396e-01 -9.49216485e-01 8.65476549e-01
4.80358899e-01 4.16710734e-01 -9.46609497e-01 6.34867966e-01
-6.56569481e-01 1.16730368e+00 -1.17145114e-01 -4.43206459e-01
3.30388039e-01 -3.02900136e-01 2.57873416e-01 9.77515638e-01
-2.80945301e-02 3.02889496e-01 -9.84827653e-02 2.92641461e-01
-2.85989493e-01 7.00193644e-01 -2.43278712e-01 3.14153470e-02
9.72435027e-02 7.82385588e-01 -2.07735404e-01 -2.77379155e-01
-3.33218753e-01 8.69570315e-01 4.41171050e-01 -9.81193110e-02
-4.98381734e-01 -4.29760754e-01 -1.30353868e-01 -3.07362646e-01
1.75065137e-02 -2.68202484e-01 -2.50257939e-01 -1.24092960e+00
-7.43520334e-02 -1.47091389e+00 5.79467416e-01 -7.81999648e-01
-1.49551749e+00 5.33565283e-01 7.02322870e-02 -6.82259977e-01
-3.56257170e-01 -8.35125029e-01 -3.32253069e-01 9.47659791e-01
-1.63728249e+00 -1.19099462e+00 -2.09560335e-01 4.38825071e-01
7.43463039e-01 -3.60468805e-01 1.23049819e+00 -1.14744026e-02
1.04967937e-01 2.97265381e-01 5.95777705e-02 3.36856633e-01
8.18985820e-01 -1.43893921e+00 2.16999441e-01 6.24709487e-01
3.21103781e-01 2.74320275e-01 6.45243585e-01 -3.09453160e-01
-1.23240209e+00 -4.06017572e-01 1.47811496e+00 -9.69751000e-01
8.28337967e-01 3.34636196e-02 -1.18510771e+00 5.29180408e-01
8.48310173e-01 -9.00724709e-01 1.04342377e+00 2.39754781e-01
-4.28123385e-01 2.40946710e-01 -1.03809309e+00 4.20357823e-01
3.60276639e-01 -6.21114850e-01 -1.64379764e+00 4.05795515e-01
4.83726203e-01 -2.76832670e-01 -1.12926805e+00 3.80384564e-01
4.25896019e-01 -8.66235495e-01 4.73674864e-01 -6.41276717e-01
1.04518497e+00 -2.17001095e-01 -5.29660463e-01 -1.53268945e+00
1.90226570e-01 -2.71653652e-01 5.67017235e-02 1.01009881e+00
6.49133205e-01 -5.57044029e-01 4.25062001e-01 2.34833583e-01
-5.42761683e-02 -6.23384833e-01 -1.21022403e+00 -2.19901532e-01
7.95807719e-01 -4.72351998e-01 1.96880341e-01 9.97497499e-01
2.49890313e-01 1.06437314e+00 -2.15021431e-01 -1.89335436e-01
2.65676469e-01 2.95437276e-01 6.83094025e-01 -1.24558365e+00
-3.88743967e-01 -1.69422448e-01 -4.55937497e-02 -1.20299768e+00
1.84376866e-01 -1.10511029e+00 1.01697542e-01 -1.85401714e+00
1.17649809e-01 -1.01933673e-01 -8.17501638e-03 5.77482164e-01
-3.67462426e-01 2.48032004e-01 2.23826766e-01 8.39345157e-02
-6.25427902e-01 6.72434807e-01 1.38070726e+00 -1.09380275e-01
1.43166944e-01 -2.36971870e-01 -6.24916732e-01 4.89929378e-01
1.24311757e+00 -1.83330953e-01 -2.27425724e-01 -6.22535050e-01
5.55619419e-01 1.92345485e-01 7.48369023e-02 -6.61103487e-01
-7.49078467e-02 -1.00835133e-02 3.64616327e-02 -6.23367846e-01
9.30004194e-02 -3.42570692e-01 -4.97914374e-01 5.66725135e-01
-5.52652717e-01 2.45383516e-01 -1.19317146e-02 7.42112994e-02
-5.65634847e-01 -4.19726700e-01 6.80181861e-01 -2.47130081e-01
-6.81273222e-01 -3.88101071e-01 -5.15858829e-01 6.39976144e-01
7.16783464e-01 1.57690138e-01 -5.88894963e-01 -6.19425714e-01
-3.15523922e-01 5.16951203e-01 1.07918039e-01 7.25436330e-01
5.35964787e-01 -7.65948474e-01 -1.62032878e+00 -1.97574258e-01
3.88464242e-01 -6.62697330e-02 1.13558464e-01 6.20864272e-01
-8.82190645e-01 6.75531387e-01 -3.62253666e-01 -4.28763419e-01
-1.22566760e+00 -1.04736621e-02 2.15721846e-01 -6.51375413e-01
-2.63308853e-01 6.04375303e-01 -4.11592305e-01 -7.25471318e-01
-2.76645541e-01 1.82249784e-01 -8.71125877e-01 2.47777551e-01
8.30074668e-01 6.94862425e-01 3.07561845e-01 -7.80993342e-01
-1.59942418e-01 3.80414993e-01 4.56987508e-03 -4.08028156e-01
1.33463478e+00 -4.57345545e-02 -5.39391458e-01 5.86675286e-01
9.49877739e-01 2.05141276e-01 -3.44534934e-01 -1.45749256e-01
5.13340533e-01 -7.33955950e-02 -2.41177574e-01 -1.49885726e+00
-1.93984196e-01 9.16441441e-01 2.01486453e-01 3.52792412e-01
8.91817927e-01 6.53697252e-02 7.37943888e-01 8.66260529e-01
5.29740930e-01 -1.08141720e+00 1.17805071e-01 1.32238853e+00
1.32265627e+00 -1.43171358e+00 -9.97783244e-02 -1.78515181e-01
-7.86463261e-01 1.17726088e+00 3.23633522e-01 -4.76831608e-02
2.02362075e-01 -4.35676873e-01 4.83285934e-01 -2.31136516e-01
-6.11371458e-01 -2.78931353e-02 4.65007097e-01 4.71582532e-01
7.89060116e-01 2.21992910e-01 -8.77378166e-01 4.68503416e-01
-7.25165367e-01 -1.28130272e-01 6.72919273e-01 7.52851427e-01
-6.41878426e-01 -1.21609378e+00 -3.19538146e-01 2.77882695e-01
-7.99538195e-01 -4.17188525e-01 -6.08033061e-01 7.44916558e-01
-2.28999078e-01 1.52446866e+00 -3.51144522e-01 3.81509773e-02
1.81389004e-01 2.64421433e-01 7.32484460e-01 -6.60956025e-01
-8.29149008e-01 -6.59424603e-01 5.08273363e-01 -2.16599911e-01
-7.86198199e-01 -5.52938223e-01 -1.13571072e+00 4.63863602e-03
-2.17599273e-01 6.28459275e-01 5.83167851e-01 1.42517853e+00
-7.04239011e-02 3.88813287e-01 3.12537581e-01 -2.64292091e-01
-7.67540932e-01 -1.52519882e+00 -1.37088194e-01 4.30618405e-01
2.27467403e-01 -2.72321403e-01 -1.81075916e-01 1.99748706e-02] | [11.376626014709473, 8.212153434753418] |
28701a37-f3a9-4f00-90cd-c8fd93b33e10 | a-deeper-look-at-saliency-feature-contrast | null | null | http://openaccess.thecvf.com/content_cvpr_2016/html/Bruce_A_Deeper_Look_CVPR_2016_paper.html | http://openaccess.thecvf.com/content_cvpr_2016/papers/Bruce_A_Deeper_Look_CVPR_2016_paper.pdf | A Deeper Look at Saliency: Feature Contrast, Semantics, and Beyond | In this paper we consider the problem of visual saliency modeling, including both human gaze prediction and salient object segmentation. The overarching goal of the paper is to identify high level considerations relevant to deriving more sophisticated visual saliency models. A deep learning model based on fully convolutional networks (FCNs) is presented, which shows very favorable performance across a wide variety of benchmarks relative to existing proposals. We also demonstrate that the manner in which training data is selected, and ground truth treated is critical to resulting model behaviour. Recent efforts have explored the relationship between human gaze and salient objects, and we also examine this point further in the context of FCNs. Close examination of the proposed and alternative models serves as a vehicle for identifying problems important to developing more comprehensive models going forward. | ['Sasa Janjic', 'Christopher Catton', 'Neil D. B. Bruce'] | 2016-06-01 | null | null | null | cvpr-2016-6 | ['eye-tracking'] | ['computer-vision'] | [ 6.14720225e-01 3.44227344e-01 -4.94956732e-01 -4.87738967e-01
-2.96572089e-01 -1.42824277e-01 7.34255075e-01 1.16018273e-01
-4.04530287e-01 5.07171154e-01 3.52852166e-01 -1.72430903e-01
-2.02320382e-01 -1.38917521e-01 -7.65170276e-01 -3.96974981e-01
2.04441044e-02 -9.37996805e-02 4.45247293e-01 -3.04940015e-01
8.81854773e-01 4.24819201e-01 -2.24810982e+00 3.61331515e-02
7.61842728e-01 8.58794749e-01 4.55697238e-01 6.93692565e-01
1.68992966e-01 1.15605724e+00 -4.39515442e-01 -3.04235429e-01
1.11607807e-02 -3.54552716e-01 -1.20729315e+00 1.67378306e-01
8.58933687e-01 -1.33879200e-01 5.24080209e-02 1.11369336e+00
2.42252991e-01 2.63414174e-01 5.98211706e-01 -1.69192684e+00
-7.30549932e-01 3.75750124e-01 -5.79163074e-01 9.71026659e-01
1.54535696e-01 1.75101459e-01 1.28639591e+00 -8.82836461e-01
5.63279092e-01 1.16835690e+00 4.12802011e-01 8.04679394e-01
-1.01136100e+00 -2.98366904e-01 3.78422350e-01 4.46541578e-01
-9.70050752e-01 -6.01990223e-01 9.11818385e-01 -3.28831971e-01
7.85697520e-01 4.96537447e-01 6.51412189e-01 8.36556196e-01
1.60316765e-01 1.26211047e+00 1.11064446e+00 -7.08400071e-01
5.61477691e-02 1.30764484e-01 5.45902371e-01 4.52018797e-01
4.72779900e-01 2.14791164e-01 -8.28422785e-01 1.10855497e-01
7.34936118e-01 -1.40421584e-01 -1.78890735e-01 -7.00754046e-01
-1.02870297e+00 9.42743599e-01 8.94875944e-01 3.18002373e-01
-3.73570859e-01 2.49234185e-01 7.74925426e-02 -2.57705331e-01
4.33558226e-01 7.72673368e-01 4.05743569e-02 2.10724100e-01
-1.30978179e+00 5.91158509e-01 3.08603406e-01 9.69486952e-01
8.12384009e-01 3.13829631e-01 -3.89545470e-01 4.32804942e-01
5.41030467e-01 1.56243265e-01 2.99485862e-01 -1.12112284e+00
7.09907711e-02 4.77309883e-01 3.08181524e-01 -1.09676647e+00
-5.31560540e-01 -2.99806923e-01 -1.41278338e-02 5.12390494e-01
2.67387927e-01 -9.46514215e-03 -1.25716233e+00 1.77396774e+00
2.37824190e-02 -5.14031993e-03 -5.62675521e-02 1.28622842e+00
1.18393087e+00 2.81660318e-01 6.20511353e-01 -7.61664733e-02
1.37714207e+00 -1.16584301e+00 -6.97592854e-01 -6.81764662e-01
3.14267159e-01 -6.99799359e-01 9.67584908e-01 -4.87045087e-02
-1.48842740e+00 -5.41194975e-01 -1.03350163e+00 -6.04345262e-01
-3.06944698e-01 -5.81065975e-02 7.24986613e-01 3.53157043e-01
-1.64811707e+00 3.21655959e-01 -5.46751738e-01 -7.95447648e-01
6.38293386e-01 5.51104188e-01 2.14921430e-01 2.53161401e-01
-1.01484704e+00 1.37241924e+00 2.03034386e-01 -2.87902635e-03
-8.67728710e-01 -4.06417191e-01 -1.07012033e+00 1.84943169e-01
2.93169975e-01 -5.48219502e-01 1.72489655e+00 -1.56598341e+00
-6.60983324e-01 1.20408273e+00 -7.19700515e-01 -7.74656594e-01
1.53821573e-01 -5.23653030e-02 -5.06469309e-01 1.83836520e-01
3.29188973e-01 1.40121138e+00 9.64952171e-01 -1.49615479e+00
-1.18883717e+00 -2.13808477e-01 3.30185533e-01 5.41407466e-01
3.50799523e-02 6.34753823e-01 -1.94605827e-01 -4.73497659e-01
-3.24690312e-01 -8.71858478e-01 -1.42875075e-01 -1.57861739e-01
-2.19501644e-01 -5.50983489e-01 8.45887840e-01 -2.80366182e-01
1.28882277e+00 -1.79124987e+00 -3.43941413e-02 -1.55310303e-01
6.23813570e-01 2.43953183e-01 2.42070723e-02 6.05691187e-02
-2.21302435e-01 3.27345468e-02 -1.99089304e-01 -3.32319587e-01
-1.06400207e-01 -2.36330882e-01 -3.17256540e-01 3.12281519e-01
5.01933277e-01 1.31639671e+00 -8.18067014e-01 -6.49656296e-01
1.38349831e-01 3.26746553e-01 -3.31721574e-01 -7.06545115e-02
-2.04800352e-01 -9.40299928e-02 -3.02344173e-01 8.14548373e-01
3.57603550e-01 -6.77914321e-01 -3.13914299e-01 -1.75915305e-02
-5.46627760e-01 4.25182998e-01 -5.03428876e-01 1.20939374e+00
3.53345454e-01 1.37708664e+00 -1.60554439e-01 -6.05556905e-01
6.33793592e-01 -3.19147408e-02 1.14880972e-01 -7.81740189e-01
3.77516270e-01 -9.64230001e-02 2.24741504e-01 -3.55093688e-01
1.14477837e+00 6.97451979e-02 2.24216282e-01 4.19669151e-01
1.64642408e-01 2.40400881e-01 2.08513230e-01 2.44220480e-01
3.82172585e-01 1.50630951e-01 4.72085595e-01 -8.79191816e-01
2.45055556e-01 6.05878413e-01 2.83050001e-01 7.21468866e-01
-9.39814389e-01 6.69733942e-01 4.45572048e-01 -4.96132493e-01
-1.03028917e+00 -6.86063170e-01 -1.38917938e-01 1.53887522e+00
7.91037202e-01 -8.60779658e-02 -1.03207266e+00 -4.80082333e-01
-1.41246170e-01 1.00122416e+00 -1.28938794e+00 -1.48656532e-01
-5.03915906e-01 -6.56327009e-01 1.47827789e-01 6.85211122e-01
8.22236091e-02 -1.67411089e+00 -1.42948699e+00 -1.93770319e-01
-1.02934673e-01 -7.24711120e-01 -2.86057860e-01 1.20882332e-01
-7.96298265e-01 -1.25754631e+00 -7.62432933e-01 -1.13225472e+00
5.59362710e-01 9.49156404e-01 1.28870893e+00 3.77949715e-01
-3.38332467e-02 5.40344775e-01 -1.22005835e-01 -1.05636477e+00
-1.62489545e-02 1.95231870e-01 -1.73837855e-01 -2.49580532e-01
1.04186571e+00 1.88901529e-01 -7.19608247e-01 1.43110454e-01
-9.11678970e-01 2.77694702e-01 3.25003982e-01 4.60432082e-01
3.25537026e-01 -6.31826699e-01 6.04740500e-01 -7.59171069e-01
6.92206621e-01 -5.72521567e-01 -4.86471534e-01 2.92090118e-01
-7.72823393e-01 -2.90721625e-01 -2.59687006e-01 -7.40140751e-02
-1.11529529e+00 -1.68736070e-01 1.69687673e-01 -3.23399633e-01
-3.06880236e-01 2.17088893e-01 3.79758060e-01 -3.47323686e-01
7.56927192e-01 4.34595579e-03 6.12890646e-02 -1.35937229e-01
1.87171012e-01 2.91553795e-01 4.53273535e-01 8.93808305e-02
4.27241534e-01 4.75410849e-01 -1.03284776e-01 -9.10892069e-01
-1.08104658e+00 -4.64753479e-01 -6.95665658e-01 -3.30735713e-01
9.44015920e-01 -8.54360819e-01 -5.23965895e-01 2.30698973e-01
-9.78812873e-01 -3.79692078e-01 -2.08310708e-01 1.23274565e-01
-7.93907046e-01 8.98862779e-02 -2.56808847e-01 -8.00397277e-01
-3.05454850e-01 -1.26900506e+00 1.04140246e+00 8.75112355e-01
-5.66201925e-01 -1.21645272e+00 -1.52114987e-01 4.59462315e-01
3.90330166e-01 1.21290848e-01 6.20265722e-01 -6.33039773e-01
-8.40372086e-01 2.73969680e-01 -5.00226498e-01 -1.38151661e-01
-1.71150751e-02 2.21546322e-01 -1.24200881e+00 -1.71709031e-01
4.29350026e-02 -3.04884285e-01 9.98584688e-01 9.06341136e-01
7.69620180e-01 1.29236877e-01 -5.54683328e-01 3.41243386e-01
1.38509631e+00 -4.71638255e-02 4.58025753e-01 6.86639249e-01
7.08529174e-01 9.07474577e-01 9.05882180e-01 -4.39701565e-02
8.11400771e-01 4.30063516e-01 7.54131973e-01 -2.63017505e-01
-3.05819988e-01 -5.04064970e-02 -8.45515653e-02 1.14234962e-01
-2.21050456e-01 -1.56057835e-01 -1.07279348e+00 1.08137429e+00
-1.91204202e+00 -1.19282174e+00 -3.24547410e-01 1.72589707e+00
4.91372496e-01 2.41691589e-01 6.48756027e-01 -6.39767721e-02
1.03111005e+00 3.63104403e-01 -6.75761878e-01 -4.01587814e-01
-1.14166461e-01 -1.14084959e-01 3.64776254e-01 4.54014629e-01
-1.33509040e+00 1.23959887e+00 8.16203880e+00 2.28729203e-01
-1.21672070e+00 -8.00092593e-02 6.23768091e-01 -3.01917076e-01
-3.05801451e-01 -2.06398107e-02 -9.69214499e-01 3.25929970e-01
8.77178907e-01 -4.01157320e-01 1.64760664e-01 9.54296827e-01
1.72972873e-01 -4.50973868e-01 -1.03246748e+00 5.95068157e-01
4.83493596e-01 -1.41829586e+00 4.85344194e-02 -6.95696026e-02
9.67043221e-01 2.37207383e-01 4.76784885e-01 9.97278020e-02
2.65349388e-01 -1.13154781e+00 1.08232021e+00 4.39079344e-01
3.93490136e-01 -7.35202014e-01 5.01658976e-01 -4.11429405e-02
-8.09867382e-01 -2.46250942e-01 -3.08430254e-01 -3.55931610e-01
7.09774643e-02 -2.89297789e-01 -8.06758106e-01 -4.67923423e-03
9.57788646e-01 7.18880057e-01 -1.01845634e+00 1.41459680e+00
-1.17513560e-01 4.91515130e-01 1.51486337e-01 -2.17864841e-01
6.43815458e-01 5.71720600e-01 5.74426413e-01 1.17729998e+00
-1.26506761e-01 -8.90908241e-02 -1.96472377e-01 1.00387073e+00
1.67357951e-01 -1.13931268e-01 -4.17256981e-01 1.83925182e-01
2.38484070e-01 1.12708795e+00 -9.56611454e-01 -2.37317622e-01
-6.29200041e-01 4.72313434e-01 5.29441774e-01 5.92049837e-01
-6.79109037e-01 -1.73896566e-01 7.93034554e-01 2.50091910e-01
4.47846860e-01 2.44241714e-01 -7.51338303e-01 -9.69738424e-01
-3.64371121e-01 -7.04615295e-01 3.75616699e-01 -1.14704704e+00
-9.09369111e-01 5.03080130e-01 2.93249220e-01 -8.56699467e-01
-3.32079560e-01 -4.13053393e-01 -8.02123249e-01 1.05798829e+00
-2.03246927e+00 -1.28016341e+00 -4.71461713e-01 5.46823025e-01
8.26784015e-01 9.96469036e-02 2.93526560e-01 -4.14663345e-01
-4.81674492e-01 3.65201563e-01 -4.06877935e-01 -1.89945146e-01
4.64149266e-01 -1.29785097e+00 7.48205066e-01 1.14630699e+00
1.52718231e-01 6.06205106e-01 1.12945998e+00 -6.30420268e-01
-7.47604489e-01 -9.00366068e-01 1.13493025e+00 -8.78478408e-01
3.83972973e-01 -4.94739190e-02 -9.66326118e-01 8.85662138e-01
7.33631730e-01 -3.42885345e-01 5.84068775e-01 1.11579724e-01
3.18959475e-01 4.92647439e-01 -1.08671570e+00 7.54124343e-01
9.52651799e-01 -4.02369827e-01 -9.75991786e-01 -1.58523306e-01
6.77957594e-01 -4.70284253e-01 -1.27488121e-01 5.02565980e-01
3.37243468e-01 -1.23494744e+00 7.97648072e-01 -8.41826200e-01
6.32080019e-01 -9.01944265e-02 1.88886821e-01 -9.58194375e-01
-7.98823774e-01 -3.75576764e-01 -3.14183176e-01 8.25372875e-01
1.83833927e-01 -1.23161264e-01 1.02687955e+00 9.66385663e-01
-1.66662216e-01 -7.41392195e-01 -4.87370789e-01 -1.27353251e-01
-1.33962139e-01 -1.61425680e-01 4.37067211e-01 8.17406178e-01
-7.34779611e-02 3.72342706e-01 -2.20714569e-01 7.99411163e-02
7.69336164e-01 2.47399956e-01 4.76804167e-01 -1.47994304e+00
5.95700800e-01 -7.78319597e-01 -4.15970445e-01 -8.51858854e-01
1.92573369e-01 -3.80860627e-01 1.65226683e-01 -1.58890176e+00
3.39695215e-01 -1.12085178e-01 -6.87037170e-01 3.07631254e-01
-5.36854327e-01 4.93705839e-01 4.91183102e-01 3.47803921e-01
-9.78219151e-01 2.66458333e-01 1.27986646e+00 2.02277690e-01
-4.10315208e-02 2.27267176e-01 -1.44088292e+00 9.28669810e-01
8.96901488e-01 -1.77795991e-01 -4.84045506e-01 -5.07532597e-01
-7.07844123e-02 -4.23688769e-01 7.08315790e-01 -8.41543496e-01
3.97686452e-01 -4.81235743e-01 4.85182375e-01 -6.92491531e-01
3.06398809e-01 -6.45821929e-01 -5.53372145e-01 1.33211583e-01
-7.77504146e-01 1.47072285e-01 3.21623653e-01 3.98880213e-01
-1.89566508e-01 -4.38376695e-01 9.59345818e-01 -1.56053990e-01
-1.47509646e+00 7.15171918e-02 -2.48046353e-01 1.18517414e-01
1.19053328e+00 -6.97439134e-01 -3.29879761e-01 -4.52929318e-01
-7.88860083e-01 3.82770091e-01 9.60957527e-01 7.56515622e-01
5.05994022e-01 -1.15175974e+00 -2.62665749e-01 1.47421211e-01
2.39723101e-01 -3.25442463e-01 9.80498940e-02 6.78340256e-01
-1.97458595e-01 8.66777480e-01 -5.95081627e-01 -6.34074152e-01
-1.49496031e+00 8.12643170e-01 2.14875847e-01 3.17876637e-01
-2.20121354e-01 8.84994686e-01 5.11607885e-01 2.77241856e-01
4.51067626e-01 -2.73087680e-01 -7.24952102e-01 7.64801279e-02
5.84289849e-01 4.97682154e-01 -2.82207400e-01 -1.25942457e+00
-3.87737513e-01 1.41881078e-01 -1.96669370e-01 2.01148286e-01
1.16847396e+00 -5.31712472e-01 -4.00071079e-03 4.76097703e-01
7.62495697e-01 -5.08190811e-01 -1.70937335e+00 -2.57821172e-01
3.49982858e-01 -4.50347602e-01 2.75604606e-01 -6.89482570e-01
-7.58677185e-01 9.22513664e-01 6.53573751e-01 4.64580894e-01
1.05381036e+00 2.63853997e-01 2.60713786e-01 -3.67157795e-02
2.13040859e-02 -9.90271032e-01 1.38637787e-02 4.21172291e-01
7.53844142e-01 -1.69281876e+00 3.66582870e-02 -1.91002965e-01
-1.09035063e+00 7.75971353e-01 1.10945070e+00 -3.67554128e-01
4.02500153e-01 -3.20480794e-01 1.83682874e-01 -5.97064912e-01
-7.57416248e-01 -6.97622538e-01 8.27151179e-01 6.49647653e-01
5.12556612e-01 -1.99369922e-01 -1.24161854e-01 1.50379732e-01
-1.12865917e-01 9.14364532e-02 6.15298808e-01 1.05903232e+00
-9.35013235e-01 -5.40176272e-01 -2.89536178e-01 5.61635077e-01
-6.47016525e-01 -1.95393234e-01 -5.32840312e-01 9.92577195e-01
6.55574631e-03 8.36559713e-01 2.90373385e-01 5.11454120e-02
6.71547577e-02 4.15159538e-02 3.31455380e-01 -7.74241686e-01
-6.33823037e-01 -9.03712213e-02 -6.87059313e-02 -3.86191875e-01
-1.09461236e+00 -9.07611609e-01 -9.84169185e-01 -1.32644281e-01
-3.74534100e-01 -8.55500996e-02 4.75998521e-01 9.45690870e-01
2.77016938e-01 4.09705579e-01 1.49660662e-01 -1.23222792e+00
-1.52961999e-01 -8.15200388e-01 -5.06932259e-01 3.50018144e-01
8.97040129e-01 -9.63440955e-01 -1.31682530e-01 1.40155643e-01] | [10.041171073913574, 1.4305429458618164] |
4c1805dd-8c21-45c7-9d51-ceb8e112a273 | exploring-visual-context-for-weakly | 2106.10506 | null | https://arxiv.org/abs/2106.10506v2 | https://arxiv.org/pdf/2106.10506v2.pdf | Exploring Visual Context for Weakly Supervised Person Search | Person search has recently emerged as a challenging task that jointly addresses pedestrian detection and person re-identification. Existing approaches follow a fully supervised setting where both bounding box and identity annotations are available. However, annotating identities is labor-intensive, limiting the practicability and scalability of current frameworks. This paper inventively considers weakly supervised person search with only bounding box annotations. We proposed to address this novel task by investigating three levels of context clues (i.e., detection, memory and scene) in unconstrained natural images. The first two are employed to promote local and global discriminative capabilities, while the latter enhances clustering accuracy. Despite its simple design, our CGPS achieves 80.0% in mAP on CUHK-SYSU, boosting the baseline model by 8.8%. Surprisingly, it even achieves comparable performance with several supervised person search models. Our code is available at https://github.com/ljpadam/CGPS | ['Ling Shao', 'Xiaokang Yang', 'Bingbing Ni', 'Jie Qin', 'Shengcai Liao', 'Jinpeng Li', 'Yichao Yan'] | 2021-06-19 | null | null | null | null | ['person-search'] | ['computer-vision'] | [-2.63434619e-01 -3.68590653e-01 -1.81395113e-01 -2.95432895e-01
-7.98813522e-01 -6.77660584e-01 7.65798271e-01 1.05503753e-01
-9.04638231e-01 8.25404823e-01 2.84399629e-01 -8.05714168e-03
3.61189961e-01 -5.96162438e-01 -5.33892632e-01 -5.89002609e-01
2.11768478e-01 3.61518085e-01 4.03331101e-01 1.91206425e-01
1.20317629e-02 2.72258878e-01 -1.78523493e+00 -7.32636228e-02
8.04353714e-01 8.99766743e-01 2.06066757e-01 6.25911772e-01
2.60224402e-01 6.26804054e-01 -3.59369546e-01 -8.84141803e-01
4.10173118e-01 6.24743402e-02 -9.06738579e-01 -9.34346020e-02
9.03717339e-01 -6.30255222e-01 -6.67999446e-01 9.88454282e-01
8.53008389e-01 2.01682776e-01 3.72483939e-01 -1.32194281e+00
-7.16392517e-01 2.95399785e-01 -6.19901180e-01 3.52702975e-01
4.74542946e-01 2.40686595e-01 9.10695195e-01 -9.24641550e-01
3.38938594e-01 1.04091322e+00 8.42528760e-01 6.27193749e-01
-1.27568865e+00 -7.97053516e-01 3.13085347e-01 4.07651186e-01
-1.86724913e+00 -7.61966288e-01 3.40777129e-01 -4.41876113e-01
8.01406145e-01 2.96241671e-01 4.03370738e-01 1.17007744e+00
-7.11943507e-01 9.51684058e-01 1.10460043e+00 -3.83144885e-01
6.83776289e-02 4.00420040e-01 4.55607533e-01 6.00750268e-01
2.93239802e-01 1.15649723e-01 -7.07713723e-01 -3.80220711e-01
7.11257160e-01 -4.63358238e-02 -1.09643996e-01 -2.58003175e-01
-9.49692726e-01 5.86182177e-01 4.76653397e-01 3.64173763e-02
-3.56732532e-02 1.44512117e-01 4.05654430e-01 -9.26041529e-02
4.51079577e-01 1.12523116e-01 -1.69459119e-01 -2.48970821e-01
-9.13012266e-01 3.44295293e-01 6.25084341e-01 1.25686800e+00
7.93031991e-01 -4.04219210e-01 -3.21856052e-01 1.01988649e+00
-5.06177396e-02 6.54253483e-01 1.95070207e-01 -8.53774071e-01
3.70614529e-01 3.91204715e-01 4.71263617e-01 -9.07301605e-01
-2.85752833e-01 -4.35922205e-01 -5.43139279e-01 -2.85876900e-01
7.67805040e-01 -1.32931516e-01 -7.19618261e-01 1.83478856e+00
4.72435236e-01 2.71512657e-01 -4.25480366e-01 1.15495384e+00
9.14693594e-01 1.41440973e-01 5.78595519e-01 2.91913331e-01
1.68100131e+00 -1.27364957e+00 -3.39552969e-01 -4.31175947e-01
5.35415113e-01 -7.22626209e-01 9.39541698e-01 -1.22254618e-01
-1.09554636e+00 -7.20660031e-01 -7.71557212e-01 -2.86691636e-01
-4.26685452e-01 5.76374233e-01 5.11326909e-01 9.39954460e-01
-1.41771924e+00 1.99174196e-01 -6.60962462e-01 -8.09913874e-01
4.52087492e-01 4.13833708e-01 -4.20996040e-01 -6.18143082e-02
-1.05416214e+00 7.22828984e-01 1.98194921e-01 -9.77736488e-02
-3.90777439e-01 -5.32087862e-01 -6.69214249e-01 -5.86121343e-02
3.91653270e-01 -7.91531742e-01 1.24386978e+00 -5.48390508e-01
-1.05060780e+00 1.26310945e+00 -6.77051902e-01 -4.60421741e-01
8.36225688e-01 -4.07555580e-01 -1.65782139e-01 1.58793524e-01
6.16232395e-01 8.78057122e-01 4.47174221e-01 -1.20632231e+00
-1.01711571e+00 -5.01966059e-01 6.41592294e-02 2.42755294e-01
-5.34473538e-01 3.07565570e-01 -9.90553856e-01 -5.58737099e-01
-1.39502987e-01 -9.34050262e-01 -1.08627044e-01 1.36983633e-01
-2.96883434e-01 -3.65863144e-01 4.78049457e-01 -9.30562735e-01
1.18052375e+00 -1.93830419e+00 -3.07501674e-01 1.07068345e-01
1.75626382e-01 2.65130043e-01 9.09761935e-02 3.06183964e-01
2.46005654e-01 1.40014309e-02 9.77174193e-02 -8.99380863e-01
2.04472378e-01 -1.57865942e-01 2.07547359e-02 4.86858457e-01
-1.25176394e-02 1.12305331e+00 -9.46470618e-01 -6.58948660e-01
2.49947712e-01 3.20579141e-01 -4.52121973e-01 7.19158500e-02
4.92551416e-01 4.35131729e-01 -2.95658529e-01 9.30817366e-01
7.62819350e-01 -3.72527450e-01 1.43723935e-01 -1.47253752e-01
-3.17550868e-01 1.30860552e-01 -1.24496531e+00 1.48233724e+00
-1.12061769e-01 5.79021573e-01 4.91240211e-02 -6.87252104e-01
5.48663914e-01 4.57136594e-02 3.11745822e-01 -7.68133342e-01
-4.08416688e-02 8.53238851e-02 -4.93363857e-01 -2.63332456e-01
6.71507478e-01 4.48774993e-01 -1.11810535e-01 3.23164046e-01
-3.04475613e-02 7.20338702e-01 2.32853308e-01 2.31090471e-01
9.70043361e-01 2.47944936e-01 3.92161816e-01 -4.23518926e-01
6.24027431e-01 3.84406485e-02 4.25354868e-01 1.11925673e+00
-8.05124760e-01 7.80551732e-01 -5.34222983e-02 -4.43988651e-01
-1.15240061e+00 -1.12429225e+00 -2.23772064e-01 1.48319757e+00
4.43064034e-01 -4.98576611e-01 -9.04166639e-01 -6.64716125e-01
1.91269457e-01 1.69302553e-01 -5.33161938e-01 1.25003204e-01
-6.64541483e-01 -7.86084175e-01 8.40525031e-01 8.43507290e-01
9.08794463e-01 -6.39930964e-01 -2.94274479e-01 -8.25690702e-02
-5.70120811e-01 -1.50842679e+00 -8.33636165e-01 -3.23683500e-01
-1.75769255e-01 -1.06749249e+00 -1.00533140e+00 -9.13144469e-01
6.98194206e-01 6.16753876e-01 1.07368255e+00 2.79180914e-01
-4.08724844e-01 6.12198114e-01 -3.64502937e-01 -1.39742509e-01
2.98312813e-01 2.08984107e-01 2.93441832e-01 5.15175201e-02
8.03203642e-01 -3.83407712e-01 -1.10782778e+00 6.03654146e-01
-1.70911893e-01 -3.08068898e-02 3.83277059e-01 7.13154018e-01
3.95637512e-01 -3.39109838e-01 4.00310487e-01 -4.95859861e-01
2.09858000e-01 -2.09433854e-01 -4.84764069e-01 3.73740047e-01
-4.15288866e-01 -4.46218818e-01 2.17114523e-01 -4.60738033e-01
-1.07500041e+00 2.62136161e-01 -1.86369047e-01 -3.97550547e-03
-5.55197716e-01 -1.59980431e-01 -2.35603169e-01 -2.56107807e-01
5.91236413e-01 3.65654469e-01 -3.67828965e-01 -4.84412611e-01
1.99221656e-01 8.01129997e-01 7.92816639e-01 -7.42895842e-01
8.40477943e-01 7.13946164e-01 -5.63246429e-01 -7.59330034e-01
-7.71932602e-01 -1.03177035e+00 -9.49069321e-01 -2.44077787e-01
9.08904016e-01 -1.37091315e+00 -9.23846602e-01 4.79456961e-01
-9.97987986e-01 -2.36032382e-01 8.32217038e-02 1.96815267e-01
-2.04560667e-01 7.06741810e-01 -6.85681224e-01 -1.02468562e+00
-4.03354913e-01 -8.20691347e-01 1.36306667e+00 3.19673479e-01
-3.98779333e-01 -5.91167152e-01 -2.42257342e-01 7.12632537e-01
2.10369036e-01 -8.65796581e-03 1.92042053e-01 -5.18016636e-01
-6.18913651e-01 -4.48420256e-01 -6.85682654e-01 -1.13040246e-01
-2.32635085e-02 -6.14092648e-01 -1.38325739e+00 -4.09045815e-01
-4.69489604e-01 -2.78774679e-01 9.71342802e-01 2.18394309e-01
9.71015453e-01 -2.16322988e-01 -6.50847316e-01 4.99531567e-01
1.17024016e+00 -8.01967755e-02 3.68976682e-01 5.64888537e-01
6.61599696e-01 5.37350357e-01 4.53420609e-01 5.15942216e-01
8.44950080e-01 1.02720606e+00 -6.32762238e-02 -1.21264391e-01
-3.55012476e-01 -3.74866039e-01 4.52300310e-02 8.86386633e-02
-5.02882600e-01 -9.53641720e-03 -1.17003202e+00 6.71511471e-01
-2.31778502e+00 -1.07696795e+00 -5.69335707e-02 2.20642686e+00
7.85502613e-01 -6.59032688e-02 6.37368917e-01 -1.42397299e-01
1.02087319e+00 -9.53006670e-02 -3.83491993e-01 4.61696982e-01
-2.20010743e-01 -1.13317788e-01 7.43664920e-01 4.72523332e-01
-1.58732736e+00 1.16725469e+00 5.69518805e+00 8.08237791e-01
-5.61019480e-01 4.19209242e-01 5.92657447e-01 -2.85428584e-01
4.31784362e-01 -2.47856930e-01 -1.30964971e+00 7.36846209e-01
5.63628197e-01 6.56346083e-02 1.60687819e-01 9.47896242e-01
1.05578616e-01 -3.32073241e-01 -1.14001870e+00 1.31698990e+00
1.23956732e-01 -9.99925256e-01 -1.90071419e-01 1.06159508e-01
5.66709220e-01 -1.21096209e-01 1.57789662e-01 2.88532376e-01
2.51776636e-01 -8.42339814e-01 8.40132415e-01 2.39512101e-01
7.81023383e-01 -6.05697393e-01 8.16473126e-01 3.33261460e-01
-1.54810047e+00 -1.78154543e-01 -2.68183202e-01 -1.94423243e-01
2.65610367e-01 9.27606151e-02 -5.04571021e-01 4.22430843e-01
1.13480949e+00 5.14691234e-01 -1.03551495e+00 1.24806392e+00
-2.16925312e-02 3.75551283e-01 -3.94632310e-01 1.16000585e-01
1.13467366e-01 2.03306135e-02 3.12539995e-01 1.49107742e+00
6.04645647e-02 1.24209121e-01 6.32324517e-01 6.41297102e-01
2.78232377e-02 -6.07038755e-03 -2.67885119e-01 5.21731496e-01
8.12589586e-01 1.13631284e+00 -7.31699824e-01 -4.11234379e-01
-6.02533937e-01 1.19331992e+00 6.41809940e-01 5.34820497e-01
-1.06943083e+00 2.23938599e-02 8.15217376e-01 3.77086848e-01
1.97349474e-01 -2.67958552e-01 -3.49636465e-01 -1.24200201e+00
2.80049741e-01 -5.44278979e-01 6.23956978e-01 -4.18859661e-01
-1.45434701e+00 4.82951880e-01 -1.20155001e-02 -8.90529335e-01
2.60045528e-02 -4.75566655e-01 -2.95113742e-01 9.32051301e-01
-1.46746230e+00 -1.71364510e+00 -6.68733418e-01 6.95762694e-01
4.22772408e-01 -1.07247710e-01 7.32370615e-01 7.92098224e-01
-7.85944819e-01 1.16066849e+00 -4.64324951e-02 5.51884711e-01
1.04991269e+00 -1.15576780e+00 6.34398222e-01 1.07679439e+00
-1.28741145e-01 7.95373559e-01 5.14177680e-01 -6.70547009e-01
-9.11633074e-01 -1.14950383e+00 1.12753880e+00 -9.14614081e-01
6.26315236e-01 -6.76611602e-01 -5.98396122e-01 6.21968687e-01
-3.10727563e-02 1.53727261e-02 7.39695966e-01 3.21813136e-01
-5.58749199e-01 6.01086468e-02 -1.01570630e+00 8.52362096e-01
1.60340643e+00 -7.28574455e-01 -1.19423382e-01 3.26360703e-01
3.13040227e-01 -3.01890671e-01 -5.98963320e-01 2.50082225e-01
6.16656959e-01 -9.58024502e-01 1.47056568e+00 -2.93032587e-01
-9.49945748e-02 -4.03421789e-01 -1.17890917e-01 -5.82714498e-01
-6.28002465e-01 -4.25812006e-01 -1.60409361e-01 1.54216003e+00
1.95994079e-01 -7.01010168e-01 1.07736027e+00 1.14918172e+00
3.02959561e-01 -4.56469983e-01 -9.25659180e-01 -9.43322480e-01
-1.21112086e-01 -2.27405623e-01 5.43533981e-01 7.85344839e-01
-3.06792948e-02 1.48884669e-01 -5.24930537e-01 4.51594830e-01
8.95910680e-01 3.00766458e-03 9.80878711e-01 -1.02325034e+00
-1.14602469e-01 -5.90389669e-01 -4.73372877e-01 -1.30233192e+00
2.10780665e-01 -7.70072401e-01 -1.74334377e-01 -1.29168308e+00
6.98891222e-01 -5.79485178e-01 -1.16917990e-01 5.38077354e-01
-5.42554855e-01 6.27125382e-01 2.86546350e-01 5.34770012e-01
-9.35575485e-01 2.84320712e-01 5.39321125e-01 -1.24492697e-01
-1.12989634e-01 1.83521003e-01 -7.71409035e-01 6.70566857e-01
1.00352609e+00 -1.59467813e-02 -2.31864825e-02 -5.93887627e-01
-3.56568396e-01 -6.26511812e-01 9.80097890e-01 -1.09127355e+00
7.99920857e-01 2.67740637e-01 7.19431639e-01 -6.13997757e-01
5.93938828e-01 -5.24653733e-01 -4.47097160e-02 2.97022194e-01
-1.51182786e-01 5.75992465e-02 6.52384236e-02 5.43512523e-01
-8.66684467e-02 2.29402874e-02 7.07545102e-01 -2.47095928e-01
-1.14519095e+00 4.73570198e-01 1.94747020e-02 -6.78156167e-02
9.55947876e-01 -4.28253502e-01 -4.21153694e-01 -2.88519382e-01
-6.68256760e-01 4.90747362e-01 6.27613425e-01 3.42518240e-01
3.81793082e-01 -1.31562412e+00 -6.87916577e-01 -5.18107712e-02
2.51466900e-01 -3.59709680e-01 3.19058716e-01 7.27628827e-01
-2.99007386e-01 5.85019469e-01 7.38564059e-02 -6.29240692e-01
-1.67476189e+00 5.20173788e-01 2.96550155e-01 -6.92001581e-02
-4.89661396e-01 9.80673552e-01 2.99902618e-01 -3.84310544e-01
6.11622691e-01 4.74059612e-01 -1.26493141e-01 -3.82907279e-02
7.41654754e-01 6.19161129e-01 -3.04495871e-01 -8.62772226e-01
-5.76365769e-01 6.24716282e-01 -2.37660870e-01 1.93084646e-02
8.54225874e-01 -5.56724489e-01 1.28132582e-01 -1.29582316e-01
9.47481036e-01 -1.69890910e-01 -1.31467450e+00 -5.18242598e-01
3.24624777e-01 -6.23859167e-01 -2.50643224e-01 -6.64059341e-01
-5.65084636e-01 5.70032835e-01 7.27602303e-01 -8.52944404e-02
8.80460262e-01 2.55888790e-01 8.73986363e-01 3.80178511e-01
7.20127404e-01 -1.29956412e+00 -8.30259025e-02 3.89804304e-01
5.22303879e-01 -1.74091327e+00 3.63574177e-02 -6.96705282e-01
-4.69427794e-01 5.83972991e-01 9.32479382e-01 2.26754338e-01
4.16033804e-01 1.39135092e-01 -8.50293785e-03 2.06382588e-01
-1.28311753e-01 -8.71892691e-01 3.27148855e-01 7.96311378e-01
2.53076255e-01 1.16050571e-01 -9.89703760e-02 7.83728242e-01
-3.10600042e-01 -2.72779554e-01 -4.23477851e-02 9.06913579e-01
-2.93644071e-01 -1.05681527e+00 -5.01151383e-01 2.07074806e-01
-4.42804068e-01 -2.91760236e-01 -5.57872117e-01 7.57252395e-01
2.28731945e-01 1.15549064e+00 6.23212159e-02 -2.60473102e-01
3.27289492e-01 1.11154556e-01 3.45487922e-01 -4.70423311e-01
-5.18195808e-01 -6.38260990e-02 2.92849988e-01 -4.83291984e-01
-4.56855893e-01 -8.63335550e-01 -7.12979257e-01 -5.77972591e-01
-2.36271024e-01 4.55740504e-02 1.61035612e-01 6.60821557e-01
4.01025206e-01 -1.23927712e-01 2.49016315e-01 -8.76105011e-01
-2.88979053e-01 -7.69565403e-01 -2.84312159e-01 4.78559673e-01
2.44838938e-01 -7.81145394e-01 -4.31742854e-02 1.99356958e-01] | [14.812477111816406, 0.8606293797492981] |
48756609-0520-404d-89c4-dceb20c94cb7 | accidental-turntables-learning-3d-pose-by | 2212.06300 | null | https://arxiv.org/abs/2212.06300v1 | https://arxiv.org/pdf/2212.06300v1.pdf | Accidental Turntables: Learning 3D Pose by Watching Objects Turn | We propose a technique for learning single-view 3D object pose estimation models by utilizing a new source of data -- in-the-wild videos where objects turn. Such videos are prevalent in practice (e.g., cars in roundabouts, airplanes near runways) and easy to collect. We show that classical structure-from-motion algorithms, coupled with the recent advances in instance detection and feature matching, provides surprisingly accurate relative 3D pose estimation on such videos. We propose a multi-stage training scheme that first learns a canonical pose across a collection of videos and then supervises a model for single-view pose estimation. The proposed technique achieves competitive performance with respect to existing state-of-the-art on standard benchmarks for 3D pose estimation, without requiring any pose labels during training. We also contribute an Accidental Turntables Dataset, containing a challenging set of 41,212 images of cars in cluttered backgrounds, motion blur and illumination changes that serves as a benchmark for 3D pose estimation. | ['Subhransu Maji', 'Matheus Gadelha', 'Zezhou Cheng'] | 2022-12-13 | null | null | null | null | ['3d-pose-estimation'] | ['computer-vision'] | [-1.30261853e-01 -2.56746352e-01 -2.66743034e-01 -5.54254234e-01
-1.07457864e+00 -8.78203630e-01 6.19313419e-01 -4.76974368e-01
-2.76818424e-01 2.57402599e-01 1.65398508e-01 4.41221260e-02
2.51544863e-01 -2.62214959e-01 -1.38825560e+00 -6.62288666e-01
-1.19028710e-01 8.60955000e-01 5.10437787e-01 -1.68213561e-01
2.02003226e-01 7.91163921e-01 -1.63816881e+00 1.17884338e-01
6.43995553e-02 1.15358758e+00 1.62621275e-01 8.48380804e-01
4.14896697e-01 6.03634655e-01 -2.42788896e-01 -6.23672724e-01
5.74910879e-01 2.55845517e-01 -7.65707552e-01 6.40468121e-01
1.38533497e+00 -7.17814207e-01 -6.19373441e-01 7.40128577e-01
2.47518361e-01 1.72046319e-01 6.79376602e-01 -1.24816716e+00
-4.80536595e-02 -1.51988432e-01 -7.85637379e-01 2.93237150e-01
8.58297467e-01 2.35596731e-01 9.61274385e-01 -1.36029005e+00
1.09519911e+00 1.52943337e+00 6.70412362e-01 4.10553128e-01
-1.05166507e+00 -4.51832414e-01 3.23126853e-01 2.73194581e-01
-1.25180566e+00 -6.12646103e-01 6.65470719e-01 -6.29106760e-01
1.00674081e+00 1.56764001e-01 7.61698306e-01 1.16549170e+00
2.49125317e-01 1.08504415e+00 6.60983980e-01 -2.11087987e-01
-1.70947194e-01 -1.71590835e-01 -2.31904685e-01 7.13716805e-01
3.08786362e-01 2.49795437e-01 -6.76903486e-01 -1.60741119e-03
6.00067556e-01 1.79013848e-01 -1.13755532e-01 -1.40836763e+00
-1.53324878e+00 4.96849477e-01 4.31447268e-01 -3.12440246e-01
-9.12304372e-02 3.39637965e-01 4.11565155e-01 1.29099831e-01
4.72086549e-01 8.17667097e-02 -6.88981056e-01 -9.48352888e-02
-7.84782767e-01 7.48124719e-01 6.60864472e-01 1.47354150e+00
9.08400118e-01 -3.14195424e-01 5.34790419e-02 3.82972389e-01
3.61321270e-01 8.66396248e-01 -1.03076622e-01 -1.31297445e+00
8.64524961e-01 2.29060844e-01 5.97714245e-01 -9.33308125e-01
-3.32116246e-01 -2.52255380e-01 -1.86126992e-01 -1.13434277e-01
6.31336570e-01 3.17548811e-01 -8.56167138e-01 1.40041554e+00
8.52036655e-01 2.64786988e-01 -2.27335975e-01 1.10901129e+00
9.79882777e-01 3.02457601e-01 -4.27614510e-01 -1.74275748e-02
1.43143129e+00 -1.13819194e+00 -3.55453163e-01 -2.94770986e-01
6.73512340e-01 -1.07491267e+00 6.43372297e-01 4.03069168e-01
-1.05796707e+00 -6.28606260e-01 -7.80908406e-01 -1.73079103e-01
-2.92918593e-01 2.26917639e-01 5.73893487e-01 5.15180528e-01
-7.69832432e-01 3.74790907e-01 -8.79136086e-01 -3.57694387e-01
6.11316562e-01 1.92056492e-01 -8.87195945e-01 -3.93315166e-01
-6.06809676e-01 1.17619598e+00 1.32380635e-01 1.27156347e-01
-1.50737631e+00 -8.32630932e-01 -1.07664323e+00 -4.76534337e-01
8.46803129e-01 -8.07146013e-01 1.52090871e+00 -3.77636522e-01
-1.12194264e+00 1.30444753e+00 -2.47246936e-01 -2.80946821e-01
9.44638610e-01 -8.51385355e-01 -1.03727847e-01 1.88553408e-01
1.21117413e-01 9.30963695e-01 1.08633363e+00 -1.37441933e+00
-6.65138006e-01 -5.03161371e-01 2.92711407e-01 2.39603028e-01
2.47333616e-01 -9.04411543e-03 -8.01447749e-01 -2.60358512e-01
2.10872516e-01 -1.37037110e+00 -1.82160750e-01 2.01288015e-01
-3.25613171e-01 -1.49101704e-01 1.15295780e+00 -4.66299593e-01
5.49247622e-01 -1.76198471e+00 3.47201228e-01 -1.17115073e-01
2.47598588e-01 7.38672540e-02 1.09857572e-02 3.24116766e-01
-4.35744487e-02 -4.14124191e-01 3.40412319e-01 -6.17763579e-01
1.90674160e-02 1.22570820e-01 -9.96203050e-02 8.86574864e-01
1.89408198e-01 9.73730385e-01 -9.32619929e-01 -4.40670431e-01
4.74418819e-01 3.14899713e-01 -7.94534266e-01 3.13139528e-01
-1.48855969e-01 4.82288152e-01 -3.85509312e-01 8.91563654e-01
9.52107668e-01 -2.46026784e-01 -2.41707489e-01 -5.95808983e-01
2.77166790e-03 1.32998198e-01 -1.11853445e+00 2.06112838e+00
-2.47414023e-01 6.06454968e-01 4.30989675e-02 -8.80090833e-01
6.27515852e-01 1.21652365e-01 5.41764140e-01 -3.62936780e-02
2.19368383e-01 1.43023282e-01 -3.65811199e-01 -6.26651704e-01
3.98874342e-01 9.39460248e-02 -2.18100503e-01 1.49414405e-01
3.52202743e-01 -5.61689019e-01 4.45006520e-01 1.82987452e-01
9.30167794e-01 5.95295906e-01 2.48127744e-01 -8.35035369e-02
6.01412833e-01 -3.90814506e-02 3.57100695e-01 2.58817375e-01
-2.87923515e-01 8.81930232e-01 3.19149226e-01 -1.01544428e+00
-1.09077096e+00 -1.15154469e+00 -6.43827915e-02 9.76263165e-01
3.16335052e-01 -3.60601187e-01 -6.14159763e-01 -1.01775801e+00
2.64048219e-01 2.08141342e-01 -6.32955074e-01 2.55787652e-02
-8.09850693e-01 -2.52517700e-01 3.42430510e-02 5.17211020e-01
4.12758499e-01 -4.81844038e-01 -7.54940629e-01 -1.87985376e-01
-2.95015246e-01 -1.78013456e+00 -7.82131016e-01 1.28514152e-02
-8.77002954e-01 -1.35237980e+00 -6.61675870e-01 -6.50787234e-01
7.91448534e-01 8.12503695e-01 1.66949904e+00 -2.45429471e-01
-2.95562148e-01 7.29129791e-01 -1.30733013e-01 -2.79826403e-01
-2.85483837e-01 -1.42417941e-02 3.11039656e-01 1.25981197e-01
4.16809022e-01 -2.78283179e-01 -7.57318974e-01 8.22819710e-01
-4.06475931e-01 -7.58406073e-02 6.98412597e-01 6.59912050e-01
5.59795976e-01 -4.80127811e-01 2.27269121e-02 -7.23443210e-01
-4.74389166e-01 -1.12883210e-01 -9.10633206e-01 9.60599855e-02
1.56875208e-01 -2.28258938e-01 1.72933385e-01 -2.34095767e-01
-9.15927768e-01 7.48189569e-01 1.45134836e-01 -9.50584054e-01
-4.32286292e-01 -2.34885082e-01 -3.10995877e-01 -4.68373090e-01
5.74844956e-01 3.08887120e-02 -1.27787385e-02 -4.37385529e-01
4.48743075e-01 2.68331289e-01 6.04096174e-01 -3.90654057e-01
1.17805636e+00 7.23571062e-01 1.89563408e-01 -7.78355122e-01
-1.07724440e+00 -9.05386806e-01 -1.16890562e+00 -4.32010591e-01
8.65143061e-01 -1.51541209e+00 -8.01599503e-01 4.09131259e-01
-1.19643021e+00 5.83899207e-02 1.61017105e-01 7.02601016e-01
-1.02354610e+00 3.01994383e-01 -3.25604081e-01 -4.53727275e-01
6.49320185e-02 -1.26448536e+00 1.80045664e+00 -2.33813256e-01
-1.29621223e-01 -6.67277038e-01 -6.26069307e-02 9.12110388e-01
-7.68809766e-02 4.53826249e-01 2.31256351e-01 -2.92491943e-01
-1.19710422e+00 -3.62301618e-01 1.21085681e-01 2.35702813e-01
-4.62118126e-02 -1.75257749e-03 -1.04944956e+00 -4.95860457e-01
-2.66910464e-01 -6.24442995e-01 8.94078135e-01 5.13230622e-01
9.32520926e-01 2.83658970e-02 -6.07110381e-01 6.52767003e-01
1.04161024e+00 -1.62891716e-01 3.09217602e-01 9.03040841e-02
8.61948669e-01 5.62638998e-01 1.04416633e+00 4.08920437e-01
4.27521169e-01 1.07872713e+00 7.89902389e-01 1.34284168e-01
3.83981988e-02 -3.30952287e-01 3.23255360e-01 5.83395362e-01
-3.11155468e-01 5.69290854e-02 -7.01680183e-01 4.67403620e-01
-1.57211435e+00 -8.69706452e-01 1.21742763e-01 2.16531658e+00
3.13800395e-01 3.60215843e-01 3.87749940e-01 -3.67273167e-02
4.71008897e-01 2.68101931e-01 -4.87776279e-01 2.86402643e-01
3.02655786e-01 -4.35978681e-01 7.58180499e-01 3.39272469e-01
-1.61859369e+00 8.10784400e-01 6.47985029e+00 5.75526774e-01
-8.60620618e-01 6.66398928e-02 5.51405191e-01 -2.91724473e-01
-3.22207063e-02 -1.43600523e-01 -1.23487926e+00 9.36041474e-02
5.03437340e-01 3.86748642e-01 1.79371163e-01 1.12673998e+00
7.13533983e-02 -1.09607048e-01 -1.58270133e+00 1.30649972e+00
3.85791212e-01 -1.38392401e+00 -6.57816455e-02 1.74382553e-02
9.51047003e-01 1.83161810e-01 1.11140780e-01 1.54453158e-01
3.49047966e-02 -7.60489941e-01 1.13034952e+00 2.63926029e-01
6.66166902e-01 -7.66594827e-01 6.26349747e-01 4.76755083e-01
-1.38542998e+00 1.58604458e-02 -3.83114904e-01 2.25440130e-01
3.12412202e-01 4.04801458e-01 -8.87985229e-01 6.48450315e-01
9.00416493e-01 1.12699413e+00 -7.22719967e-01 1.01620758e+00
8.20618719e-02 1.33533066e-03 -3.11205596e-01 8.26707184e-02
2.62768477e-01 4.93379161e-02 7.15491176e-01 1.09516025e+00
4.05585319e-01 -2.64982849e-01 3.11388314e-01 3.87568116e-01
-1.25027508e-01 -2.88439900e-01 -9.73809481e-01 3.12646627e-01
2.38929778e-01 1.34859395e+00 -7.74750769e-01 -1.92616120e-01
-7.50501394e-01 6.40838921e-01 1.84542999e-01 2.25031018e-01
-9.86576378e-01 1.98215574e-01 8.88941824e-01 5.05915940e-01
8.42252195e-01 -4.39792663e-01 6.76302254e-01 -1.29874134e+00
2.02585444e-01 -8.04006398e-01 5.51276170e-02 -8.16186965e-01
-9.74785388e-01 2.83735871e-01 6.03994846e-01 -1.71294498e+00
-4.64977980e-01 -9.63694930e-01 -1.90305650e-01 1.60244852e-01
-1.39648080e+00 -1.32272577e+00 -4.87980068e-01 4.53916907e-01
8.71432006e-01 -1.19962238e-01 3.43873978e-01 3.10770631e-01
-6.46960735e-02 4.37580198e-01 -1.50967360e-01 4.41559218e-02
8.31246078e-01 -1.04818916e+00 7.13275135e-01 5.30676842e-01
5.02044082e-01 3.66196126e-01 7.40903676e-01 -1.71712235e-01
-2.01795149e+00 -1.04445934e+00 7.42181063e-01 -1.22057164e+00
3.14868867e-01 -8.87206197e-01 -3.50299299e-01 9.30854559e-01
-1.34606212e-01 6.80962265e-01 1.87217265e-01 -1.32300854e-01
-4.07159179e-01 -2.94540375e-01 -9.08892930e-01 3.14035475e-01
1.51191163e+00 -3.76828909e-01 -5.79993367e-01 5.94140232e-01
6.17915630e-01 -1.18898177e+00 -5.73665738e-01 4.02731568e-01
7.32552052e-01 -1.06444776e+00 1.52956021e+00 -6.67909980e-01
3.20162416e-01 -5.10561347e-01 -3.71511489e-01 -1.14859104e+00
-1.70476511e-01 -8.72590661e-01 -3.55877429e-01 6.66445673e-01
1.26280233e-01 -1.20135546e-01 1.06960630e+00 1.21325396e-01
-2.10645616e-01 -7.30752051e-01 -1.11111367e+00 -8.30935419e-01
-2.87826955e-01 -3.09972137e-01 3.88405979e-01 2.76780069e-01
-6.64751709e-01 3.66760612e-01 -5.56906044e-01 6.16057515e-02
7.91551232e-01 2.99496651e-01 1.48710346e+00 -1.10355592e+00
-1.50884002e-01 5.76845892e-02 -1.01788270e+00 -1.66062188e+00
2.63839543e-01 -5.87388635e-01 3.30799669e-02 -9.89849865e-01
3.44347745e-01 9.05393511e-02 2.37902105e-01 -1.63703844e-01
9.61017609e-02 4.37132746e-01 2.54800081e-01 -1.78635940e-02
-1.16833055e+00 4.48928803e-01 1.53594148e+00 -3.96061279e-02
2.80806780e-01 2.89685875e-01 -2.17609197e-01 9.18953180e-01
1.19225517e-01 -3.80171210e-01 -4.25028116e-01 -3.04505527e-01
1.83684230e-01 8.92044008e-02 5.94583213e-01 -9.53386843e-01
-4.74170819e-02 1.03548486e-02 6.42964900e-01 -1.29333007e+00
8.61567318e-01 -1.11938596e+00 2.00021937e-01 4.18009520e-01
-4.54320246e-03 2.35668659e-01 1.32921087e-02 9.06769156e-01
-2.22715601e-01 3.19718271e-02 6.33433044e-01 -4.15081084e-01
-9.86572921e-01 5.51159322e-01 4.60706428e-02 3.66450518e-01
1.25048351e+00 -3.58766109e-01 -9.29926112e-02 -4.77090746e-01
-6.73337042e-01 2.88123429e-01 5.63079119e-01 7.12005675e-01
6.12873852e-01 -1.58666968e+00 -7.74704874e-01 3.28196049e-01
4.66137648e-01 3.16297233e-01 1.67605013e-01 9.68323290e-01
-7.24819899e-01 6.49640679e-01 -1.33259222e-01 -1.31350136e+00
-1.50814569e+00 6.33826494e-01 2.63968855e-01 -1.29253000e-01
-5.26851356e-01 9.64960217e-01 5.56567788e-01 -5.52569807e-01
2.53376245e-01 -6.35643780e-01 1.09007187e-01 1.22684591e-01
4.77388471e-01 3.25986862e-01 3.46954048e-01 -1.12914932e+00
-6.47085667e-01 1.03611267e+00 -1.18422985e-01 1.99355617e-01
1.28140187e+00 -2.24374890e-01 4.34786409e-01 3.19110841e-01
1.48091114e+00 -1.33380711e-01 -1.69452798e+00 -2.27686599e-01
-1.71158239e-01 -1.06368887e+00 -2.91803658e-01 -9.99581069e-02
-1.07189119e+00 8.11848521e-01 5.11912525e-01 -3.61572444e-01
5.75921774e-01 3.52351755e-01 6.86561048e-01 8.13612938e-01
6.98017478e-01 -9.68188167e-01 5.50046325e-01 6.07924700e-01
9.98042583e-01 -1.55413425e+00 2.60152042e-01 -5.92276633e-01
-5.72174609e-01 1.10767233e+00 5.95739543e-01 -2.76982397e-01
7.20340610e-01 1.06628060e-01 -2.95104627e-02 -1.72154471e-01
-9.41404283e-01 1.48781892e-02 4.90590483e-01 6.78115785e-01
2.50557274e-01 -1.14186719e-01 5.13324916e-01 -5.67597970e-02
-1.68837875e-01 -2.58388728e-01 2.86614418e-01 9.38159466e-01
-3.30406815e-01 -6.69657826e-01 -5.31482577e-01 2.09723592e-01
-3.25882077e-01 3.86612624e-01 -2.53387719e-01 1.03020966e+00
1.44465029e-01 5.97675502e-01 6.53233528e-02 -3.53260726e-01
5.72331131e-01 1.83678772e-02 9.80592966e-01 -5.30805051e-01
-9.00428295e-02 2.07690284e-01 2.55487114e-01 -1.11631608e+00
-9.31925237e-01 -1.11778057e+00 -6.34192288e-01 -1.63753226e-01
-4.43336755e-01 -2.96918005e-01 3.65577012e-01 8.72232020e-01
2.05748215e-01 8.74986351e-02 7.08421469e-01 -1.79095733e+00
-6.39578164e-01 -8.41198385e-01 -4.07444745e-01 5.78969419e-01
6.73043370e-01 -1.32809246e+00 -4.38998759e-01 3.25425804e-01] | [7.590003490447998, -2.6413393020629883] |
67d47309-6e2e-412a-94c8-fe37d360bca3 | dropout-training-of-matrix-factorization-and | 1512.04483 | null | http://arxiv.org/abs/1512.04483v1 | http://arxiv.org/pdf/1512.04483v1.pdf | Dropout Training of Matrix Factorization and Autoencoder for Link Prediction in Sparse Graphs | Matrix factorization (MF) and Autoencoder (AE) are among the most successful
approaches of unsupervised learning. While MF based models have been
extensively exploited in the graph modeling and link prediction literature, the
AE family has not gained much attention. In this paper we investigate both MF
and AE's application to the link prediction problem in sparse graphs. We show
the connection between AE and MF from the perspective of multiview learning,
and further propose MF+AE: a model training MF and AE jointly with shared
parameters. We apply dropout to training both the MF and AE parts, and show
that it can significantly prevent overfitting by acting as an adaptive
regularization. We conduct experiments on six real world sparse graph datasets,
and show that MF+AE consistently outperforms the competing methods, especially
on datasets that demonstrate strong non-cohesive structures. | ['Zhongfei Zhang', 'Shuangfei Zhai'] | 2015-12-14 | null | null | null | null | ['multiview-learning'] | ['computer-vision'] | [-1.33092687e-01 3.14310223e-01 -4.71352428e-01 -2.03040078e-01
-1.19525477e-01 -1.55838534e-01 4.84007120e-01 3.19861710e-01
6.68949336e-02 5.67240596e-01 4.54802692e-01 -2.02701345e-01
-4.31341618e-01 -8.39152753e-01 -8.65386426e-01 -5.12873590e-01
-4.44382310e-01 4.88758236e-01 -4.13363501e-02 -8.24993476e-02
-2.82737672e-01 1.93243161e-01 -1.15660226e+00 5.09617664e-02
1.03921235e+00 5.24367571e-01 -6.63505867e-02 2.96313792e-01
2.52669416e-02 1.10904479e+00 -2.96011657e-01 -7.91662693e-01
5.95712736e-02 -2.23806471e-01 -7.86972940e-01 2.41255075e-01
6.01107895e-01 -1.47503614e-01 -7.69443154e-01 9.24306870e-01
2.18365982e-01 7.57910088e-02 6.75546169e-01 -1.35664403e+00
-7.61935890e-01 1.05723894e+00 -8.05821896e-01 1.36690438e-01
2.89732605e-01 -5.51230371e-01 1.52324843e+00 -8.92335236e-01
9.18879569e-01 1.14892375e+00 9.07566249e-01 1.34730592e-01
-1.69080067e+00 -4.96620715e-01 4.33383167e-01 3.36669058e-01
-1.46333683e+00 -1.87026888e-01 1.06925309e+00 -5.65741718e-01
8.94159675e-01 -2.81566530e-01 6.85787559e-01 1.19503212e+00
1.28931865e-01 1.04941773e+00 7.47982681e-01 -3.20620447e-01
9.62173939e-02 -6.03140779e-02 1.88360944e-01 1.02924597e+00
3.81235182e-01 -4.18968648e-02 -7.22008169e-01 -3.47979128e-01
8.21399510e-01 -4.94345324e-04 -3.05574358e-01 -7.14903235e-01
-9.21933293e-01 1.28361642e+00 6.89986169e-01 2.66785532e-01
-3.50526750e-01 2.29625702e-01 2.36610919e-01 5.19755840e-01
7.11515307e-01 6.94306986e-03 -1.53084934e-01 4.29794490e-01
-7.07865059e-01 4.69845496e-02 9.68653023e-01 9.00380254e-01
9.27148700e-01 2.64878094e-01 1.69117063e-01 9.53064799e-01
3.79782170e-01 -4.24369937e-03 7.95149356e-02 -7.80452013e-01
4.37507302e-01 6.25995815e-01 -4.15283710e-01 -1.49223518e+00
-3.48465234e-01 -8.51011395e-01 -1.12282360e+00 -1.67339414e-01
7.43884519e-02 -2.21134022e-01 -5.04008055e-01 1.94047105e+00
1.27119809e-01 6.91958547e-01 -1.67001024e-01 6.07332766e-01
1.12405682e+00 5.02281368e-01 1.18027940e-01 -2.26728439e-01
6.54201627e-01 -1.01875818e+00 -7.56753743e-01 -1.79531619e-01
8.13399255e-01 -3.50248694e-01 6.05256200e-01 2.60409206e-01
-9.75240707e-01 -3.99234056e-01 -9.99784291e-01 7.03649372e-02
-8.17174539e-02 1.31280497e-01 1.13604796e+00 1.75953731e-01
-1.17684865e+00 8.21872532e-01 -6.86679840e-01 -4.67260033e-01
5.66557765e-01 4.23416764e-01 -6.74149036e-01 -2.14466006e-01
-1.13871253e+00 5.79302847e-01 1.30912155e-01 -4.85189492e-03
-8.22456002e-01 -5.83169580e-01 -1.04100072e+00 3.79057527e-01
6.00661874e-01 -1.11692262e+00 5.85687220e-01 -1.01860714e+00
-1.12972426e+00 4.59091038e-01 -1.46743506e-01 -7.23533273e-01
4.99210544e-02 -3.61494124e-01 -4.78821993e-01 1.80111483e-01
-1.51466817e-01 5.11026680e-01 1.14161563e+00 -1.36224389e+00
-1.25482678e-01 -1.41747653e-01 1.37874022e-01 -3.88576686e-02
-8.04992199e-01 -2.79386610e-01 -2.83526361e-01 -8.74038219e-01
3.50541100e-02 -7.34827220e-01 -3.08315545e-01 -1.73342690e-01
-5.36275327e-01 -3.98515552e-01 8.05196285e-01 -5.43888330e-01
1.57416666e+00 -1.92201412e+00 7.06260204e-01 5.16248703e-01
5.93755186e-01 7.47781247e-02 -3.61406088e-01 7.30941653e-01
-3.24310929e-01 -1.15626361e-02 -2.78182924e-01 -5.95599651e-01
-1.82378128e-01 5.54498971e-01 -1.55813922e-03 4.37531382e-01
2.21860275e-01 8.58142853e-01 -7.67199814e-01 -4.36524242e-01
1.11629264e-02 6.61835909e-01 -9.71074939e-01 2.59154379e-01
-1.48766905e-01 1.94298923e-01 -3.80496055e-01 5.20106971e-01
4.15621817e-01 -8.02081764e-01 5.91615021e-01 -5.40749729e-01
4.18134689e-01 -1.55661730e-02 -1.30378377e+00 1.70276606e+00
-1.15504518e-01 5.65595686e-01 4.11843359e-01 -1.38159275e+00
7.97514260e-01 3.23310524e-01 6.55427098e-01 -3.37813377e-01
6.18664827e-03 -2.48060133e-02 1.51349425e-01 -3.99285316e-01
2.23917365e-01 -2.27783713e-02 4.43143398e-01 5.01690924e-01
6.40029013e-01 4.25024211e-01 3.67061257e-01 8.34911823e-01
1.24076128e+00 -4.22536582e-02 1.41752645e-01 -3.61046195e-01
5.36802649e-01 -5.37242770e-01 5.67520618e-01 5.93158841e-01
1.37848765e-01 3.48098904e-01 6.33405864e-01 -3.45821977e-01
-7.73972929e-01 -8.71022522e-01 1.59790248e-01 9.74176288e-01
-1.11476012e-01 -1.03113878e+00 -4.22904968e-01 -9.37502801e-01
2.69114822e-01 2.68707514e-01 -6.27716362e-01 -2.25193471e-01
-4.59536016e-01 -6.73007846e-01 7.56962597e-02 6.59007549e-01
1.94311649e-01 -6.36178672e-01 3.39228392e-01 2.44358808e-01
-3.12084462e-02 -1.06178093e+00 -2.56768465e-01 2.79478967e-01
-9.38315630e-01 -1.13631117e+00 -3.44809085e-01 -8.66495609e-01
7.47877300e-01 2.90609419e-01 1.43236244e+00 3.91208202e-01
-1.13098398e-01 7.89388180e-01 -3.95372063e-01 1.00395590e-01
-1.95097193e-01 2.25822136e-01 1.97563946e-01 3.19892466e-01
1.32405441e-02 -1.25722218e+00 -3.33348542e-01 3.00016860e-03
-8.54668021e-01 -2.38789413e-02 4.86771166e-01 1.08807778e+00
3.16226155e-01 -1.38712406e-01 7.49594033e-01 -1.31303465e+00
7.96627760e-01 -7.48573959e-01 -3.96892726e-01 8.23056549e-02
-7.98865914e-01 5.32820299e-02 4.82318193e-01 -1.83574647e-01
-7.41145849e-01 2.13604346e-02 -1.13231383e-01 -7.38059878e-01
1.70396760e-01 1.09215546e+00 7.87649676e-03 -3.40086877e-01
3.46009880e-01 -1.86654642e-01 8.44988376e-02 -5.63139319e-01
4.47946697e-01 6.47350401e-02 2.33864099e-01 -5.15575707e-01
9.60059047e-01 3.91504824e-01 2.09427312e-01 -7.59109139e-01
-7.76784778e-01 -4.23883259e-01 -4.77325439e-01 -2.62313783e-01
6.01732016e-01 -1.19687676e+00 -5.89135706e-01 -1.64628342e-01
-9.34435189e-01 -9.56585780e-02 -1.42833203e-01 5.09895146e-01
-4.32772785e-01 5.20385444e-01 -9.02460456e-01 -4.35319185e-01
-1.41976863e-01 -8.38056087e-01 6.46692634e-01 2.95000076e-02
-2.01105908e-01 -1.48982525e+00 2.36523941e-01 1.89286605e-01
2.48597831e-01 1.73391208e-01 1.19909096e+00 -6.51999831e-01
-5.16703904e-01 -1.04903199e-01 -3.85802127e-02 3.09976578e-01
-1.71613153e-02 1.17756262e-01 -6.62484288e-01 -4.88926619e-01
-3.97797525e-01 -4.12495255e-01 1.26293325e+00 5.01588762e-01
1.01529062e+00 -2.28871644e-01 -4.55003202e-01 6.84691012e-01
1.60070550e+00 -3.93586934e-01 3.91702801e-01 5.86527865e-03
1.04410207e+00 5.09339452e-01 -6.15731301e-03 3.93972814e-01
5.05926549e-01 5.06642461e-01 6.30059838e-01 -1.76057398e-01
-1.44616351e-01 -4.27566886e-01 3.43515396e-01 1.41553116e+00
-2.58763462e-01 -3.94223213e-01 -6.87689662e-01 5.04251301e-01
-2.22588348e+00 -9.99875903e-01 -2.48305663e-01 1.81728768e+00
4.96455997e-01 5.45818184e-04 3.10296386e-01 3.74372527e-02
6.93690956e-01 3.91198307e-01 -5.09868450e-02 -2.91110396e-01
-3.49439114e-01 3.09689760e-01 5.86423352e-02 6.76062822e-01
-1.27833688e+00 8.47761333e-01 6.78664160e+00 6.90189302e-01
-5.89070439e-01 -1.33746676e-02 4.65017468e-01 2.04123631e-01
-6.27515137e-01 1.51497319e-01 -4.22688544e-01 6.08576611e-02
7.22401023e-01 6.40981048e-02 4.07823741e-01 7.91082859e-01
-1.22238614e-01 2.26028576e-01 -1.15857875e+00 8.85231018e-01
9.32386369e-02 -1.58784270e+00 2.87847072e-01 -3.82903926e-02
1.01635051e+00 2.10484415e-01 -5.37242703e-02 3.41595739e-01
5.28744400e-01 -1.14044118e+00 2.30861306e-01 7.02935457e-01
6.33823127e-02 -8.25729549e-01 6.82089329e-01 2.17064589e-01
-1.31402647e+00 -1.02863230e-01 -4.08061177e-01 -9.33782831e-02
2.20362395e-01 9.01716888e-01 -3.06761652e-01 9.36862051e-01
5.96444249e-01 1.48508394e+00 -7.21313357e-01 9.99916971e-01
-2.79468089e-01 9.19248104e-01 -2.53548682e-01 2.11380050e-01
9.52825397e-02 -5.78854382e-01 7.02369809e-01 1.01472569e+00
1.92296191e-03 -1.98484108e-01 4.12613541e-01 8.94990265e-01
-2.85512060e-01 1.39811397e-01 -8.09250772e-01 -2.80813664e-01
3.45433116e-01 1.30870879e+00 -4.33298886e-01 -1.83967859e-01
-1.02500057e+00 7.44682610e-01 9.32995141e-01 5.93524456e-01
-5.34931064e-01 1.24260597e-02 5.11263788e-01 6.86339512e-02
5.72682917e-01 -2.88759172e-01 1.96190178e-01 -1.43979394e+00
-1.53157905e-01 -8.53174746e-01 7.37131953e-01 -3.46048504e-01
-1.76804137e+00 4.01660025e-01 -1.91059515e-01 -1.02400684e+00
-1.82364345e-01 -5.32785952e-01 -6.87924802e-01 4.55584198e-01
-1.54531991e+00 -1.27006602e+00 -2.14449972e-01 8.13902020e-01
1.00988783e-01 -2.72579759e-01 9.98237729e-01 5.94258964e-01
-8.51496935e-01 4.43817317e-01 2.25294814e-01 1.45912617e-01
5.91714382e-01 -1.21369028e+00 9.97584909e-02 6.76678061e-01
7.81936288e-01 7.70532489e-01 6.03571594e-01 -7.24901080e-01
-1.51346195e+00 -1.05142188e+00 7.07565188e-01 -1.74467489e-02
9.81689274e-01 -2.98977733e-01 -1.00018382e+00 1.03093231e+00
5.06297410e-01 3.43748510e-01 8.85389030e-01 5.54790020e-01
-5.38016796e-01 -5.03688827e-02 -5.89426517e-01 4.52167779e-01
1.25475466e+00 -6.54223382e-01 -2.03102842e-01 2.91452438e-01
4.44821805e-01 2.16000870e-01 -1.29849243e+00 5.28302073e-01
2.55149066e-01 -1.19230509e+00 1.03616941e+00 -6.83930695e-01
5.26881158e-01 -1.27781138e-01 -9.90362540e-02 -1.53510690e+00
-7.30982065e-01 -6.07854486e-01 -8.50323260e-01 1.41996181e+00
3.22801590e-01 -4.73143846e-01 9.51495588e-01 -1.48015216e-01
-4.84267157e-03 -9.20380294e-01 -7.33442008e-01 -6.50924861e-01
-1.78324003e-02 -1.55474305e-01 7.37982839e-02 1.28827226e+00
-5.84547222e-02 7.59016395e-01 -6.39695108e-01 7.13427784e-03
7.46767521e-01 2.87268966e-01 7.70004630e-01 -1.54904616e+00
-6.84148967e-01 -3.46680254e-01 -2.57836342e-01 -9.47933197e-01
5.97673953e-01 -1.35813689e+00 -5.15577435e-01 -1.59424913e+00
3.14264297e-01 -1.17732048e-01 -2.06241950e-01 3.14188957e-01
-2.34339461e-01 4.03575562e-02 9.83551610e-03 1.80484176e-01
-7.06585586e-01 9.09770250e-01 9.15246487e-01 -3.48154932e-01
9.22866538e-02 -6.03594966e-02 -5.61400115e-01 7.65652776e-01
4.97866124e-01 -3.16348463e-01 -6.17125988e-01 -5.28873146e-01
6.07505441e-01 -1.23278186e-01 3.38061959e-01 -8.11326325e-01
3.53060752e-01 2.90327400e-01 4.15045053e-01 -3.19763452e-01
6.46992102e-02 -9.37851489e-01 1.93316460e-01 3.66423607e-01
-1.92490980e-01 8.61914754e-02 -8.57829601e-02 1.12655377e+00
-4.35683757e-01 -1.01811945e-01 2.41879731e-01 5.91391027e-02
-6.08148277e-01 6.76651299e-01 -2.52990007e-01 -1.06470406e-01
5.65640569e-01 8.86540785e-02 -1.77726477e-01 -7.78632224e-01
-1.21817327e+00 4.10140187e-01 1.20319128e-01 2.26889312e-01
7.04517186e-01 -1.56745529e+00 -7.22453773e-01 1.82857007e-01
1.86653938e-02 -3.22063863e-01 9.53819603e-02 1.09583533e+00
-9.84253734e-02 3.09965108e-02 1.54531166e-01 -4.82050627e-01
-1.08184910e+00 7.74788022e-01 6.11412451e-02 -4.43470538e-01
-7.66267478e-01 8.83405805e-01 7.72562847e-02 -2.58090317e-01
4.52750266e-01 2.85359651e-01 -4.38798368e-01 1.77854970e-01
2.54318975e-02 3.28243881e-01 -2.70328194e-01 -6.34549856e-01
-1.57148302e-01 4.86108243e-01 -1.26248613e-01 3.96501899e-01
1.80100322e+00 -1.25282705e-01 -3.71238887e-01 3.69171202e-01
1.21386790e+00 5.89760616e-02 -1.05976176e+00 -4.63255733e-01
2.93592185e-01 -7.25475699e-02 6.19744174e-02 -2.00779945e-01
-1.39816523e+00 7.49564350e-01 1.36590824e-01 4.61150676e-01
9.27466452e-01 2.03255624e-01 5.08966327e-01 2.20330790e-01
2.66955793e-01 -8.05530846e-01 3.20827693e-01 5.81016660e-01
7.48271644e-01 -1.18324685e+00 2.77139544e-01 -8.78634036e-01
-3.02136064e-01 1.13886285e+00 5.86820960e-01 -6.26043677e-01
9.75145459e-01 2.81332612e-01 -4.55975115e-01 -4.04595524e-01
-9.44512784e-01 -3.20414305e-01 5.44211149e-01 5.92305303e-01
6.01469457e-01 -1.60296842e-01 -3.95251542e-01 6.79644883e-01
1.79970995e-01 -2.20597357e-01 2.67821014e-01 7.90463746e-01
-1.55353993e-01 -1.36956823e+00 1.89641029e-01 7.47974932e-01
-3.24669302e-01 -1.94596246e-01 -6.09896600e-01 8.17428052e-01
-2.02431306e-02 7.20607102e-01 -6.99461922e-02 -5.90833604e-01
-2.04033423e-02 6.58634752e-02 6.09903634e-01 -5.73216319e-01
-6.81400299e-01 2.22724646e-01 3.56064647e-01 -6.08825922e-01
-9.26517844e-01 -4.92445230e-01 -8.33995104e-01 -4.59977388e-01
-5.81425190e-01 3.19566995e-01 1.92341078e-02 8.26590657e-01
4.01077628e-01 6.34286404e-01 6.00613534e-01 -5.94695032e-01
-3.42895746e-01 -9.50470328e-01 -7.40200520e-01 5.56658030e-01
2.57416755e-01 -7.93266118e-01 -5.06462455e-01 -1.24592192e-01] | [7.109405040740967, 6.1716461181640625] |
fffe11ee-1acc-4491-9354-b6005faa6db3 | detecting-stance-of-authorities-towards | 2301.05863 | null | https://arxiv.org/abs/2301.05863v1 | https://arxiv.org/pdf/2301.05863v1.pdf | Detecting Stance of Authorities towards Rumors in Arabic Tweets: A Preliminary Study | A myriad of studies addressed the problem of rumor verification in Twitter by either utilizing evidence from the propagation networks or external evidence from the Web. However, none of these studies exploited evidence from trusted authorities. In this paper, we define the task of detecting the stance of authorities towards rumors in tweets, i.e., whether a tweet from an authority agrees, disagrees, or is unrelated to the rumor. We believe the task is useful to augment the sources of evidence utilized by existing rumor verification systems. We construct and release the first Authority STance towards Rumors (AuSTR) dataset, where evidence is retrieved from authority timelines in Arabic Twitter. Due to the relatively limited size of our dataset, we study the usefulness of existing datasets for stance detection in our task. We show that existing datasets are somewhat useful for the task; however, they are clearly insufficient, which motivates the need to augment them with annotated data constituting stance of authorities from Twitter. | ['Tamer Elsayed', 'Fatima Haouari'] | 2023-01-14 | null | null | null | null | ['stance-detection'] | ['natural-language-processing'] | [-2.30277002e-01 3.69730562e-01 -4.90174204e-01 -3.83890212e-01
-4.66566145e-01 -8.77985477e-01 1.26778853e+00 5.93692124e-01
-4.00310338e-01 9.22296047e-01 6.88655794e-01 -5.07155597e-01
3.34433585e-01 -7.03350246e-01 -4.65489775e-01 -3.75773787e-01
2.42141277e-01 5.46944678e-01 6.24119818e-01 -8.38649511e-01
7.49031365e-01 1.89325824e-01 -1.23251688e+00 4.23429370e-01
6.34135127e-01 6.61729991e-01 -6.01504028e-01 2.47055709e-01
9.66284126e-02 1.27339315e+00 -6.17277920e-01 -5.48365772e-01
-2.48300582e-02 -2.95524955e-01 -1.28911281e+00 7.21279979e-02
3.89610052e-01 -4.50461060e-01 -2.17911780e-01 1.07131112e+00
1.46858707e-01 -5.35998940e-01 4.21388507e-01 -1.42716634e+00
-3.73720050e-01 1.33886993e+00 -3.31829101e-01 5.59503317e-01
6.37710333e-01 -2.36466736e-01 1.09796107e+00 -8.27482104e-01
8.54321301e-01 9.00736570e-01 8.26276898e-01 1.75813451e-01
-7.78479576e-01 -5.24443030e-01 -7.48483390e-02 2.21550494e-01
-1.02749789e+00 -7.60241151e-01 7.02793002e-01 -5.63703239e-01
3.71802777e-01 3.21230263e-01 4.23395157e-01 1.31455755e+00
-1.35879859e-01 7.52549827e-01 1.88842928e+00 -3.85370016e-01
6.47748187e-02 5.97004294e-01 6.00889266e-01 7.07539499e-01
4.37312245e-01 -3.15638244e-01 -9.10224557e-01 -8.06974471e-01
2.97163844e-01 -5.96069872e-01 -4.34433162e-01 2.28322059e-01
-1.28379107e+00 1.11274028e+00 4.06131953e-01 3.87246221e-01
-3.79291058e-01 -4.24021751e-01 5.96511900e-01 7.44870424e-01
9.69279587e-01 4.65914577e-01 -2.86536783e-01 -4.81868237e-02
-1.21489048e+00 4.47860628e-01 1.51280153e+00 6.94033861e-01
5.72988987e-01 -2.36360505e-01 4.75908548e-01 4.01421547e-01
5.81340015e-01 4.16815013e-01 4.12255675e-01 -6.27615392e-01
5.16259313e-01 6.40923500e-01 6.30524933e-01 -1.35608137e+00
-4.94027466e-01 -1.11975014e-01 -3.25243056e-01 -1.50425360e-01
7.69027352e-01 -1.20094977e-01 -3.28253627e-01 1.29805028e+00
6.22788668e-01 -1.69458538e-01 -5.68953529e-02 1.23782158e+00
9.42951500e-01 1.31171390e-01 -3.83974195e-01 -3.11103851e-01
1.61556554e+00 -6.04762673e-01 -9.56404328e-01 2.38887239e-02
5.03907084e-01 -1.13109624e+00 6.14823580e-01 3.16082597e-01
-1.11445343e+00 2.34970227e-01 -1.05272615e+00 2.05741182e-01
-4.76449847e-01 -4.27994281e-01 4.48037714e-01 5.68711698e-01
-7.83869863e-01 4.42876488e-01 -5.90002000e-01 -3.00010115e-01
6.51053041e-02 -1.66334078e-01 -2.18378633e-01 1.27161309e-01
-2.00530100e+00 1.14113379e+00 1.35794967e-01 -1.65516347e-01
-1.04503918e+00 -5.26392795e-02 -6.93866849e-01 -4.65392768e-01
6.08283401e-01 -3.21436465e-01 1.49556148e+00 -9.39799070e-01
-1.11654723e+00 1.07631540e+00 -1.07404981e-02 -8.10702264e-01
9.57503676e-01 -1.70287192e-01 -4.81949925e-01 1.90018818e-01
5.35699725e-01 -2.07282245e-01 1.10706949e+00 -1.32668459e+00
-5.67694724e-01 -5.93207061e-01 1.20622806e-01 -1.27894267e-01
-2.06911832e-01 7.21121550e-01 9.18402523e-02 -4.22495872e-01
3.27075869e-01 -1.12341428e+00 1.15077645e-01 -8.78608048e-01
-8.44044566e-01 -3.75606298e-01 8.84123206e-01 -7.42460310e-01
1.35309732e+00 -1.54198444e+00 -4.40147042e-01 3.38300288e-01
5.65116942e-01 2.04507485e-01 5.02619505e-01 8.50633621e-01
2.06293672e-01 6.86526775e-01 1.14298992e-01 1.40935732e-02
4.06620204e-02 1.32578507e-01 -8.09961438e-01 9.37747180e-01
-3.42373922e-02 6.05586648e-01 -1.07976902e+00 -6.41030967e-01
-5.55987895e-01 1.37195989e-01 -1.55318016e-02 -1.27220273e-01
-1.64033726e-01 4.21008140e-01 -6.94715559e-01 7.22218871e-01
5.07281303e-01 -4.29381460e-01 3.18066984e-01 -1.00687183e-01
-5.00337243e-01 1.25286734e+00 -8.45311522e-01 6.15176141e-01
-3.24381888e-02 7.74072647e-01 5.10536313e-01 -2.65514791e-01
7.58313715e-01 7.93558180e-01 1.94220245e-01 -2.34455302e-01
3.89056206e-01 6.33730650e-01 -1.14982121e-01 -4.08601880e-01
9.07666028e-01 -2.34950066e-01 -2.42474467e-01 1.28591037e+00
-5.07532179e-01 2.09015980e-02 1.33532360e-01 5.52762508e-01
9.93959844e-01 -3.51420909e-01 5.77540874e-01 -3.27816188e-01
5.16217411e-01 4.19027269e-01 2.12719828e-01 8.16574872e-01
-4.11994517e-01 2.14902729e-01 6.11738026e-01 -4.19880748e-01
-1.13839614e+00 -5.05622387e-01 -2.76033312e-01 1.19448400e+00
-7.16462210e-02 -5.92636406e-01 -6.80449665e-01 -9.89074171e-01
1.23157976e-02 4.25798684e-01 -6.79629743e-01 5.36921859e-01
-6.38220608e-01 -6.86393797e-01 9.84162867e-01 5.43840490e-02
5.32219291e-01 -6.35534108e-01 -4.41157609e-01 3.05477947e-01
-7.92593539e-01 -1.20175004e+00 -2.06982717e-01 -1.50665328e-01
-7.67590702e-01 -1.43899763e+00 -1.81895614e-01 -1.60428092e-01
4.87333119e-01 4.27297920e-01 1.26075876e+00 5.85718751e-01
8.12837183e-01 3.79791632e-02 -5.34467757e-01 -5.00485122e-01
-8.61611009e-01 2.20051840e-01 4.69305307e-01 6.14128709e-02
3.52985710e-01 -3.77080530e-01 -2.35861227e-01 9.20294046e-01
-8.76111567e-01 -2.35944316e-01 6.13838397e-02 5.37230194e-01
-2.59068519e-01 -2.63459057e-01 7.08657801e-01 -1.33504510e+00
1.16062164e+00 -9.52879608e-01 -3.83458883e-01 -3.69805962e-01
-8.05114925e-01 -1.59628332e-01 2.99345911e-01 8.90228373e-04
-8.01658332e-01 -7.30241239e-01 6.42278465e-03 2.43391246e-01
-1.42497169e-02 1.01592243e+00 3.67682993e-01 5.25702983e-02
8.92781079e-01 -1.04449391e-01 4.69089597e-02 -3.05232078e-01
2.91585922e-01 1.18961573e+00 2.74194747e-01 -5.18487632e-01
7.51376152e-01 8.60975266e-01 -4.58388537e-01 -5.89149773e-01
-1.37890971e+00 -8.46793830e-01 -6.18965447e-01 -2.30884686e-01
2.04593599e-01 -8.30161929e-01 -6.27673566e-01 3.85797560e-01
-1.20205998e+00 1.29289716e-01 3.50507975e-01 1.67756096e-01
-2.43591741e-01 6.50207460e-01 -1.20334029e+00 -8.60335588e-01
-3.55699688e-01 -8.77074897e-01 4.00912046e-01 -3.62567186e-01
-7.63149619e-01 -1.05222869e+00 2.63172328e-01 9.40616965e-01
5.73431253e-01 2.12979674e-01 4.12015796e-01 -1.21822453e+00
-2.56008774e-01 -3.33925813e-01 -1.48010552e-01 -3.90219912e-02
2.09730044e-01 1.33912385e-01 -9.70173776e-01 -1.47713274e-01
3.87967139e-01 -6.36853456e-01 8.34304452e-01 -2.96990663e-01
1.81637287e-01 -1.00805056e+00 -1.20600842e-01 -4.30662870e-01
9.36581671e-01 -7.58952260e-01 3.65576386e-01 1.08837593e+00
2.37443998e-01 7.88819075e-01 6.91196799e-01 7.02652037e-01
7.01871157e-01 6.14159942e-01 6.27810895e-01 2.79825598e-01
2.56050348e-01 -2.70877421e-01 5.49552441e-01 9.76496279e-01
-3.21710706e-01 -1.25282735e-01 -1.29561090e+00 7.76282012e-01
-1.83467472e+00 -1.09239745e+00 -8.55792820e-01 1.79109585e+00
1.26970744e+00 3.23669970e-01 3.70782286e-01 3.57255697e-01
8.22152257e-01 2.63033360e-01 1.48402125e-01 -3.80091220e-01
-1.32244930e-01 -5.21575630e-01 4.39350724e-01 7.69885004e-01
-1.12317705e+00 7.71989524e-01 6.52328634e+00 1.46324277e-01
-1.01561189e+00 4.28429395e-01 2.26417810e-01 2.02460662e-01
-3.95747274e-01 1.91306815e-01 -9.34249163e-01 5.30998647e-01
1.13751054e+00 -1.16244532e-01 8.82796049e-02 5.26865065e-01
5.42834580e-01 -1.44898906e-01 -9.26822186e-01 3.54244448e-02
1.83776796e-01 -1.20377684e+00 -3.26162070e-01 2.02292353e-01
6.99514687e-01 6.69789374e-01 -1.57404333e-01 1.07461102e-02
6.83079422e-01 -7.76385307e-01 9.66166139e-01 8.12835917e-02
7.20487311e-02 -3.69846314e-01 1.26640403e+00 7.26777136e-01
-3.13704431e-01 2.50516623e-01 4.65723425e-02 -3.44828546e-01
5.06470978e-01 7.60124564e-01 -1.66818428e+00 4.73194927e-01
4.05280799e-01 5.37990093e-01 -5.82032025e-01 6.18368089e-01
-7.84161747e-01 1.27846873e+00 -3.60337079e-01 -5.44947609e-02
5.60627460e-01 6.12548105e-02 9.08243656e-01 1.29128468e+00
-3.01497400e-01 -1.44832060e-01 3.22222948e-01 4.88116503e-01
-4.21963871e-01 1.85777212e-03 -6.85478687e-01 -1.42319009e-01
6.12799823e-01 1.31724536e+00 -5.09526610e-01 -4.80251163e-01
-2.44250774e-01 3.98254275e-01 2.69605696e-01 -4.79882471e-02
-7.86969244e-01 2.81723112e-01 1.23810872e-01 6.52180612e-01
-1.08631626e-01 -3.08925450e-01 -2.54014969e-01 -1.19534624e+00
-1.93775624e-01 -1.39775777e+00 5.50257325e-01 -5.73554933e-01
-1.61929023e+00 6.56368017e-01 -5.77344671e-02 -1.03586221e+00
-3.65810037e-01 -1.36869699e-01 -4.46940005e-01 7.03275084e-01
-1.89563930e+00 -1.16351521e+00 -1.97555751e-01 5.76371729e-01
1.58986256e-01 2.93908585e-02 3.94547015e-01 3.87925319e-02
-2.02018350e-01 -4.82792109e-02 -4.81410265e-01 4.95466679e-01
1.21275580e+00 -1.03064263e+00 3.06895852e-01 5.20791769e-01
-2.79847700e-02 1.03670907e+00 1.21528208e+00 -9.21001077e-01
-1.05687511e+00 -6.90472841e-01 1.44099367e+00 -1.04815423e+00
1.57801807e+00 1.30194938e-02 -1.16090500e+00 8.72647047e-01
5.89990735e-01 -3.76338273e-01 8.55383456e-01 3.37354898e-01
-7.18938529e-01 6.63634121e-01 -1.08954465e+00 4.50680375e-01
5.13234198e-01 -6.84643328e-01 -1.22159803e+00 5.71511745e-01
3.70659262e-01 -2.69424945e-01 -9.18972373e-01 5.14998958e-02
3.01960021e-01 -8.64140868e-01 5.11691689e-01 -9.25127804e-01
6.59035325e-01 -4.01798099e-01 -1.51313409e-01 -1.26372659e+00
5.88539205e-02 -5.66633224e-01 -2.65246332e-01 1.37112772e+00
5.91053605e-01 -8.70991051e-01 5.79932213e-01 3.62296641e-01
9.12309363e-02 -2.66568542e-01 -8.44630599e-01 -2.90202558e-01
5.38237132e-02 -2.08369926e-01 4.76731271e-01 1.77492499e+00
5.75608790e-01 5.67181408e-01 -3.98725957e-01 2.79431969e-01
5.70869803e-01 3.28405619e-01 7.76578307e-01 -1.35528040e+00
1.60776883e-01 -1.29728928e-01 -3.25350314e-02 -7.00138986e-01
2.62335211e-01 -8.14139485e-01 -3.80432718e-02 -1.32600033e+00
2.55783617e-01 -6.28479242e-01 8.84582698e-02 5.46365023e-01
-5.21945283e-02 4.25079912e-01 2.23359633e-02 1.02750671e+00
-6.35017812e-01 2.39370197e-01 1.05964351e+00 -1.55721635e-01
1.14112429e-01 1.31291077e-01 -8.27316463e-01 1.22739279e+00
1.10378063e+00 -8.35452795e-01 1.98144794e-01 -3.91842276e-02
9.55973208e-01 -1.23402374e-02 4.75262493e-01 -2.31174529e-01
4.66488212e-01 -8.88854265e-02 -2.34059528e-01 -7.14337051e-01
-3.13856453e-02 -4.79859143e-01 -4.66509461e-01 3.08866799e-01
-2.76011735e-01 3.25449705e-01 -2.43534938e-01 6.48047149e-01
-5.16062558e-01 -4.03790146e-01 5.05678535e-01 -5.03788531e-01
-3.10808778e-01 -1.90178901e-01 -8.22253525e-01 2.48586163e-01
3.83022934e-01 2.51353860e-01 -1.00077057e+00 -7.46678412e-01
-6.05927587e-01 1.32244721e-01 8.68027270e-01 1.36917591e-01
4.77063805e-01 -1.05812311e+00 -1.23354781e+00 -3.39740008e-01
1.35868654e-01 -5.54973006e-01 -6.46945000e-01 1.54929638e+00
-2.71308810e-01 3.55577409e-01 -3.40051241e-02 -3.92999649e-01
-1.42275000e+00 4.18244302e-01 1.13485577e-02 -3.09461325e-01
-2.95139760e-01 1.93734497e-01 -5.95798016e-01 -5.30701220e-01
-3.71435404e-01 1.82584748e-01 -4.78354514e-01 5.26532471e-01
7.84277916e-01 5.15885353e-01 3.88406813e-01 -9.57871437e-01
-3.20014447e-01 -3.03248942e-01 -4.32586312e-01 -4.85972285e-01
1.21090090e+00 -3.87798727e-01 -8.16487789e-01 6.16984904e-01
5.29772401e-01 7.17109203e-01 -3.76232564e-01 -4.83648539e-01
4.56112504e-01 -5.97447336e-01 6.73569217e-02 -6.38686955e-01
-3.67248535e-01 2.28935972e-01 -5.08068562e-01 1.09114265e+00
2.70651877e-01 1.92434162e-01 6.95763111e-01 3.26214939e-01
5.49147427e-01 -1.10454810e+00 1.11978136e-01 1.05729795e+00
9.82685387e-01 -1.51285172e+00 3.04669291e-01 -5.81009746e-01
-7.03264594e-01 1.07456207e+00 7.67674819e-02 1.35155931e-01
4.36817229e-01 1.02872707e-01 4.68813747e-01 -8.16548288e-01
-8.22798193e-01 -3.58489864e-02 -1.14712794e-03 1.87194034e-01
8.11686814e-01 -2.39717513e-02 -6.90430641e-01 4.65194315e-01
-5.08126020e-01 -4.55475040e-02 1.27583635e+00 1.08398449e+00
-7.39714444e-01 -9.10925627e-01 -7.60582805e-01 1.18077137e-01
-8.41687262e-01 -2.98572063e-01 -1.14605606e+00 9.39148247e-01
-3.52075338e-01 1.41811931e+00 -4.69625950e-01 -2.51996934e-01
-1.40555516e-01 3.79818939e-02 1.37628941e-03 -5.16984046e-01
-1.07838559e+00 -2.46953994e-01 1.11029947e+00 -2.53773808e-01
-9.05218899e-01 -8.48616719e-01 -9.56392944e-01 -9.50862229e-01
-6.44148529e-01 6.12339258e-01 7.13418186e-01 1.17224228e+00
-3.51240486e-02 -1.68413118e-01 7.32668638e-01 -3.54515880e-01
-9.01418626e-01 -1.19634140e+00 -3.67326677e-01 4.94880915e-01
4.83721495e-01 -5.86413682e-01 -6.45105660e-01 3.13713141e-02] | [8.327198028564453, 10.06649398803711] |
97fd39e3-4c3a-42f3-ba17-b53821aeab39 | episodic-camn-contextual-attention-based | null | null | http://openaccess.thecvf.com/content_cvpr_2017/html/Abdulnabi_Episodic_CAMN_Contextual_CVPR_2017_paper.html | http://openaccess.thecvf.com/content_cvpr_2017/papers/Abdulnabi_Episodic_CAMN_Contextual_CVPR_2017_paper.pdf | Episodic CAMN: Contextual Attention-Based Memory Networks With Iterative Feedback for Scene Labeling | Scene labeling can be seen as a sequence-sequence prediction task (pixels-labels), and it is quite important to leverage relevant context to enhance the performance of pixel classification. In this paper, we introduce an episodic attention-based memory network to achieve the goal. We present a unified framework that mainly consists of a Convolutional Neural Network (CNN), specifically, Fully Convolutional Network (FCN) and an attention-based memory module with feedback connections to perform context selection and refinement. The full model produces context-aware representation for each target patch by aggregating the activated context and its original local representation produced by the convolution layers. We evaluate our model on PASCAL Context, SIFT Flow and PASCAL VOC 2011 datasets and achieve competitive results to other state-of-the-art methods in scene labeling.
| ['Stefan Winkler', 'Abrar H. Abdulnabi', 'Gang Wang', 'Bing Shuai'] | 2017-07-01 | null | null | null | cvpr-2017-7 | ['scene-labeling'] | ['computer-vision'] | [ 6.69310689e-01 -2.79190123e-01 -3.16546440e-01 -6.49744868e-01
-3.72370541e-01 -1.89982355e-01 6.09609067e-01 2.00615469e-02
-4.73898441e-01 6.29003763e-01 4.41787660e-01 -1.91016823e-01
5.24395049e-01 -8.00306559e-01 -7.68516362e-01 -6.70574367e-01
1.78483933e-01 -1.66307062e-01 5.89793563e-01 -4.40899171e-02
5.33835590e-01 3.62911910e-01 -1.74470270e+00 9.05120969e-01
8.34253013e-01 1.15711248e+00 6.90421402e-01 7.87826419e-01
-3.71068269e-01 1.61725688e+00 -4.60611045e-01 -3.16594844e-04
-7.08023384e-02 -4.41970080e-01 -1.27874362e+00 1.14255443e-01
8.63956511e-01 -2.25666374e-01 -4.71940696e-01 9.67616618e-01
2.80515552e-01 5.33749163e-01 3.63625050e-01 -8.43649507e-01
-8.23340237e-01 3.50457460e-01 -3.98745924e-01 5.01247883e-01
2.18797147e-01 4.63106692e-01 8.48624229e-01 -9.71212864e-01
8.65623474e-01 1.08100343e+00 4.80524451e-01 4.99497294e-01
-9.81446326e-01 -1.58550054e-01 8.25449884e-01 6.57366812e-01
-1.06449795e+00 -6.36816844e-02 8.42174768e-01 -5.40404439e-01
1.43882823e+00 1.07805701e-02 8.83447349e-01 1.13848948e+00
1.93115324e-01 1.13680482e+00 9.28436100e-01 -3.64735186e-01
2.57683694e-01 -2.86594063e-01 2.20596388e-01 5.90319574e-01
-2.37828851e-01 1.96540833e-01 -4.32712227e-01 2.83600152e-01
7.51100302e-01 1.82859868e-01 -2.39548296e-01 -1.93151549e-01
-1.14154005e+00 6.63653970e-01 1.01906431e+00 1.13099255e-01
-5.96771538e-01 6.59903944e-01 4.76963520e-01 -9.01034176e-02
2.26607621e-01 4.03027236e-01 -5.38749039e-01 1.23834036e-01
-7.88222313e-01 2.06945941e-01 3.06513011e-01 9.18123901e-01
9.81733918e-01 1.75402731e-01 -6.93317533e-01 7.41004705e-01
1.11914821e-01 6.73326477e-02 4.05000806e-01 -9.76518214e-01
2.17984468e-01 6.99488580e-01 -1.25648633e-01 -6.32262170e-01
-2.43560672e-01 -5.08054972e-01 -7.30800569e-01 2.14588821e-01
-3.49883027e-02 -4.14343104e-02 -1.36655581e+00 1.67131102e+00
2.93332934e-01 8.75390053e-01 -1.65549479e-02 1.01995790e+00
1.09319901e+00 8.33543539e-01 9.55870092e-01 2.33431503e-01
1.16867399e+00 -1.73941731e+00 -5.76239407e-01 -5.87294519e-01
5.16378880e-01 -6.52909160e-01 1.01403522e+00 -3.39837410e-02
-8.36814404e-01 -1.28260326e+00 -1.06764877e+00 -4.03043956e-01
-6.26002729e-01 6.97624981e-02 7.42443621e-01 1.23756111e-01
-1.08034766e+00 6.31946862e-01 -4.43174392e-01 -4.46540743e-01
7.32719600e-01 6.82910234e-02 -1.71174362e-01 -3.30676407e-01
-1.10940993e+00 7.99155533e-01 5.17723680e-01 1.60984844e-01
-1.31083846e+00 -7.15583444e-01 -1.13625443e+00 1.38225511e-01
1.96179688e-01 -8.69141877e-01 1.15792155e+00 -1.30853879e+00
-1.26151156e+00 8.74598682e-01 -3.37292790e-01 -6.25037193e-01
1.92824170e-01 -3.13082337e-01 -3.43023688e-01 2.30619401e-01
8.64740312e-02 1.29884672e+00 8.86954308e-01 -1.17152452e+00
-1.16198397e+00 1.09607488e-01 2.40796223e-01 2.43389934e-01
1.94158584e-01 -1.44554392e-01 -5.28841197e-01 -7.46146321e-01
-2.07526460e-01 -7.83443451e-01 -7.00101972e-01 -1.02567829e-01
-2.24650368e-01 -2.65449256e-01 1.12059116e+00 -5.19167840e-01
1.06450546e+00 -2.02490783e+00 -8.80058184e-02 -3.10831100e-01
-2.68971562e-01 5.14034629e-01 -4.58761811e-01 1.77924097e-01
-3.17839772e-01 -1.09252691e-01 -4.71767575e-01 -4.15815383e-01
-3.22907120e-01 3.00494790e-01 -5.83713114e-01 2.15345621e-01
7.47416914e-01 1.18971729e+00 -1.22763610e+00 -4.83035982e-01
7.35655963e-01 5.65916717e-01 -5.70910752e-01 3.28581065e-01
-6.95923448e-01 6.23240888e-01 -9.44071785e-02 8.45815003e-01
5.02159178e-01 -3.47874224e-01 1.33194834e-01 -2.61964500e-01
-2.27440730e-01 2.90193886e-01 -9.19000030e-01 2.11945367e+00
-5.35189927e-01 8.68225574e-01 -3.14777434e-01 -8.87056112e-01
8.06394219e-01 2.09831335e-02 1.40056074e-01 -8.46198857e-01
8.95169750e-02 -1.99355751e-01 -2.28848398e-01 -6.62776470e-01
7.69620657e-01 4.77017909e-01 1.77332118e-01 1.23870827e-01
3.92308831e-01 2.15074182e-01 1.12168655e-01 -1.56997293e-02
9.51797128e-01 6.25741065e-01 2.50332683e-01 -3.77953321e-01
9.08576131e-01 2.50476420e-01 6.06054187e-01 8.08700919e-01
-6.29484773e-01 5.71267486e-01 1.97589785e-01 -9.34689581e-01
-8.89503956e-01 -6.78158581e-01 -3.41062434e-02 1.43607438e+00
4.67549115e-01 -2.76111364e-01 -6.12081349e-01 -9.76324677e-01
-1.48892939e-01 7.56039023e-01 -1.03493714e+00 -1.95578948e-01
-7.89358616e-01 -3.76986384e-01 1.66070536e-01 1.11941421e+00
1.10363877e+00 -1.76498115e+00 -9.61150348e-01 2.39606082e-01
-2.38979578e-01 -1.07306194e+00 -4.37604457e-01 4.12033886e-01
-7.50469983e-01 -1.14088798e+00 -3.52412552e-01 -1.15232515e+00
5.32851815e-01 3.07068616e-01 1.39916384e+00 2.71800607e-01
-5.65586567e-01 2.73465753e-01 -2.84423083e-01 -1.79691315e-01
8.67837295e-02 9.22948197e-02 -8.04731011e-01 1.25807539e-01
4.72586840e-01 -2.68115580e-01 -1.06411266e+00 1.82049960e-01
-8.46731186e-01 2.22615868e-01 6.05576694e-01 9.41034198e-01
9.75407362e-01 -3.39273810e-01 5.04287839e-01 -1.01940107e+00
1.94315668e-02 -3.21771204e-01 -5.02031684e-01 3.21017951e-01
-2.12712228e-01 1.03704415e-01 5.08942604e-01 -1.83031216e-01
-1.39477968e+00 6.88446045e-01 -2.70329356e-01 -5.28207004e-01
-6.02841794e-01 3.11323583e-01 -1.59708440e-01 -1.90292016e-01
5.55433512e-01 3.44181716e-01 -6.70381665e-01 -3.24722916e-01
8.44798982e-01 4.06972498e-01 9.53325152e-01 -4.68089551e-01
3.22121233e-02 5.19851863e-01 -7.06391484e-02 -5.78710198e-01
-1.21947169e+00 -5.67418993e-01 -9.34376299e-01 -4.68995363e-01
1.27671659e+00 -1.03421128e+00 -4.70965147e-01 4.03321832e-01
-1.22479212e+00 -8.32629383e-01 -5.58524966e-01 1.08381458e-01
-6.88016772e-01 3.35553475e-02 -6.98203206e-01 -5.52312493e-01
-3.63743216e-01 -1.24652290e+00 1.13673246e+00 5.38077116e-01
-8.01389813e-02 -1.09637845e+00 1.77684352e-01 1.60231724e-01
5.13501406e-01 4.56226200e-01 7.34479725e-01 -2.13156283e-01
-1.01115847e+00 1.52469561e-01 -5.15700638e-01 2.66809106e-01
-1.41846240e-01 2.50559095e-02 -1.47399282e+00 5.24716675e-02
-2.07978338e-01 -4.67169404e-01 1.66235089e+00 5.41165054e-01
1.67677081e+00 1.31539796e-02 -4.88031000e-01 8.52396727e-01
1.67035556e+00 2.28523776e-01 9.91587520e-01 3.61638069e-01
7.60165274e-01 4.71607268e-01 8.02208781e-01 1.69987336e-01
2.61267692e-01 2.80837059e-01 7.19203949e-01 -1.85226813e-01
-7.31144488e-01 -3.58435422e-01 -1.35509465e-02 1.67767495e-01
6.74395114e-02 -4.07083817e-02 -7.47081339e-01 9.46841002e-01
-2.13328242e+00 -1.26080644e+00 -1.56419277e-01 1.85369122e+00
5.71910739e-01 1.09969482e-01 -2.42790759e-01 -1.93687811e-01
8.17917049e-01 5.46285272e-01 -6.88382387e-01 -5.81539512e-01
-2.38255024e-01 2.67886668e-01 3.07344466e-01 5.49747109e-01
-1.67378986e+00 1.38210177e+00 6.97041750e+00 6.57615185e-01
-1.28223038e+00 2.90917698e-02 8.99608195e-01 2.19198585e-01
3.77825025e-04 1.59082517e-01 -6.84910536e-01 3.41380775e-01
6.57545984e-01 3.36813450e-01 9.26723108e-02 1.33284104e+00
-1.27350062e-01 -3.44356805e-01 -1.01306021e+00 8.69837821e-01
1.85460687e-01 -1.70656931e+00 2.06488192e-01 -3.19068491e-01
1.21965945e+00 2.89182484e-01 1.13902695e-01 4.90486801e-01
3.68181616e-01 -1.12601459e+00 5.79989612e-01 5.79831541e-01
6.63702190e-01 -6.05776548e-01 6.73256040e-01 -1.34416014e-01
-1.54284334e+00 -3.17201972e-01 -4.80322480e-01 -1.62799299e-01
1.79367542e-01 3.75837117e-01 -5.32638848e-01 2.90481836e-01
7.17211962e-01 1.18485224e+00 -7.82290757e-01 1.26609826e+00
-5.79081059e-01 5.74824750e-01 2.58224487e-01 2.06512988e-01
6.74291611e-01 8.64595696e-02 5.82948513e-02 1.72718549e+00
-2.19830602e-01 1.64335907e-01 4.28049445e-01 8.79169822e-01
-1.70782611e-01 -1.43168062e-01 -4.54134166e-01 3.69792998e-01
2.77130455e-01 1.56649709e+00 -8.20806980e-01 -8.37284207e-01
-6.17083907e-01 1.17696774e+00 6.60270989e-01 5.31002402e-01
-8.71782005e-01 -3.45833391e-01 8.98767710e-01 -2.67266959e-01
6.77854180e-01 2.64048781e-02 -4.65760231e-01 -9.97202635e-01
-4.15013641e-01 -3.78513902e-01 6.97229981e-01 -1.06937456e+00
-1.09350109e+00 8.26898277e-01 -4.25187379e-01 -1.10128093e+00
3.76424901e-02 -5.50022602e-01 -9.84859705e-01 8.54223013e-01
-2.04422116e+00 -1.26750970e+00 -7.81842589e-01 6.68914676e-01
9.57814991e-01 2.81111687e-01 6.88157916e-01 3.35594088e-01
-5.90095580e-01 2.64187932e-01 -4.45462972e-01 2.42983505e-01
4.81136590e-01 -1.36118984e+00 5.81199110e-01 1.10426819e+00
2.14561790e-01 3.39062989e-01 2.69530624e-01 -8.10725451e-01
-7.87440300e-01 -1.94686282e+00 8.52198303e-01 -3.27727228e-01
2.95313507e-01 -6.77102506e-02 -8.17712128e-01 7.95232534e-01
6.52423084e-01 4.93778080e-01 4.58564371e-01 -1.55229941e-01
-4.12556916e-01 1.38511434e-01 -8.35366249e-01 6.83699787e-01
1.49284530e+00 -6.28816426e-01 -6.24794126e-01 2.45494083e-01
9.11952913e-01 -4.19270217e-01 -5.15007138e-01 5.71115971e-01
2.04128966e-01 -9.53806579e-01 1.10729396e+00 -7.65615225e-01
6.87415540e-01 -5.46320498e-01 -2.84378529e-01 -1.01712286e+00
-7.20961094e-01 -1.51861876e-01 -1.00820765e-01 1.02131569e+00
1.64852381e-01 -1.33153439e-01 8.30461264e-01 4.09325480e-01
-5.54804623e-01 -9.20792639e-01 -5.36019087e-01 -3.23792756e-01
-1.78777248e-01 -3.85040224e-01 5.84827662e-01 8.52685452e-01
-3.46089065e-01 4.98856574e-01 -3.11260700e-01 4.72379513e-02
4.23606455e-01 6.24601960e-01 4.96930242e-01 -7.67705917e-01
-1.11751720e-01 -4.42672998e-01 -4.06246901e-01 -1.24054611e+00
2.28026494e-01 -9.42648828e-01 2.33474255e-01 -1.64354300e+00
2.24899292e-01 -2.88748354e-01 -6.35012150e-01 6.29905760e-01
-5.68626404e-01 3.69715422e-01 3.20995510e-01 -3.04543171e-02
-1.14439309e+00 6.19437397e-01 1.23085463e+00 -3.82224500e-01
-1.96245372e-01 -3.73179674e-01 -4.25576001e-01 6.05724394e-01
6.84069395e-01 -1.94350779e-01 -3.18921447e-01 -6.06414795e-01
-2.00361609e-01 -3.05654317e-01 5.00446141e-01 -1.28668165e+00
3.49750161e-01 -2.59100556e-01 8.23656440e-01 -9.06510711e-01
2.23855570e-01 -6.60443902e-01 -9.71113965e-02 5.88567019e-01
-6.59845591e-01 -1.88653424e-01 2.06129402e-01 6.50577068e-01
-4.50627357e-01 -1.67284355e-01 8.23644578e-01 -3.07058781e-01
-1.77014029e+00 4.94899333e-01 -3.47202063e-01 -1.01516746e-01
1.02774489e+00 -8.59501064e-02 -5.74263275e-01 -9.02499780e-02
-7.64902592e-01 3.73813450e-01 3.66309404e-01 5.10810494e-01
7.43371010e-01 -1.41301346e+00 -4.28547531e-01 2.27924347e-01
3.60825241e-01 1.08771734e-01 6.33158565e-01 4.60903555e-01
-5.87188005e-01 4.52665299e-01 -3.60723674e-01 -7.73805678e-01
-9.80835319e-01 1.02918231e+00 4.46033061e-01 -1.73995093e-01
-5.24898171e-01 1.11599445e+00 3.91816497e-01 -1.67232975e-01
3.12190443e-01 -5.31031847e-01 -5.06228864e-01 -2.22693056e-01
8.64272058e-01 6.56200796e-02 -1.09676078e-01 -6.58499420e-01
-2.98317999e-01 6.65788531e-01 2.62505095e-02 2.86823452e-01
1.06549907e+00 -1.34648740e-01 7.29162693e-02 8.19539875e-02
1.16097796e+00 -6.19134009e-01 -1.89581585e+00 -2.54096270e-01
-1.13166850e-02 -4.16969955e-01 5.35395928e-02 -9.59775746e-01
-1.27550983e+00 1.07302749e+00 6.73029125e-01 -3.09924513e-01
1.32701313e+00 -1.09124899e-01 7.32108116e-01 3.12717646e-01
2.43017133e-02 -1.11385107e+00 2.79682875e-01 6.96969211e-01
7.23792553e-01 -1.33116126e+00 -2.31829092e-01 -4.51859474e-01
-8.33161354e-01 1.01848817e+00 1.02816355e+00 -5.90793490e-01
7.26856411e-01 7.11660013e-02 1.29385695e-01 -1.58118591e-01
-9.87733364e-01 -8.46107006e-01 3.42986912e-01 7.14110672e-01
4.17025328e-01 -1.45161420e-01 5.60934804e-02 4.88458037e-01
4.01227832e-01 1.20622754e-01 1.85193628e-01 1.02568483e+00
-6.91137195e-01 -9.33477998e-01 3.97124775e-02 2.34918267e-01
-2.69879907e-01 -2.54307508e-01 -3.48881960e-01 5.30365288e-01
4.79307771e-01 8.73404503e-01 3.53027165e-01 -3.99586856e-01
1.29124507e-01 2.74700046e-01 2.45846406e-01 -6.86774611e-01
-9.05145645e-01 -1.12142354e-01 1.04887057e-02 -1.06306875e+00
-7.51127422e-01 -3.97232234e-01 -1.17570186e+00 3.57162841e-02
-2.14909036e-02 -3.12943220e-01 4.83161449e-01 8.61471295e-01
5.13635397e-01 9.25237954e-01 3.35939437e-01 -9.52376187e-01
2.37190768e-01 -7.80822396e-01 -9.81891081e-02 7.38436639e-01
2.46769890e-01 -5.63226283e-01 1.57810166e-01 4.38969731e-01] | [9.572493553161621, 0.4084725081920624] |
5e67b718-259f-4190-9183-ac4cecbf67b3 | mlanet-multi-level-attention-network-with-sub | 2303.01396 | null | https://arxiv.org/abs/2303.01396v1 | https://arxiv.org/pdf/2303.01396v1.pdf | MLANet: Multi-Level Attention Network with Sub-instruction for Continuous Vision-and-Language Navigation | Vision-and-Language Navigation (VLN) aims to develop intelligent agents to navigate in unseen environments only through language and vision supervision. In the recently proposed continuous settings (continuous VLN), the agent must act in a free 3D space and faces tougher challenges like real-time execution, complex instruction understanding, and long action sequence prediction. For a better performance in continuous VLN, we design a multi-level instruction understanding procedure and propose a novel model, Multi-Level Attention Network (MLANet). The first step of MLANet is to generate sub-instructions efficiently. We design a Fast Sub-instruction Algorithm (FSA) to segment the raw instruction into sub-instructions and generate a new sub-instruction dataset named ``FSASub". FSA is annotation-free and faster than the current method by 70 times, thus fitting the real-time requirement in continuous VLN. To solve the complex instruction understanding problem, MLANet needs a global perception of the instruction and observations. We propose a Multi-Level Attention (MLA) module to fuse vision, low-level semantics, and high-level semantics, which produce features containing a dynamic and global comprehension of the task. MLA also mitigates the adverse effects of noise words, thus ensuring a robust understanding of the instruction. To correctly predict actions in long trajectories, MLANet needs to focus on what sub-instruction is being executed every step. We propose a Peak Attention Loss (PAL) to improve the flexible and adaptive selection of the current sub-instruction. PAL benefits the navigation agent by concentrating its attention on the local information, thus helping the agent predict the most appropriate actions. We train and test MLANet in the standard benchmark. Experiment results show MLANet outperforms baselines by a significant margin. | ['Qijun Chen', 'Chengju Liu', 'Qingqing Yan', 'Shu Li', 'Liuyi Wang', 'Zongtao He'] | 2023-03-02 | null | null | null | null | ['vision-and-language-navigation'] | ['robots'] | [ 2.44989574e-01 -1.93976924e-01 -6.47396073e-02 -4.63596374e-01
-3.84978473e-01 -3.64557683e-01 6.20451987e-01 -1.43860117e-01
-6.64509535e-01 1.74385235e-01 1.72709659e-01 -6.05449557e-01
2.10323796e-01 -8.42981160e-01 -9.94495451e-01 -5.35784066e-01
2.19467551e-01 6.25999629e-01 4.05742735e-01 -4.65203851e-01
4.07582939e-01 2.35257700e-01 -1.66309261e+00 2.25603178e-01
1.19770825e+00 7.52898276e-01 1.09825075e+00 7.86812305e-01
-2.45946780e-01 8.51765215e-01 -4.89968151e-01 1.71917677e-01
1.93871886e-01 -5.28774440e-01 -7.77732134e-01 -4.27976623e-02
3.57499629e-01 -5.14937341e-01 -2.58361787e-01 1.13148081e+00
2.08555385e-01 4.88007158e-01 3.70151222e-01 -1.14240730e+00
-6.13249302e-01 5.60034871e-01 -2.93443173e-01 5.80092855e-02
5.94201744e-01 8.14645827e-01 8.15293670e-01 -7.77376592e-01
5.01791775e-01 1.69936097e+00 2.07634434e-01 8.54570746e-01
-7.32973695e-01 -5.31040847e-01 7.85682261e-01 5.32996714e-01
-8.89752328e-01 -2.77240813e-01 4.77645427e-01 -2.82918274e-01
1.13644683e+00 5.07334098e-02 5.52360952e-01 1.35720646e+00
3.96623939e-01 1.10952532e+00 8.78360987e-01 -2.31768474e-01
3.34653646e-01 -4.84404474e-01 3.95180106e-01 1.17541552e+00
-8.74130949e-02 3.06025505e-01 -5.17155945e-01 4.00863677e-01
6.85525715e-01 1.94630027e-01 -4.44449723e-01 -3.23246062e-01
-1.50898767e+00 7.76920974e-01 6.35794222e-01 1.40012980e-01
-4.31057930e-01 1.82884946e-01 4.78427202e-01 1.47770748e-01
-3.35517794e-01 3.75319928e-01 -5.38741589e-01 -3.59565794e-01
-2.20318303e-01 9.41632763e-02 6.63985729e-01 1.04270756e+00
7.49684095e-01 2.10330993e-01 -3.41256440e-01 4.61149871e-01
3.52537453e-01 6.30601227e-01 8.80796850e-01 -1.07543278e+00
5.52146196e-01 7.37899303e-01 -6.38957694e-02 -6.52527809e-01
-5.46286881e-01 -3.66486907e-01 -5.50843000e-01 4.78838831e-01
3.47445577e-01 3.98113988e-02 -1.26033103e+00 1.92921674e+00
3.58161986e-01 3.04355115e-01 2.15957895e-01 1.01794577e+00
7.11614847e-01 8.38826597e-01 -3.86944115e-02 -8.31358954e-02
1.34693706e+00 -1.72485232e+00 -6.97861671e-01 -8.76244426e-01
9.04594302e-01 -2.69986629e-01 1.54768980e+00 4.23109025e-01
-6.78213716e-01 -1.02916300e+00 -1.11584985e+00 -2.02948138e-01
-2.81618416e-01 -8.81539509e-02 5.58320165e-01 9.46579725e-02
-9.33795571e-01 1.81411117e-01 -9.74746525e-01 -2.29456216e-01
3.02188963e-01 4.42862809e-01 -2.93459088e-01 -3.28166962e-01
-1.01714659e+00 8.26685786e-01 4.78065997e-01 3.02480757e-01
-1.49984384e+00 -4.27015692e-01 -1.43792510e+00 3.08727976e-02
8.72992277e-01 -6.70251667e-01 1.39973891e+00 -9.09622967e-01
-1.83139050e+00 3.84593189e-01 -4.65651512e-01 -4.35402155e-01
2.65332967e-01 -4.69052702e-01 -7.03954697e-02 -1.88227177e-01
2.21456990e-01 7.34373987e-01 7.59020686e-01 -1.25496376e+00
-8.02387416e-01 -4.73022997e-01 2.97971308e-01 5.02718270e-01
1.79611966e-01 -6.17542207e-01 -7.68223464e-01 -2.16346636e-01
4.37052408e-03 -1.00760162e+00 -3.59289706e-01 -4.61864024e-01
-2.85188049e-01 -4.56724554e-01 1.00177753e+00 -5.48569143e-01
9.30819273e-01 -2.03947210e+00 3.51629823e-01 -2.54294991e-01
1.99106142e-01 4.90709513e-01 -5.98179042e-01 -2.84582209e-02
2.74695307e-01 -4.09168482e-01 -2.04279974e-01 -3.63185197e-01
1.73151642e-01 6.12982154e-01 -2.35344663e-01 1.34628907e-01
-2.16369126e-02 1.25005126e+00 -1.16040158e+00 -1.93627298e-01
4.11840081e-01 1.64401293e-01 -6.94890141e-01 5.37698150e-01
-7.59440243e-01 6.84415221e-01 -6.91032112e-01 6.95281386e-01
2.79032141e-01 -1.76948830e-01 -1.98607653e-01 -1.15769237e-01
-9.49485600e-02 1.93636104e-01 -8.10309827e-01 2.14363217e+00
-9.10868168e-01 4.26368147e-01 1.98244765e-01 -9.44982350e-01
8.52936625e-01 -2.87264407e-01 -1.25413120e-01 -1.22830462e+00
6.07832223e-02 5.17416969e-02 1.67444229e-01 -7.60865688e-01
1.67013034e-01 4.71071184e-01 -2.97463924e-01 2.66396224e-01
-1.85469776e-01 1.21824391e-01 1.52082995e-01 1.26349911e-01
1.22227383e+00 5.82064271e-01 2.12079599e-01 -6.09091669e-02
9.42852259e-01 1.97590869e-02 6.86168373e-01 9.76770818e-01
-5.71440518e-01 -1.37795908e-02 1.19365454e-01 -7.01814055e-01
-4.50341940e-01 -7.07402110e-01 5.86687028e-01 1.58626413e+00
5.70297539e-01 -2.00551376e-01 -1.00950944e+00 -1.11535001e+00
-2.66915411e-01 1.19258881e+00 -6.17985129e-01 -3.72479767e-01
-9.87542033e-01 -3.49542290e-01 4.74507362e-02 5.28845072e-01
7.96458006e-01 -1.54592800e+00 -1.10089254e+00 2.16497898e-01
-2.44954422e-01 -1.23087692e+00 -9.96898055e-01 4.40225363e-01
-4.20411974e-01 -1.06578779e+00 -6.40995242e-03 -9.94742155e-01
6.69781506e-01 3.40844542e-01 1.16924703e+00 2.61604309e-01
-1.16688646e-01 3.59044969e-01 -3.58143389e-01 -2.05980897e-01
-4.90372151e-01 -4.25897390e-02 9.18852761e-02 -6.83636665e-02
6.48436666e-01 -1.37547597e-01 -5.64022243e-01 2.49762207e-01
-6.48247063e-01 3.32601517e-01 9.25253749e-01 1.09311855e+00
6.20989323e-01 -2.04965889e-01 4.97314632e-01 -6.72695339e-01
5.50860107e-01 -2.81875819e-01 -8.35540414e-01 1.87459677e-01
-6.76446378e-01 4.18378621e-01 1.04674602e+00 -4.14243579e-01
-1.15558124e+00 2.58299202e-01 -2.85191536e-01 -4.84164596e-01
-5.78872859e-01 2.70108342e-01 -5.62435031e-01 7.75016844e-02
4.09811139e-01 7.18145907e-01 -3.77606712e-02 -2.30476171e-01
4.58408177e-01 3.99268270e-01 7.35367119e-01 -5.37284374e-01
5.81522763e-01 1.38419047e-01 -1.55395925e-01 -6.43619537e-01
-9.49915946e-01 -2.28644311e-01 -5.58183432e-01 -7.53752515e-02
1.13573456e+00 -6.91798866e-01 -1.25873649e+00 4.36052531e-01
-1.15404463e+00 -9.48636174e-01 -3.49590369e-02 2.81277120e-01
-7.35946774e-01 2.67061800e-01 -4.10431206e-01 -4.93069202e-01
-2.53678650e-01 -1.96762121e+00 1.11931479e+00 4.37467486e-01
-2.27921084e-02 -8.85030448e-01 -8.22584555e-02 6.40966594e-01
2.35745251e-01 -1.38234884e-01 9.06870961e-01 -8.07898402e-01
-8.88683259e-01 2.90845722e-01 -1.46339029e-01 6.83527291e-02
1.32639974e-01 -4.67103034e-01 -7.88515806e-01 -3.36280465e-01
2.17564717e-01 -5.04125059e-01 9.01376545e-01 2.89602458e-01
1.29959035e+00 -2.43645102e-01 -3.43860507e-01 8.68176460e-01
1.20912468e+00 7.41264880e-01 2.33729273e-01 4.87925291e-01
1.04664040e+00 5.26575923e-01 9.72467184e-01 9.88817289e-02
7.93564022e-01 6.33008122e-01 1.09045672e+00 2.40424007e-01
-1.67460263e-01 -3.66398841e-01 8.98997962e-01 9.67826188e-01
3.08162004e-01 -3.21589351e-01 -8.96808386e-01 4.59980667e-01
-2.03304863e+00 -7.06427932e-01 9.38146841e-03 2.03164196e+00
6.06536508e-01 4.59480017e-01 -3.40658307e-01 -3.50365579e-01
3.67473751e-01 2.28949070e-01 -1.10740829e+00 -5.51028788e-01
2.79294759e-01 -7.54234195e-02 2.52474576e-01 9.42434490e-01
-9.84290898e-01 1.27817273e+00 4.76352453e+00 8.62027466e-01
-9.95730102e-01 -1.18512079e-01 4.01148230e-01 1.78336471e-01
-1.74417883e-01 -2.02709273e-01 -1.09235573e+00 5.12036204e-01
9.71739054e-01 1.03696547e-01 6.26837730e-01 9.52563703e-01
3.77742738e-01 -2.72621185e-01 -1.37339008e+00 9.20177579e-01
2.65750498e-01 -1.03872466e+00 2.64815867e-01 -1.55983940e-01
4.29505795e-01 1.81869492e-01 8.52525011e-02 9.21818435e-01
4.95037705e-01 -1.15793884e+00 5.16692519e-01 3.88798922e-01
3.48706961e-01 -7.63547242e-01 6.96819365e-01 8.26353014e-01
-1.31786692e+00 -4.49084908e-01 -7.96997473e-02 2.89462041e-02
1.28845200e-01 -3.89483795e-02 -7.27540731e-01 4.78836238e-01
5.45705020e-01 7.05426157e-01 -5.16266525e-01 4.93387699e-01
-4.79528546e-01 2.26867467e-01 -1.70842614e-02 -2.15130910e-01
8.20997119e-01 -4.02644634e-01 5.61580360e-01 1.00250852e+00
1.17766552e-01 1.57977268e-01 7.86673307e-01 7.98317671e-01
7.83515126e-02 -2.02335820e-01 -5.57973921e-01 1.27210751e-01
3.65194052e-01 9.92422462e-01 -4.71431583e-01 -4.05111104e-01
-5.84478319e-01 1.25614524e+00 4.94641155e-01 4.21943069e-01
-9.64367032e-01 -3.02089006e-01 8.10574472e-01 -3.61300468e-01
3.55553240e-01 -3.29284430e-01 -4.27272990e-02 -9.52412367e-01
-5.19337058e-02 -1.18007457e+00 1.50510505e-01 -7.06899226e-01
-6.44082308e-01 8.02567005e-01 -3.92126054e-01 -8.72895300e-01
-3.73164803e-01 -7.39951134e-01 -6.20124221e-01 6.97985649e-01
-1.71312845e+00 -9.78307545e-01 -6.67727709e-01 6.65903628e-01
1.52232897e+00 -1.22738622e-01 7.26916909e-01 -1.04672313e-01
-8.45270097e-01 4.43363339e-01 -4.15957332e-01 -5.08901402e-02
5.13473988e-01 -1.35867202e+00 5.90581417e-01 8.91835392e-01
2.45553050e-02 5.17802596e-01 6.78789854e-01 -7.76301265e-01
-1.89721012e+00 -1.20751607e+00 3.60018194e-01 -5.67862272e-01
5.06628215e-01 -3.16455036e-01 -9.67984557e-01 9.13896263e-01
1.13717064e-01 -1.96658865e-01 3.03017676e-01 -3.63034308e-01
-2.76866853e-01 7.51584768e-02 -7.03207016e-01 8.26847911e-01
1.17914557e+00 -3.91064167e-01 -8.76299083e-01 4.49083328e-01
1.30973804e+00 -5.84391117e-01 -2.03759298e-01 4.03335482e-01
2.93639779e-01 -1.11876333e+00 8.64702582e-01 -4.03193623e-01
2.63581574e-01 -4.86329436e-01 -5.62005639e-02 -1.29907691e+00
-3.05142730e-01 -5.48711002e-01 -2.51148283e-01 5.33401370e-01
1.77008927e-01 -4.85616297e-01 6.65626347e-01 3.24214458e-01
-7.18806744e-01 -1.02561033e+00 -6.31386220e-01 -6.25051498e-01
-2.98044980e-01 -4.91224080e-01 5.92895329e-01 5.19673884e-01
-3.85498494e-01 5.29063284e-01 -1.67148948e-01 4.56047326e-01
6.53654039e-01 3.49126726e-01 8.79837573e-01 -8.93948257e-01
-3.40746224e-01 -5.27633369e-01 -6.19635023e-02 -1.86537659e+00
6.48583829e-01 -7.09530890e-01 5.80655277e-01 -1.50827932e+00
-3.51601318e-02 -5.99036478e-02 -1.61549732e-01 4.93329138e-01
-4.00493801e-01 -4.73153472e-01 1.28133833e-01 -7.74109364e-02
-1.00027716e+00 8.63182366e-01 1.64420700e+00 -3.44144732e-01
-3.65374446e-01 -1.29001155e-01 -3.32332253e-01 1.00701272e+00
4.95875657e-01 4.58368063e-02 -7.73217440e-01 -7.07030773e-01
-2.35408023e-01 9.48444977e-02 1.56685799e-01 -9.96833622e-01
5.40138721e-01 -4.53366548e-01 1.27980486e-01 -8.83145452e-01
3.21925789e-01 -8.64370644e-01 -4.01799738e-01 8.01581085e-01
-2.51359284e-01 2.44984701e-01 1.01225920e-01 7.40020990e-01
-6.84276223e-02 -1.87068284e-01 6.73570752e-01 -3.30441475e-01
-1.51339436e+00 3.69942486e-01 -3.37994695e-01 1.81563739e-02
1.17957306e+00 -1.79776937e-01 -4.14147466e-01 -2.60055870e-01
-6.39608920e-01 9.36419904e-01 4.85666662e-01 7.89568007e-01
9.44630504e-01 -1.06504393e+00 -3.99159074e-01 6.20686352e-01
1.65631846e-01 4.50648397e-01 3.61906886e-01 5.07486820e-01
-6.09566391e-01 6.41291261e-01 -1.04476333e-01 -6.40372276e-01
-9.56683755e-01 8.67629647e-01 4.22149658e-01 -2.74936557e-01
-6.72617674e-01 1.00751650e+00 8.72658312e-01 -6.50870442e-01
5.03875673e-01 -5.18279374e-01 -4.68432665e-01 -3.57025445e-01
7.79907882e-01 1.29667267e-01 -1.52852952e-01 -6.06912613e-01
-2.76713371e-01 6.36841238e-01 -2.88214803e-01 2.69010931e-01
1.00202096e+00 -3.60598922e-01 -3.35969194e-03 4.46813434e-01
8.39788914e-01 -2.88887054e-01 -1.71180081e+00 -9.76191089e-02
2.08720732e-02 -1.15424089e-01 5.58370575e-02 -8.72232437e-01
-9.36587393e-01 9.12514448e-01 5.90125382e-01 -1.10543840e-01
1.01319945e+00 -2.18082577e-01 1.12385559e+00 7.46383190e-01
5.11707246e-01 -8.73675764e-01 4.51050967e-01 1.15737450e+00
8.58835518e-01 -1.47772896e+00 -5.25543392e-01 -5.74911572e-02
-5.83907068e-01 9.63866532e-01 1.43259323e+00 1.26680985e-01
7.41399080e-02 1.38815582e-01 1.38024241e-01 -2.70856470e-01
-9.21826720e-01 -2.23516285e-01 1.57125905e-01 6.02144659e-01
-1.22330435e-01 -4.53277230e-02 2.36452013e-01 6.44759476e-01
-1.53338194e-01 -2.60270774e-01 4.17332709e-01 8.94107699e-01
-7.76734233e-01 -9.37310338e-01 -3.64518166e-01 2.39499718e-01
2.37602174e-01 -9.88001376e-03 -1.07145861e-01 5.83653331e-01
3.14208895e-01 9.02857423e-01 -3.37271616e-02 -5.33497453e-01
4.37082767e-01 1.10119238e-01 8.41724053e-02 -7.61216223e-01
-4.35672253e-01 -9.64268968e-02 -2.06991836e-01 -1.23500931e+00
-9.82930288e-02 -4.20418471e-01 -1.80985808e+00 1.21926805e-02
-1.14510015e-01 -8.96393880e-02 5.85767448e-01 1.27969038e+00
3.47282439e-01 1.05916345e+00 4.77045566e-01 -9.76971567e-01
-6.41667783e-01 -7.00312257e-01 1.72318190e-01 3.87364239e-01
7.02414751e-01 -6.83287680e-01 -4.02703643e-01 -9.22840983e-02] | [4.466393947601318, 0.5192986130714417] |
568877d9-0df0-4c82-aaf0-e20c27c46904 | deep-reinforcement-learning-with-explicitly | 1911.08756 | null | https://arxiv.org/abs/1911.08756v5 | https://arxiv.org/pdf/1911.08756v5.pdf | Hierarchical Multiple-Instance Data Classification with Costly Features | We motivate our research with a real-world problem of classifying malicious web domains using a remote service that provides various information. Crucially, some of the information can be further analyzed into a certain depth and this process sequentially creates a tree of hierarchically structured multiple-instance data. Each request sent to the remote service is associated with a cost (e.g., time or another cost per request) and the objective is to maximize the accuracy, constrained with a budget. We present a generic framework able to work with a class of similar problems. Our method is based on Classification with Costly Features (CwCF), Hierarchical Multiple-Instance Learning (HMIL) and hierarchical decomposition of the action space. It works with samples described as partially-observed trees of features of various types (similar to a JSON/XML file), which allows to model data with complex structure. The process is modeled as a Markov Decision Process (MDP), where a state represents acquired features, and actions select yet unknown ones. The policy is trained with deep reinforcement learning and we demonstrate our method with both real-world and synthetic data. | ['Viliam Lisý', 'Tomáš Pevný', 'Jaromír Janisch'] | 2019-11-20 | null | null | null | null | ['classification-with-costly-features'] | ['miscellaneous'] | [ 4.38104719e-01 1.34258151e-01 -5.33501506e-01 -5.90080142e-01
-8.91345799e-01 -4.99292195e-01 8.65125060e-01 1.59018010e-01
-3.28602999e-01 7.33939886e-01 -3.38816494e-01 -1.88294947e-01
-2.90485591e-01 -1.11633027e+00 -7.78316021e-01 -7.97438443e-01
-3.18628818e-01 1.17573786e+00 6.46594763e-01 1.68157205e-01
4.14041340e-01 4.97940391e-01 -1.75985515e+00 7.59285212e-01
6.81394756e-01 1.29083240e+00 1.98432162e-01 7.57927358e-01
-2.50570685e-01 1.08536184e+00 -7.11882651e-01 -3.87374163e-01
3.74770969e-01 -4.13599014e-02 -1.42399001e+00 6.19260669e-01
-1.76741749e-01 -5.19768834e-01 2.33149558e-01 9.22323346e-01
-8.68955702e-02 -7.73498788e-02 6.16947532e-01 -1.74498713e+00
-3.89740639e-03 5.72678089e-01 -8.21413007e-03 -1.95804983e-02
4.56365019e-01 1.72668725e-01 1.17788339e+00 -2.54749745e-01
7.99777150e-01 1.27873600e+00 1.30141601e-01 5.23389161e-01
-1.32923162e+00 -1.00426994e-01 3.31394106e-01 7.02155650e-01
-6.52753949e-01 -5.85394353e-02 5.18662930e-01 -3.75699162e-01
9.62922871e-01 4.30936247e-01 6.88507676e-01 1.52452266e+00
2.14776531e-01 1.20640743e+00 1.30952227e+00 -2.03874514e-01
7.95702279e-01 2.82164842e-01 2.68447816e-01 6.09576881e-01
-6.68704361e-02 3.97215858e-02 -3.16138506e-01 -7.02477157e-01
2.60027111e-01 4.17891651e-01 1.63560957e-01 -6.59364283e-01
-6.17785752e-01 1.09455645e+00 -6.64301440e-02 2.29899973e-01
-5.35031438e-01 -2.10331455e-01 4.53686833e-01 7.77389526e-01
2.05794781e-01 1.76584497e-01 -8.49987805e-01 -2.08547831e-01
-6.62512004e-01 1.92847729e-01 1.44439757e+00 7.05998898e-01
8.28828812e-01 -2.96571046e-01 2.43377551e-01 5.05591989e-01
3.18809211e-01 1.75943926e-01 6.08111978e-01 -1.17900681e+00
3.12321097e-01 5.29582441e-01 2.68807113e-01 -3.71111393e-01
-1.23799942e-01 -1.38246850e-03 -5.30433059e-01 5.03126979e-01
5.83848476e-01 9.82952788e-02 -6.47548318e-01 1.43860042e+00
5.35795987e-01 -1.90306641e-02 -2.57778205e-02 5.69377482e-01
3.31497081e-02 8.20111454e-01 4.95285504e-02 -5.23547828e-01
1.28966129e+00 -8.54411602e-01 -4.33047265e-01 4.49631177e-02
3.65004510e-01 -2.46783987e-01 9.48660851e-01 1.01537645e+00
-9.87650454e-01 -2.22308531e-01 -5.21028578e-01 4.45951492e-01
-6.82006121e-01 -4.99807209e-01 2.96843648e-01 6.16933405e-01
-9.96119440e-01 9.36001003e-01 -7.79041171e-01 -3.29823822e-01
1.51389405e-01 3.98165047e-01 -2.32686520e-01 -3.02251503e-02
-1.30205226e+00 6.37745321e-01 3.42772961e-01 -4.72791821e-01
-1.70624328e+00 -5.89399152e-02 -5.93159258e-01 2.58781761e-01
8.75595987e-01 -4.30153787e-01 1.31840158e+00 -1.10010219e+00
-1.65927792e+00 7.50465631e-01 1.02961667e-01 -7.17406511e-01
6.69689000e-01 -1.07784033e-01 -3.88339162e-01 2.12494642e-01
2.77532581e-02 3.62281390e-02 1.56870317e+00 -1.49818861e+00
-9.82045174e-01 -6.77905560e-01 2.90049136e-01 -1.46588907e-01
-2.89252400e-01 3.25138062e-01 3.68986875e-02 -2.63120621e-01
-1.33818030e-01 -8.84468794e-01 -4.63583797e-01 -4.98782605e-01
-4.46947396e-01 -3.88609022e-01 9.78415847e-01 -7.78549910e-01
1.04355097e+00 -1.80044734e+00 2.85071462e-01 3.74260992e-01
4.96929958e-02 -9.34827514e-03 -1.57804921e-01 7.59491384e-01
9.85273421e-02 4.72820342e-01 -4.40577626e-01 -2.58049548e-01
7.14155957e-02 6.40246034e-01 -1.26327828e-01 2.33400092e-01
1.57153130e-01 3.07616115e-01 -7.96915472e-01 -5.78956366e-01
-2.13721767e-02 -5.04425121e-03 -5.47063470e-01 4.86353070e-01
-7.53519058e-01 4.75509375e-01 -8.78704667e-01 7.55336583e-01
6.08331501e-01 -5.32142818e-01 6.39652431e-01 4.00586784e-01
1.65370060e-04 3.30767453e-01 -1.40859163e+00 1.08683777e+00
-5.36753297e-01 -1.16884984e-01 4.68211383e-01 -1.40231335e+00
6.85946882e-01 2.97134191e-01 6.84476197e-01 -3.33489776e-01
-7.94432387e-02 1.82352766e-01 -2.45062992e-01 -7.79183805e-01
-2.48821750e-02 1.57580748e-01 3.15268035e-03 7.98640966e-01
3.49296927e-02 -4.42063948e-03 1.61934137e-01 1.28869578e-01
1.60887551e+00 -9.96221229e-03 5.22606015e-01 -7.08631799e-03
7.11845219e-01 -2.65848059e-02 4.27560151e-01 9.88767803e-01
-1.65187329e-01 -1.80469885e-01 1.07261467e+00 -6.11185908e-01
-1.04653406e+00 -8.85946989e-01 -1.81864686e-02 1.37201548e+00
-4.17752191e-02 -3.17730963e-01 -7.70160139e-01 -1.40181196e+00
8.95395130e-02 7.41395772e-01 -5.05104721e-01 1.02273062e-01
-4.21876490e-01 -5.51252544e-01 -1.53452367e-01 -7.73770176e-03
2.36761332e-01 -1.58487439e+00 -7.11950302e-01 4.91441220e-01
-2.09480207e-02 -1.06851053e+00 2.61834860e-01 4.97754246e-01
-1.02808833e+00 -1.32745814e+00 1.12837046e-01 -6.53171241e-01
4.08537805e-01 -1.74123749e-01 1.19647110e+00 -6.86626276e-03
-3.65655035e-01 6.17869377e-01 -6.64247215e-01 -5.62827066e-02
-7.90332258e-01 1.23562209e-01 -3.73376906e-02 6.20272100e-01
2.24565402e-01 -6.90222085e-01 -2.22873271e-01 2.19689891e-01
-1.37392986e+00 -4.72118825e-01 6.09857917e-01 1.04270899e+00
4.15690005e-01 4.38759863e-01 2.38255262e-01 -1.20548296e+00
7.01893210e-01 -8.06934357e-01 -8.00046563e-01 3.30275744e-01
-5.89411795e-01 3.17410380e-01 1.07611465e+00 -4.40020144e-01
-9.57615495e-01 1.44274339e-01 -5.81050813e-02 -1.28530294e-01
-6.61216021e-01 1.33636713e-01 -4.15182978e-01 2.42999583e-01
4.73273337e-01 6.64690673e-01 1.93197757e-01 -6.97398722e-01
1.12148635e-01 9.56930399e-01 -3.69088978e-01 -7.90677130e-01
6.20719075e-01 4.68483120e-01 -3.26243341e-02 -4.89384860e-01
-5.77013910e-01 -2.93386042e-01 -7.70735323e-01 -2.08863884e-01
5.92529893e-01 -2.33968630e-01 -1.19029474e+00 5.14952242e-01
-8.52485299e-01 -5.86694956e-01 -3.32928658e-01 3.14034410e-02
-8.65313709e-01 5.21974444e-01 -9.06617701e-01 -9.96501982e-01
-7.51199424e-02 -1.34437120e+00 1.10923970e+00 -2.32960775e-01
2.64119804e-01 -7.34433353e-01 -3.69574875e-02 4.73740071e-01
3.01422864e-01 8.11341964e-03 1.27358544e+00 -1.42087030e+00
-8.72924268e-01 -3.17352891e-01 1.66538894e-01 6.40113354e-01
-3.63309048e-02 7.24963546e-02 -7.73202360e-01 -3.98355365e-01
3.65238547e-01 -6.69915438e-01 4.98327553e-01 1.74784381e-02
1.79987800e+00 -8.61750245e-01 -2.15122506e-01 -2.42085271e-02
1.42837965e+00 4.08005804e-01 3.31745088e-01 6.32369459e-01
1.72856957e-01 8.37774038e-01 8.09428453e-01 9.29985523e-01
2.54131794e-01 6.99458480e-01 1.06777477e+00 4.22092110e-01
5.95955849e-01 1.38620526e-01 5.18867314e-01 3.90964419e-01
-5.86708193e-04 -1.43790498e-01 -7.70005286e-01 2.07615867e-01
-1.90640891e+00 -1.15658021e+00 1.15513109e-01 2.40870643e+00
7.94864893e-01 3.38469893e-01 4.76721346e-01 3.54926616e-01
7.72119522e-01 1.75929099e-01 -8.21730316e-01 -5.78475773e-01
4.39068526e-01 5.46275377e-02 2.54177034e-01 7.01618612e-01
-1.17627156e+00 7.42087543e-01 5.94186211e+00 8.39003623e-01
-8.29296827e-01 1.06202289e-02 7.05198288e-01 1.72401041e-01
-1.50208563e-01 7.69899487e-02 -8.62148345e-01 6.38267636e-01
1.30496907e+00 4.43684347e-02 1.07096159e+00 1.31363237e+00
-9.49585885e-02 -4.26116884e-02 -1.28458929e+00 4.46657419e-01
-2.52694428e-01 -9.57317472e-01 1.83511108e-01 3.21832478e-01
3.08295906e-01 5.75993508e-02 -6.43514330e-03 5.84753931e-01
8.78514767e-01 -5.54091513e-01 2.93483377e-01 3.15514475e-01
4.17129725e-01 -6.42206252e-01 3.91939133e-01 9.87794638e-01
-7.78640568e-01 -8.06476235e-01 -2.56708652e-01 2.43183911e-01
-9.17261615e-02 5.90395808e-01 -1.02525485e+00 5.71487308e-01
7.03005195e-01 4.99997705e-01 -3.32804799e-01 7.50233710e-01
-1.02693126e-01 6.05474949e-01 -2.43902847e-01 -2.16316372e-01
2.99743116e-01 -3.40033978e-01 5.59345484e-01 9.06021833e-01
7.49291256e-02 -2.91366637e-01 7.83444822e-01 3.01238269e-01
1.50440708e-01 2.40726098e-01 -5.39136112e-01 1.51377991e-01
3.51726502e-01 1.31632757e+00 -4.38181430e-01 -7.00422943e-01
-5.00892818e-01 8.97624016e-01 4.84586954e-01 1.99211523e-01
-7.05484390e-01 -5.31083755e-02 5.25957525e-01 3.04935902e-01
3.45236450e-01 2.22368296e-02 2.45416343e-01 -1.27347326e+00
5.31579517e-02 -1.38448811e+00 7.58650482e-01 -3.43585253e-01
-1.59609485e+00 6.41865790e-01 -2.12390721e-01 -1.27666593e+00
-6.23003602e-01 -7.20585585e-01 -1.42896101e-01 5.00441670e-01
-1.68094170e+00 -6.66800737e-01 -9.81050506e-02 1.11610103e+00
8.93670559e-01 -2.81121880e-01 1.00920939e+00 -6.08467832e-02
-4.76215988e-01 -1.43817402e-02 2.50554889e-01 -7.92139173e-02
3.15186054e-01 -1.57213330e+00 1.65702865e-01 2.70998836e-01
-1.36671677e-01 2.14788198e-01 5.13601005e-01 -6.41410649e-01
-1.43410754e+00 -9.96918201e-01 7.26194799e-01 -4.38984513e-01
1.02145088e+00 -3.51984173e-01 -8.97926092e-01 8.31863225e-01
-1.37907817e-04 5.90550900e-02 6.49512053e-01 -6.17922097e-02
-3.96208107e-01 -2.89814472e-01 -1.70771551e+00 1.31313816e-01
9.04791892e-01 -2.98953533e-01 -4.23040271e-01 7.64645100e-01
6.21572375e-01 1.77596077e-01 -1.12263620e+00 2.35101357e-01
2.94080019e-01 -1.06937838e+00 6.94140196e-01 -1.40560341e+00
3.75082493e-01 1.50761142e-01 -2.59670317e-01 -1.20387435e+00
-3.53260726e-01 -5.65118313e-01 -6.52991891e-01 7.74780273e-01
3.12416703e-01 -7.43659616e-01 9.01321352e-01 6.10450506e-01
4.18455750e-01 -7.80419171e-01 -9.78542984e-01 -8.96582246e-01
-2.69014746e-01 -2.14694992e-01 6.49017036e-01 8.78085613e-01
-7.57436380e-02 3.37516852e-02 -4.44835484e-01 2.46312454e-01
9.48444784e-01 4.53723311e-01 6.13281071e-01 -1.63289344e+00
-6.76789582e-01 -1.02812126e-01 -1.13296203e-01 -8.63593578e-01
3.54994506e-01 -7.36296952e-01 -1.74039677e-01 -1.31409431e+00
2.76409477e-01 -4.92600769e-01 -2.64684469e-01 5.53245723e-01
2.21569985e-01 -7.41698921e-01 2.57452548e-01 4.01436746e-01
-1.07769334e+00 2.52744198e-01 1.00733900e+00 -1.44276973e-02
-2.35254057e-02 7.19392002e-01 -3.99384387e-02 7.31264770e-01
9.09397602e-01 -7.39808321e-01 -1.42499149e-01 1.91877618e-01
-4.11781855e-03 7.70360351e-01 1.07651360e-01 -7.80453205e-01
-8.37327633e-03 -5.52143455e-01 2.89040744e-01 -3.75644505e-01
4.49101657e-01 -1.31735969e+00 -6.92346320e-02 9.39801574e-01
-6.27237141e-01 -7.30779320e-02 -7.79754519e-01 7.42383122e-01
-1.96320027e-01 -6.88085616e-01 7.61902273e-01 -6.37472153e-01
-4.97651905e-01 4.42580432e-01 -6.56535923e-01 -1.50946081e-01
1.35382402e+00 1.97347298e-01 -1.25158429e-01 -5.67853510e-01
-1.48949766e+00 1.66613027e-01 3.81976277e-01 3.50706458e-01
3.07906359e-01 -1.01954579e+00 -5.49001694e-01 3.58589381e-01
-2.08304599e-02 -3.11485797e-01 -4.06108312e-02 3.30969721e-01
-1.07333325e-01 2.11439759e-01 -2.48808533e-01 -6.21924877e-01
-1.33578503e+00 1.13670373e+00 1.32346317e-01 -1.13445449e+00
-3.01866233e-01 3.17009807e-01 -4.49157655e-01 -6.73946857e-01
4.20731932e-01 2.39218865e-02 -5.15012324e-01 2.76220471e-01
4.91841376e-01 4.31477547e-01 1.14744209e-01 -6.81306794e-02
-9.70512629e-02 5.14997430e-02 -2.82507092e-01 -1.05148621e-01
1.66320205e+00 1.12178668e-01 -3.93670142e-01 3.31506759e-01
1.18048418e+00 -2.47844651e-01 -1.27867162e+00 -4.53264713e-01
4.26889241e-01 -6.46860063e-01 -5.30272305e-01 -7.55618334e-01
-9.18825507e-01 6.03123486e-01 3.97757441e-01 1.35501897e+00
9.70666587e-01 3.65914069e-02 5.30785859e-01 6.27367258e-01
9.38933671e-01 -1.19638419e+00 4.29755867e-01 5.85879803e-01
7.32341409e-01 -1.24623394e+00 -3.26756865e-01 -3.63742441e-01
-8.44921231e-01 1.15628695e+00 5.38339138e-01 -1.80358231e-01
7.86911070e-01 3.87313902e-01 -3.41406971e-01 7.87579715e-02
-1.24636149e+00 -2.20556244e-01 -4.54201102e-01 6.63716197e-01
-2.13982701e-01 9.73223522e-02 -3.59641969e-01 4.37007427e-01
1.47022337e-01 -6.18444718e-02 5.11641622e-01 1.15892482e+00
-8.31436098e-01 -1.62455499e+00 -5.91354549e-01 7.62759387e-01
-6.88841641e-01 1.48930013e-01 -2.34733716e-01 5.22867024e-01
-8.14188793e-02 8.79320860e-01 -1.10812090e-01 -3.20125103e-01
1.24051586e-01 4.57321227e-01 1.35472506e-01 -6.56665623e-01
-5.77496350e-01 1.71616971e-02 2.92247951e-01 -9.80544746e-01
-1.24449283e-01 -1.09354806e+00 -9.00876701e-01 -7.15408325e-02
8.37055072e-02 3.05133790e-01 9.22527134e-01 8.75816166e-01
2.51782686e-01 3.45836669e-01 1.45388055e+00 -6.93371475e-01
-1.21564019e+00 -7.87556767e-01 -7.72574186e-01 6.65338576e-01
2.51796752e-01 -3.82659048e-01 -5.86107850e-01 8.90169740e-02] | [4.156050682067871, 2.0414702892303467] |
ff0e4262-32cb-4ec3-98b1-0c6e73175d7a | contactpose-a-dataset-of-grasps-with-object | 2007.09545 | null | https://arxiv.org/abs/2007.09545v1 | https://arxiv.org/pdf/2007.09545v1.pdf | ContactPose: A Dataset of Grasps with Object Contact and Hand Pose | Grasping is natural for humans. However, it involves complex hand configurations and soft tissue deformation that can result in complicated regions of contact between the hand and the object. Understanding and modeling this contact can potentially improve hand models, AR/VR experiences, and robotic grasping. Yet, we currently lack datasets of hand-object contact paired with other data modalities, which is crucial for developing and evaluating contact modeling techniques. We introduce ContactPose, the first dataset of hand-object contact paired with hand pose, object pose, and RGB-D images. ContactPose has 2306 unique grasps of 25 household objects grasped with 2 functional intents by 50 participants, and more than 2.9 M RGB-D grasp images. Analysis of ContactPose data reveals interesting relationships between hand pose and contact. We use this data to rigorously evaluate various data representations, heuristics from the literature, and learning methods for contact modeling. Data, code, and trained models are available at https://contactpose.cc.gatech.edu. | ['Christopher D. Twigg', 'Samarth Brahmbhatt', 'James Hays', 'Chengcheng Tang', 'Charles C. Kemp'] | 2020-07-19 | null | https://www.ecva.net/papers/eccv_2020/papers_ECCV/html/1889_ECCV_2020_paper.php | https://www.ecva.net/papers/eccv_2020/papers_ECCV/papers/123580358.pdf | eccv-2020-8 | ['human-grasp-contact-prediction'] | ['miscellaneous'] | [-6.82826638e-02 -1.55762240e-01 -1.92745760e-01 -3.50998640e-01
-3.93739045e-01 -8.15185130e-01 2.69123644e-01 -2.34245598e-01
1.59649566e-01 2.04340488e-01 2.60427505e-01 -3.11546102e-02
-2.34487146e-01 -4.39854383e-01 -8.57247353e-01 -3.29914063e-01
-2.56467223e-01 9.51532304e-01 1.30049393e-01 -4.10042070e-02
3.11309606e-01 9.67523277e-01 -1.34748101e+00 6.00447178e-01
4.62522537e-01 1.12935424e+00 6.62352622e-01 4.97024328e-01
1.77985251e-01 1.18161902e-01 -1.78041160e-01 -1.85245216e-01
4.77309227e-01 3.87842357e-01 -8.32434952e-01 -2.98759788e-01
6.42021716e-01 -1.08739769e+00 -5.74291587e-01 3.27476799e-01
7.52198458e-01 8.90039001e-03 7.49348104e-01 -1.31927931e+00
-6.83623672e-01 5.24583638e-01 -4.44244742e-01 -4.31780607e-01
8.61200154e-01 4.10042673e-01 5.82834959e-01 -1.07512224e+00
1.04978144e+00 1.59816325e+00 6.33862078e-01 8.94898057e-01
-1.13184524e+00 -4.83845711e-01 -6.29861280e-02 -4.62216139e-02
-9.85687912e-01 -2.28165746e-01 6.46409869e-01 -6.00917459e-01
1.11181808e+00 2.65884876e-01 9.38515842e-01 1.67979479e+00
3.95651519e-01 1.03227293e+00 9.65975761e-01 -3.42815518e-01
-1.29142672e-01 -2.28959456e-01 1.17489658e-01 5.20299971e-01
1.55860588e-01 2.22603008e-01 -4.89002049e-01 -4.79170382e-01
1.52343774e+00 2.28923246e-01 -2.26210311e-01 -7.54841506e-01
-1.33188498e+00 1.23721221e-02 6.59247220e-01 -9.89925042e-02
-6.26373589e-01 5.01854181e-01 1.00283548e-01 -1.91341955e-02
1.24259874e-01 3.40435296e-01 -6.99960649e-01 -5.05107939e-01
-1.25408411e-01 7.19231784e-01 1.14319098e+00 1.75338888e+00
1.08359903e-01 -5.77066422e-01 -1.34992287e-01 5.87537766e-01
3.25648785e-01 7.70489097e-01 -5.02470315e-01 -1.23555315e+00
6.87298477e-01 5.53623796e-01 5.40014148e-01 -9.53139544e-01
-5.68655193e-01 6.22035444e-01 -1.83355093e-01 4.43739176e-01
5.47323108e-01 1.05028100e-01 -1.08836663e+00 1.17987740e+00
2.18306407e-01 -6.43142700e-01 -4.69411671e-01 1.05494523e+00
9.13144886e-01 5.00184344e-03 1.11262985e-01 3.99404883e-01
1.08831024e+00 -6.59024954e-01 -6.96848452e-01 1.35846836e-02
7.71316066e-02 -8.85769725e-01 1.33533001e+00 4.93938923e-01
-1.44564676e+00 -2.39476979e-01 -6.09015405e-01 -4.88783896e-01
-1.89868629e-01 3.71821433e-01 1.05170512e+00 1.35436252e-01
-5.51160216e-01 1.01379740e+00 -1.25312018e+00 -4.71958458e-01
5.21437109e-01 5.55716693e-01 -3.41527045e-01 -3.18811476e-01
-5.30243278e-01 1.24132454e+00 -3.10720559e-02 2.90965229e-01
-7.67753959e-01 -8.97117913e-01 -3.11328769e-01 -4.22172546e-01
3.50782216e-01 -3.81797642e-01 1.29038656e+00 1.38274744e-01
-1.38525045e+00 8.83584023e-01 -1.22697726e-02 4.71075624e-01
9.33307469e-01 -8.72049987e-01 3.43223572e-01 3.12672675e-01
-4.20136243e-01 6.61429465e-01 7.71137476e-01 -1.82057083e+00
2.47723207e-01 -7.35866606e-01 1.63049385e-01 1.89493652e-02
1.55876905e-01 2.63603032e-02 -4.32823181e-01 -5.68028390e-01
4.82982010e-01 -1.07053530e+00 1.09780453e-01 8.66417170e-01
-4.85407054e-01 -3.06412548e-01 8.84984016e-01 -1.09911060e+00
2.42684498e-01 -1.90464675e+00 5.73336899e-01 3.12853485e-01
8.29650909e-02 -4.73949732e-03 -9.57043990e-02 7.38027930e-01
4.13527489e-01 9.31842029e-02 3.19562882e-01 -2.85345018e-01
2.07077697e-01 2.28142291e-01 -3.52441251e-01 3.52304637e-01
-3.71076986e-02 1.17176473e+00 -7.15171099e-01 -5.66504776e-01
3.52158964e-01 7.39129364e-01 -6.53573036e-01 5.46381235e-01
-4.38431442e-01 6.12882137e-01 -7.75303960e-01 1.46881700e+00
8.55018854e-01 2.27492094e-01 8.98301601e-02 -1.01878035e+00
5.27499840e-02 -1.82119068e-02 -9.02882278e-01 1.95873940e+00
-2.28194401e-01 1.82031408e-01 5.70546746e-01 -7.78060108e-02
8.94191563e-01 3.90997738e-01 9.62698400e-01 -1.47006795e-01
5.15181363e-01 3.83470327e-01 -3.02317858e-01 -8.40234101e-01
1.81460410e-01 3.49461675e-01 4.02085841e-01 5.50862610e-01
-3.08016669e-02 -9.09992993e-01 -4.01267737e-01 -2.39392877e-01
7.80789554e-01 8.56118500e-01 -3.53465140e-01 -3.79565269e-01
-6.40839338e-01 3.24331149e-02 -7.57572576e-02 5.69578528e-01
3.80430184e-02 7.38967955e-01 1.66274697e-01 -4.17869717e-01
-1.23479247e+00 -1.50653398e+00 -4.30290073e-01 8.52641046e-01
4.76786375e-01 -1.29816905e-01 -6.15582228e-01 3.85831036e-02
9.36854064e-01 1.93067223e-01 -6.32236421e-01 -5.97190224e-02
-7.00750411e-01 8.46308917e-02 3.15501124e-01 1.20350885e+00
1.46738082e-01 -1.41829431e+00 -8.52085590e-01 1.09401144e-01
-1.91159233e-01 -1.01731646e+00 -3.68230611e-01 -2.27096062e-02
-1.19921005e+00 -1.34427977e+00 -7.55098164e-01 -5.29874682e-01
7.75069475e-01 1.42575443e-01 1.08755875e+00 2.57941633e-01
-9.80806231e-01 1.03703332e+00 -6.66109979e-01 -5.81329107e-01
-1.84046790e-01 -5.94928041e-02 4.61260647e-01 -1.07559979e+00
1.61010444e-01 -5.48449695e-01 -6.17059886e-01 6.05601013e-01
-5.24648070e-01 5.08145466e-02 7.02465475e-01 4.32535082e-01
5.08223355e-01 -8.89345050e-01 1.08938180e-01 2.05476791e-01
5.58418632e-01 -2.56469220e-01 -1.22668736e-01 5.87986946e-01
1.59726769e-01 -3.51320177e-01 -3.14513296e-01 -8.80363405e-01
-1.03999352e+00 8.66373703e-02 2.41859183e-01 -7.91217387e-01
-5.37794158e-02 1.50248289e-01 -1.78628087e-01 -1.85611695e-01
6.70542419e-01 -4.17140633e-01 4.37250555e-01 -1.04470265e+00
3.45067501e-01 8.91596675e-01 5.75960875e-01 -1.52795386e+00
4.14301604e-01 3.69397432e-01 -7.78517947e-02 -7.40484893e-01
-7.82886222e-02 -6.45774603e-02 -1.48784614e+00 -2.51140028e-01
7.16779172e-01 -3.38306755e-01 -1.45748723e+00 7.29146421e-01
-1.43272638e+00 -7.04086959e-01 -1.94671471e-02 4.72040147e-01
-9.96783972e-01 -1.80856772e-02 -7.86607921e-01 -9.84294236e-01
-4.79346484e-01 -1.20295727e+00 1.34787762e+00 7.84170721e-03
-4.66262311e-01 -2.76790559e-01 -6.58969760e-01 2.80668467e-01
3.66427511e-01 5.77064097e-01 8.66294205e-01 8.87827054e-02
-7.83652425e-01 -4.84960288e-01 -2.51832902e-01 7.80775174e-02
5.82325280e-01 2.10071415e-01 -7.60090113e-01 -7.24791586e-01
-4.79027003e-01 -8.18743944e-01 2.44977489e-01 3.63924205e-01
1.39671254e+00 -1.53486490e-01 -6.91360116e-01 2.55451322e-01
1.05291092e+00 1.93401650e-01 4.56638932e-01 -1.40431285e-01
8.42601597e-01 7.41331398e-01 8.76278758e-01 5.17302930e-01
-8.00913274e-02 7.81552196e-01 5.67678273e-01 2.95857370e-01
-2.69278605e-02 -2.58018941e-01 -8.19689408e-02 3.45002890e-01
-1.10014582e+00 4.26285937e-02 -1.13499153e+00 2.71009445e-01
-1.28918576e+00 -5.76942325e-01 2.92044189e-02 2.00002027e+00
1.07568061e+00 -2.82718331e-01 1.09546311e-01 -1.60062477e-01
1.64138630e-01 -4.69174504e-01 -9.91810262e-01 -4.16853204e-02
3.39066148e-01 3.99270296e-01 3.69778574e-01 3.62267166e-01
-6.84978247e-01 9.93402123e-01 6.54763937e+00 -3.86628732e-02
-9.42032099e-01 -1.79019958e-01 -3.16966236e-01 -3.38324904e-01
-1.80755392e-01 -3.67933244e-01 -4.21643823e-01 -4.24198546e-02
-1.80895686e-01 1.27341151e-01 9.17562842e-01 9.84608471e-01
-2.01388702e-01 -1.32187203e-01 -1.74575484e+00 8.74099374e-01
-2.01222599e-01 -8.42551529e-01 -6.07209243e-02 5.67112176e-04
2.08905801e-01 3.70385125e-02 -8.31679478e-02 -1.11191489e-01
1.19001038e-01 -1.14961624e+00 9.09894645e-01 8.74777019e-01
1.14149833e+00 -5.07803708e-02 2.21652240e-01 1.68312624e-01
-1.18484521e+00 -1.63919345e-01 -3.53547931e-01 1.12015031e-01
2.31245100e-01 -9.77776647e-02 -5.36983132e-01 4.58993942e-01
1.26424789e+00 4.89233136e-01 -1.55146629e-01 6.97709441e-01
2.31518403e-01 -2.00787410e-01 -6.57653689e-01 -9.88238603e-02
-4.59598869e-01 1.33326232e-01 5.98168254e-01 7.30461776e-01
8.33016112e-02 6.08549476e-01 2.60320514e-01 1.29777443e+00
1.76778510e-01 -4.21577811e-01 -5.63499868e-01 -3.71704221e-01
7.91731894e-01 8.35794747e-01 -4.87507135e-01 9.65290964e-02
1.47985891e-01 1.04789150e+00 9.73659158e-02 5.42702258e-01
-3.81791860e-01 -5.38867228e-02 8.56633544e-01 3.15934867e-01
-1.90823764e-01 -9.50015008e-01 -4.43992764e-01 -9.02340353e-01
7.14762211e-01 -6.34134233e-01 -4.76273686e-01 -1.03120303e+00
-1.56932807e+00 1.59414351e-01 6.93075538e-01 -9.49413836e-01
5.06725125e-02 -1.25655282e+00 -2.14266434e-01 1.06531119e+00
-5.34334600e-01 -1.67644346e+00 -1.02090132e+00 3.99172336e-01
2.86569715e-01 3.01144123e-01 1.10496235e+00 -1.53197259e-01
1.24233775e-01 3.85277361e-01 -2.39880651e-01 -8.53949636e-02
6.60309553e-01 -9.23179984e-01 5.19620299e-01 -3.54117095e-01
-6.32006109e-01 1.06375229e+00 5.53596795e-01 -1.14308679e+00
-2.41299438e+00 -3.05705458e-01 -3.11149899e-02 -1.18476343e+00
1.90851316e-01 -5.54319978e-01 -7.92617202e-01 1.06405568e+00
-3.08833778e-01 9.21150893e-02 3.91420200e-02 -2.05752086e-02
-2.13399127e-01 4.57357168e-01 -1.66444135e+00 6.11503482e-01
1.78566217e+00 -5.85739732e-01 -6.19732320e-01 4.42757756e-01
4.62398887e-01 -1.16870296e+00 -1.53189504e+00 7.58703291e-01
1.61613047e+00 -3.32332790e-01 1.55498374e+00 -8.71847332e-01
5.26622832e-01 3.98154765e-01 -5.79234123e-01 -1.08018243e+00
-1.61425069e-01 -3.46266598e-01 -5.81108451e-01 8.88262391e-01
-2.24914327e-01 -3.37610304e-01 7.73894608e-01 1.37737846e+00
-1.63298011e-01 -1.17890251e+00 -6.90243602e-01 -9.49804425e-01
4.35206085e-01 -1.60515904e-01 6.77969277e-01 5.89883804e-01
2.86289185e-01 -8.21636438e-01 2.45008972e-02 1.05371274e-01
8.38318169e-01 2.94619590e-01 8.03501904e-01 -1.45553613e+00
-2.84997690e-02 -3.47096533e-01 -1.02789868e-02 -1.07661605e+00
2.10049693e-02 -5.93050241e-01 3.02522540e-01 -1.87165082e+00
2.17494145e-01 -1.17929828e+00 3.98626983e-01 8.81603420e-01
1.32014602e-01 -3.01864803e-01 3.71820241e-01 5.98814726e-01
4.53542531e-01 4.32403415e-01 1.88625360e+00 -7.38831237e-02
-2.87688851e-01 -8.21372271e-02 1.34111539e-01 5.93920171e-01
8.38310301e-01 9.26922113e-02 -1.27732605e-01 -8.21770132e-01
-3.34330648e-01 2.53658891e-01 8.70623052e-01 -7.25665689e-01
-5.22853248e-02 -5.89381993e-01 7.21832693e-01 -7.36969411e-01
8.37429583e-01 -1.18907738e+00 2.29166716e-01 5.38894117e-01
-2.97272295e-01 -1.25595450e-01 4.27053362e-01 1.50523439e-01
7.06439555e-01 -7.01949745e-02 2.21017376e-01 -4.41469222e-01
-2.60422051e-01 5.79606414e-01 6.86074495e-02 -3.28209043e-01
9.87800479e-01 -3.11069071e-01 -2.18379080e-01 -2.20605470e-02
-1.06363392e+00 2.03073204e-01 7.48666763e-01 1.01307249e+00
7.92388856e-01 -1.43274069e+00 -4.01554108e-01 1.62373409e-01
-5.56688234e-02 4.75370586e-01 6.70110583e-02 5.66604078e-01
-6.24794364e-01 1.78966001e-01 -6.49710417e-01 -7.39307880e-01
-1.29592538e+00 3.20175916e-01 1.72745302e-01 7.95607686e-01
-8.66652906e-01 9.20161843e-01 -5.13562679e-01 -8.04715633e-01
7.58249879e-01 -8.29757392e-01 4.91836578e-01 -4.21424925e-01
2.38551944e-01 8.00609171e-01 -1.12984285e-01 -1.57777786e-01
-4.13446844e-01 7.91731060e-01 7.75995553e-02 -3.16280685e-02
1.38284528e+00 3.90874624e-01 -2.23137751e-01 2.81727195e-01
9.48619425e-01 -3.05626184e-01 -1.59221733e+00 -4.45372146e-03
-2.72035688e-01 -9.56161916e-01 -4.75450039e-01 -1.14537239e+00
-7.98577964e-01 1.03488231e+00 5.93118191e-01 -4.09783691e-01
4.63691443e-01 5.81771910e-01 7.53970981e-01 8.52185428e-01
1.35534251e+00 -1.00204325e+00 4.81569231e-01 3.26343983e-01
2.13731289e+00 -1.04025710e+00 7.35834911e-02 -9.69963431e-01
-3.30678433e-01 1.13246715e+00 9.90436673e-01 -2.84712818e-02
8.57877076e-01 7.72360623e-01 -5.36610410e-02 -4.31419581e-01
-7.78265074e-02 5.55422008e-01 2.35976651e-01 9.25497830e-01
4.06437546e-01 3.46320242e-01 1.46143541e-01 6.29987180e-01
-2.78108329e-01 4.64074403e-01 -4.11032140e-02 1.91251957e+00
-3.34266238e-02 -1.13526595e+00 -4.30879772e-01 6.85577095e-01
-1.24973329e-02 4.00300294e-01 -6.18786216e-01 7.82649219e-01
-3.12452435e-01 5.58194995e-01 9.66553017e-02 -6.03071511e-01
7.50578821e-01 -3.32429260e-02 1.55470562e+00 -4.46171075e-01
-3.70856166e-01 -4.26210880e-01 -1.66061074e-01 -1.05948710e+00
-1.32848635e-01 -8.14769864e-01 -1.52940321e+00 -5.54323316e-01
-4.10973608e-01 -6.90329790e-01 9.78733122e-01 5.86794794e-01
4.53371286e-01 1.52297616e-01 1.40166789e-01 -2.18975329e+00
-1.09275413e+00 -1.37511253e+00 -5.30108333e-01 4.55930382e-01
7.13179633e-02 -1.42468762e+00 3.02314851e-02 1.10969385e-02] | [5.978238105773926, -0.9090684652328491] |
096c91bd-3c2a-434d-a451-9bef27f48a55 | joint-modelling-of-spoken-language | 2305.00926 | null | https://arxiv.org/abs/2305.00926v1 | https://arxiv.org/pdf/2305.00926v1.pdf | Joint Modelling of Spoken Language Understanding Tasks with Integrated Dialog History | Most human interactions occur in the form of spoken conversations where the semantic meaning of a given utterance depends on the context. Each utterance in spoken conversation can be represented by many semantic and speaker attributes, and there has been an interest in building Spoken Language Understanding (SLU) systems for automatically predicting these attributes. Recent work has shown that incorporating dialogue history can help advance SLU performance. However, separate models are used for each SLU task, leading to an increase in inference time and computation cost. Motivated by this, we aim to ask: can we jointly model all the SLU tasks while incorporating context to facilitate low-latency and lightweight inference? To answer this, we propose a novel model architecture that learns dialog context to jointly predict the intent, dialog act, speaker role, and emotion for the spoken utterance. Note that our joint prediction is based on an autoregressive model and we need to decide the prediction order of dialog attributes, which is not trivial. To mitigate the issue, we also propose an order agnostic training method. Our experiments show that our joint model achieves similar results to task-specific classifiers and can effectively integrate dialog context to further improve the SLU performance. | ['Shinji Watanabe', 'Brian Yan', 'Emiru Tsunoo', 'Hayato Futami', 'Siddhant Arora'] | 2023-05-01 | null | null | null | null | ['spoken-language-understanding', 'spoken-language-understanding'] | ['natural-language-processing', 'speech'] | [ 1.64666578e-01 2.95826018e-01 -9.17176083e-02 -1.14052546e+00
-5.17024934e-01 -5.36389351e-01 7.71682203e-01 1.50994122e-01
-2.27686048e-01 6.11823142e-01 8.77457440e-01 -3.13728869e-01
4.18931812e-01 -5.19491076e-01 -2.14511514e-01 -2.02860415e-01
1.34948999e-01 5.97624719e-01 1.57405674e-01 -4.90996987e-01
1.39339700e-01 -5.24747856e-02 -1.30258667e+00 4.78682935e-01
8.17513406e-01 1.00355935e+00 3.06020081e-01 9.06702101e-01
-6.29011035e-01 1.15291870e+00 -6.98569715e-01 -9.31563377e-02
-2.29057878e-01 -7.59057522e-01 -1.33427060e+00 2.25377083e-01
-3.12145031e-03 -5.79038918e-01 -8.49541649e-02 5.10500133e-01
3.62939209e-01 4.10939872e-01 6.80237770e-01 -1.23932970e+00
-1.60097003e-01 8.55547488e-01 1.89590007e-01 -9.22977179e-02
6.66670322e-01 8.03176593e-03 1.29050171e+00 -6.81445897e-01
2.33813748e-01 1.64456391e+00 2.51763791e-01 8.00430417e-01
-1.07148325e+00 -3.54518920e-01 6.99561775e-01 8.42158124e-02
-9.16082740e-01 -7.38367438e-01 6.50470555e-01 -2.51829863e-01
1.01994896e+00 3.35260749e-01 3.24548811e-01 1.05866206e+00
-1.11319378e-01 1.19463134e+00 9.69966471e-01 -5.82686067e-01
3.65681767e-01 3.13321263e-01 6.38112605e-01 4.32892770e-01
-8.18138361e-01 -5.74227691e-01 -5.53594291e-01 -3.10406536e-01
3.49853814e-01 -1.42234927e-02 -2.35652670e-01 5.78365773e-02
-8.00479352e-01 1.03199279e+00 1.78096175e-01 2.31394172e-01
-1.14117377e-01 -2.21025348e-01 4.30742353e-01 4.98185575e-01
4.97746170e-01 3.83457780e-01 -7.00618804e-01 -5.42415082e-01
-1.78726777e-01 7.00748265e-02 1.60120511e+00 7.75574744e-01
9.00273323e-01 -2.73261130e-01 -2.26232320e-01 1.35585833e+00
3.45584273e-01 1.42459482e-01 4.33767617e-01 -1.14640832e+00
1.71889707e-01 7.57161558e-01 9.90993306e-02 -4.67320383e-01
-4.75396454e-01 3.76057386e-01 -4.95446950e-01 -2.36450553e-01
3.99024129e-01 -4.41578239e-01 -5.87761343e-01 1.89674532e+00
4.24515665e-01 1.91559404e-01 3.09449375e-01 8.15362632e-01
7.62593985e-01 8.87599051e-01 3.32314312e-01 -2.27118373e-01
1.51284420e+00 -9.31930363e-01 -9.18987393e-01 -6.25472486e-01
8.01523566e-01 -7.39393532e-01 1.29173183e+00 9.58813075e-03
-7.99953580e-01 -3.33962858e-01 -8.65775645e-01 -2.72206783e-01
-2.48002633e-02 -1.08905107e-01 6.28415108e-01 4.23510849e-01
-9.33304310e-01 1.15175746e-01 -5.70255995e-01 -5.05373180e-01
-4.38081831e-01 3.58266354e-01 3.66142057e-02 8.48468170e-02
-1.60785627e+00 9.78991568e-01 2.45298058e-01 -9.03165957e-04
-4.39838290e-01 -1.46148026e-01 -1.01704359e+00 3.58579725e-01
6.54452622e-01 -4.12790358e-01 1.82859647e+00 -8.23659480e-01
-2.12333560e+00 3.14086676e-01 -6.92575872e-01 -3.72139037e-01
-6.44861013e-02 -1.12649553e-01 -6.88076541e-02 -1.72129899e-01
-3.42851341e-01 6.51779950e-01 5.45503497e-01 -1.05828702e+00
-8.48875880e-01 -3.24921906e-01 5.35128713e-01 6.12659276e-01
-5.78415930e-01 1.34331554e-01 -3.84405226e-01 -2.03950360e-01
3.02533172e-02 -9.90327597e-01 -2.19099298e-01 -4.26973253e-01
-7.86537398e-03 -8.91416490e-01 9.11733329e-01 -6.12886965e-01
1.37000287e+00 -2.14358687e+00 5.74256778e-02 -6.80146813e-02
4.82088402e-02 1.10183366e-01 -3.31443883e-02 5.47310948e-01
6.26138449e-01 -3.12988497e-02 -2.05491573e-01 -5.66470444e-01
9.45499465e-02 6.45818532e-01 -5.60669065e-01 -3.48719776e-01
1.24675930e-01 6.43271387e-01 -7.84996152e-01 -4.21573907e-01
1.12744153e-01 3.30337048e-01 -6.78515971e-01 9.21996832e-01
-7.66330540e-01 6.19220793e-01 -6.67991519e-01 2.54226029e-01
1.32690221e-01 -3.56902957e-01 5.74719429e-01 1.36642218e-01
1.21923909e-01 9.25631881e-01 -9.84781623e-01 1.45805800e+00
-8.91178787e-01 3.79104018e-01 5.87595165e-01 -7.36839235e-01
1.03422856e+00 5.35906494e-01 6.84918687e-02 -2.38924935e-01
2.01447323e-01 -3.81103680e-02 1.80511177e-01 -4.07369375e-01
4.78837937e-01 -6.39108717e-02 -3.90315503e-01 8.85976493e-01
-3.94455669e-03 -1.86526462e-01 -1.66841179e-01 2.82358646e-01
8.36889029e-01 -2.65883893e-01 4.58816737e-01 -3.53009552e-02
7.11223423e-01 -2.94239372e-01 5.50242364e-01 6.39534235e-01
-3.52522165e-01 6.35385588e-02 6.29220009e-01 -4.01991159e-01
-4.40815479e-01 -6.31987095e-01 8.74034017e-02 1.78072107e+00
1.19644001e-01 -5.39135218e-01 -8.84904265e-01 -6.20141625e-01
-2.58965582e-01 1.02143812e+00 -2.36055940e-01 -4.99031991e-02
-7.00871170e-01 -3.53355050e-01 1.28734455e-01 6.55849576e-01
4.43082899e-01 -1.14823401e+00 -1.73757419e-01 4.21880186e-01
-5.34399748e-01 -1.41533053e+00 -7.69342959e-01 5.56759983e-02
-6.05253160e-01 -7.11806536e-01 -1.39404401e-01 -7.05388069e-01
3.34454000e-01 1.83980212e-01 1.00535953e+00 3.47284786e-02
3.08759809e-01 6.19233966e-01 -6.11384392e-01 -5.03774464e-01
-9.48008239e-01 2.09743857e-01 1.45521387e-01 1.61546901e-01
6.00655019e-01 -4.84173030e-01 -5.50446749e-01 5.69282115e-01
-6.58454478e-01 1.35890603e-01 3.54759246e-01 9.28847849e-01
-9.86095667e-02 -3.46859366e-01 8.90877306e-01 -8.89682233e-01
9.47529912e-01 -3.43822837e-01 -2.60542214e-01 4.71705168e-01
-4.07704949e-01 2.43953705e-01 6.50694251e-01 -3.89862448e-01
-1.36615229e+00 1.99400410e-01 -3.47977370e-01 1.43517911e-01
-3.95322919e-01 5.86254239e-01 -3.63897711e-01 2.72348374e-01
1.86668500e-01 1.64931804e-01 3.82659465e-01 -4.41464841e-01
4.61617976e-01 1.26614046e+00 1.49948895e-01 -7.23891675e-01
-3.17255855e-02 5.85798845e-02 -5.88230014e-01 -1.23857272e+00
-1.15201354e+00 -6.97678626e-01 -5.46222031e-01 -1.88070223e-01
8.93400073e-01 -9.34093773e-01 -1.07412314e+00 2.37132072e-01
-1.35044265e+00 -6.21379733e-01 3.17836463e-01 4.14371908e-01
-4.35746342e-01 4.72052693e-01 -9.02083576e-01 -1.27013433e+00
-3.95871997e-01 -1.27849460e+00 9.21635747e-01 1.04309000e-01
-8.13199282e-01 -9.95567083e-01 -1.78426579e-01 6.96892321e-01
5.07297039e-01 -6.96746588e-01 1.02782774e+00 -1.24006045e+00
-4.61170793e-01 5.55568747e-03 -1.12209499e-01 3.28641474e-01
2.67952234e-01 -4.40693021e-01 -1.28714633e+00 -1.24722071e-01
3.85651320e-01 -6.63528144e-01 4.65060800e-01 1.24040328e-01
9.42660511e-01 -5.82984269e-01 -5.95749989e-02 -4.86001559e-02
5.01762629e-01 4.22640711e-01 9.97207761e-02 -3.28876913e-01
4.78203028e-01 1.03891754e+00 5.97490788e-01 5.65863907e-01
9.53365266e-01 7.92580545e-01 1.75123496e-04 2.92626113e-01
2.80899942e-01 -1.03352062e-01 5.50678849e-01 9.93460834e-01
3.77326906e-01 -2.27825657e-01 -8.29887271e-01 3.62074137e-01
-2.02657247e+00 -5.80512583e-01 2.64833897e-01 1.95952880e+00
1.13870692e+00 3.25737037e-02 1.42944917e-01 -2.56343842e-01
4.23327416e-01 3.23860675e-01 -4.38880950e-01 -5.98634779e-01
3.53780061e-01 -2.23763824e-01 -2.91587830e-01 1.05292034e+00
-9.27516639e-01 1.14320743e+00 6.34037352e+00 4.46112543e-01
-1.10215843e+00 1.36077506e-02 7.92206347e-01 2.40976393e-01
-2.08016813e-01 1.72664687e-01 -1.02995646e+00 3.12262625e-01
1.08484340e+00 -2.03522414e-01 3.78088355e-01 7.91079879e-01
2.28358001e-01 -2.20820576e-01 -1.39920926e+00 5.37189603e-01
1.84198737e-01 -8.08453262e-01 -3.24508408e-04 -1.55141026e-01
2.54601598e-01 -3.34149450e-01 -4.08922553e-01 6.63841367e-01
5.87909758e-01 -7.40549386e-01 2.14915257e-02 1.33504450e-01
1.91255271e-01 -6.02080047e-01 6.94285750e-01 7.20099866e-01
-1.08302462e+00 -1.60293654e-01 -1.77368879e-01 -4.78104174e-01
2.22590968e-01 9.31314528e-02 -1.61151969e+00 1.09980274e-02
3.29558820e-01 1.65140033e-01 1.13035925e-02 2.68160939e-01
-2.45986328e-01 8.12751353e-01 -3.25189680e-01 -4.63930488e-01
2.50228822e-01 -2.20430404e-01 3.80866468e-01 1.05034304e+00
8.25455040e-02 6.60491228e-01 7.93427467e-01 4.61853564e-01
-1.82914704e-01 2.37588465e-01 -5.49288690e-01 4.19889577e-02
7.68471479e-01 1.01261175e+00 -3.67803037e-01 -4.88406867e-01
-7.82919884e-01 1.03229558e+00 3.17294568e-01 2.55450755e-01
-2.68458456e-01 1.47074580e-01 8.58677208e-01 -2.02082843e-01
-4.60905284e-02 -3.39058220e-01 -2.27791861e-01 -1.16394389e+00
-1.43981963e-01 -7.58482575e-01 4.61723715e-01 -3.97979140e-01
-1.29446626e+00 6.86640799e-01 -6.01117611e-02 -6.98465228e-01
-9.77795124e-01 -4.45990026e-01 -8.35391164e-01 8.35615873e-01
-1.28009748e+00 -1.05379975e+00 -1.66509286e-01 3.30023646e-01
1.15701425e+00 -5.34564666e-02 1.17145622e+00 -2.21904010e-01
-4.68051970e-01 4.39729005e-01 -3.85688007e-01 3.07451099e-01
1.05952811e+00 -1.29016447e+00 2.76101828e-01 3.04611981e-01
-6.79178908e-02 6.29006863e-01 8.53065073e-01 -4.51054007e-01
-1.25641918e+00 -7.35637963e-01 1.36281061e+00 -5.83381534e-01
6.49757147e-01 -5.40861070e-01 -1.22762311e+00 8.24618220e-01
4.50608522e-01 -5.57695925e-01 1.09370005e+00 7.20988512e-01
-1.05357260e-01 6.18190505e-02 -7.04039812e-01 7.67130435e-01
7.39265144e-01 -7.63435841e-01 -8.90021503e-01 1.80672079e-01
9.68115866e-01 -2.35413179e-01 -7.97625899e-01 3.06138903e-01
3.88357311e-01 -7.30904639e-01 6.94610596e-01 -4.63172197e-01
2.55651444e-01 1.40044138e-01 -2.03058124e-01 -1.28423119e+00
1.46898687e-01 -5.91349304e-01 -3.01813632e-01 1.30866075e+00
5.35832703e-01 -6.23478830e-01 6.60142004e-01 1.61413169e+00
-6.83625937e-02 -7.67829955e-01 -7.80387938e-01 -3.02448601e-01
-1.36032537e-01 -2.73635060e-01 6.04342580e-01 8.26154590e-01
6.27319992e-01 1.30845380e+00 -5.87171257e-01 1.25309721e-01
7.26014078e-02 4.03686315e-01 8.90901208e-01 -1.32437801e+00
-2.93717206e-01 -2.79947191e-01 7.30048344e-02 -1.78004515e+00
5.37930548e-01 -5.07180870e-01 3.93063188e-01 -1.38118136e+00
6.11519553e-02 -3.26072812e-01 4.98258770e-02 5.65149724e-01
-6.66743934e-01 -7.05367863e-01 1.90997466e-01 2.47904986e-01
-6.92898750e-01 9.45216656e-01 9.45396602e-01 -4.08411175e-02
-4.74978954e-01 3.75603735e-01 -5.10676861e-01 7.39472091e-01
6.60298884e-01 2.02421129e-01 -6.63321793e-01 -3.07159036e-01
-2.89982706e-01 5.55194914e-01 -1.67812511e-01 -5.05618811e-01
5.55886030e-01 -2.55987316e-01 -2.18208730e-01 -3.19877237e-01
7.90901780e-01 -7.83380568e-01 -7.23935783e-01 1.10497005e-01
-8.11532259e-01 -3.58190298e-01 -8.68996754e-02 5.13113618e-01
-5.22154093e-01 -3.01886201e-01 5.22154450e-01 -1.02991477e-01
-8.94566655e-01 8.31693336e-02 -7.71086037e-01 -1.86614543e-01
7.19965398e-01 1.31499603e-01 6.65207626e-03 -1.19383681e+00
-7.67631412e-01 8.17257464e-01 1.44474208e-01 4.91903663e-01
4.98417288e-01 -8.60215306e-01 -5.55749655e-01 1.88343540e-01
8.03163350e-02 3.41071710e-02 2.04406872e-01 3.82907987e-01
1.93976521e-01 4.75042880e-01 1.91648528e-01 -4.78232235e-01
-1.46966445e+00 2.11048290e-01 1.06037594e-01 -3.28092664e-01
-2.63745546e-01 8.94629896e-01 5.92709422e-01 -8.04846644e-01
5.66511691e-01 -2.20893443e-01 -5.48287272e-01 1.62791431e-01
6.70378149e-01 -4.51240130e-02 -2.28946254e-01 -3.97224218e-01
-7.58138970e-02 -4.11689766e-02 -3.94310832e-01 -2.35048726e-01
1.01020145e+00 -5.68914890e-01 -8.97683576e-02 7.92141080e-01
1.01102257e+00 -2.01643899e-01 -1.23970962e+00 -6.45705879e-01
1.83260128e-01 -3.91186029e-02 -1.93993777e-01 -7.86598623e-01
-3.64835113e-01 9.70571816e-01 7.39076734e-02 4.69230562e-01
9.14859951e-01 2.10405275e-01 1.12231004e+00 7.67481029e-01
3.93395185e-01 -1.18799090e+00 2.26144984e-01 1.30811906e+00
9.45825040e-01 -1.62058520e+00 -3.76420885e-01 -7.12306798e-01
-1.13468206e+00 1.00953424e+00 1.05530274e+00 4.51745957e-01
6.91401839e-01 2.24601403e-01 4.94070590e-01 5.53884171e-02
-1.21770036e+00 -1.47873074e-01 1.08041249e-01 5.38624078e-02
7.89040327e-01 4.24078247e-03 -1.35989398e-01 7.41382420e-01
-1.34392157e-01 -2.95790017e-01 3.51708084e-01 9.35100853e-01
-7.24746764e-01 -1.58471417e+00 -1.97238415e-01 3.92957807e-01
-1.74670875e-01 -3.99515182e-02 -8.16694736e-01 2.42673725e-01
-6.49354577e-01 1.35726511e+00 7.53207803e-02 -4.90302265e-01
2.06281632e-01 7.68962622e-01 -5.22769019e-02 -8.79319608e-01
-6.27199709e-01 1.71703234e-01 5.56316674e-01 -3.56807292e-01
-2.57818669e-01 -3.29362333e-01 -1.35934472e+00 -8.17715898e-02
-2.25033790e-01 4.90124255e-01 4.45068598e-01 1.30742323e+00
3.54497641e-01 2.62033820e-01 1.05603278e+00 -6.27277076e-01
-8.67373586e-01 -1.31002975e+00 -1.76474333e-01 3.64203036e-01
2.22530946e-01 -3.79219711e-01 -3.93075138e-01 -1.26173357e-02] | [12.7003755569458, 7.706036567687988] |
b7f23b73-93a3-4954-ad7e-a730d202138b | deploying-a-retrieval-based-response-model | 2210.14379 | null | https://arxiv.org/abs/2210.14379v1 | https://arxiv.org/pdf/2210.14379v1.pdf | Deploying a Retrieval based Response Model for Task Oriented Dialogues | Task-oriented dialogue systems in industry settings need to have high conversational capability, be easily adaptable to changing situations and conform to business constraints. This paper describes a 3-step procedure to develop a conversational model that satisfies these criteria and can efficiently scale to rank a large set of response candidates. First, we provide a simple algorithm to semi-automatically create a high-coverage template set from historic conversations without any annotation. Second, we propose a neural architecture that encodes the dialogue context and applicable business constraints as profile features for ranking the next turn. Third, we describe a two-stage learning strategy with self-supervised training, followed by supervised fine-tuning on limited data collected through a human-in-the-loop platform. Finally, we describe offline experiments and present results of deploying our model with human-in-the-loop to converse with live customers online. | ['Patrick Ernst', 'Pavel Danchenko', 'Jorge Balazs', 'Cheng Wang', 'György Szarvas', 'Lahari Poddar'] | 2022-10-25 | null | null | null | null | ['task-oriented-dialogue-systems'] | ['natural-language-processing'] | [ 2.71184146e-01 5.31314194e-01 -3.00470255e-02 -1.02360654e+00
-8.94092441e-01 -7.18779087e-01 6.02855444e-01 -2.15984941e-01
-1.90346241e-01 9.21903372e-01 4.38316762e-01 -2.96661168e-01
-1.10376082e-01 -4.25938994e-01 -1.19566247e-02 7.09533468e-02
-5.99829145e-02 1.29399550e+00 1.20662220e-01 -6.54263020e-01
4.06802475e-01 1.37850881e-01 -1.38093102e+00 7.87829459e-01
6.85511410e-01 8.39783549e-01 4.30469513e-01 8.97186041e-01
-2.12962896e-01 8.63293767e-01 -8.13192308e-01 -4.33543772e-01
1.44858152e-01 -6.31911099e-01 -1.38752234e+00 4.59396452e-01
-1.41055435e-01 -4.73328412e-01 -5.61098717e-02 2.61016190e-01
5.07067561e-01 3.50172758e-01 3.07584494e-01 -1.22556210e+00
-2.65466273e-01 8.76795232e-01 3.98494929e-01 -1.69790134e-01
8.58488619e-01 4.14608985e-01 1.01558280e+00 -7.05190897e-01
7.76289105e-01 1.20571780e+00 4.98781353e-01 9.60193694e-01
-1.05239916e+00 -8.06078240e-02 1.51162043e-01 -2.11660564e-02
-6.31583095e-01 -6.57721043e-01 8.42258334e-01 -3.51269394e-01
1.24717033e+00 5.34702063e-01 5.86196959e-01 1.29336989e+00
-2.71186799e-01 7.76635945e-01 9.77090061e-01 -5.02707481e-01
1.57290027e-01 6.60855889e-01 1.79758370e-01 5.06706536e-01
-6.82510197e-01 -3.88358116e-01 -8.30436647e-01 -4.94108826e-01
4.69256580e-01 -3.84413391e-01 -7.44434446e-02 -2.33423874e-01
-1.01094925e+00 8.58273625e-01 -2.23084852e-01 2.27355406e-01
-3.55944723e-01 -4.03141201e-01 5.96881628e-01 7.33375728e-01
3.72699350e-01 8.60651076e-01 -8.26437950e-01 -7.67218530e-01
-3.72041643e-01 2.85288662e-01 1.84601879e+00 1.22370553e+00
6.23358607e-01 -3.88359576e-01 -1.94171652e-01 1.50011659e+00
8.89526680e-02 -2.47952100e-02 4.77345318e-01 -1.30052602e+00
6.26424134e-01 5.77534199e-01 3.89425904e-01 -5.12416124e-01
-5.94800353e-01 -9.27138422e-03 -3.88879150e-01 -3.50277334e-01
5.04882574e-01 -5.31393170e-01 -1.00139201e-01 1.51224899e+00
2.77293384e-01 -6.04016542e-01 1.27307430e-01 8.13172936e-01
6.23641372e-01 5.55566013e-01 -3.94759029e-01 -5.48359275e-01
1.20555651e+00 -1.40401375e+00 -7.01146483e-01 -3.15284193e-01
6.51024580e-01 -6.78761959e-01 1.48178375e+00 5.22783518e-01
-1.24234760e+00 -4.87567484e-01 -7.77078331e-01 2.09158733e-01
-1.18839219e-01 -2.38648579e-02 8.96773338e-01 6.88186467e-01
-1.02368236e+00 5.12612581e-01 -1.87984720e-01 -7.07698286e-01
-4.01893824e-01 5.77751577e-01 -1.39648110e-01 3.26315731e-01
-1.29012835e+00 1.09705698e+00 1.65421918e-01 1.96991652e-01
-6.67334259e-01 -3.56653705e-02 -6.35283649e-01 -1.05693825e-02
5.40887177e-01 -6.22157216e-01 2.13532782e+00 -1.13249254e+00
-2.49179268e+00 7.19205201e-01 -7.72614311e-03 -2.92940676e-01
6.13446712e-01 -1.16928564e-02 -3.07261378e-01 -5.30862659e-02
-1.55203536e-01 3.65845501e-01 1.91185042e-01 -1.20085609e+00
-1.00050998e+00 3.15832272e-02 4.98766005e-01 5.26247501e-01
-4.42160368e-01 1.09527349e-01 -6.49189651e-01 1.09355263e-01
-9.28471163e-02 -1.05668390e+00 -3.54425609e-01 -7.13397145e-01
-3.98020893e-01 -4.61025715e-01 3.60700965e-01 -4.06373709e-01
1.17775524e+00 -1.58287609e+00 6.12417720e-02 3.53921413e-01
-1.80441782e-01 7.09698498e-02 -3.57901245e-01 1.04961133e+00
3.76848966e-01 -1.75778255e-01 1.57734305e-01 -3.19126397e-01
2.86195040e-01 7.14483187e-02 4.99548465e-02 -1.57962024e-01
1.78588971e-01 6.09994233e-01 -9.10698354e-01 -3.46394002e-01
3.69454660e-02 -1.85280442e-01 -7.46370316e-01 8.28944445e-01
-4.67191428e-01 5.72841287e-01 -4.47043031e-01 3.97577852e-01
-1.51675288e-02 -2.19504833e-01 7.70851552e-01 2.13692188e-01
-4.93992586e-04 8.56182635e-01 -8.76147985e-01 1.43824542e+00
-1.02921927e+00 3.81520480e-01 4.71481711e-01 -6.61355674e-01
1.22376966e+00 3.49960059e-01 5.01439452e-01 -6.83616281e-01
2.50853449e-01 2.81256169e-01 -9.17423423e-03 -7.89203107e-01
8.17883670e-01 3.04514408e-01 -5.38530052e-01 1.02483153e+00
1.63735271e-01 -2.33735248e-01 4.03512776e-01 3.04133482e-02
1.14125931e+00 -1.86029449e-01 1.48211762e-01 1.37943476e-01
7.58577645e-01 6.30327687e-02 3.49086046e-01 8.77067983e-01
-1.66838169e-01 3.11176687e-01 7.03371108e-01 -6.16865456e-01
-9.13149714e-01 -2.87966758e-01 2.42290676e-01 1.73604989e+00
-1.43604100e-01 -2.69448996e-01 -8.48178625e-01 -8.01730275e-01
-2.55230814e-01 8.43474030e-01 -2.52207935e-01 1.19833998e-01
-6.97719038e-01 -2.74927706e-01 2.78795242e-01 1.61767334e-01
5.59222043e-01 -1.41998672e+00 -3.69371295e-01 6.46319807e-01
-6.66084111e-01 -1.02946866e+00 -6.87888384e-01 4.02961731e-01
-4.12813604e-01 -9.61137414e-01 -2.24836022e-01 -1.04525828e+00
3.99997056e-01 3.85127515e-02 1.22022843e+00 2.15306543e-02
1.42948031e-01 4.50996011e-01 -5.24077415e-01 -2.30736062e-01
-9.76041138e-01 5.72246194e-01 -3.68715301e-02 -1.37961119e-01
4.66389954e-01 -4.75025386e-01 -2.88507104e-01 9.91463840e-01
-2.20262468e-01 1.80078760e-01 3.94743830e-01 1.11874664e+00
-5.57073355e-02 -3.27128261e-01 1.04208934e+00 -1.26207995e+00
1.56338453e+00 -2.43038088e-01 -2.08767831e-01 4.09247994e-01
-5.59496403e-01 -1.73597127e-01 6.04926705e-01 -5.00620067e-01
-1.21235216e+00 3.48364234e-01 -2.79865623e-01 3.53882521e-01
-2.14085445e-01 5.22520185e-01 -2.69800037e-01 3.33730340e-01
7.55587459e-01 -1.02904979e-02 1.77591950e-01 -3.34053785e-01
3.91804129e-01 1.44998550e+00 3.27807009e-01 -8.46619964e-01
3.47953916e-01 -1.50853470e-01 -8.46339166e-01 -5.87318778e-01
-6.52611315e-01 -7.14821398e-01 -7.26947069e-01 -5.08789837e-01
2.36060858e-01 -5.93100667e-01 -1.27413952e+00 3.03148597e-01
-1.18276441e+00 -9.96926069e-01 2.98422221e-02 1.54282734e-01
-1.09413898e+00 1.61127776e-01 -9.06972289e-01 -1.23222733e+00
-6.14132762e-01 -8.28775823e-01 5.75787306e-01 2.32794896e-01
-9.15806293e-01 -8.12148213e-01 2.49066893e-02 9.39505160e-01
5.85670054e-01 -3.73492628e-01 5.74358046e-01 -1.31400347e+00
-2.67976224e-01 -3.17016959e-01 1.96148857e-01 2.76073724e-01
9.57447514e-02 -2.13680550e-01 -8.47770751e-01 -1.99918002e-01
-8.62122700e-02 -8.39771569e-01 9.55522507e-02 -1.77866220e-01
9.15355206e-01 -7.63063431e-01 -8.69583786e-02 5.82299614e-03
4.72666532e-01 4.12314415e-01 1.81630209e-01 4.12694693e-01
1.22965649e-01 1.24541402e+00 1.01004732e+00 5.93220413e-01
7.42940485e-01 8.66920888e-01 2.62683694e-04 1.73044458e-01
4.08939481e-01 -2.87430614e-01 3.44988376e-01 1.12447798e+00
1.50628507e-01 -4.31385010e-01 -6.45313203e-01 5.20383954e-01
-2.18589640e+00 -9.60998714e-01 2.81401128e-01 1.89441121e+00
1.10196781e+00 3.87425363e-01 6.32158697e-01 -2.89646268e-01
7.37021804e-01 -2.07902625e-01 -4.86362427e-01 -9.88533556e-01
2.64448404e-01 -2.30337188e-01 1.54858911e-02 6.79168701e-01
-7.84578145e-01 9.89290893e-01 6.70587349e+00 2.44346410e-01
-8.58229399e-01 3.90393659e-02 7.73858726e-01 -8.30629468e-02
-3.10219884e-01 -1.20020233e-01 -6.75246358e-01 3.38222384e-01
1.28207839e+00 -1.73618674e-01 8.75585139e-01 1.05331159e+00
4.62311566e-01 5.44704311e-02 -1.49644423e+00 6.04017615e-01
1.31968679e-02 -1.32139206e+00 -3.91776621e-01 2.72729024e-02
4.17046160e-01 -1.80681780e-01 -3.55584919e-01 6.85662150e-01
7.53601611e-01 -6.95081234e-01 4.41019088e-01 3.03601652e-01
5.84795892e-01 -7.08297789e-01 9.24592972e-01 7.98258603e-01
-5.41181147e-01 -3.67227525e-01 -1.45073175e-01 -1.84959590e-01
3.69730800e-01 5.89945614e-02 -1.64050984e+00 2.21914694e-01
3.52071524e-01 3.51749286e-02 5.53633012e-02 6.03526056e-01
-6.08384423e-02 3.51550698e-01 -1.06288522e-01 -7.69924045e-01
1.11685492e-01 -2.05135345e-01 2.69345969e-01 1.41836154e+00
5.52688278e-02 1.80768952e-01 5.19504428e-01 4.76210177e-01
-9.13764834e-02 2.62842327e-01 -4.70364869e-01 -4.80891280e-02
8.07429135e-01 1.38710153e+00 -3.91365230e-01 -6.40147924e-02
-2.97116280e-01 9.43593323e-01 4.14408594e-01 2.24607661e-01
-4.86328304e-01 -5.82290828e-01 4.17645663e-01 8.68132561e-02
1.85876451e-02 -5.69845214e-02 1.78149180e-03 -8.45345318e-01
1.20893598e-01 -1.48256588e+00 2.81285584e-01 -5.42389452e-01
-1.47853172e+00 9.77759600e-01 -2.11792171e-01 -9.83264625e-01
-1.04528093e+00 -5.19809008e-01 -5.87964594e-01 7.94937968e-01
-9.49799895e-01 -1.14004111e+00 -2.37206668e-01 4.28232223e-01
1.07675838e+00 -4.84044373e-01 1.21487164e+00 8.79527405e-02
-4.94781226e-01 6.46573961e-01 -1.03170633e-01 4.49462086e-02
8.46735418e-01 -1.06843400e+00 3.88822287e-01 2.94921473e-02
-2.70527273e-01 9.40353453e-01 8.19864690e-01 -3.51829916e-01
-1.32920575e+00 -6.78231478e-01 1.14616477e+00 -4.33452606e-01
6.87371969e-01 -7.96885550e-01 -5.94470739e-01 6.00455523e-01
4.37499225e-01 -9.28205192e-01 9.32501435e-01 8.15625489e-01
2.00954471e-02 -1.83370888e-01 -1.21331227e+00 5.55833817e-01
1.02704811e+00 -7.10212350e-01 -5.76175988e-01 6.63441539e-01
6.38792872e-01 -7.45203495e-01 -8.20353925e-01 -2.53931377e-02
7.41678119e-01 -8.47386420e-01 4.67631906e-01 -6.22792006e-01
3.02956253e-01 2.41186231e-01 6.34629428e-02 -1.39505601e+00
-1.61971986e-01 -1.55532205e+00 1.07616089e-01 1.36201489e+00
8.65234017e-01 -6.89143658e-01 8.46684515e-01 1.08988631e+00
-2.38353729e-01 -7.77398348e-01 -6.50151193e-01 -4.15669858e-01
-2.78268456e-01 -1.86577544e-01 7.27092683e-01 7.60360718e-01
1.02325630e+00 8.54928792e-01 -6.12258136e-01 -4.49168354e-01
-1.53150618e-01 1.42817363e-01 1.28986216e+00 -1.15515018e+00
-5.16826689e-01 -2.44860873e-01 3.30043525e-01 -1.52865982e+00
3.06484252e-01 -3.80045950e-01 5.30424714e-01 -1.55071747e+00
1.36627480e-01 -5.51189482e-01 6.08982891e-02 4.43768114e-01
1.81891695e-01 -2.45057091e-01 4.91917431e-02 2.29295194e-01
-8.06778550e-01 3.22767377e-01 1.04641950e+00 -2.85983901e-03
-7.49477267e-01 6.89661264e-01 -8.52172673e-01 4.35356498e-01
9.68029559e-01 -4.13014330e-02 -6.34257972e-01 -6.10644594e-02
1.60720035e-01 5.73716819e-01 -3.48272443e-01 -4.44322735e-01
4.05005366e-01 -4.14710045e-01 -1.68458998e-01 -4.88487214e-01
4.90188599e-01 -6.14351273e-01 -2.61709929e-01 -1.80021934e-02
-9.87298369e-01 6.63470430e-03 -6.43551648e-02 3.40962261e-01
-1.02308981e-01 -4.27678913e-01 2.89377481e-01 -2.55311579e-01
-5.63096225e-01 -2.77709186e-01 -8.52828741e-01 -1.50722399e-01
8.12172174e-01 -2.28278413e-01 -2.52771348e-01 -1.09743989e+00
-7.75752664e-01 7.49287486e-01 2.28031859e-01 7.13793993e-01
2.64250219e-01 -8.75942290e-01 -5.84708452e-01 6.90068230e-02
2.49831557e-01 -3.39613289e-01 -2.47482527e-02 2.50517935e-01
-3.70862782e-01 6.30586267e-01 -2.02641174e-01 -3.84829223e-01
-1.35822189e+00 1.79244146e-01 3.46897691e-01 -5.67137122e-01
-1.73937783e-01 8.42016697e-01 -3.88452858e-01 -1.27943897e+00
6.18472219e-01 1.02018826e-02 -2.79875994e-01 2.22451296e-02
3.43053997e-01 6.41613081e-02 1.32018641e-01 -2.66021192e-01
-1.85719609e-01 -2.42073238e-01 -1.11631759e-01 -5.85619509e-01
1.28601003e+00 -4.46406424e-01 2.22789170e-03 6.02920353e-01
8.89883816e-01 -1.15943074e-01 -1.08875573e+00 -3.57827395e-01
2.69080013e-01 -2.56712198e-01 -6.25423670e-01 -1.18856335e+00
-4.37210590e-01 4.65964228e-01 4.96557169e-02 7.23848045e-01
7.48728573e-01 -8.28951523e-02 7.68234074e-01 1.19659388e+00
6.36507213e-01 -1.84913051e+00 5.31286538e-01 9.50776935e-01
1.05832005e+00 -1.25736070e+00 -3.77463907e-01 -5.09425938e-01
-1.32041180e+00 1.34980226e+00 8.76075983e-01 5.21338642e-01
3.21021378e-01 1.39238715e-01 5.62778533e-01 -1.11775301e-01
-1.37534320e+00 4.33456823e-02 -1.08705953e-01 6.64435327e-01
6.34262502e-01 4.85842042e-02 -3.57497066e-01 6.47968233e-01
-4.86022532e-01 -9.72287580e-02 7.20738292e-01 8.55552852e-01
-5.38192570e-01 -1.43054748e+00 1.08738035e-01 3.77495497e-01
4.17486951e-02 2.12938666e-01 -1.05334032e+00 5.25259912e-01
-4.11238849e-01 1.62828112e+00 -3.61124836e-02 -7.61606216e-01
8.36606145e-01 4.15816784e-01 1.16606645e-01 -9.04014349e-01
-1.18418610e+00 1.55684382e-01 1.31262147e+00 -4.47563678e-01
-1.45728394e-01 -7.73956120e-01 -9.31682467e-01 -3.39920133e-01
-5.08414984e-01 5.60352981e-01 5.69238663e-01 7.87830651e-01
4.18861747e-01 2.87135363e-01 1.35923374e+00 -7.34848857e-01
-9.61115539e-01 -1.45501506e+00 -3.16210121e-01 3.92169654e-01
-1.72253788e-01 -1.39104337e-01 -2.06527069e-01 -1.75750092e-01] | [12.86872673034668, 7.96024751663208] |
548c6b6e-51ee-4311-952e-0d21dd2f5c16 | distinguishing-unseen-from-seen-for | null | null | http://openaccess.thecvf.com//content/CVPR2022/html/Su_Distinguishing_Unseen_From_Seen_for_Generalized_Zero-Shot_Learning_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Su_Distinguishing_Unseen_From_Seen_for_Generalized_Zero-Shot_Learning_CVPR_2022_paper.pdf | Distinguishing Unseen From Seen for Generalized Zero-Shot Learning | Generalized zero-shot learning (GZSL) aims to recognize samples whose categories may not have been seen at training. Recognizing unseen classes as seen ones or vice versa often leads to poor performance in GZSL. Therefore, distinguishing seen and unseen domains is naturally an effective yet challenging solution for GZSL. In this paper, we present a novel method which leverages both visual and semantic modalities to distinguish seen and unseen categories. Specifically, our method deploys two variational autoencoders to generate latent representations for visual and semantic modalities in a shared latent space, in which we align latent representations of both modalities by Wasserstein distance and reconstruct two modalities with the representations of each other. In order to learn a clearer boundary between seen and unseen classes, we propose a two-stage training strategy which takes advantage of seen and unseen semantic descriptions and searches a threshold to separate seen and unseen visual samples. At last, a seen expert and an unseen expert are used for final classification. Extensive experiments on five widely used benchmarks verify that the proposed method can significantly improve the results of GZSL. For instance, our method correctly recognizes more than 99% samples when separating domains and improves the final classification accuracy from 72.6% to 82.9% on AWA1. | ['Ke Lu', 'Lei Zhu', 'Zhi Chen', 'Jingjing Li', 'Hongzu Su'] | 2022-01-01 | null | null | null | cvpr-2022-1 | ['generalized-zero-shot-learning', 'generalized-zero-shot-learning'] | ['computer-vision', 'methodology'] | [ 2.75996745e-01 -6.54346496e-02 -2.82364070e-01 -3.99137020e-01
-8.78080845e-01 -5.75189948e-01 7.50821173e-01 -1.32093847e-01
-6.73243329e-02 5.20975471e-01 1.45565122e-02 1.14871450e-01
5.85973402e-03 -8.03780615e-01 -4.70595151e-01 -9.52947676e-01
6.05403602e-01 4.51178104e-01 2.62186974e-01 1.88750342e-01
1.22385100e-02 1.57984331e-01 -1.91698992e+00 5.12609541e-01
8.24944496e-01 1.08196163e+00 1.87007293e-01 3.85454208e-01
-3.38875771e-01 6.54513776e-01 -4.75143880e-01 -1.85738280e-01
2.27405846e-01 -4.54937667e-01 -6.49646521e-01 5.68602562e-01
5.41912735e-01 -3.96676838e-01 -2.47987419e-01 1.16960287e+00
2.68584847e-01 5.28166056e-01 1.05614996e+00 -1.47487831e+00
-9.70886350e-01 1.58164367e-01 -5.94998777e-01 3.66668031e-02
2.07906947e-01 1.86107308e-01 1.08219373e+00 -1.08941019e+00
4.97991174e-01 1.38040376e+00 1.48483798e-01 8.55044425e-01
-1.32804537e+00 -6.31630301e-01 2.91864425e-01 3.59537721e-01
-1.31784058e+00 -5.93808055e-01 8.22848499e-01 -6.35378361e-01
6.32758319e-01 2.00098172e-01 3.26123863e-01 1.48963130e+00
-1.43397316e-01 1.06926298e+00 1.18705571e+00 -2.99773514e-01
5.15841782e-01 3.52374256e-01 3.30490202e-01 3.32447916e-01
2.14691564e-01 3.40740606e-02 -5.21036088e-01 -9.74712446e-02
5.24375618e-01 4.91490930e-01 -2.69462317e-01 -8.10591578e-01
-1.13230503e+00 8.60777974e-01 3.92514229e-01 1.55520439e-01
-1.29988924e-01 -2.84347653e-01 2.21831337e-01 1.31081447e-01
4.22355533e-01 -6.62131384e-02 1.73154324e-02 2.49239698e-01
-7.20379770e-01 -2.01949179e-01 3.61192435e-01 8.76364410e-01
8.30178320e-01 5.79243153e-02 -3.46110046e-01 1.10129070e+00
4.18438315e-01 7.20144391e-01 6.07732773e-01 -7.92153060e-01
3.22828889e-01 8.21551085e-01 1.76223561e-01 -9.13121164e-01
3.27518374e-01 -3.19655538e-01 -8.89460146e-01 1.70459479e-01
3.23374540e-01 2.31884494e-01 -1.30537605e+00 1.52341008e+00
3.40281963e-01 5.51558256e-01 6.13553345e-01 1.11434495e+00
9.96681690e-01 8.53073359e-01 2.03518510e-01 7.50398859e-02
1.35668910e+00 -9.78633225e-01 -6.65054321e-01 -5.29141903e-01
1.26343533e-01 -5.18327355e-01 9.41308558e-01 2.08343059e-01
-7.60385036e-01 -8.68740499e-01 -1.12280416e+00 -4.43440787e-02
-5.39124250e-01 2.38369644e-01 1.79188743e-01 4.07818258e-01
-5.65233886e-01 5.08525670e-01 -9.92168009e-01 -2.98961878e-01
6.02294207e-01 -1.64689064e-01 -3.22358429e-01 -3.73319417e-01
-1.11515915e+00 4.72966462e-01 6.32434309e-01 -7.49054477e-02
-1.37999737e+00 -2.48875841e-01 -1.11600995e+00 1.83850601e-01
4.79211092e-01 -4.61753130e-01 9.77898359e-01 -9.03999448e-01
-1.22016144e+00 1.09917092e+00 -3.18667322e-01 -2.29808480e-01
4.94554609e-01 3.67801115e-02 -6.39288187e-01 2.05345050e-01
3.56014043e-01 4.59699780e-01 1.25454712e+00 -1.45372498e+00
-8.05134714e-01 -5.42820275e-01 -1.02224953e-01 2.49734491e-01
-4.69271511e-01 -5.95561862e-01 -3.75320315e-01 -3.97227496e-01
4.51764435e-01 -6.55365348e-01 3.18244815e-01 7.46620595e-02
-2.46252790e-01 -5.57806492e-01 1.02302563e+00 -4.59868759e-01
6.52583718e-01 -2.43974733e+00 2.06701979e-01 -2.95902472e-02
4.60266382e-01 2.28499159e-01 -1.72242820e-01 2.42187351e-01
8.93420801e-02 -2.14641184e-01 -1.68051884e-01 -4.75660712e-01
2.54428536e-01 4.96701986e-01 -5.45903325e-01 4.49941933e-01
2.71838546e-01 7.80387700e-01 -1.16594517e+00 -3.56048048e-01
5.12303650e-01 5.66362143e-01 -4.32142168e-02 3.67009133e-01
1.72590688e-02 3.50405872e-01 -3.75708818e-01 8.06017816e-01
7.29630709e-01 -5.48002839e-01 1.28953174e-01 -8.57851505e-02
3.18829715e-01 -2.09785014e-01 -1.18317223e+00 1.44005334e+00
-3.48271728e-01 5.89046419e-01 -2.33844519e-01 -1.12822080e+00
1.06040227e+00 2.78528482e-01 1.25286551e-02 -5.94043493e-01
1.87148198e-01 1.64707780e-01 -5.20343602e-01 -4.99870777e-01
8.47577304e-02 -4.08350855e-01 -5.12333773e-02 1.73424095e-01
3.45964700e-01 2.95469344e-01 -8.15290660e-02 1.40345886e-01
5.74203968e-01 -5.06803812e-03 3.31683874e-01 5.94743267e-02
5.59376717e-01 -9.23491120e-02 8.60920012e-01 8.12796235e-01
-6.01523161e-01 7.35583127e-01 2.61778325e-01 -2.56603777e-01
-8.82545829e-01 -1.75648081e+00 -8.24746490e-02 1.04489791e+00
7.38585651e-01 1.63144935e-02 -4.94221956e-01 -9.05330896e-01
8.20731446e-02 9.50142145e-01 -7.20048904e-01 -5.17873764e-01
4.86509465e-02 -3.75482917e-01 2.01704815e-01 7.19198227e-01
4.78351355e-01 -8.59865487e-01 -4.86248374e-01 -1.34031534e-01
-1.27627596e-01 -1.01601207e+00 -3.64857092e-02 3.75066996e-02
-6.40400112e-01 -1.21852493e+00 -9.91686165e-01 -9.85047579e-01
7.48113155e-01 6.60521626e-01 8.75187516e-01 -3.02320421e-01
-1.30244389e-01 4.13513064e-01 -5.01855016e-01 -8.67878124e-02
-3.05332392e-01 -3.08420390e-01 7.50991106e-02 5.80412269e-01
8.36789429e-01 -2.81441629e-01 -5.79680204e-01 5.40261805e-01
-8.25633943e-01 -1.31638739e-02 5.08717418e-01 1.03931797e+00
7.47223735e-01 1.16005592e-01 5.30873835e-01 -7.66060412e-01
2.33742818e-01 -7.53979981e-01 -4.24101710e-01 4.82758105e-01
-4.22278702e-01 8.37978125e-02 7.72704482e-01 -5.15983582e-01
-1.14768577e+00 -8.65240991e-02 3.37744504e-01 -1.00570047e+00
-6.31469309e-01 9.67551619e-02 -3.05098683e-01 4.40375000e-01
4.74048704e-01 4.80644554e-01 1.38490018e-03 -4.67174768e-01
4.21746373e-01 1.06793463e+00 4.97486413e-01 -4.47109580e-01
8.69196057e-01 7.97929764e-01 -6.11611366e-01 -7.59521008e-01
-1.02542770e+00 -7.94117868e-01 -5.59870958e-01 -9.95908305e-02
1.05230153e+00 -1.03958774e+00 -4.27290082e-01 4.02194113e-01
-7.79222906e-01 9.30680335e-03 -4.22649175e-01 4.05417025e-01
-3.64115387e-01 4.39759016e-01 -3.26496303e-01 -8.94002140e-01
-5.37157655e-02 -1.05440724e+00 1.42265546e+00 3.72011304e-01
4.46270080e-03 -1.01956558e+00 -1.15432411e-01 5.50920486e-01
1.47722796e-01 3.29635382e-01 7.60695100e-01 -9.04552758e-01
-4.28853899e-01 -1.94600850e-01 -3.94214749e-01 5.41925490e-01
3.44617397e-01 -1.76171377e-01 -1.35519040e+00 -6.02899969e-01
4.54355367e-02 -5.00315964e-01 1.01748848e+00 1.14445440e-01
9.63009298e-01 4.79222946e-02 -3.79750401e-01 3.21596056e-01
1.46967769e+00 4.08295602e-01 4.06375825e-01 8.74543861e-02
6.65257335e-01 5.98647296e-01 6.20938718e-01 3.08679521e-01
8.01311210e-02 2.43279904e-01 3.68975788e-01 2.71812022e-01
-8.55488256e-02 -3.48067433e-01 3.99691939e-01 7.78804660e-01
2.19618574e-01 -2.93030769e-01 -8.62501860e-01 7.84544587e-01
-1.91248989e+00 -1.07585549e+00 2.55092263e-01 2.20303345e+00
3.53800982e-01 1.67047322e-01 -1.23502128e-01 1.30075157e-01
1.05948400e+00 3.94084305e-01 -9.20783401e-01 5.28472438e-02
-5.29891253e-02 -7.90426284e-02 1.69660330e-01 2.12683961e-01
-1.19206309e+00 8.10515404e-01 5.22745943e+00 1.11475217e+00
-1.04654634e+00 1.33208811e-01 5.09084582e-01 -3.02654561e-02
-2.87552655e-01 -6.66391402e-02 -7.47109652e-01 5.36763847e-01
5.62510967e-01 -1.78290740e-01 3.53829354e-01 1.18351364e+00
-1.33746654e-01 2.55818576e-01 -1.21955812e+00 1.35343468e+00
2.40962222e-01 -9.94440377e-01 3.66653830e-01 -1.17596395e-01
7.36509860e-01 -2.74127603e-01 3.27359587e-01 6.23649120e-01
2.43074819e-01 -9.25020218e-01 6.45926237e-01 5.81623614e-01
7.20980346e-01 -6.24691427e-01 6.53986871e-01 5.53986192e-01
-1.29578722e+00 -1.53152108e-01 -7.78329849e-01 1.18406201e-02
-1.74109429e-01 3.41386646e-01 -5.88122189e-01 6.00314021e-01
6.86480224e-01 1.17483830e+00 -4.04390395e-01 7.20706880e-01
-2.57289410e-01 4.51745659e-01 1.13785282e-01 1.05931006e-01
2.81060696e-01 -1.82502612e-01 5.27287483e-01 6.91508353e-01
2.76647180e-01 -1.69370901e-02 3.87710601e-01 1.15280771e+00
2.88190264e-02 -2.05573678e-01 -6.24592245e-01 -2.21141994e-01
4.99765724e-01 1.00908196e+00 -5.97062051e-01 -7.30898678e-01
-4.85239416e-01 1.28445196e+00 2.80091554e-01 6.71245277e-01
-8.45059812e-01 -3.93208504e-01 6.22569859e-01 -1.71438769e-01
3.58318895e-01 1.63325146e-01 8.82647336e-02 -1.66682482e+00
-8.16489756e-02 -5.90875566e-01 7.21430719e-01 -9.47465181e-01
-1.84772742e+00 6.54505551e-01 -1.76923480e-02 -1.65866184e+00
-1.71216443e-01 -7.63030410e-01 -4.45591629e-01 8.56862128e-01
-1.52539766e+00 -1.05210435e+00 -7.32331336e-01 6.01636469e-01
9.60922956e-01 -3.39023858e-01 7.55022466e-01 5.99013045e-02
-5.90684414e-01 4.24285293e-01 5.97457170e-01 3.31189573e-01
5.83667457e-01 -1.42463577e+00 1.71493590e-01 7.75877595e-01
2.68766135e-01 5.59239209e-01 3.95730257e-01 -4.87279385e-01
-1.02799165e+00 -1.13797665e+00 4.54862237e-01 -3.31576198e-01
5.84074855e-01 -3.68903011e-01 -1.10233438e+00 7.62126982e-01
1.50820956e-01 1.69219539e-01 7.71368623e-01 -1.01285756e-01
-6.23142004e-01 -4.55004387e-02 -1.11578953e+00 4.67211783e-01
8.65711153e-01 -8.58808398e-01 -1.12427258e+00 9.80709493e-02
4.76346701e-01 -1.77859902e-01 -6.58025682e-01 3.30473274e-01
4.15940136e-01 -9.21474159e-01 9.86766875e-01 -7.08690524e-01
3.74546438e-01 -3.82911861e-01 -4.48326230e-01 -1.39856112e+00
-2.89680630e-01 8.51337537e-02 -2.17207193e-01 1.27758920e+00
-6.54805824e-02 -8.31832647e-01 8.20813835e-01 3.83018494e-01
1.25304475e-01 -5.31181455e-01 -7.50793457e-01 -9.82356846e-01
-6.38678074e-02 -2.10015103e-01 4.33389753e-01 1.24695814e+00
-3.30366284e-01 3.90383035e-01 -3.80955219e-01 3.44861388e-01
9.82273400e-01 5.37498593e-01 5.16250789e-01 -1.39485407e+00
-1.44987792e-01 -1.88733876e-01 -5.64553976e-01 -1.14884126e+00
3.36717874e-01 -1.02801323e+00 1.60241891e-02 -1.58540201e+00
4.72282469e-01 -1.40065119e-01 -7.29563415e-01 4.51474816e-01
-2.30929062e-01 2.08963349e-01 8.01459253e-02 3.39004934e-01
-7.89435148e-01 8.93632531e-01 1.08427286e+00 -5.93610585e-01
4.26420122e-02 -1.79465100e-01 -6.06142819e-01 7.13682830e-01
6.62711263e-01 -3.08584064e-01 -7.51021922e-01 -3.97689909e-01
-4.08755749e-01 -5.95049709e-02 6.95468724e-01 -1.03616357e+00
-2.34647393e-02 -2.81699538e-01 5.62096536e-01 -5.69242597e-01
4.70705837e-01 -9.44641829e-01 -7.11820647e-02 2.30931357e-01
-3.61089855e-01 -7.09005654e-01 -9.35547054e-02 1.08552492e+00
-3.89606774e-01 -2.22073779e-01 9.26536798e-01 -7.04501644e-02
-1.26896226e+00 2.09211126e-01 -1.32411823e-01 5.39269149e-02
1.26842511e+00 -5.21545887e-01 -4.41257328e-01 -1.82507753e-01
-1.12908185e+00 3.97075057e-01 5.07527292e-01 6.97070181e-01
9.91191924e-01 -1.58637154e+00 -4.81083244e-01 4.92944449e-01
5.99303722e-01 -3.62603925e-02 5.66863418e-01 4.64394569e-01
-1.60757855e-01 9.46510807e-02 -2.58131474e-01 -9.64985192e-01
-1.20902801e+00 6.83864176e-01 3.20147187e-01 1.90567017e-01
-5.89791179e-01 7.38401771e-01 4.19108123e-01 -3.87010008e-01
2.97854275e-01 -8.37022811e-02 -3.52967262e-01 2.65124053e-01
6.77467942e-01 4.52954680e-01 -3.17361563e-01 -8.46977174e-01
-3.77402008e-01 5.91407537e-01 -1.68627813e-01 1.93536997e-01
1.18634498e+00 -2.36419737e-01 2.88647234e-01 9.37463224e-01
1.51401889e+00 -3.15756708e-01 -1.51205611e+00 -4.28320825e-01
-1.67138487e-01 -7.47944772e-01 -3.21970284e-02 -5.41978717e-01
-9.89795506e-01 1.34531200e+00 8.68658900e-01 4.49612796e-01
9.66217279e-01 2.33189434e-01 8.75530660e-01 1.64137095e-01
2.93848485e-01 -1.01379669e+00 4.36013728e-01 2.32520133e-01
4.65505034e-01 -1.63241482e+00 -4.59933996e-01 -3.83352131e-01
-8.03377926e-01 9.86876965e-01 6.65091634e-01 -3.15536141e-01
2.33472303e-01 -5.85335612e-01 1.22496754e-01 -1.90392077e-01
-5.69793522e-01 -2.88880140e-01 6.28673553e-01 5.73384821e-01
-6.42959848e-02 1.67207837e-01 3.14934134e-01 6.21758819e-01
3.31888318e-01 -2.75528222e-01 1.75564989e-01 9.72913802e-01
-6.34117842e-01 -6.59760177e-01 -4.67664331e-01 3.90987545e-01
2.50784289e-02 2.92489707e-01 -3.52488488e-01 5.74658573e-01
1.98127866e-01 1.04668427e+00 2.31539547e-01 -4.37816083e-01
1.82534173e-01 3.02200079e-01 1.74537718e-01 -7.98341155e-01
2.53898084e-01 1.78334191e-01 -3.70586336e-01 -4.07840610e-01
-3.91693175e-01 -4.50351298e-01 -1.16270816e+00 -5.17830020e-03
-1.27650574e-01 2.06429169e-01 1.07023954e-01 9.64298964e-01
2.46542454e-01 5.39728165e-01 6.60263658e-01 -6.99664772e-01
-8.77547085e-01 -5.39175212e-01 -8.46772790e-01 8.94014418e-01
5.39120495e-01 -1.16383600e+00 -8.57748270e-01 3.24147083e-02] | [9.902298927307129, 2.3995718955993652] |
b45338e3-8e1f-4ce6-adfc-24cc87e929af | is-attention-always-needed-a-case-study-on | 2110.03427 | null | https://arxiv.org/abs/2110.03427v2 | https://arxiv.org/pdf/2110.03427v2.pdf | Is Attention always needed? A Case Study on Language Identification from Speech | Language Identification (LID), a recommended initial step to Automatic Speech Recognition (ASR), is used to detect a spoken language from audio specimens. In state-of-the-art systems capable of multilingual speech processing, however, users have to explicitly set one or more languages before using them. LID, therefore, plays a very important role in situations where ASR based systems cannot parse the uttered language in multilingual contexts causing failure in speech recognition. We propose an attention based convolutional recurrent neural network (CRNN with Attention) that works on Mel-frequency Cepstral Coefficient (MFCC) features of audio specimens. Additionally, we reproduce some state-of-the-art approaches, namely Convolutional Neural Network (CNN) and Convolutional Recurrent Neural Network (CRNN), and compare them to our proposed method. We performed extensive evaluation on thirteen different Indian languages and our model achieves classification accuracy over 98%. Our LID model is robust to noise and provides 91.2% accuracy in a noisy scenario. The proposed model is easily extensible to new languages. | ['Sudip Kumar Naskar', 'Mahidas Bhattacharya', 'Indranil Dutta', 'Santanu Pal', 'Atanu Mandal'] | 2021-10-05 | null | null | null | null | ['classification', 'spoken-language-identification'] | ['methodology', 'speech'] | [-2.90302392e-02 -2.89425343e-01 8.67572278e-02 -1.29753992e-01
-1.10161734e+00 -4.46127504e-01 2.72157162e-01 -1.50130808e-01
-6.87844396e-01 3.92378360e-01 4.35765594e-01 -7.54778922e-01
3.00498337e-01 -3.78855824e-01 -3.48314226e-01 -4.77059484e-01
6.98453039e-02 3.10478151e-01 -9.00643691e-02 -4.50208932e-01
4.44329046e-02 6.88065827e-01 -1.46058452e+00 2.33263269e-01
4.09829289e-01 6.88313186e-01 3.51137519e-01 1.13027346e+00
-2.09258631e-01 8.58336151e-01 -8.92237604e-01 -1.38006181e-01
-2.53513843e-01 -1.14991084e-01 -1.08030033e+00 -7.47069269e-02
1.39863849e-01 -9.14171785e-02 -4.61527109e-01 1.07701313e+00
8.45433474e-01 1.84534103e-01 3.34312320e-01 -7.49379396e-01
-8.73134851e-01 1.12278569e+00 -2.92990115e-02 4.94562596e-01
6.08266950e-01 -9.10333246e-02 8.89873564e-01 -1.18923724e+00
1.49468064e-01 1.42205524e+00 6.61851346e-01 8.04536343e-01
-9.08251882e-01 -7.52524555e-01 1.16472892e-01 1.75318435e-01
-1.59760284e+00 -9.18962240e-01 8.32293272e-01 -1.82975009e-01
1.37643158e+00 2.70998746e-01 2.10554570e-01 1.12855852e+00
-1.63988188e-01 9.18104410e-01 8.49635303e-01 -7.04899669e-01
7.06065297e-02 7.35659599e-02 1.09751187e-01 3.97109002e-01
-4.02904004e-01 -1.46540746e-01 -6.40690982e-01 -1.74675342e-02
5.47645807e-01 -3.55450183e-01 -3.33943337e-01 5.84708273e-01
-1.18128347e+00 7.77760684e-01 2.38383651e-01 6.88823640e-01
-3.78598332e-01 -1.21594019e-01 6.58689260e-01 3.98538858e-01
4.61208552e-01 2.12412938e-01 -4.36656028e-01 -2.89940268e-01
-9.69638467e-01 -3.73769313e-01 6.62494719e-01 6.95762515e-01
1.88293442e-01 6.97526813e-01 1.26232848e-01 1.59065187e+00
3.42963547e-01 5.70352256e-01 1.14749300e+00 -4.30404454e-01
2.68363386e-01 1.89847291e-01 -4.21118528e-01 -5.69342315e-01
-3.90451998e-01 -3.55393648e-01 -9.39878523e-01 -2.26000041e-01
-1.37569115e-01 -1.29736394e-01 -1.05353236e+00 1.54205012e+00
-1.45985499e-01 5.03754653e-02 5.29923916e-01 8.05835009e-01
1.15775335e+00 7.25862741e-01 -8.80129933e-02 -2.57716417e-01
1.23438799e+00 -8.15080404e-01 -8.45758736e-01 -2.01572388e-01
3.95251960e-01 -1.13580978e+00 1.11970305e+00 4.96828258e-01
-8.19486678e-01 -6.01314008e-01 -9.84726787e-01 -2.13539079e-02
-6.32471204e-01 5.60594559e-01 2.17806712e-01 9.67684150e-01
-1.29083395e+00 1.07715398e-01 -6.75282657e-01 -5.92589140e-01
-3.23634595e-01 6.57809675e-01 -5.54225087e-01 3.50826532e-01
-1.29960477e+00 6.49401724e-01 1.58944592e-01 5.68673968e-01
-1.00726306e+00 -2.03427210e-01 -9.79101598e-01 -8.41526538e-02
6.78781793e-02 1.49491563e-01 1.55731368e+00 -8.50560784e-01
-1.97257006e+00 8.61125529e-01 -2.86232769e-01 -6.83406711e-01
5.78845069e-02 -3.32466543e-01 -9.60237026e-01 1.10615097e-01
-1.35887086e-01 4.88534421e-01 7.65419364e-01 -8.61569583e-01
-5.09218514e-01 -1.70264512e-01 -4.05296117e-01 2.85508018e-02
-4.56522733e-01 6.88992798e-01 -3.39935720e-01 -9.54118431e-01
2.23449141e-01 -8.67498040e-01 2.53136605e-02 -5.56401074e-01
-3.66030514e-01 -5.29568017e-01 9.67720032e-01 -1.06744277e+00
1.46442890e+00 -2.24949789e+00 -3.44790779e-02 -1.00640468e-01
-2.84529567e-01 7.23721981e-01 -2.50928909e-01 3.30493540e-01
-2.01722234e-01 2.24917829e-01 -1.91723153e-01 -6.39078498e-01
-2.31719226e-01 2.49109626e-01 -5.02551913e-01 4.30852383e-01
3.25704843e-01 3.49433869e-01 -6.56006753e-01 -2.39002049e-01
4.02825892e-01 8.33580375e-01 -2.92196989e-01 4.28018302e-01
2.23573089e-01 3.16014528e-01 1.06763363e-01 8.40066612e-01
2.79468745e-01 4.11657065e-01 8.91194940e-02 8.22199732e-02
-3.86243254e-01 8.74218881e-01 -1.13243449e+00 1.48008502e+00
-9.29285347e-01 8.98539782e-01 3.94131094e-01 -9.06079471e-01
1.23691642e+00 8.89942288e-01 -5.14714867e-02 -3.78209949e-01
1.01217084e-01 5.60789824e-01 2.11024672e-01 -4.07388628e-01
8.19294393e-01 2.93112308e-01 -1.28225207e-01 3.90057147e-01
5.05933106e-01 8.13125893e-02 -7.74265006e-02 8.02173465e-02
8.43384802e-01 -4.88867521e-01 3.57385010e-01 -1.65610537e-01
1.00338387e+00 -6.24656141e-01 5.56885660e-01 6.42309904e-01
-2.86353141e-01 7.65453756e-01 -4.40046862e-02 -4.18755472e-01
-9.55505669e-01 -7.72899568e-01 -4.50885557e-02 1.36803663e+00
-6.52663469e-01 -3.53070080e-01 -6.60242438e-01 -2.32406974e-01
-4.35513169e-01 5.10785580e-01 -2.66855091e-01 3.98382433e-02
-6.84816420e-01 -5.03237247e-01 1.14884853e+00 5.46129942e-01
4.08030480e-01 -1.46141160e+00 -7.27993324e-02 4.81882453e-01
-3.04838330e-01 -1.20032251e+00 -5.76483011e-01 3.90911311e-01
-2.46146008e-01 -6.45303786e-01 -9.15691078e-01 -1.20654094e+00
2.43360415e-01 7.80316070e-02 7.79381216e-01 -7.85002783e-02
-1.29286319e-01 4.25228864e-01 -6.05278313e-01 -2.21738651e-01
-8.31222951e-01 5.01618266e-01 6.17873847e-01 1.93058327e-01
5.02335608e-01 -3.82049054e-01 -9.23366696e-02 1.29063264e-01
-5.81300259e-01 -5.11005878e-01 3.98994416e-01 9.77967143e-01
3.89779270e-01 -1.90653667e-01 9.72979307e-01 -4.62250739e-01
1.00009811e+00 -2.54441142e-01 -4.65924293e-01 3.70572984e-01
-5.95119409e-02 -1.21217072e-02 8.58758450e-01 -5.77364564e-01
-7.23772407e-01 2.26686269e-01 -8.56024027e-01 -5.51322043e-01
-4.40981925e-01 6.15960717e-01 -1.20139368e-01 -4.59101424e-03
4.00896996e-01 5.41683733e-01 -1.90884277e-01 -7.77951300e-01
1.53116718e-01 1.70702982e+00 8.75796020e-01 -2.98666924e-01
2.17697740e-01 -1.33681148e-01 -7.26625443e-01 -1.60272753e+00
-4.09287989e-01 -7.64659584e-01 -7.63723612e-01 -2.41687104e-01
6.95827425e-01 -1.04614854e+00 -8.55351627e-01 6.58943474e-01
-1.31261981e+00 -6.59557506e-02 1.68591082e-01 6.80805147e-01
-3.46495777e-01 3.46893102e-01 -8.87854993e-01 -1.50831592e+00
-8.27054322e-01 -1.42165887e+00 1.11551762e+00 1.02458678e-01
-2.61657864e-01 -8.68682206e-01 -1.46039659e-02 2.20814481e-01
6.01333916e-01 -5.87389827e-01 5.07053375e-01 -1.09770739e+00
1.94204986e-01 -1.39849588e-01 2.36294016e-01 5.89143693e-01
4.10333782e-01 2.14244306e-01 -1.48338401e+00 -3.78370881e-01
-2.75990844e-01 -3.61356020e-01 9.17535543e-01 1.89114228e-01
1.10275054e+00 -3.09038311e-01 2.21643701e-01 3.67835492e-01
7.22417891e-01 4.57000583e-01 5.56718647e-01 1.04963571e-01
5.98573208e-01 3.83487344e-01 5.40681668e-02 1.88301340e-01
2.22528711e-01 5.45824051e-01 -4.74145971e-02 -3.78311127e-02
-2.41670489e-01 -2.76432455e-01 8.25785100e-01 1.83171475e+00
1.95036277e-01 -2.11948499e-01 -1.14013171e+00 8.95714700e-01
-1.45350277e+00 -7.62998044e-01 3.05870026e-01 2.28976297e+00
8.48761499e-01 -1.18669821e-02 2.64286071e-01 6.48558855e-01
9.75441694e-01 -3.14838742e-03 -2.15727538e-01 -9.38215017e-01
-3.69433522e-01 4.06706631e-01 2.58613080e-01 7.02342272e-01
-1.26363587e+00 1.30531788e+00 6.78073692e+00 8.36919785e-01
-1.72001123e+00 2.59710681e-02 4.82256830e-01 2.98946589e-01
6.20278344e-02 -4.92513806e-01 -9.81443048e-01 1.20088622e-01
1.43467081e+00 3.02216560e-02 6.55122697e-01 9.37968433e-01
1.43058807e-01 3.10348183e-01 -8.60793829e-01 1.20197320e+00
3.35312009e-01 -9.63506043e-01 1.34731844e-01 -2.53097087e-01
1.29371434e-01 3.55649233e-01 1.59266442e-01 4.45573956e-01
3.93617302e-01 -1.19187105e+00 6.84511244e-01 3.08542773e-02
1.10154963e+00 -1.08829820e+00 8.69347334e-01 1.24131821e-01
-1.48217261e+00 -3.48551050e-02 -3.45994323e-01 1.16811961e-01
-7.08278343e-02 1.26601800e-01 -1.39554060e+00 2.45433778e-01
8.54656994e-01 4.74929214e-01 -5.26627481e-01 6.93676651e-01
-1.82230845e-01 9.64261472e-01 -3.46755594e-01 -1.83301747e-01
2.72026777e-01 3.89862478e-01 5.90698123e-01 1.62258160e+00
3.97271186e-01 -2.40511075e-01 1.12058885e-01 4.43223953e-01
-2.55178034e-01 3.84895563e-01 -6.20368421e-01 -3.71161431e-01
7.43084788e-01 1.09253824e+00 -5.85674644e-01 -3.75049636e-02
-3.32137853e-01 9.57051218e-01 3.33041966e-01 2.47453317e-01
-1.56515598e-01 -6.73988044e-01 8.45915914e-01 -3.89234394e-01
1.90223917e-01 -3.45176190e-01 8.56018960e-02 -1.09887695e+00
-1.85526028e-01 -1.05894852e+00 3.29957575e-01 -6.84257030e-01
-1.37381136e+00 1.27266300e+00 -6.19920015e-01 -1.13029146e+00
-5.10576665e-01 -8.08843851e-01 -4.52604234e-01 9.84957993e-01
-1.48708630e+00 -1.30633223e+00 2.37267196e-01 6.37937188e-01
1.17477548e+00 -7.96572924e-01 1.30158913e+00 5.61268270e-01
-7.21402049e-01 9.40794885e-01 -1.21651649e-01 7.04501629e-01
7.23629296e-01 -1.06828928e+00 5.62291086e-01 1.08062601e+00
4.55162466e-01 8.13698351e-01 3.55399877e-01 -4.78699952e-01
-1.36336350e+00 -1.11349511e+00 1.18253326e+00 7.39163458e-02
8.26782465e-01 -5.79553902e-01 -9.56988037e-01 6.52010381e-01
4.65858340e-01 -9.21408534e-02 8.54709327e-01 2.59069413e-01
-4.40161943e-01 -1.11227743e-01 -7.11625993e-01 5.81601322e-01
4.67020243e-01 -1.37810743e+00 -5.04822135e-01 1.22359686e-01
1.07951891e+00 -2.73061424e-01 -7.29619503e-01 2.61600256e-01
4.89630342e-01 -3.62758785e-01 7.49898851e-01 -4.14435983e-01
-1.21564321e-01 -3.19147050e-01 -5.37550569e-01 -1.42775440e+00
-6.93398118e-02 -9.07979310e-01 2.08623141e-01 1.52239072e+00
6.44694030e-01 -5.52472174e-01 6.55757710e-02 5.09203374e-02
-3.73594522e-01 -5.94363250e-02 -1.28541076e+00 -7.37410009e-01
9.25356746e-02 -8.00137699e-01 5.45577884e-01 8.54545832e-01
8.35370198e-02 4.09516096e-01 -6.28782213e-01 5.10011017e-01
1.06228858e-01 -4.69861180e-01 4.61340547e-01 -9.49098229e-01
1.79658663e-02 -4.02148306e-01 -5.76680899e-01 -7.55471528e-01
4.70800608e-01 -7.89217472e-01 1.57934606e-01 -9.63053226e-01
-3.19222957e-01 -2.50417471e-01 -4.78257358e-01 6.29555881e-01
1.88419521e-01 2.50940919e-01 2.31668934e-01 1.73666049e-02
-3.13708991e-01 4.25964415e-01 4.19104040e-01 -3.92003119e-01
-3.04558009e-01 6.13316484e-02 -3.07648599e-01 6.08473182e-01
8.92032087e-01 -2.54525810e-01 1.86637957e-02 -4.17593211e-01
-6.16167001e-02 4.56523485e-02 -1.38803590e-02 -1.15327048e+00
6.87758684e-01 4.15728062e-01 1.28254414e-01 -7.74723887e-01
3.80643249e-01 -5.64662814e-01 -1.61764756e-01 3.26464057e-01
-5.72031915e-01 2.35552251e-01 2.69483924e-01 3.60126048e-02
-7.16444016e-01 -1.92400202e-01 7.19725311e-01 -9.88164619e-02
-7.63272047e-01 -5.88025078e-02 -1.04090786e+00 -3.29456270e-01
2.62060016e-01 1.00822784e-01 -1.41077250e-01 -5.00961840e-01
-9.46046174e-01 -5.09143211e-02 -9.99882892e-02 8.70029211e-01
8.88957560e-01 -1.25221682e+00 -8.50278437e-01 6.65846646e-01
1.61063775e-01 -3.31424892e-01 7.46426731e-02 4.64016527e-01
-5.89490891e-01 7.89110124e-01 1.46210447e-01 -5.10969102e-01
-1.55920553e+00 4.56790984e-01 4.83848214e-01 1.21300660e-01
-2.70459682e-01 8.71878803e-01 -3.38521540e-01 -9.03001845e-01
7.15762258e-01 -3.89374405e-01 -6.23152792e-01 4.15572338e-02
7.98534691e-01 8.90128687e-02 5.67165613e-01 -1.14439666e+00
-5.62256217e-01 3.30852091e-01 -2.65228987e-01 -4.03650671e-01
1.21011436e+00 -1.27189666e-01 -1.15439571e-01 9.67579007e-01
1.09835231e+00 1.24329329e-01 -4.02446121e-01 -3.37629676e-01
3.91899161e-02 1.10994406e-01 3.91284436e-01 -5.73518693e-01
-9.62929189e-01 1.20450056e+00 8.06791365e-01 3.51956248e-01
1.11099935e+00 -1.55465305e-01 7.42182851e-01 6.25111759e-01
2.25143358e-01 -1.32785404e+00 -2.60918111e-01 9.69156146e-01
1.12281442e+00 -1.28166628e+00 -6.04463100e-01 -1.21707313e-01
-5.23170710e-01 1.39635861e+00 4.39531922e-01 2.28770301e-02
8.14226210e-01 5.08361936e-01 4.40353841e-01 2.62101561e-01
-8.46559107e-01 -3.33597541e-01 1.84583560e-01 6.86820090e-01
9.72395718e-01 2.98675954e-01 -9.20451526e-03 6.42287493e-01
-5.73516548e-01 -3.88114780e-01 5.59463203e-01 7.90677249e-01
-3.30768913e-01 -9.83000875e-01 -7.22246587e-01 5.10830991e-03
-8.71690631e-01 -3.42144370e-01 -6.20958209e-01 3.24569196e-01
-2.77766019e-01 1.32483566e+00 2.93650270e-01 -6.67830050e-01
1.23882309e-01 2.33914971e-01 -1.58434033e-01 -6.41453207e-01
-8.98101509e-01 4.71020967e-01 8.59390795e-02 -9.90351364e-02
-4.74945873e-01 -5.32226503e-01 -1.00703537e+00 -8.74862671e-02
-4.10932332e-01 2.46840119e-01 8.45370710e-01 6.85507596e-01
1.40624866e-01 6.66545749e-01 8.28662038e-01 -6.45914853e-01
-2.35847756e-01 -1.39936674e+00 -5.08467555e-01 -1.35198012e-01
7.33147860e-01 -2.48858944e-01 -3.80995125e-01 1.93929449e-02] | [14.220562934875488, 6.624608516693115] |
1b2c51e5-3beb-43ce-8b7b-1e4a8eb7b1ee | lighting-the-darkness-in-the-deep-learning | 2104.10729 | null | https://arxiv.org/abs/2104.10729v3 | https://arxiv.org/pdf/2104.10729v3.pdf | Low-Light Image and Video Enhancement Using Deep Learning: A Survey | Low-light image enhancement (LLIE) aims at improving the perception or interpretability of an image captured in an environment with poor illumination. Recent advances in this area are dominated by deep learning-based solutions, where many learning strategies, network structures, loss functions, training data, etc. have been employed. In this paper, we provide a comprehensive survey to cover various aspects ranging from algorithm taxonomy to open issues. To examine the generalization of existing methods, we propose a low-light image and video dataset, in which the images and videos are taken by different mobile phones' cameras under diverse illumination conditions. Besides, for the first time, we provide a unified online platform that covers many popular LLIE methods, of which the results can be produced through a user-friendly web interface. In addition to qualitative and quantitative evaluation of existing methods on publicly available and our proposed datasets, we also validate their performance in face detection in the dark.This survey together with the proposed dataset and online platform could serve as a reference source for future study and promote the development of this research field. The proposed platform and dataset as well as the collected methods, datasets, and evaluation metrics are publicly available and will be regularly updated. | ['Chen Change Loy', 'Jinwei Gu', 'Ming-Ming Cheng', 'Jun Jiang', 'Linghao Han', 'Chunle Guo', 'Chongyi Li'] | 2021-04-21 | null | null | null | null | ['video-enhancement'] | ['computer-vision'] | [ 4.22294766e-01 -5.37890375e-01 -1.11481465e-01 -4.18973505e-01
-3.69651645e-01 -2.63628870e-01 2.62378424e-01 -1.89005628e-01
-4.48918521e-01 7.87878036e-01 -1.97045892e-01 -1.84393907e-03
-7.14396983e-02 -7.08737850e-01 -4.15729672e-01 -9.09328878e-01
3.97313386e-02 -3.72100800e-01 -7.60247037e-02 -9.79997888e-02
2.86815882e-01 4.98050332e-01 -2.02328610e+00 2.70006835e-01
6.67191088e-01 1.18969429e+00 2.64391661e-01 4.50322092e-01
7.81485960e-02 4.73624349e-01 -4.97490644e-01 -8.60312283e-01
3.62520218e-01 -1.70080990e-01 -3.52172732e-01 3.36980760e-01
8.01695764e-01 -5.13047934e-01 -2.82374650e-01 1.30876648e+00
1.02919829e+00 -4.02239226e-02 2.77173758e-01 -1.31376922e+00
-7.89321542e-01 -5.29455058e-02 -6.20266914e-01 2.57998139e-01
3.97490948e-01 2.93409616e-01 5.93661249e-01 -1.16944480e+00
3.46064001e-01 1.07517219e+00 7.14837968e-01 5.68024099e-01
-6.13739789e-01 -8.35545659e-01 -7.98337162e-02 6.78628802e-01
-1.33204234e+00 -6.16374850e-01 9.71818328e-01 -2.38974929e-01
5.39422512e-01 2.78104991e-01 5.56289911e-01 1.19601178e+00
1.27351016e-01 8.29209864e-01 1.27812278e+00 -6.43151164e-01
-3.14581692e-02 4.42387849e-01 7.56785879e-03 9.09129500e-01
4.85516042e-01 2.72949189e-01 -4.67549324e-01 1.22264713e-01
5.16739905e-01 1.14802308e-01 -3.80843312e-01 -1.48752972e-01
-5.84163845e-01 5.61313987e-01 2.69872844e-01 1.74024850e-01
-1.08401380e-01 -3.64904702e-01 3.21836889e-01 1.43462539e-01
5.08000016e-01 -1.03843905e-01 -2.95226067e-01 1.23860300e-01
-6.21970236e-01 9.07807872e-02 4.71847594e-01 8.30323756e-01
8.36061716e-01 4.61013392e-02 -1.51712984e-01 9.93592680e-01
3.81067723e-01 6.14254355e-01 3.81329387e-01 -9.91397202e-01
3.86285216e-01 4.19788808e-01 1.27159476e-01 -1.36706698e+00
-2.95457572e-01 -3.91839623e-01 -8.15938175e-01 2.52852261e-01
1.36250019e-01 -4.04869437e-01 -3.98123533e-01 1.48100531e+00
3.50013524e-01 2.19998538e-01 -2.22302899e-02 8.82685542e-01
1.39197302e+00 5.77250600e-01 -1.25764102e-01 -5.61205506e-01
1.48683095e+00 -9.89978015e-01 -1.10100639e+00 -2.49337796e-02
1.40342444e-01 -1.07179487e+00 1.22215092e+00 6.24349475e-01
-1.11617362e+00 -8.62715006e-01 -8.94754112e-01 -1.83888987e-01
-5.25974333e-01 6.46831036e-01 6.28142059e-01 9.45742130e-01
-1.22262406e+00 2.26523489e-01 -4.84234124e-01 -6.33128405e-01
6.54613316e-01 2.63988346e-01 -2.58203149e-01 -2.39542603e-01
-1.09378719e+00 6.33097172e-01 -1.65144410e-02 4.22237128e-01
-9.07602608e-01 -3.09890956e-01 -7.94257522e-01 -2.69224375e-01
4.64791745e-01 -3.60740125e-01 1.08943343e+00 -9.54324305e-01
-1.41683245e+00 1.03855693e+00 -1.65983945e-01 -1.06438398e-01
3.70879084e-01 -1.75545692e-01 -5.61828136e-01 2.78869450e-01
-2.45129541e-01 4.18112636e-01 8.37441742e-01 -1.42260015e+00
-7.55412161e-01 -5.25487363e-01 2.84958899e-01 2.17278436e-01
-8.23629260e-01 3.98266017e-01 -7.93138266e-01 -5.14571607e-01
-4.46259797e-01 -6.55074775e-01 2.61407673e-01 2.62933552e-01
-1.53820172e-01 -1.17265560e-01 1.10088301e+00 -6.39046669e-01
1.36777103e+00 -2.10606718e+00 -3.33763123e-01 -2.56190568e-01
4.48318571e-02 7.37851381e-01 -1.22918695e-01 2.95223862e-01
4.63232026e-02 -1.07521541e-01 -1.88029304e-01 -4.81579959e-01
-8.87934417e-02 -2.56016683e-02 2.48365209e-01 6.41600311e-01
8.03090706e-02 4.30414259e-01 -6.73354924e-01 -6.87516809e-01
5.37423015e-01 8.19240391e-01 -3.38709205e-01 2.29960725e-01
2.82139510e-01 3.42013508e-01 -9.01535153e-02 9.93860543e-01
1.04007626e+00 9.77198314e-03 -1.20287314e-01 -6.91379368e-01
-2.91055799e-01 -3.53414416e-01 -1.34203446e+00 1.24881029e+00
-6.07090473e-01 8.94752800e-01 2.49818206e-01 -8.54045987e-01
8.89501631e-01 3.01248401e-01 4.69300568e-01 -7.85225213e-01
3.76055390e-01 -1.41002564e-02 -3.50201845e-01 -1.07551479e+00
4.29789782e-01 1.77484289e-01 6.92220449e-01 2.57657081e-01
-6.56660795e-02 4.46451813e-01 4.65233117e-01 -2.33035460e-01
5.21625161e-01 8.60531032e-02 2.44329914e-01 1.36471272e-01
1.04009259e+00 -4.22455668e-01 5.82413137e-01 5.56586921e-01
-6.25902832e-01 3.03801447e-01 -2.93215066e-02 -5.47642112e-01
-8.20128143e-01 -7.67207623e-01 -5.81574500e-01 1.03388095e+00
3.38479787e-01 -2.64161408e-01 -8.33258867e-01 -3.94688427e-01
-2.72485435e-01 1.40674546e-01 -4.49374974e-01 1.97295547e-01
-3.63674700e-01 -1.17079771e+00 3.24437588e-01 2.95584708e-01
1.08895314e+00 -1.18276036e+00 -5.03472686e-01 -2.44131953e-01
-3.65737259e-01 -1.21647096e+00 -1.77642226e-01 -2.91192055e-01
-7.34333932e-01 -1.35401750e+00 -6.44329011e-01 -9.95559752e-01
6.54999614e-01 6.64008081e-01 9.70682621e-01 3.50896150e-01
-5.12945592e-01 4.32550877e-01 -3.63144964e-01 -5.74663341e-01
3.04868724e-02 -3.70713532e-01 7.32918084e-02 2.80890733e-01
6.50099039e-01 -2.15405628e-01 -9.23681557e-01 3.19017529e-01
-9.90919650e-01 -1.90381750e-01 5.90532482e-01 6.75800323e-01
5.10090947e-01 3.85228604e-01 3.19308102e-01 -7.36488998e-01
6.47968292e-01 -2.33927831e-01 -7.44118690e-01 3.07067424e-01
-6.92576170e-01 -5.93932927e-01 4.91912216e-01 2.65136664e-03
-1.57379079e+00 -1.84850216e-01 -3.63304734e-01 -3.27129513e-02
-3.77681702e-01 1.63340002e-01 -4.19244140e-01 -6.02392614e-01
5.35816610e-01 2.63696939e-01 -1.64896235e-01 -6.28022492e-01
2.40151118e-02 1.07449234e+00 5.09445548e-01 -2.19533533e-01
5.61150730e-01 6.92120612e-01 -8.99402797e-03 -1.03981400e+00
-7.42193162e-01 -4.22347248e-01 -3.87675345e-01 -5.87141275e-01
6.32412136e-01 -9.28607166e-01 -8.49531412e-01 8.45609009e-01
-9.92058575e-01 3.81081440e-02 3.60896796e-01 4.86983240e-01
-1.98573172e-01 6.16476417e-01 -6.56923532e-01 -9.70294237e-01
-3.97537500e-01 -1.23478019e+00 1.07521629e+00 6.58473134e-01
4.97380704e-01 -1.13239300e+00 -9.18097198e-02 6.39570892e-01
4.60669369e-01 8.68363902e-02 4.95373011e-01 -3.41155864e-02
-4.41211075e-01 -1.06338203e-01 -4.90664124e-01 7.27135479e-01
3.16686392e-01 2.76259691e-01 -1.34229827e+00 -6.06388927e-01
1.66034341e-01 -3.52771491e-01 6.93317533e-01 4.97085780e-01
1.58670199e+00 -3.66223216e-01 -2.41342887e-01 7.92640746e-01
1.77214170e+00 3.95378292e-01 7.20157325e-01 4.60327506e-01
4.86945540e-01 8.07442248e-01 5.84545434e-01 6.32024825e-01
2.63114631e-01 7.40239918e-01 6.70826793e-01 -4.22144204e-01
-2.75884867e-01 2.11336821e-01 4.07111347e-01 6.56320095e-01
-3.53932530e-01 -2.93118060e-01 -6.22913063e-01 2.47930482e-01
-1.36463261e+00 -1.18897569e+00 -1.97815344e-01 2.12945580e+00
6.07239544e-01 -2.40542695e-01 1.86794832e-01 2.95787692e-01
8.74334157e-01 1.71092451e-01 -3.85888189e-01 -2.46716470e-01
-2.68684119e-01 1.60058308e-02 2.54248679e-01 3.61457199e-01
-1.22250378e+00 6.41875207e-01 6.67953873e+00 8.82826567e-01
-1.28763938e+00 2.70413667e-01 6.16788447e-01 -8.70199800e-02
1.91413462e-01 -4.42720890e-01 -1.05328536e+00 6.43048942e-01
5.86092353e-01 8.04070092e-04 4.49559808e-01 7.43887901e-01
6.84036314e-01 -2.02420324e-01 -7.38191843e-01 1.52887666e+00
5.00603080e-01 -1.08705008e+00 -1.57983258e-01 -1.58614442e-02
7.50558913e-01 -2.28504255e-01 4.29775000e-01 1.22871637e-01
-3.96155775e-01 -7.43331194e-01 1.96836650e-01 4.10251796e-01
7.18722165e-01 -7.12762058e-01 9.75340426e-01 1.02568360e-03
-1.16189635e+00 -3.25461596e-01 -5.07082939e-01 -1.04620352e-01
-1.53743640e-01 4.20727640e-01 -3.38650286e-01 7.01405704e-01
1.01554954e+00 8.77755761e-01 -8.89289677e-01 1.35040331e+00
-1.69333503e-01 5.50313592e-01 6.55345023e-02 -2.89945770e-02
-1.10193804e-01 -2.27943063e-01 3.21884334e-01 1.38865054e+00
2.35779241e-01 -8.14282894e-02 8.19205716e-02 5.42531610e-01
-1.87866852e-01 2.77012557e-01 -6.78000331e-01 2.29412749e-01
3.05404484e-01 1.64085329e+00 -4.18777645e-01 -2.35227525e-01
-8.71449649e-01 7.06643164e-01 -9.09719020e-02 4.80745107e-01
-8.93816769e-01 -6.76196516e-01 5.37759006e-01 -5.06818518e-02
-1.21438675e-01 4.75691035e-02 -1.49569854e-01 -1.17154551e+00
1.43694043e-01 -9.84602332e-01 4.22372013e-01 -9.98674333e-01
-1.25050104e+00 7.54042029e-01 1.52402192e-01 -1.27023768e+00
2.53090799e-01 -9.18919325e-01 -4.79138732e-01 5.44770181e-01
-1.91533494e+00 -9.14601624e-01 -9.13747430e-01 8.45491350e-01
7.04038918e-01 -5.54497480e-01 4.78228986e-01 8.89904022e-01
-9.94163275e-01 7.10848987e-01 2.94539452e-01 6.76485226e-02
8.44545603e-01 -7.86917627e-01 -3.08012545e-01 9.82751608e-01
2.68649850e-02 5.61722457e-01 5.96215248e-01 -2.61468679e-01
-1.29304898e+00 -1.07733679e+00 6.50589406e-01 -1.19841851e-01
2.42194131e-01 -1.61610410e-01 -7.05607414e-01 4.36572731e-01
6.00015581e-01 6.18052669e-02 8.22784662e-01 -1.34789929e-01
2.33689725e-01 -6.56504512e-01 -1.41236544e+00 5.16037524e-01
9.46118355e-01 -2.68312514e-01 -2.22395658e-01 4.54164565e-01
1.18268475e-01 -2.38418400e-01 -5.44758737e-01 5.48865020e-01
5.42821944e-01 -1.37318277e+00 8.49322438e-01 -1.63524300e-01
3.49448889e-01 -3.06119680e-01 -1.97030067e-01 -8.85110736e-01
-1.79085776e-01 -4.37918901e-01 -6.63486794e-02 1.26561022e+00
-5.95628917e-02 -5.12486100e-01 7.93535233e-01 1.85903758e-01
-4.54193838e-02 -9.05257761e-01 -4.33382750e-01 -5.98878086e-01
-5.00684917e-01 -4.83024567e-01 5.34357250e-01 5.68067133e-01
-5.28465211e-01 3.98035049e-02 -6.06473684e-01 2.31536165e-01
9.85966563e-01 -6.24196138e-03 6.79251313e-01 -9.76210237e-01
9.85596925e-02 -1.86295494e-01 -3.46774012e-01 -6.32945180e-01
3.08447424e-03 -5.72399557e-01 -2.02385038e-01 -1.43890631e+00
5.71675062e-01 -9.06229541e-02 -3.38795930e-01 4.04629439e-01
-1.46054849e-01 9.54872787e-01 2.08007004e-02 2.18725920e-01
-6.95885062e-01 5.13261020e-01 1.31761587e+00 -7.63936937e-02
8.96664057e-03 -2.62859203e-02 -7.47235596e-01 1.01459157e+00
8.02767515e-01 -1.42851144e-01 -4.27669317e-01 -6.62291467e-01
7.28633255e-02 -4.69694644e-01 5.12195528e-01 -1.05136490e+00
2.02286959e-01 -1.24429911e-01 6.36574388e-01 -5.77201784e-01
4.48984504e-01 -9.47580278e-01 -3.26377973e-02 3.83503884e-01
-1.89808384e-01 9.31308419e-02 2.32621998e-01 3.64048630e-01
-3.86766106e-01 -3.34255904e-01 1.08774924e+00 -1.51793584e-01
-9.66451943e-01 4.87553477e-01 -1.67184174e-01 -2.15032831e-01
1.16590166e+00 -3.92488390e-01 -5.17538071e-01 -3.80558461e-01
-5.37921309e-01 -3.76431271e-02 4.10477936e-01 3.60772043e-01
8.31002831e-01 -1.47779298e+00 -5.93582094e-01 4.14783299e-01
1.40956238e-01 -5.75756073e-01 4.66139883e-01 8.33714366e-01
-5.50460875e-01 4.50672388e-01 -6.67976081e-01 -4.02183950e-01
-1.74379647e+00 5.28134346e-01 3.37930351e-01 2.04128131e-01
-2.04873294e-01 6.75839603e-01 1.18402191e-01 -1.43475682e-01
5.66603780e-01 1.35574609e-01 -5.80108941e-01 -6.49456754e-02
8.95821035e-01 5.96378863e-01 2.08715186e-01 -7.85447717e-01
-4.05265868e-01 7.04057157e-01 1.92017108e-01 2.94819951e-01
1.32988298e+00 -5.88046432e-01 -1.92881346e-01 1.65345520e-01
1.23924780e+00 1.50236443e-01 -1.18206644e+00 -1.87715039e-01
-5.16568720e-01 -9.59053636e-01 1.81316957e-01 -7.31168807e-01
-1.50232530e+00 9.51789320e-01 1.37963963e+00 1.30779654e-01
1.71747196e+00 -3.36262584e-01 6.89009666e-01 2.92806506e-01
3.03741574e-01 -1.11362207e+00 3.12849343e-01 1.15758434e-01
6.78641737e-01 -1.63654160e+00 2.49444358e-02 -4.21748996e-01
-3.78193170e-01 1.09705245e+00 8.61475050e-01 3.08691084e-01
5.67455053e-01 2.46493936e-01 2.17171296e-01 -5.62293753e-02
-5.28077066e-01 -4.67264950e-01 1.05110720e-01 9.55359817e-01
7.18430638e-01 -4.35941726e-01 -5.44520736e-01 3.62069607e-01
5.00961468e-02 3.10909063e-01 3.90993416e-01 7.13746965e-01
-4.76770043e-01 -1.18707049e+00 -6.98919356e-01 2.84460753e-01
-7.71564066e-01 4.89914566e-02 -1.65663555e-01 7.09002435e-01
6.68885231e-01 1.34505558e+00 -3.04293543e-01 -2.66275227e-01
2.59462655e-01 -4.02407318e-01 5.62720299e-01 -3.05517882e-01
-2.20501259e-01 -6.04658462e-02 -1.30971000e-01 -4.49972391e-01
-9.33197796e-01 -5.16036212e-01 -4.59878802e-01 -4.82299984e-01
-2.25577384e-01 -1.65657416e-01 7.93625355e-01 6.23701692e-01
1.80900350e-01 3.22056115e-01 7.93270171e-01 -9.18982983e-01
-1.75444096e-01 -8.41553628e-01 -6.03698969e-01 3.83089483e-01
3.26254219e-01 -8.64121020e-01 -3.67289603e-01 2.96533585e-01] | [10.874138832092285, -2.3174889087677] |
0e462ce7-6e2c-4cbc-a046-3ede895505af | cost-volume-pyramid-network-with-multi | 2207.12032 | null | https://arxiv.org/abs/2207.12032v1 | https://arxiv.org/pdf/2207.12032v1.pdf | Cost Volume Pyramid Network with Multi-strategies Range Searching for Multi-view Stereo | Multi-view stereo is an important research task in computer vision while still keeping challenging. In recent years, deep learning-based methods have shown superior performance on this task. Cost volume pyramid network-based methods which progressively refine depth map in coarse-to-fine manner, have yielded promising results while consuming less memory. However, these methods fail to take fully consideration of the characteristics of the cost volumes in each stage, leading to adopt similar range search strategies for each cost volume stage. In this work, we present a novel cost volume pyramid based network with different searching strategies for multi-view stereo. By choosing different depth range sampling strategies and applying adaptive unimodal filtering, we are able to obtain more accurate depth estimation in low resolution stages and iteratively upsample depth map to arbitrary resolution. We conducted extensive experiments on both DTU and BlendedMVS datasets, and results show that our method outperforms most state-of-the-art methods. | ['Zhaoqi Wang', 'Zhaoxin Li', 'Shiyu Gao'] | 2022-07-25 | null | null | null | null | ['stereo-matching-1'] | ['computer-vision'] | [ 2.41068929e-01 -3.89466107e-01 9.74819437e-02 -3.02873105e-01
-5.35421312e-01 -3.08208048e-01 5.08529961e-01 -2.14488223e-01
-5.12094557e-01 8.32323909e-01 2.06990466e-01 2.64518827e-01
-3.24827790e-01 -1.09654307e+00 -5.63145936e-01 -6.34311199e-01
2.77003884e-01 5.29526711e-01 7.83624172e-01 -8.20646137e-02
6.27320647e-01 5.80158710e-01 -1.76469672e+00 4.14643049e-01
8.06037545e-01 8.71865332e-01 4.85276550e-01 4.48033333e-01
1.43372759e-01 5.38891256e-01 -9.98479426e-02 -1.95941538e-01
5.25827408e-01 -1.31160408e-01 -7.58113503e-01 6.48288131e-02
9.07862365e-01 -6.33456349e-01 -4.72387820e-01 1.21758556e+00
6.83984160e-01 6.77737966e-02 5.17188191e-01 -6.28459573e-01
-2.13147119e-01 2.17354462e-01 -1.02645493e+00 4.82947052e-01
2.70306915e-01 -1.41189024e-01 8.41935396e-01 -8.67999673e-01
7.52466440e-01 1.35773253e+00 5.86714923e-01 4.40688670e-01
-1.23331225e+00 -5.27271867e-01 1.56027988e-01 5.15770853e-01
-1.12437224e+00 -1.72108322e-01 1.09011650e+00 -2.93459833e-01
1.03362083e+00 4.54621427e-02 8.22521687e-01 6.66632831e-01
3.57773244e-01 6.39099002e-01 1.46571910e+00 -2.12686643e-01
9.03728306e-02 -2.97908992e-01 -1.44166678e-01 6.61080718e-01
8.81796405e-02 3.05837005e-01 -6.67383194e-01 -9.82225779e-03
1.37320364e+00 1.32086381e-01 -6.35638714e-01 -7.10016668e-01
-1.15711832e+00 9.45761204e-01 5.29504776e-01 3.19676548e-01
-5.37048161e-01 3.86312045e-02 3.13921660e-01 7.25044161e-02
6.42892003e-01 1.09894000e-01 -5.31613410e-01 1.09755255e-01
-9.83109951e-01 3.05004150e-01 3.40537459e-01 5.86235940e-01
1.01517284e+00 -3.62087637e-02 5.24242744e-02 9.89493549e-01
4.31115925e-02 2.35958338e-01 1.75126210e-01 -1.46771610e+00
5.44650912e-01 5.30691445e-01 -2.53609922e-02 -1.12374735e+00
-6.23832762e-01 -3.51708919e-01 -1.19016278e+00 5.08571625e-01
6.07361913e-01 1.63428187e-01 -9.61502135e-01 1.32935119e+00
3.46099764e-01 -9.45772231e-02 -1.35298103e-01 1.01533473e+00
7.01041639e-01 5.51825106e-01 -5.04852593e-01 -2.53417403e-01
1.18484211e+00 -9.77905452e-01 -4.04021084e-01 -1.79869115e-01
-8.97860620e-03 -7.13625491e-01 9.02253568e-01 9.43196476e-01
-1.47143042e+00 -6.95192695e-01 -9.29018080e-01 -1.92629755e-01
-3.16995793e-05 -2.69290745e-01 7.92896092e-01 4.48966771e-01
-1.35121095e+00 9.28240895e-01 -7.01500654e-01 -2.66895205e-01
4.49359119e-01 3.17254096e-01 -4.10277307e-01 -5.46221197e-01
-9.54094648e-01 5.90773404e-01 3.61273915e-01 5.15231565e-02
-7.62563646e-01 -6.50380492e-01 -8.85716915e-01 -2.02532429e-02
3.50468278e-01 -1.09258544e+00 9.60420370e-01 -5.07870495e-01
-1.21747220e+00 8.95141840e-01 -3.27422351e-01 -3.52659434e-01
6.62306190e-01 1.10868807e-03 3.58774364e-02 4.63660985e-01
1.22280367e-01 9.80094969e-01 8.04317236e-01 -1.26767433e+00
-9.76442695e-01 -8.56441140e-01 3.10126841e-01 5.25489390e-01
-2.01954812e-01 -4.05635178e-01 -4.83361065e-01 -3.26727599e-01
6.46543741e-01 -4.18854177e-01 -3.44446659e-01 1.40724003e-01
-1.97989065e-02 -1.47807553e-01 6.06092691e-01 -4.16454017e-01
9.60525453e-01 -1.74078727e+00 3.28811616e-01 -1.24305375e-01
4.87193435e-01 -7.26526752e-02 2.02459931e-01 1.54719755e-01
1.59707919e-01 -3.11473250e-01 -1.52914971e-01 -3.00686926e-01
-4.56551284e-01 2.30816334e-01 2.36209244e-01 5.99751472e-01
-3.76023561e-01 5.44902325e-01 -8.45670283e-01 -6.33993268e-01
6.87217116e-01 6.61448658e-01 -8.60690296e-01 -2.56597670e-03
-3.05244084e-02 5.09426355e-01 -1.73422918e-01 5.77523351e-01
1.10155737e+00 -2.24118650e-01 -8.82940441e-02 -6.45995677e-01
-4.00265813e-01 -1.55279949e-01 -1.27304256e+00 1.86762667e+00
-6.21909738e-01 6.72417521e-01 2.59314209e-01 -9.56992984e-01
7.22137630e-01 -3.93373594e-02 5.21485448e-01 -1.02120149e+00
4.39337976e-02 1.66762605e-01 -9.48808640e-02 -2.69809186e-01
5.87032199e-01 -2.81960964e-01 4.10197824e-01 3.92615609e-02
-9.94167626e-02 -4.13917989e-01 2.32926056e-01 -1.81338310e-01
7.20413625e-01 -2.72317547e-02 3.57507944e-01 -1.16399787e-01
8.95564914e-01 -7.19515607e-02 4.32647049e-01 6.32643819e-01
-2.03388676e-01 1.11490607e+00 1.96343079e-01 -6.77206755e-01
-1.15621853e+00 -9.74603176e-01 -2.76187986e-01 6.97171509e-01
3.67799103e-01 1.51453048e-01 -6.06780827e-01 -2.49567762e-01
-1.98663965e-01 1.22157112e-01 -4.76322681e-01 4.60800320e-01
-8.46013308e-01 -6.89600468e-01 8.89110342e-02 5.87058067e-01
9.95559216e-01 -9.28666234e-01 -7.94396281e-01 3.56311291e-01
-3.62145990e-01 -1.25927186e+00 -3.95883054e-01 1.47610202e-01
-1.32759881e+00 -1.06676817e+00 -1.17246449e+00 -9.12717760e-01
5.72309375e-01 5.65892398e-01 1.19087732e+00 -2.93627799e-01
-2.03761831e-01 2.63494998e-01 7.66638368e-02 1.90576106e-01
6.76645041e-02 2.55679456e-03 -2.26458207e-01 -1.35308594e-01
4.04803678e-02 -9.95238543e-01 -1.16153562e+00 3.47168386e-01
-9.48977172e-01 2.67030030e-01 4.37428504e-01 8.99177492e-01
8.60305369e-01 2.42794171e-01 9.06738415e-02 -7.80570388e-01
4.02521998e-01 7.83026740e-02 -9.48614717e-01 -1.24340318e-01
-6.45524561e-01 2.45271511e-02 7.02211022e-01 -3.97285596e-02
-1.12173688e+00 8.89454484e-02 -4.58573908e-01 -5.82398057e-01
-2.93870121e-01 1.42683908e-01 -3.00048348e-02 -3.98404062e-01
4.42834347e-01 5.90830564e-01 -1.37697324e-01 -4.07948732e-01
1.35805622e-01 3.81976932e-01 4.53999341e-01 -1.03738233e-01
5.45336366e-01 1.16697657e+00 2.64894515e-01 -8.30825806e-01
-8.45185280e-01 -5.75008154e-01 -7.64151812e-01 -4.34839070e-01
1.04296970e+00 -1.03001714e+00 -7.70990610e-01 8.08368325e-01
-1.15088546e+00 3.04807574e-02 1.02199391e-01 4.13301826e-01
-7.37031698e-01 7.03772962e-01 -7.56445348e-01 -4.50685620e-01
-3.80638778e-01 -1.19741142e+00 1.05145156e+00 3.29861760e-01
2.39266649e-01 -1.15807283e+00 5.61906435e-02 5.36280572e-01
5.04141152e-01 1.05015576e-01 8.05569589e-01 1.03347063e-01
-1.03512740e+00 3.53431910e-01 -5.79141259e-01 2.14752138e-01
4.11261059e-02 -3.93656492e-01 -1.10912728e+00 -2.54345328e-01
2.58035660e-01 -2.03186825e-01 1.10543859e+00 1.04717636e+00
1.26865458e+00 1.96487918e-01 -1.95902228e-01 1.10197604e+00
1.92461872e+00 1.62790835e-01 7.79525697e-01 7.11413085e-01
7.40972221e-01 6.63735986e-01 5.63519716e-01 4.49546784e-01
4.64164704e-01 6.17022097e-01 6.92220509e-01 -1.13659754e-01
-9.91971642e-02 -1.16613298e-03 -1.80504128e-01 5.44387758e-01
-3.75910997e-01 -1.44771799e-01 -6.57177687e-01 6.01471484e-01
-1.64764822e+00 -1.08379066e+00 -2.87174493e-01 2.09564090e+00
5.11281729e-01 3.19598615e-01 -1.25840619e-01 2.65065938e-01
5.87576151e-01 5.28137505e-01 -5.61228752e-01 -1.56993583e-01
-1.76623493e-01 7.93281719e-02 6.55566096e-01 6.65661931e-01
-1.03719318e+00 7.41894186e-01 6.42634964e+00 8.98648798e-01
-1.03315985e+00 7.08856508e-02 6.84687376e-01 -1.28447533e-01
-4.90654260e-01 -2.92165011e-01 -8.08471978e-01 1.08686961e-01
1.22661248e-01 1.62986308e-01 4.62859958e-01 6.84652746e-01
2.08196089e-01 -5.98697186e-01 -9.82575536e-01 1.50675726e+00
1.33343577e-01 -1.55114198e+00 3.46454680e-02 1.05035663e-01
1.02740347e+00 2.87304610e-01 -6.61666840e-02 -1.60953596e-01
-1.40845254e-01 -8.55697870e-01 3.88943434e-01 2.61324078e-01
4.92134869e-01 -9.11528587e-01 6.13812864e-01 4.23388034e-01
-1.35244691e+00 -7.16933832e-02 -6.84526205e-01 -2.37690717e-01
3.03761631e-01 7.06946611e-01 -9.50639173e-02 6.83164656e-01
9.55226660e-01 8.22542250e-01 -3.67459178e-01 1.12060642e+00
1.80464968e-01 -1.30510941e-01 -1.91189393e-01 2.07773000e-01
3.13670754e-01 -3.83411705e-01 3.81311566e-01 7.33161867e-01
2.69335181e-01 -1.32273138e-02 -8.74670818e-02 9.14790094e-01
1.74254060e-01 -1.31038278e-01 -7.48995841e-01 6.98190331e-01
5.88025749e-02 1.16728628e+00 -9.02002513e-01 -2.73852229e-01
-6.38631284e-01 1.10699356e+00 3.77525866e-01 1.85384080e-01
-5.49804926e-01 -5.09137660e-02 3.87139589e-01 2.65404671e-01
5.45422196e-01 -2.27964759e-01 -3.48740160e-01 -1.19420290e+00
5.68480827e-02 -6.25482857e-01 3.70181948e-01 -8.99875641e-01
-1.12300670e+00 6.78842306e-01 9.94810686e-02 -1.40288842e+00
-2.40889251e-01 -5.72006583e-01 -2.68252283e-01 8.96589160e-01
-1.96002841e+00 -8.74875844e-01 -6.21838868e-01 9.13409173e-01
9.75703716e-01 1.16713934e-01 3.19652826e-01 4.57190126e-01
-1.11094885e-01 1.36052206e-01 1.67089388e-01 -2.84484595e-01
4.65025246e-01 -1.03310156e+00 2.34727085e-01 6.86045170e-01
-1.72889948e-01 2.03062505e-01 4.95853245e-01 -4.03215289e-01
-1.17563355e+00 -7.57171690e-01 5.38893938e-01 -6.38225153e-02
2.98813075e-01 -5.28775156e-02 -9.02762294e-01 1.60326391e-01
4.62828934e-01 5.25137745e-02 3.15502199e-04 -2.31322557e-01
1.71831008e-02 -2.53307253e-01 -1.26112020e+00 4.45030421e-01
1.27131820e+00 -3.11197877e-01 -4.01874065e-01 -1.05335284e-03
4.23238456e-01 -5.81589699e-01 -7.33496308e-01 6.05519116e-01
6.19302988e-01 -2.01289725e+00 1.32908654e+00 2.44999275e-01
6.22657180e-01 -2.55010813e-01 -1.67893440e-01 -1.26009989e+00
-4.56585824e-01 -1.33039309e-02 -1.01431813e-02 7.85877526e-01
2.51576919e-02 -7.21169293e-01 1.07111931e+00 1.47481486e-01
-1.80322662e-01 -1.00455022e+00 -1.04426885e+00 -5.49763739e-01
-1.23490453e-01 -2.23736092e-01 1.02323420e-01 6.49137616e-01
-5.92042387e-01 3.12767357e-01 -3.94000977e-01 1.99563488e-01
1.17017412e+00 4.42643195e-01 5.05801320e-01 -1.33216345e+00
-3.61624181e-01 -5.44069171e-01 -3.51741493e-01 -1.34617198e+00
-2.69070268e-01 -3.45708042e-01 -1.36675522e-01 -1.94175959e+00
4.28850919e-01 -8.61159340e-03 7.31060207e-02 -1.22272469e-01
-1.41255051e-01 5.89822650e-01 -6.91383705e-02 9.78954285e-02
-2.27259740e-01 4.72774923e-01 1.76674652e+00 8.94126762e-03
-1.79907620e-01 -9.84415710e-02 -3.73136014e-01 1.15303671e+00
6.42450869e-01 -1.19185254e-01 -6.00040853e-01 -6.40855670e-01
1.42569929e-01 4.15398747e-01 2.94352561e-01 -1.24593830e+00
3.02588820e-01 -9.02419090e-02 6.84205532e-01 -1.25321746e+00
7.68733740e-01 -8.05555284e-01 1.21772811e-02 5.31845391e-01
1.38693720e-01 1.69454828e-01 7.00286180e-02 5.12266755e-01
-4.58873123e-01 -1.12893097e-01 1.17839062e+00 -5.98855972e-01
-1.14155388e+00 5.38832903e-01 -1.42272934e-01 1.93786882e-02
7.65776455e-01 -7.66079068e-01 -4.39308211e-02 -3.53565603e-01
-6.45899534e-01 1.70228481e-01 7.54650295e-01 1.32749468e-01
9.46464002e-01 -1.09786546e+00 -5.63710630e-01 3.53533953e-01
-8.84586349e-02 3.39998603e-01 7.02587008e-01 8.25297058e-01
-8.75263810e-01 5.58234453e-01 -5.41539907e-01 -9.93245304e-01
-1.30659568e+00 3.95806193e-01 5.18215954e-01 -4.34359103e-01
-7.58755982e-01 8.82930696e-01 5.82040429e-01 -3.59137148e-01
3.31586391e-01 -3.63690734e-01 -5.65038145e-01 5.99926934e-02
4.35855955e-01 5.29701352e-01 7.73097873e-02 -3.76117527e-01
-2.14412913e-01 1.36663365e+00 -5.19333780e-02 -2.04778284e-01
1.46313345e+00 -5.04803956e-01 -9.31523070e-02 1.49305701e-01
1.22148275e+00 -2.40111172e-01 -1.46565676e+00 -6.33182451e-02
-4.53173369e-01 -9.16138113e-01 3.52347553e-01 -1.46395594e-01
-1.43330324e+00 1.01520562e+00 8.10170233e-01 1.61145225e-01
1.52219462e+00 -1.51110470e-01 8.36497962e-01 4.05437872e-02
7.68104374e-01 -9.33345735e-01 1.04975775e-01 6.34139478e-01
6.73799634e-01 -1.45540464e+00 1.67793453e-01 -6.37909412e-01
-1.64505690e-01 1.28980088e+00 9.85027492e-01 -2.15560675e-01
5.21479309e-01 2.26037763e-02 -5.73799610e-02 -2.67112374e-01
-5.96148431e-01 -2.98216045e-01 1.59787804e-01 7.21888006e-01
2.89821118e-01 -4.76136506e-01 -3.69151920e-01 -2.24721864e-01
6.30660281e-02 -2.78615449e-02 5.80511451e-01 6.75970674e-01
-6.19943142e-01 -9.40973341e-01 -5.55540681e-01 4.20579940e-01
-4.45522100e-01 -2.57009268e-01 8.99173990e-02 7.78137207e-01
1.18642673e-01 6.63181067e-01 1.51299596e-01 -2.29462191e-01
4.91787702e-01 -4.15972352e-01 1.01573575e+00 -3.36599827e-01
-2.68249363e-01 2.77893662e-01 -8.37673545e-02 -8.30744505e-01
-8.44402194e-01 -5.84732294e-01 -9.59339917e-01 -3.88099879e-01
-1.40406385e-01 -2.84898698e-01 2.20151916e-01 7.20723450e-01
5.23418076e-02 4.07882094e-01 6.96902990e-01 -1.40340352e+00
-2.98355132e-01 -7.74766684e-01 -6.49112582e-01 1.85365632e-01
3.86312574e-01 -6.89335406e-01 -4.04369444e-01 -2.25001216e-01] | [9.065885543823242, -2.5014808177948] |
73a4dd4e-dd18-4317-ab0c-2494f7fc2e45 | comment-reflections-on-the-deconfounder | 1910.08042 | null | https://arxiv.org/abs/1910.08042v1 | https://arxiv.org/pdf/1910.08042v1.pdf | Comment: Reflections on the Deconfounder | The aim of this comment (set to appear in a formal discussion in JASA) is to draw out some conclusions from an extended back-and-forth I have had with Wang and Blei regarding the deconfounder method proposed in "The Blessings of Multiple Causes" [arXiv:1805.06826]. I will make three points here. First, in my role as the critic in this conversation, I will summarize some arguments about the lack of causal identification in the bulk of settings where the "informal" message of the paper suggests that the deconfounder could be used. This is a point that is discussed at length in D'Amour 2019 [arXiv:1902.10286], which motivated the results concerning causal identification in Theorems 6--8 of "Blessings". Second, I will argue that adding parametric assumptions to the working model in order to obtain identification of causal parameters (a strategy followed in Theorem 6 and in the experimental examples) is a risky strategy, and should only be done when extremely strong prior information is available. Finally, I will consider the implications of the nonparametric identification results provided for a narrow, but non-trivial, set of causal estimands in Theorems 7 and 8. I will highlight that these results may be even more interesting from the perspective of detecting causal identification from observed data, under relatively weak assumptions about confounders. | ["Alexander D'Amour"] | 2019-10-17 | null | null | null | null | ['causal-identification'] | ['reasoning'] | [ 2.70729780e-01 4.30157334e-01 -4.02995944e-01 -3.11177284e-01
-4.32707220e-01 -4.65924472e-01 6.20398879e-01 3.01352143e-01
-4.47317123e-01 1.02446401e+00 5.50615072e-01 -8.97110045e-01
-9.43469822e-01 -6.12474620e-01 -8.44385028e-01 -5.34655690e-01
-2.51202881e-01 -1.22972265e-01 -1.53340504e-01 -2.41215900e-02
5.43736219e-01 2.35668272e-01 -1.35874033e+00 -5.58068231e-02
8.33490133e-01 9.92940143e-02 -1.81473702e-01 5.72472811e-01
3.44085842e-01 7.61373222e-01 -5.83631754e-01 -6.49726987e-01
-3.30290683e-02 -7.63134122e-01 -9.26727951e-01 -3.03269893e-01
1.27998918e-01 -2.41527066e-01 -2.99435884e-01 9.40863609e-01
2.75363594e-01 -5.56538254e-02 8.30694795e-01 -1.25813317e+00
-4.61601883e-01 9.33184922e-01 -5.75533569e-01 4.70952451e-01
7.99757168e-02 -3.88658121e-02 9.17933702e-01 -5.22652805e-01
3.03120613e-01 1.54708934e+00 6.16796911e-01 3.31585556e-01
-1.18350041e+00 -8.56919706e-01 2.49102667e-01 8.63138470e-04
-1.10552263e+00 -5.34197628e-01 3.12310189e-01 -6.46091580e-01
3.08256000e-01 6.39776230e-01 2.31944263e-01 1.30829072e+00
1.93132132e-01 2.89566070e-01 1.31750739e+00 -6.04834378e-01
5.61287552e-02 1.18462956e-02 3.92607301e-01 3.04961801e-01
9.79026556e-01 5.65191150e-01 -4.26371455e-01 -5.90834022e-01
8.56447279e-01 -4.72225815e-01 -2.14750364e-01 -1.49675444e-01
-1.22067690e+00 1.23159695e+00 -5.49527742e-02 3.88948321e-01
-1.94853008e-01 2.09485978e-01 3.69163364e-01 3.73906374e-01
5.48775196e-01 2.92942584e-01 -3.42812717e-01 -1.54569805e-01
-5.91283679e-01 3.87750328e-01 8.82170439e-01 5.14799297e-01
1.30120844e-01 -1.86075941e-01 1.92521766e-01 4.53068376e-01
4.61465210e-01 4.81996477e-01 5.15576117e-02 -1.12021708e+00
2.89501011e-01 -2.07894612e-02 4.41059321e-01 -9.75701451e-01
-5.75925887e-01 -2.90461302e-01 -6.81374788e-01 7.71633722e-03
8.48184466e-01 -6.76491857e-01 -1.70691982e-01 1.90342379e+00
8.26101005e-02 9.04586837e-02 -2.85634011e-01 7.27991521e-01
3.89402360e-01 1.22363493e-01 4.15967911e-01 -5.42161584e-01
1.10559773e+00 2.80936956e-02 -6.93761408e-01 1.81714565e-01
7.30465531e-01 -7.80745804e-01 7.98554063e-01 4.11337376e-01
-1.01306319e+00 -3.82612646e-01 -8.77570748e-01 2.60659516e-01
-1.99458096e-02 -2.84461409e-01 1.11657047e+00 9.86269653e-01
-6.13886714e-01 6.33817554e-01 -5.82842767e-01 -7.87540853e-01
-3.09538487e-02 2.72863239e-01 -1.62036270e-02 2.60600984e-01
-1.41398335e+00 6.07253969e-01 -3.48034240e-02 -6.56913267e-03
-5.20261705e-01 -7.83775091e-01 -4.55308259e-01 2.81040780e-02
5.91104984e-01 -7.93221056e-01 1.05695748e+00 -9.53750670e-01
-9.18431282e-01 6.50165558e-01 -2.25439087e-01 -1.44661754e-01
6.26060784e-01 -1.87030777e-01 -5.30389786e-01 1.85559783e-02
3.63026589e-01 1.30375512e-02 3.06514353e-01 -1.23454559e+00
-5.90948641e-01 -4.53106135e-01 2.63230652e-01 -1.03236392e-01
1.67266384e-01 5.53969502e-01 2.64099419e-01 -8.59641194e-01
4.73688617e-02 -8.45676839e-01 -2.03406010e-02 -5.73113203e-01
-6.02684200e-01 -4.77918744e-01 2.02942699e-01 -3.12109292e-01
1.26462185e+00 -2.04558229e+00 -3.58071685e-01 3.86514455e-01
3.35558020e-02 -2.99901634e-01 2.60466993e-01 7.21163034e-01
-7.94771850e-01 7.71541357e-01 -1.36674613e-01 1.35117546e-01
2.14789391e-01 -1.12765640e-01 -4.74323839e-01 8.71023118e-01
-1.00471914e-01 8.10858548e-01 -7.53318071e-01 -3.35591882e-01
3.71473804e-02 1.92697778e-01 -2.68335104e-01 -3.21860701e-01
4.89142060e-01 4.27137017e-01 -3.23926270e-01 2.08409101e-01
4.82100517e-01 -2.52560794e-01 4.43956703e-01 4.66389060e-02
-6.45491421e-01 4.46694762e-01 -1.22377753e+00 1.00760078e+00
2.12555483e-01 4.26247120e-01 -7.19852075e-02 -1.36934376e+00
3.88471544e-01 4.83853191e-01 3.44599873e-01 -4.05909151e-01
2.15525061e-01 1.75109893e-01 5.09432614e-01 -5.00794828e-01
1.28313556e-01 -5.15480697e-01 -3.65272105e-01 4.17280495e-01
-3.39056373e-01 1.36175662e-01 1.51184708e-01 6.76679090e-02
8.51098478e-01 -7.98988417e-02 3.15799177e-01 -7.98287749e-01
7.14940205e-02 -8.78799781e-02 5.93849659e-01 1.33757961e+00
3.13602947e-02 2.28800908e-01 9.87409055e-01 -6.67645335e-02
-8.48714113e-01 -1.01123786e+00 -5.74209273e-01 6.78319156e-01
-2.28126660e-01 -1.08054362e-01 -4.45544451e-01 -4.98594791e-01
1.85100004e-01 1.11883497e+00 -1.00983393e+00 8.99207070e-02
-4.32901770e-01 -1.14374304e+00 6.69854045e-01 3.30473095e-01
2.66180933e-02 -4.74040329e-01 -6.17082596e-01 -1.22860990e-01
-1.88468069e-01 -4.56476122e-01 1.47576272e-01 2.40683511e-01
-9.63595212e-01 -1.45068049e+00 -6.20143950e-01 -1.22985944e-01
3.77358437e-01 2.30273753e-01 6.36799932e-01 2.23582000e-01
2.46665031e-01 5.16867518e-01 -3.71281981e-01 -7.31169999e-01
-6.18728042e-01 -3.26826483e-01 1.11899130e-01 -3.92097861e-01
5.25754333e-01 -6.81805313e-01 -4.01697010e-01 2.09116846e-01
-7.50416696e-01 -2.32647657e-01 4.20030713e-01 7.91442633e-01
-1.51083767e-01 7.72706792e-02 9.10304725e-01 -1.19820893e+00
4.08438146e-01 -7.30350673e-01 -6.00116074e-01 1.42374128e-01
-7.37369418e-01 -9.67045426e-02 1.78206041e-01 -8.52247700e-02
-1.11903751e+00 -8.47817719e-01 1.84328198e-01 1.87360123e-01
-3.13279450e-01 6.96923494e-01 -9.29874033e-02 2.24705562e-01
7.87816644e-01 -4.44306850e-01 1.29497454e-01 -6.59754395e-01
3.53417248e-02 4.14533287e-01 2.65084922e-01 -6.75343096e-01
7.26769626e-01 4.65900689e-01 3.00671875e-01 -7.20176339e-01
-9.20731306e-01 -1.99766770e-01 -3.74564677e-01 2.15662524e-01
5.84687769e-01 -7.94879973e-01 -1.20022011e+00 1.00700133e-01
-9.30783212e-01 -6.45155549e-01 -1.58595353e-01 1.09875619e+00
-7.01250136e-01 4.18925166e-01 -4.62791026e-01 -9.85440969e-01
5.31791270e-01 -7.81950831e-01 3.26200932e-01 3.61469239e-02
-6.69642568e-01 -1.30793893e+00 6.14830256e-02 2.97611356e-01
-3.98524180e-02 3.08000833e-01 1.21992385e+00 -6.16820872e-01
1.85261182e-02 5.19986227e-02 -1.70992985e-01 -2.79934436e-01
2.60240138e-01 1.92912400e-01 -9.58683491e-01 -4.47197370e-02
1.65519252e-01 -7.43409023e-02 6.79021835e-01 8.44550252e-01
1.09028697e+00 -4.04651374e-01 -3.88726234e-01 2.87736077e-02
1.43934178e+00 2.40136996e-01 4.86934692e-01 2.05398440e-01
3.56518924e-01 1.18394387e+00 4.62806553e-01 5.64627826e-01
4.28485602e-01 5.50832689e-01 1.21199928e-01 9.69799161e-02
-1.20125236e-02 -4.50293988e-01 1.28531218e-01 2.68121928e-01
-7.64254257e-02 -3.48395556e-01 -6.21827960e-01 6.55189931e-01
-1.83613420e+00 -1.04294586e+00 -9.21126723e-01 2.68773389e+00
7.80132353e-01 3.97950634e-02 6.94559097e-01 3.46984677e-02
9.60785031e-01 -2.24053204e-01 -8.43001828e-02 -5.41127443e-01
-2.55617291e-01 1.06324151e-01 7.89689481e-01 7.45343864e-01
-7.24327743e-01 1.18390299e-01 7.31091595e+00 4.58093554e-01
-6.09640718e-01 1.27005965e-01 8.27171743e-01 9.71497595e-02
-6.26001239e-01 5.38126588e-01 -4.38410729e-01 5.87964952e-01
1.46038485e+00 -4.35752094e-01 -9.92595553e-02 -1.74004529e-02
9.08338785e-01 -7.27222860e-01 -1.19687974e+00 3.72076392e-01
-2.71733195e-01 -8.81346941e-01 -3.32380146e-01 3.48910451e-01
5.39155245e-01 -4.23196644e-01 4.06440422e-02 5.65323383e-02
5.59683979e-01 -1.27366352e+00 5.95439672e-01 4.98922229e-01
4.89754826e-01 -8.19485664e-01 6.58324540e-01 4.21886653e-01
-2.95332134e-01 -1.44679368e-01 -4.00434315e-01 -6.10772312e-01
6.01556599e-02 9.64842319e-01 -4.29223090e-01 8.23772490e-01
5.70035338e-01 1.09025128e-01 -2.86865383e-01 1.12949526e+00
-3.45428795e-01 1.22757113e+00 -2.06579775e-01 5.34254573e-02
-6.29818290e-02 1.19385280e-01 6.03789985e-01 1.05740249e+00
3.26268643e-01 2.96593696e-01 -5.88183224e-01 9.78137314e-01
4.29684609e-01 8.71088132e-02 -6.54414177e-01 -8.68861228e-02
3.95973295e-01 6.36097133e-01 -6.75698221e-01 -1.15095317e-01
-7.14165628e-01 2.35457316e-01 -1.84361920e-01 2.83733636e-01
-7.70428956e-01 -3.11594814e-01 1.44696593e-01 3.69466186e-01
-1.76347777e-01 6.62820265e-02 -7.41442025e-01 -9.25098777e-01
-3.38012934e-01 -7.52318680e-01 5.26808023e-01 -5.34024596e-01
-1.16985464e+00 -5.38153112e-01 8.20526958e-01 -6.14037812e-01
-2.41848066e-01 -4.33637589e-01 -6.65498972e-01 1.14178002e+00
-7.84458101e-01 -5.70179999e-01 3.09890687e-01 1.93381637e-01
8.12815651e-02 4.83411312e-01 7.56414711e-01 1.01854220e-01
-3.72857720e-01 5.66806257e-01 3.73023674e-02 -3.05164382e-02
9.37144339e-01 -1.38173842e+00 1.26897529e-01 8.01275551e-01
-1.63743228e-01 1.13086700e+00 1.20296562e+00 -8.99198174e-01
-1.02673125e+00 -4.81974453e-01 1.27120638e+00 -4.44869846e-01
1.02634835e+00 -1.48486316e-01 -5.87737381e-01 8.35862637e-01
3.20968509e-01 -7.85668612e-01 8.19951713e-01 7.11177647e-01
-1.54551463e-02 3.03082079e-01 -9.09658253e-01 6.94408417e-01
9.76617932e-01 -1.46911982e-02 -8.37312222e-01 3.77241045e-01
4.50568914e-01 1.05790548e-01 -1.01894784e+00 4.18177456e-01
6.91727638e-01 -1.16645992e+00 6.92015946e-01 -6.79010153e-01
3.49552751e-01 1.27816033e-02 -4.02841605e-02 -1.00577700e+00
-3.67129713e-01 -7.53443301e-01 7.35821068e-01 1.29134798e+00
4.21937346e-01 -9.44020867e-01 4.99717146e-01 7.39091575e-01
2.21698403e-01 -3.36061090e-01 -8.62759531e-01 -7.29016066e-01
7.58461535e-01 -5.06750464e-01 2.74040729e-01 1.37904954e+00
3.92522067e-01 2.04228789e-01 -2.12832272e-01 1.02877997e-01
8.66149127e-01 -1.55240893e-01 7.88757443e-01 -1.61857212e+00
-4.04600382e-01 -1.94924131e-01 -7.92081654e-03 -5.75180411e-01
-5.81648611e-02 -3.48437816e-01 -4.18764532e-01 -1.05281222e+00
6.29460752e-01 -5.52731216e-01 -2.86185175e-01 2.36059397e-01
-2.05945656e-01 -2.73595601e-01 1.21885024e-01 2.67060608e-01
4.32435088e-02 9.58913714e-02 1.06847429e+00 2.54621416e-01
-3.91367115e-02 3.13803345e-01 -1.38770568e+00 8.06616366e-01
7.58711100e-01 -6.25013709e-01 -4.73976165e-01 1.00723803e-01
6.26668334e-01 5.84823489e-01 1.22020876e+00 -3.32101196e-01
-8.53653997e-02 -5.85209489e-01 3.13176364e-01 -3.70192736e-01
-8.97936225e-02 -4.18245524e-01 4.15738493e-01 6.07610583e-01
-4.98552293e-01 1.25191256e-01 3.77305329e-01 4.46615309e-01
2.18652755e-01 -6.05825722e-01 3.71170521e-01 -1.72931567e-01
-1.14882372e-01 -3.74877751e-01 -3.97447973e-01 1.10918671e-01
6.40146077e-01 -4.33960967e-02 -4.95661050e-01 -5.93389630e-01
-7.37672508e-01 6.86932029e-03 3.57618302e-01 2.61299968e-01
-4.91336770e-02 -1.06473410e+00 -8.97760451e-01 -4.06151146e-01
-1.45859301e-01 -7.02165902e-01 4.90288615e-01 1.52086663e+00
8.16938952e-02 7.27858663e-01 2.82674670e-01 -1.05142437e-01
-1.09764814e+00 6.73120618e-01 1.73817039e-01 1.24538355e-01
-5.14249444e-01 5.61401129e-01 7.14763224e-01 1.24803387e-01
-4.51413281e-02 -1.96461305e-01 3.44245657e-02 6.38231216e-03
4.52888191e-01 7.48567164e-01 -3.29012156e-01 -3.51003379e-01
-4.10813123e-01 1.16075706e-02 1.59381509e-01 -4.48269188e-01
1.30465209e+00 -4.80966777e-01 -3.36641878e-01 1.06365228e+00
9.51325417e-01 6.48910701e-01 -8.65182042e-01 2.68583894e-01
1.53824538e-01 -2.60018587e-01 -3.74924302e-01 -9.66579020e-01
-2.08741993e-01 6.15424573e-01 3.71746480e-01 5.38346589e-01
6.93547666e-01 -1.73093043e-02 -3.51980120e-01 -3.73838022e-02
1.86489627e-01 -8.47879171e-01 -5.29696703e-01 -1.65008027e-02
8.60953927e-01 -8.88197601e-01 2.76953012e-01 -3.86258394e-01
-2.83601999e-01 1.07435608e+00 2.04295099e-01 -3.06216236e-02
5.23558795e-01 2.84714103e-02 -3.34495753e-01 -4.06154662e-01
-6.39102578e-01 7.33748153e-02 6.58566889e-04 4.50151384e-01
6.18506312e-01 1.65419728e-01 -1.12200880e+00 6.52000964e-01
-2.79656798e-01 -4.96381847e-03 1.02731681e+00 5.16503990e-01
-3.52427363e-01 -1.29633737e+00 -8.59474063e-01 4.52788115e-01
-1.01790309e+00 -1.65043965e-01 -4.55071658e-01 1.61707807e+00
-3.29543347e-03 1.55423915e+00 -4.75466661e-02 6.71436489e-02
1.31198958e-01 1.84011713e-01 4.74125504e-01 -4.04883832e-01
-3.65617037e-01 3.88918281e-01 3.67515981e-01 -2.54902214e-01
-8.73137593e-01 -1.09739983e+00 -6.75447047e-01 -8.22350085e-01
-3.01225871e-01 4.71047401e-01 2.38194793e-01 1.15934730e+00
-1.59766093e-01 4.11384344e-01 4.46723998e-01 -8.82949457e-02
-7.22417057e-01 -9.90823805e-01 -8.83468270e-01 -4.40970389e-03
1.49308205e-01 -7.58161366e-01 -7.86454916e-01 -2.34146923e-01] | [8.057106018066406, 5.297098159790039] |
3e6c1f23-e44d-44cb-b24b-f6b3afa12126 | shall-we-pretrain-autoregressive-language | 2304.06762 | null | https://arxiv.org/abs/2304.06762v1 | https://arxiv.org/pdf/2304.06762v1.pdf | Shall We Pretrain Autoregressive Language Models with Retrieval? A Comprehensive Study | Large decoder-only language models (LMs) can be largely improved in terms of perplexity by retrieval (e.g., RETRO), but its impact on text generation quality and downstream task accuracy is unclear. Thus, it is still an open question: shall we pretrain large autoregressive LMs with retrieval? To answer it, we perform a comprehensive study on a scalable pre-trained retrieval-augmented LM (i.e., RETRO) compared with standard GPT and retrieval-augmented GPT incorporated at fine-tuning or inference stages. We first provide the recipe to reproduce RETRO up to 9.5B parameters while retrieving a text corpus with 330B tokens. Based on that, we have the following novel findings: i) RETRO outperforms GPT on text generation with much less degeneration (i.e., repetition), moderately higher factual accuracy, and slightly lower toxicity with a nontoxic retrieval database. ii) On the LM Evaluation Harness benchmark, RETRO largely outperforms GPT on knowledge-intensive tasks, but is on par with GPT on other tasks. Furthermore, we introduce a simple variant of the model, RETRO++, which largely improves open-domain QA results of original RETRO (e.g., EM score +8.6 on Natural Question) and significantly outperforms retrieval-augmented GPT across different model sizes. Our findings highlight the promising direction of pretraining autoregressive LMs with retrieval as future foundation models. We release our implementation at: https://github.com/NVIDIA/Megatron-LM#retro | ['Bryan Catanzaro', 'Anima Anandkumar', 'Chaowei Xiao', 'Bo Li', 'Oleksii Kuchaiev', 'Yi Dong', 'Mohammad Shoeybi', 'Zihan Liu', 'Lawrence McAfee', 'Peng Xu', 'Wei Ping', 'Boxin Wang'] | 2023-04-13 | null | null | null | null | ['open-question'] | ['natural-language-processing'] | [ 8.07390511e-02 6.18434884e-02 -5.53543605e-02 -3.03833038e-02
-1.72633648e+00 -8.12253773e-01 7.27086306e-01 -2.52072662e-01
-6.35280430e-01 9.19493496e-01 4.42764044e-01 -6.52915001e-01
3.57161202e-02 -5.64924061e-01 -9.84009087e-01 -5.13223648e-01
4.58607793e-01 7.26348519e-01 8.29137489e-02 -4.98856127e-01
1.44019961e-01 4.27931361e-02 -9.17410135e-01 3.46718490e-01
1.11696541e+00 4.84847397e-01 4.43341255e-01 8.69109035e-01
1.93999723e-01 7.17722237e-01 -7.24767685e-01 -7.18885541e-01
-1.99670792e-02 -2.87753791e-01 -9.42203462e-01 -5.51352084e-01
4.84138191e-01 -5.29857218e-01 -4.42591906e-01 4.40288872e-01
1.16379583e+00 3.04448009e-01 8.43645334e-01 -7.12806582e-01
-9.65352416e-01 8.80176425e-01 -3.43533337e-01 1.39798403e-01
3.19559991e-01 5.03701687e-01 1.24770522e+00 -1.01475716e+00
5.66617072e-01 1.35871029e+00 5.78186333e-01 6.92902863e-01
-1.21576345e+00 -6.33474946e-01 7.31565431e-02 -1.43398568e-01
-1.42394865e+00 -7.64189124e-01 -9.88882873e-03 -1.76885240e-02
1.64830661e+00 2.90765107e-01 7.10186511e-02 1.49953878e+00
3.11855257e-01 9.19255733e-01 9.66466129e-01 -4.46033299e-01
-1.00647353e-01 1.29162803e-01 5.39506786e-02 4.07911569e-01
2.22495764e-01 2.26259883e-02 -6.45594299e-01 -7.47821629e-02
3.32664132e-01 -6.63053215e-01 -4.28781688e-01 4.97049987e-01
-1.23851848e+00 8.53443086e-01 4.92662005e-02 7.15633109e-02
-3.33598763e-01 5.11159360e-01 2.82812119e-01 3.90472054e-01
5.85818589e-01 7.34966695e-01 -7.26657689e-01 -4.81614560e-01
-9.66377258e-01 3.49878103e-01 8.86669219e-01 1.11673629e+00
5.04520893e-01 2.19172865e-01 -7.13853419e-01 1.22325909e+00
3.09806526e-01 1.05402172e+00 7.41598248e-01 -7.75367618e-01
7.84673214e-01 -8.96194130e-02 -6.18396401e-02 -3.51284772e-01
-3.32826763e-01 -5.69501162e-01 -5.86750031e-01 -7.05245376e-01
3.02756280e-01 -6.61869287e-01 -7.66360044e-01 1.92696869e+00
-1.15736686e-01 -6.24722466e-02 2.38006443e-01 5.79935133e-01
1.18350339e+00 1.10206354e+00 2.36060888e-01 -1.02430927e-02
1.33890665e+00 -1.21190548e+00 -6.95201099e-01 -4.34068173e-01
1.14971352e+00 -1.03525579e+00 1.37432837e+00 3.56102526e-01
-1.27633059e+00 -4.63781506e-01 -6.69240773e-01 -4.33271199e-01
-2.24028423e-01 4.15995330e-01 5.68537116e-01 7.89861798e-01
-1.54235959e+00 3.99589360e-01 -6.53400004e-01 -3.33983630e-01
-1.29775569e-01 2.99339175e-01 -1.55231804e-01 -8.68689939e-02
-1.68314433e+00 1.05868459e+00 4.25574660e-01 -1.89745463e-02
-8.28859985e-01 -7.84711957e-01 -7.44706213e-01 1.10536955e-01
1.93990916e-01 -1.12857580e+00 1.55773985e+00 -4.06831026e-01
-1.94312668e+00 4.38763648e-01 -3.70512217e-01 -5.11018693e-01
2.66175985e-01 -5.51386356e-01 -2.39650145e-01 -7.76254386e-02
-1.02197044e-01 1.06447625e+00 6.35356665e-01 -9.38992798e-01
-3.68345976e-01 1.02358252e-01 -1.38527742e-02 5.70906222e-01
-3.85042995e-01 1.91409718e-02 -8.00174594e-01 -7.18133211e-01
-3.86855274e-01 -1.20478725e+00 2.75838692e-02 -7.94338346e-01
-4.89378542e-01 -3.72833937e-01 3.84314656e-02 -9.39541876e-01
1.46341014e+00 -1.81992888e+00 -1.04911653e-02 -1.26228660e-01
-2.21125141e-01 3.97293627e-01 -7.71306515e-01 6.42182648e-01
6.93296548e-03 4.23517406e-01 -1.44464988e-03 -4.42245811e-01
2.27437869e-01 -1.55291647e-01 -6.36884630e-01 -5.26346713e-02
3.91495615e-01 1.49973249e+00 -6.92076981e-01 -3.69675577e-01
-1.60876527e-01 5.68945229e-01 -7.47498453e-01 7.99448695e-03
-5.26562512e-01 2.71052092e-01 -3.89662594e-01 4.77966398e-01
3.96892399e-01 -3.98294538e-01 1.84960254e-02 1.21982634e-01
1.28803432e-01 8.39608192e-01 -6.98194325e-01 1.75984907e+00
-6.67506576e-01 7.68971145e-01 -4.11801815e-01 -3.34867388e-01
6.35202944e-01 4.42044288e-01 -1.20941341e-01 -1.04694688e+00
-8.51510167e-02 2.71921724e-01 6.43942431e-02 -3.32960129e-01
1.16299820e+00 -6.79068938e-02 -8.04886520e-02 7.15254188e-01
1.25849828e-01 -3.88194740e-01 2.14534223e-01 4.00358945e-01
1.19745004e+00 2.56479919e-01 -9.58516821e-02 -3.47071923e-02
2.08869055e-01 -2.48770919e-02 6.27273545e-02 1.10148954e+00
4.03617024e-01 6.56453490e-01 3.29458266e-01 5.53081274e-01
-8.14809740e-01 -9.76507604e-01 -1.55673534e-01 1.51056969e+00
-3.89854968e-01 -6.59115255e-01 -7.64634728e-01 -3.56458813e-01
-2.25070938e-01 1.15720701e+00 -1.01052299e-01 -3.06256682e-01
-6.87577367e-01 -1.13885474e+00 1.08727562e+00 4.71047729e-01
2.92861611e-01 -1.18822038e+00 1.46868721e-01 1.36712968e-01
-5.29914677e-01 -1.03624880e+00 -6.33804858e-01 -2.01497376e-02
-9.94247973e-01 -2.96999902e-01 -9.70634222e-01 -5.72839618e-01
2.60412365e-01 2.29957506e-01 1.46577322e+00 -4.08142544e-02
3.21085870e-01 5.60032845e-01 -5.35614014e-01 -3.67031127e-01
-5.42762756e-01 6.93957865e-01 -2.14464501e-01 -6.98468029e-01
6.25942796e-02 -1.19360715e-01 -6.21217430e-01 1.59414724e-01
-8.26219916e-01 -7.70316496e-02 1.07501662e+00 1.06486511e+00
4.72393364e-01 -3.23575467e-01 8.67684186e-01 -9.33607817e-01
1.17515206e+00 -5.85425138e-01 -4.27143097e-01 4.64051634e-01
-6.16458535e-01 2.80912727e-01 3.55711341e-01 -5.65860212e-01
-1.33540785e+00 -5.72994769e-01 -5.29240131e-01 5.55256940e-02
2.78080583e-01 7.88808525e-01 7.10788295e-02 3.47926855e-01
7.45450437e-01 3.93017918e-01 -2.69340008e-01 -3.75200808e-01
7.73689091e-01 8.31234515e-01 1.85391292e-01 -8.20274532e-01
5.70679307e-01 -1.05480663e-01 -4.75805193e-01 -6.92045987e-01
-6.29406929e-01 -3.43178660e-01 -2.64103934e-02 2.79019535e-01
6.49963081e-01 -1.40939033e+00 -5.33033907e-01 4.19103354e-01
-1.26210582e+00 -1.03981698e+00 -6.27771094e-02 5.52566230e-01
-5.18852949e-01 3.20543021e-01 -9.61046934e-01 -6.74865007e-01
-9.30703878e-01 -1.08180034e+00 1.40582430e+00 -7.77696371e-02
-4.45021003e-01 -1.17025900e+00 2.48093560e-01 6.73531592e-01
5.15470862e-01 -6.25153899e-01 1.06478643e+00 -7.65116751e-01
-5.09771347e-01 7.08339969e-03 -1.74406067e-01 4.69874084e-01
-2.89133966e-01 -1.06185265e-01 -1.13043714e+00 -4.48360920e-01
-2.54165620e-01 -5.47186494e-01 1.01800227e+00 3.91006887e-01
8.86317909e-01 -2.52372265e-01 -1.88515782e-01 4.74928170e-01
1.09348059e+00 6.85991067e-03 8.26258659e-01 1.89705536e-01
3.83548588e-01 3.81752014e-01 6.46426022e-01 2.02244863e-01
3.90850157e-01 6.69508576e-01 -2.11609393e-01 1.53160676e-01
-3.97990257e-01 -4.55987573e-01 1.01596487e+00 1.26602697e+00
1.49469361e-01 -1.03599489e+00 -7.97655046e-01 3.31830263e-01
-1.57519519e+00 -6.35545135e-01 -2.79327154e-01 2.19517064e+00
1.13578510e+00 1.06885307e-01 -2.36994341e-01 -4.31348890e-01
4.19042021e-01 2.71800440e-02 -4.69573885e-01 -4.95771348e-01
-4.15410638e-01 6.32454932e-01 4.44963306e-01 6.98735833e-01
-5.67815244e-01 1.36811280e+00 6.65437746e+00 1.49282813e+00
-1.00901484e+00 2.88744181e-01 6.95811391e-01 -3.26088667e-01
-5.24804294e-01 6.14690520e-02 -1.24375963e+00 4.78157163e-01
1.67106509e+00 -2.78722495e-01 4.41067070e-01 4.67438281e-01
2.36992508e-01 -1.18147358e-01 -1.02333188e+00 6.72370613e-01
1.20609581e-01 -1.11421645e+00 3.62435043e-01 -2.21449994e-02
8.17306101e-01 4.10116345e-01 4.95828480e-01 1.10804355e+00
5.05707264e-01 -1.24534178e+00 5.62931716e-01 5.54650843e-01
9.24898267e-01 -4.59915012e-01 8.29231679e-01 4.69593883e-01
-5.46875954e-01 1.59070343e-01 -5.04964769e-01 1.40722215e-01
3.25590402e-01 3.69276553e-01 -1.11015141e+00 6.15286648e-01
3.06445241e-01 3.98403436e-01 -7.34597027e-01 7.91921556e-01
-4.28789854e-01 1.04376435e+00 -3.45444769e-01 -2.51896262e-01
2.44101405e-01 -4.41962853e-02 3.03601056e-01 1.38054669e+00
5.44128716e-01 1.46841640e-02 -6.38903856e-01 7.47737706e-01
-4.22164977e-01 2.65836954e-01 -4.00837481e-01 -1.48679614e-01
6.44987345e-01 1.22996998e+00 2.79904250e-02 -4.66377318e-01
-3.44450265e-01 9.60588872e-01 3.89499724e-01 5.98752856e-01
-9.86905754e-01 -1.45220160e-01 2.14825526e-01 -9.04032663e-02
1.39204457e-01 -3.15144472e-02 -4.45018448e-02 -1.00237918e+00
-2.37947870e-02 -1.19750655e+00 3.39300603e-01 -1.26674902e+00
-1.30374730e+00 6.24109805e-01 -1.46844417e-01 -8.85265768e-01
-5.32843769e-01 -5.04309475e-01 -3.72260630e-01 1.18798292e+00
-1.64961958e+00 -1.01732612e+00 1.55979559e-01 3.74997556e-01
6.52670920e-01 -5.52499630e-02 8.72104466e-01 3.31528187e-01
-6.86067224e-01 1.08085668e+00 5.98110735e-01 -4.84290943e-02
1.38056898e+00 -1.09414697e+00 5.76163292e-01 5.78792393e-01
1.46646500e-01 1.03667462e+00 3.74284357e-01 -6.43672347e-01
-1.52316308e+00 -1.27755845e+00 1.16492784e+00 -1.05723834e+00
7.10613787e-01 -2.80414164e-01 -7.81915367e-01 8.81939828e-01
4.99100834e-01 -7.27997482e-01 5.91005623e-01 2.19396755e-01
-9.56630185e-02 1.23086005e-01 -5.96247792e-01 9.49864507e-01
8.31519604e-01 -5.86495042e-01 -5.42275190e-01 7.08581984e-01
1.21867287e+00 -5.74524879e-01 -9.92068231e-01 2.98073322e-01
4.05773491e-01 -4.35830683e-01 9.96744931e-01 -5.42150438e-01
5.02049863e-01 1.40091941e-01 -6.61142170e-02 -1.44098163e+00
-2.14424610e-01 -7.91192055e-01 -1.18335113e-01 1.26494277e+00
1.03387511e+00 -8.78577232e-01 4.86131251e-01 5.81509173e-01
-4.03853565e-01 -7.79851019e-01 -7.38333106e-01 -8.70588720e-01
8.08416307e-01 -5.18562913e-01 3.06601971e-01 5.09729743e-01
-1.66498572e-01 8.90872777e-01 -3.19747508e-01 -1.14226267e-01
-8.02320316e-02 -3.52629423e-01 8.41847837e-01 -5.49956322e-01
-5.98076522e-01 -4.43555206e-01 4.50552434e-01 -1.67246354e+00
3.79585147e-01 -1.12865996e+00 5.19940890e-02 -1.50643802e+00
3.36510926e-01 -5.09299815e-01 -2.94236634e-02 5.06036878e-01
-5.67256987e-01 2.54171938e-01 1.37977093e-01 2.57731408e-01
-6.94120228e-01 8.03059042e-01 1.29088855e+00 -4.71487120e-02
-1.55685514e-01 -1.40375048e-01 -9.97492790e-01 1.58489019e-01
8.85361791e-01 -3.71699899e-01 -4.90802199e-01 -8.65239084e-01
5.99850416e-01 2.65175134e-01 5.82941212e-02 -6.76310778e-01
2.11170495e-01 2.61937231e-01 1.98224172e-01 -5.77472866e-01
5.65004051e-01 -5.97090274e-02 -7.86266774e-02 2.24539101e-01
-5.55075943e-01 2.78036386e-01 5.63510120e-01 3.51097792e-01
-2.56725661e-02 -4.44560409e-01 2.66894251e-01 1.71392914e-02
-3.39867324e-01 6.26151487e-02 -5.51610887e-01 4.28707123e-01
2.45115831e-01 1.74869046e-01 -8.98590922e-01 -7.02823460e-01
-3.24461699e-01 3.35782290e-01 1.24283589e-01 4.39980537e-01
4.21678543e-01 -1.03274643e+00 -1.11752021e+00 -2.13020131e-01
-2.98010129e-02 -5.62661998e-02 3.15305710e-01 1.03577983e+00
-3.26128602e-01 1.15514660e+00 4.74036723e-01 -2.83900976e-01
-1.08142579e+00 8.65889192e-02 2.01191559e-01 -7.22054601e-01
-9.44375694e-02 9.38873112e-01 3.33648950e-01 -6.86150134e-01
-1.95101136e-03 -2.06788778e-01 -1.15268435e-02 -4.25323807e-02
3.12384576e-01 4.00449783e-01 3.81683409e-01 -3.45524937e-01
-1.54861435e-02 3.48722249e-01 -3.22086245e-01 -5.37181914e-01
1.00298369e+00 -1.33490652e-01 -8.01303163e-02 2.46457756e-01
1.05944073e+00 2.23759666e-01 -8.78908932e-01 -2.34845459e-01
-1.33771896e-01 6.55428097e-02 5.88493347e-02 -1.26813936e+00
-7.29808271e-01 8.96303177e-01 2.69405484e-01 -4.39812839e-01
9.20610368e-01 -6.48969319e-03 9.10938442e-01 9.04481292e-01
1.30033255e-01 -1.09106410e+00 2.78126627e-01 1.04734313e+00
1.06054711e+00 -1.15286815e+00 6.66380599e-02 -2.16381140e-02
-8.68152201e-01 6.61570489e-01 5.69736660e-01 2.94061184e-01
2.17487484e-01 5.76014956e-03 -5.92128858e-02 6.18059747e-03
-1.37570381e+00 -1.57967553e-01 4.03550833e-01 3.66923511e-01
9.95425165e-01 -6.22684620e-02 -3.35063845e-01 6.60283267e-01
-6.56467557e-01 -1.84298068e-01 2.67286062e-01 4.48690385e-01
-3.03416640e-01 -1.17549443e+00 -2.69022077e-01 6.05824888e-01
-5.95886409e-01 -1.03986526e+00 -3.23414057e-01 8.37822020e-01
-5.33960879e-01 1.18867683e+00 -9.73750353e-02 -2.22856000e-01
2.61962026e-01 3.03555906e-01 3.98380995e-01 -7.34065711e-01
-8.34062159e-01 1.50879651e-01 4.99783039e-01 -2.89603025e-01
1.13027982e-01 -5.14424980e-01 -9.78481770e-01 -3.93128157e-01
-7.69970357e-01 1.80608019e-01 6.04504347e-01 6.78393781e-01
6.58816934e-01 5.44094861e-01 2.21829146e-01 -3.74212712e-01
-8.61329436e-01 -1.47885501e+00 -1.42347887e-01 9.42183062e-02
-7.53870159e-02 -1.53230175e-01 -3.72847646e-01 -8.73370692e-02] | [11.630195617675781, 8.803160667419434] |
56382222-a8e8-4e2b-83ee-2bdb9ddc3128 | estimating-metric-poses-of-dynamic-objects | 1808.06753 | null | http://arxiv.org/abs/1808.06753v1 | http://arxiv.org/pdf/1808.06753v1.pdf | Estimating Metric Poses of Dynamic Objects Using Monocular Visual-Inertial Fusion | A monocular 3D object tracking system generally has only up-to-scale pose
estimation results without any prior knowledge of the tracked object. In this
paper, we propose a novel idea to recover the metric scale of an arbitrary
dynamic object by optimizing the trajectory of the objects in the world frame,
without motion assumptions. By introducing an additional constraint in the time
domain, our monocular visual-inertial tracking system can obtain continuous six
degree of freedom (6-DoF) pose estimation without scale ambiguity. Our method
requires neither fixed multi-camera nor depth sensor settings for scale
observability, instead, the IMU inside the monocular sensing suite provides
scale information for both camera itself and the tracked object. We build the
proposed system on top of our monocular visual-inertial system (VINS) to obtain
accurate state estimation of the monocular camera in the world frame. The whole
system consists of a 2D object tracker, an object region-based visual bundle
adjustment (BA), VINS and a correlation analysis-based metric scale estimator.
Experimental comparisons with ground truth demonstrate the tracking accuracy of
our 3D tracking performance while a mobile augmented reality (AR) demo shows
the feasibility of potential applications. | ['Tong Qin', 'Kejie Qiu', 'Hongwen Xie', 'Shaojie Shen'] | 2018-08-21 | null | null | null | null | ['3d-object-tracking'] | ['computer-vision'] | [-5.11516809e-01 -3.11256737e-01 2.11127624e-02 7.05525279e-02
-1.42759353e-01 -9.08971369e-01 5.11377752e-01 -5.44165254e-01
-4.53854948e-01 4.36834574e-01 -4.37569797e-01 -1.04768880e-01
2.00940192e-01 -3.54349852e-01 -8.67305875e-01 -6.53039932e-01
2.65136331e-01 5.78336239e-01 4.17149186e-01 9.38486010e-02
-2.47418545e-02 9.06031072e-01 -9.94419217e-01 -9.65817451e-01
4.67115313e-01 1.17168069e+00 5.92383921e-01 9.06288028e-01
5.04199922e-01 3.89940500e-01 -3.14846486e-01 2.08895937e-01
7.04786301e-01 1.50124788e-01 2.55555212e-01 5.96935213e-01
6.25187278e-01 -6.61741376e-01 -7.31234074e-01 1.29282141e+00
4.09246206e-01 -1.17189497e-01 1.97510839e-01 -1.30362129e+00
-3.70000750e-01 -2.59537578e-01 -6.75437272e-01 2.66271122e-02
6.14335835e-01 2.18757287e-01 3.06099385e-01 -9.96322811e-01
9.30063784e-01 9.84104216e-01 8.32911313e-01 2.58723289e-01
-6.54095471e-01 -4.43560839e-01 -2.31961627e-02 3.98512259e-02
-1.70856929e+00 -3.20308775e-01 6.91276968e-01 -6.43691719e-01
4.62804586e-01 1.74093246e-01 9.85350311e-01 5.83185971e-01
5.26036918e-01 8.92440677e-02 8.11898470e-01 -1.42289504e-01
8.88966769e-02 3.50443184e-01 7.36648068e-02 7.12381661e-01
8.62869918e-01 1.57397285e-01 -1.99068308e-01 -1.63680404e-01
1.26922858e+00 5.22321761e-01 -4.17848200e-01 -1.26182556e+00
-1.61009932e+00 3.05023074e-01 4.97959435e-01 3.81909087e-02
-2.69020170e-01 3.12933981e-01 -1.40787706e-01 2.22099215e-01
1.02203384e-01 -2.08247802e-03 -3.52851421e-01 -4.14310098e-02
-3.81316781e-01 -6.68241009e-02 4.63945419e-01 1.75455058e+00
6.38110578e-01 2.97487736e-01 3.53012204e-01 -1.15561381e-01
8.62685859e-01 1.34960687e+00 3.52311879e-01 -8.70412707e-01
3.83942962e-01 6.55707419e-01 8.77156794e-01 -1.03605592e+00
-5.80282211e-01 -6.11550510e-01 -4.52755332e-01 3.16413045e-01
1.87463462e-01 -3.30708116e-01 -4.61370915e-01 1.36391377e+00
1.09933662e+00 1.11638486e-01 -1.42998442e-01 1.40389979e+00
3.33288938e-01 2.96615601e-01 -7.63089001e-01 -4.31415081e-01
1.29498446e+00 -5.18874943e-01 -8.76447618e-01 -2.52903819e-01
4.94949490e-01 -1.01069450e+00 4.12190169e-01 9.99315307e-02
-8.69189203e-01 -6.88825011e-01 -1.33127689e+00 1.43513724e-01
1.10801004e-01 5.37958205e-01 2.39156470e-01 6.39784276e-01
-7.49054432e-01 -4.49303985e-02 -1.37978590e+00 -5.75358510e-01
-5.01380622e-01 5.47046602e-01 -4.99608964e-01 3.03832620e-01
-7.02267706e-01 1.22475636e+00 2.75949597e-01 3.87108237e-01
-7.13950932e-01 -3.50487918e-01 -7.86183834e-01 -5.01422167e-01
5.73031723e-01 -1.13241506e+00 1.09774661e+00 -3.62673312e-01
-1.67914128e+00 7.79684782e-01 -1.77394629e-01 -1.73410267e-01
7.72909164e-01 -5.19871891e-01 -1.74132198e-01 6.69113174e-02
2.13639438e-02 -1.64197050e-02 7.33072340e-01 -1.27939224e+00
-5.48942685e-01 -8.67955744e-01 3.31417173e-02 7.65641034e-01
1.42497569e-01 -3.42984378e-01 -6.01255834e-01 -1.57056034e-01
1.02761853e+00 -1.41441262e+00 -6.87160194e-02 5.38066149e-01
-3.83052155e-02 3.69632572e-01 1.04483235e+00 -4.07399237e-01
7.35841691e-01 -2.16706228e+00 2.58986682e-01 -2.13908941e-01
1.69338807e-01 -8.51200074e-02 4.83778477e-01 1.34841502e-01
3.57778937e-01 -6.98416471e-01 6.21028543e-01 -4.07289922e-01
-2.01046467e-01 -3.16939177e-03 -2.33101681e-01 1.42933059e+00
-6.21329069e-01 6.81456923e-01 -7.47028708e-01 -3.13442767e-01
5.68404496e-01 5.70742548e-01 -1.66062444e-01 2.36569464e-01
1.12283260e-01 6.14344299e-01 -6.74311161e-01 8.15212190e-01
8.89773190e-01 -1.88495547e-01 4.70890179e-02 -4.59811002e-01
-5.08538008e-01 -9.15503055e-02 -1.79586101e+00 1.83369792e+00
-3.26971740e-01 3.68980616e-01 5.14229774e-01 -1.55772120e-01
9.71333325e-01 2.55585730e-01 6.31480038e-01 -2.04577073e-01
3.50392252e-01 1.29329905e-01 -2.62025654e-01 -1.92071185e-01
6.34817362e-01 3.94404493e-02 1.61075398e-01 6.75403625e-02
-1.33136377e-01 -2.35506728e-01 -4.24483567e-01 8.83495063e-02
8.61295104e-01 4.04487073e-01 5.98018289e-01 -4.90929671e-02
5.47911823e-01 7.48262228e-03 7.88798094e-01 3.07179093e-01
-2.84543365e-01 3.44775736e-01 -5.41751206e-01 -1.90810427e-01
-1.15233433e+00 -1.15683925e+00 -1.81841955e-01 1.13185078e-01
1.03324163e+00 -3.67574543e-02 -1.93556711e-01 -2.00348794e-01
5.24107933e-01 -1.15693361e-01 -1.96046531e-01 7.47221783e-02
-6.39322042e-01 -2.09335908e-01 -4.79493625e-02 3.68917793e-01
4.85172898e-01 7.60301501e-02 -1.07333124e+00 1.43687889e-01
5.05182929e-02 -1.48878455e+00 -6.27757609e-01 -3.24117005e-01
-9.94879901e-01 -1.06909883e+00 -3.87424886e-01 -5.13183177e-01
7.00180531e-01 1.01880682e+00 3.23203206e-01 -2.67762214e-01
1.64810345e-01 8.86440575e-01 -3.01860785e-03 -2.49864042e-01
-7.25530833e-02 -4.97893900e-01 7.92792618e-01 4.53236289e-02
6.62347004e-02 -2.42409065e-01 -6.16653383e-01 9.01697099e-01
-1.25289559e-01 5.06236739e-02 2.02541098e-01 3.74200612e-01
5.59028924e-01 -3.30555439e-01 -1.01361953e-01 1.31443635e-01
-4.43355978e-01 -6.89238682e-02 -1.55894029e+00 -9.26245749e-03
-3.71994674e-01 -5.29369533e-01 2.02520862e-01 -7.61995971e-01
-7.05109656e-01 7.51437545e-01 5.84112823e-01 -9.81531680e-01
1.73962116e-01 1.02953762e-01 -4.37835872e-01 -5.99332333e-01
5.41866720e-01 2.52608836e-01 1.53649569e-01 -5.29197991e-01
3.88675302e-01 6.16908193e-01 7.02762544e-01 -3.46178338e-02
1.37935793e+00 9.36223209e-01 4.40618157e-01 -9.28468883e-01
-4.25124377e-01 -8.38207603e-01 -9.91830051e-01 -5.44511914e-01
8.08719456e-01 -1.62968171e+00 -1.09367430e+00 4.36807990e-01
-1.29813671e+00 2.74217814e-01 -1.50314108e-01 1.20260823e+00
-6.62370443e-01 6.32459342e-01 -3.82226706e-01 -9.80097890e-01
-9.21181887e-02 -1.09415209e+00 1.29397810e+00 2.79640198e-01
2.74869442e-01 -7.52777994e-01 1.44695759e-01 1.98206589e-01
2.71097124e-02 4.77169603e-01 -3.77451509e-01 3.14423591e-01
-1.21383452e+00 -4.53959882e-01 6.15049563e-02 -2.14298919e-01
1.74175531e-01 -1.12583190e-01 -5.22080064e-01 -8.67749929e-01
7.03368723e-01 3.73769820e-01 6.31963164e-02 4.38396096e-01
-6.21601753e-02 -5.65209575e-02 -7.36163318e-01 9.42987680e-01
1.66608906e+00 5.27279675e-01 -2.49231942e-02 5.93551517e-01
1.09859920e+00 -1.21753000e-01 1.18200505e+00 5.49122870e-01
5.72332919e-01 1.26118124e+00 7.51835108e-01 2.56979227e-01
1.79709475e-02 -9.53130648e-02 6.99804902e-01 1.00075638e+00
2.30166689e-02 2.01509520e-01 -7.44800091e-01 2.30810344e-01
-1.90440452e+00 -4.48214293e-01 -4.44965065e-01 2.49822450e+00
2.84960508e-01 -1.20599508e-01 -1.28024682e-01 -2.34710529e-01
9.03689921e-01 -1.90378189e-01 -8.63398552e-01 5.09986222e-01
7.52573088e-02 -8.67699325e-01 1.07359028e+00 7.43964016e-01
-8.83437872e-01 7.85105944e-01 5.02431965e+00 -2.36260369e-02
-1.25138712e+00 1.46957368e-01 -7.00407028e-01 -1.76231697e-01
3.93462509e-01 3.00360918e-01 -1.47306561e+00 2.70165592e-01
4.68826622e-01 -3.53447050e-01 2.18575567e-01 1.18012309e+00
1.13004267e-01 -9.12666917e-02 -1.00969934e+00 1.21469033e+00
1.37604401e-01 -9.63926971e-01 -5.49086511e-01 3.68528068e-01
5.88719547e-01 2.80857801e-01 -2.20134825e-01 -3.21631223e-01
-9.55812633e-03 1.21049836e-01 1.15573418e+00 6.13975048e-01
5.20854175e-01 -1.98971435e-01 4.82268542e-01 7.75661647e-01
-1.62205040e+00 1.00992189e-03 -6.85894728e-01 -3.80937964e-01
5.33195674e-01 5.42041063e-01 -9.57837939e-01 7.91968465e-01
4.21716928e-01 8.37088585e-01 -4.98840392e-01 1.25714839e+00
1.06197363e-02 -6.46319464e-02 -6.29166782e-01 1.21001408e-01
-2.19472021e-01 -4.26785111e-01 1.20075905e+00 4.84712422e-01
6.48990273e-01 3.06526870e-01 5.00185311e-01 4.64566320e-01
3.32000047e-01 -2.67209083e-01 -7.22215772e-01 5.26776195e-01
5.03053606e-01 1.27627361e+00 -6.29266620e-01 -3.25743645e-01
-4.81280804e-01 7.70170569e-01 -2.71888375e-01 1.71660393e-01
-1.01584816e+00 -2.99665518e-02 7.19931006e-01 1.16711721e-01
4.81969267e-01 -1.04525518e+00 -2.18127787e-01 -1.90388680e+00
3.28009158e-01 -4.79287952e-01 -7.34471828e-02 -1.20687306e+00
-6.27699971e-01 3.44531238e-01 -3.28579664e-01 -2.12031817e+00
-1.42132059e-01 -5.25803030e-01 1.11891359e-01 8.88282955e-01
-9.70254362e-01 -1.16559362e+00 -7.41022289e-01 6.89267874e-01
2.35472292e-01 -3.81051600e-02 1.82273701e-01 2.06459045e-01
-1.07115515e-01 3.14195976e-02 2.21176431e-01 -4.56077047e-02
7.59490252e-01 -9.72017407e-01 2.85792589e-01 1.06108034e+00
3.26303542e-02 7.31242776e-01 9.07652915e-01 -1.01094627e+00
-2.43039227e+00 -9.23605919e-01 4.88259047e-01 -1.00180340e+00
8.63524377e-01 -5.50976813e-01 -4.62079436e-01 1.27813184e+00
-4.07148778e-01 3.22159022e-01 -2.33386815e-01 -4.16220665e-01
4.42042612e-02 -3.10613662e-01 -9.03705835e-01 2.94984937e-01
1.13642180e+00 -4.46021020e-01 -6.12681091e-01 4.12772685e-01
9.63077784e-01 -1.37163806e+00 -9.10577774e-01 4.09858346e-01
8.59682620e-01 -5.06024420e-01 1.14276218e+00 7.28815868e-02
-8.36301923e-01 -1.27740204e+00 -4.41350639e-01 -8.48451734e-01
-3.99472237e-01 -6.23833656e-01 -4.44064975e-01 9.04342294e-01
-4.60738510e-01 -7.79086232e-01 8.58387232e-01 2.70006806e-01
2.36434583e-02 7.21494248e-03 -1.27331614e+00 -1.14352620e+00
-7.83327103e-01 -1.50498316e-01 2.39851758e-01 7.35265493e-01
-2.95946151e-01 4.47992623e-01 -5.53912938e-01 1.30522645e+00
1.13139760e+00 2.61967391e-01 1.45809186e+00 -1.25680017e+00
-3.34992111e-01 3.25584948e-01 -9.68313575e-01 -1.70007765e+00
-2.12997243e-01 -3.60043049e-01 -1.25911430e-01 -1.09887242e+00
-1.97318476e-03 -2.40125582e-01 2.35611901e-01 -2.78464794e-01
2.01821417e-01 1.03990406e-01 4.36681330e-01 5.12276828e-01
-5.58344603e-01 5.78935087e-01 1.25951815e+00 2.13557005e-01
-1.65013149e-01 3.02624732e-01 -3.76752578e-04 8.45866144e-01
7.06626698e-02 -5.05361676e-01 -1.98892489e-01 -6.58872843e-01
1.95711195e-01 6.85402155e-01 6.44282520e-01 -1.16310418e+00
5.84657729e-01 -7.96768442e-03 5.05250096e-01 -1.18925202e+00
5.17597377e-01 -1.50033629e+00 8.38381946e-01 8.32932174e-01
4.89129722e-01 3.20930541e-01 -4.56968881e-02 5.66671431e-01
1.16234727e-01 1.32698305e-02 6.83579147e-01 1.57914653e-01
-4.47506309e-01 4.30981457e-01 7.24817589e-02 -7.50293434e-01
1.18798172e+00 -5.11082590e-01 -2.64666796e-01 -3.00399721e-01
-8.31583142e-01 1.34158298e-01 1.27441859e+00 4.08374012e-01
3.58789623e-01 -1.62575924e+00 -1.73784330e-01 1.96319386e-01
2.76705846e-02 1.33251235e-01 8.88028741e-02 9.72701550e-01
-6.60280168e-01 6.98274493e-01 -1.06365412e-01 -1.36564577e+00
-1.47339797e+00 8.19743872e-01 4.99572873e-01 2.82583326e-01
-6.11981690e-01 7.31139481e-02 2.08539531e-01 -4.93918687e-01
9.27916616e-02 -6.65211737e-01 1.92119807e-01 -2.41440758e-01
4.42034453e-01 5.76617777e-01 -1.03445590e-01 -1.11631787e+00
-5.83880961e-01 1.33586490e+00 3.86204422e-01 -2.84272522e-01
8.59827220e-01 -8.68101656e-01 1.51870191e-01 7.12988615e-01
1.01549983e+00 4.63680848e-02 -1.62851107e+00 -3.40576172e-01
-3.31612349e-01 -7.05926359e-01 8.45833868e-02 -1.50065035e-01
-7.57985771e-01 5.36793649e-01 1.04516423e+00 -7.66674206e-02
7.39715099e-01 -1.32605791e-01 3.35431188e-01 5.94570220e-01
1.04542232e+00 -5.79782784e-01 -2.61936486e-01 5.94118416e-01
8.00261259e-01 -1.21777332e+00 4.68279630e-01 -4.87088710e-01
-1.55233473e-01 1.00120354e+00 5.79438508e-01 -3.21131945e-01
6.42994344e-01 2.16035321e-01 2.44022250e-01 1.80870265e-01
-4.38982636e-01 -6.39848933e-02 2.57811338e-01 4.84566748e-01
-1.11702725e-01 7.60604069e-03 -2.00487748e-01 -1.06372543e-01
3.21369525e-03 -6.69464096e-02 6.42634749e-01 1.08234692e+00
-6.81597352e-01 -5.97421825e-01 -8.97574246e-01 -4.92596775e-01
5.40345795e-02 5.51221490e-01 -3.67877670e-02 9.99038577e-01
-7.39574507e-02 5.76463044e-01 1.19229026e-01 -3.57594252e-01
5.73756278e-01 -3.67401123e-01 7.87539899e-01 -3.97661239e-01
-2.75885135e-01 5.37907124e-01 -3.47411215e-01 -7.71597862e-01
-3.34275186e-01 -1.01876533e+00 -1.22249699e+00 -1.33531049e-01
-9.12531555e-01 -1.02915309e-01 1.19706011e+00 7.35341728e-01
3.09313625e-01 -2.71560730e-05 8.44709814e-01 -1.02153361e+00
-6.77472949e-01 -7.63432920e-01 -7.54797101e-01 -4.98251989e-02
7.45598257e-01 -9.11602557e-01 -3.88125718e-01 1.64561737e-02] | [7.285168170928955, -2.1327545642852783] |
06ff675b-806c-4b31-bdf9-19f2d38a0ce2 | unlocking-temporal-question-answering-for | 2305.15014 | null | https://arxiv.org/abs/2305.15014v1 | https://arxiv.org/pdf/2305.15014v1.pdf | Unlocking Temporal Question Answering for Large Language Models Using Code Execution | Large language models (LLMs) have made significant progress in natural language processing (NLP), and are utilized extensively in various applications. Recent works, such as chain-of-thought (CoT), have shown that intermediate reasoning steps can improve the performance of LLMs for complex reasoning tasks, such as math problems and symbolic question-answering tasks. However, we notice the challenge that LLMs face when it comes to temporal reasoning. Our preliminary experiments show that generating intermediate reasoning steps does not always boost the performance of complex temporal question-answering tasks. Therefore, we propose a novel framework that combines the extraction capability of LLMs and the logical reasoning capability of a Python solver to tackle this issue. Extensive experiments and analysis demonstrate the effectiveness of our framework in handling intricate time-bound reasoning tasks. | ['Lidong Bing', 'Shafiq Joty', 'Hwee Tou Ng', 'Qingyu Tan', 'Liying Cheng', 'Xingxuan Li'] | 2023-05-24 | null | null | null | null | ['logical-reasoning'] | ['reasoning'] | [-2.36201912e-01 2.35067204e-01 -2.09113583e-02 -3.53275120e-01
-4.94394660e-01 -5.99432647e-01 6.91957831e-01 3.91284466e-01
-2.40960270e-01 5.52033305e-01 -2.89373398e-02 -9.57022607e-01
-1.90231353e-01 -1.18191278e+00 -5.74588835e-01 1.47959724e-01
-2.42990181e-01 4.88431007e-01 9.40254509e-01 -3.93799484e-01
1.39782548e-01 3.84847045e-01 -1.41424847e+00 6.96983933e-01
9.44913507e-01 6.45892084e-01 -3.84904236e-01 6.70924604e-01
-7.06818759e-01 1.69938052e+00 -3.94525379e-01 -5.15950620e-01
4.24810350e-02 -2.61033326e-01 -1.50519216e+00 -4.83968019e-01
-3.54785509e-02 -3.00081968e-01 -1.61461860e-01 7.49712050e-01
-1.31493136e-01 3.87179255e-01 4.32685874e-02 -1.72441363e+00
9.40582156e-02 8.91235173e-01 -1.54779851e-01 3.00216138e-01
9.37240422e-01 4.23188865e-01 9.64670241e-01 -2.52916425e-01
6.51327372e-01 1.58548236e+00 4.96146441e-01 3.07629287e-01
-1.08253622e+00 -2.69855738e-01 5.06226659e-01 8.21722150e-01
-1.34334314e+00 -1.44502789e-01 4.02698845e-01 -1.77542835e-01
1.62384677e+00 4.04740900e-01 4.38931584e-01 6.04345083e-01
4.95834082e-01 9.26358819e-01 1.19704080e+00 -5.22883236e-01
4.95021492e-01 -1.21453747e-01 5.23706615e-01 8.26875567e-01
-3.47461760e-01 -2.73333549e-01 -5.52805722e-01 -4.22910303e-01
4.44137901e-01 -3.73597473e-01 3.70004535e-01 6.33616894e-02
-1.06475091e+00 8.11800003e-01 5.24674878e-02 5.07621348e-01
-1.55296857e-02 3.81795555e-01 5.36102772e-01 4.17431116e-01
-6.46475377e-03 5.06138980e-01 -4.25283372e-01 -2.86030293e-01
-8.80248010e-01 6.52407169e-01 1.34452331e+00 7.83899248e-01
3.35920751e-01 -4.71199870e-01 -3.29694480e-01 1.97082520e-01
2.15397164e-01 1.67971134e-01 8.85170624e-02 -1.42918694e+00
4.94532585e-01 9.92938936e-01 1.75733209e-01 -9.75800335e-01
-7.23986149e-01 8.24750066e-02 -3.19567770e-01 -1.28168151e-01
6.31495357e-01 1.06939338e-01 -3.24006021e-01 1.71942782e+00
6.11907184e-01 2.61427820e-01 1.13740452e-01 7.06860542e-01
3.98032188e-01 9.08337653e-01 2.86197543e-01 -4.80954111e-01
1.48974419e+00 -1.01032770e+00 -7.91937530e-01 -4.51444834e-01
6.64331198e-01 -4.02828932e-01 1.01066566e+00 4.68506813e-01
-1.46988523e+00 -7.33134076e-02 -7.92972147e-01 -2.88367987e-01
-1.37942731e-01 -5.63148022e-01 1.35110915e+00 7.12558627e-02
-8.36181343e-01 4.52937931e-01 -1.10734713e+00 -2.71104306e-01
-2.48041749e-02 2.73575544e-01 9.44279060e-02 -2.33764932e-01
-1.43691742e+00 1.02915156e+00 4.75711495e-01 1.31984800e-01
-5.43460906e-01 -6.95921600e-01 -8.14737022e-01 1.44029334e-01
1.06876373e+00 -6.92671657e-01 1.77677524e+00 -4.45599318e-01
-1.59130692e+00 5.24403274e-01 -4.02931690e-01 -7.87949622e-01
7.76409924e-01 -1.97524637e-01 -4.41106826e-01 4.08622652e-01
1.24620989e-01 3.94280940e-01 4.08282280e-01 -3.84709388e-01
-3.35230052e-01 -1.87953666e-01 8.78417790e-01 -2.81056881e-01
1.56677961e-01 5.71644664e-01 -3.06311399e-01 -2.07142867e-02
2.35358134e-01 -9.17156577e-01 -6.52935982e-01 -6.19245209e-02
-3.07868235e-02 -6.82965994e-01 3.62488687e-01 -5.06449521e-01
1.43850338e+00 -1.97846472e+00 1.57738820e-01 1.57330826e-01
-1.29343003e-01 1.12691924e-01 -2.06053760e-02 7.22117305e-01
-1.70792211e-02 5.93493879e-02 -1.46757498e-01 3.09572756e-01
3.88250798e-01 5.57707965e-01 -7.79188991e-01 -2.90917344e-02
4.58358675e-01 1.30889511e+00 -9.55799460e-01 -1.01434135e+00
1.00235224e-01 -4.08380121e-01 -7.31817722e-01 2.80284792e-01
-1.36598945e+00 1.29233271e-01 -4.73029524e-01 4.02620435e-01
4.31194633e-01 -4.50318754e-01 4.95322704e-01 5.67051172e-01
-5.02638109e-02 7.38286376e-01 -1.44319439e+00 1.72183120e+00
-6.24487579e-01 3.09226692e-01 -1.67992562e-01 -5.60522318e-01
3.46757174e-01 3.64562899e-01 8.36747214e-02 -9.28658485e-01
-1.55879274e-01 -4.34223115e-02 1.94558516e-01 -8.60269785e-01
2.64624566e-01 -3.89617294e-01 -2.78087109e-01 8.74406517e-01
-3.17038834e-01 -5.08667052e-01 7.22821832e-01 4.70566541e-01
1.40913606e+00 3.78791183e-01 3.98449749e-01 -9.90603771e-03
9.82508898e-01 5.97183883e-01 6.78730071e-01 6.91709161e-01
-1.87120512e-01 -1.55592769e-01 1.14326835e+00 -7.49163330e-01
-6.52359307e-01 -7.60411561e-01 1.89399585e-01 1.10669017e+00
-5.39116887e-03 -9.62207019e-01 -6.22126758e-01 -5.29356658e-01
-2.40446955e-01 1.18693805e+00 8.77921209e-02 -1.65889353e-01
-9.02742743e-01 -4.16322976e-01 1.01521873e+00 6.12392962e-01
4.59240943e-01 -1.02299607e+00 -1.00421691e+00 4.45310563e-01
-5.70016205e-01 -1.55559063e+00 2.52414882e-01 -1.32772475e-01
-1.02969587e+00 -1.16842175e+00 2.99092114e-01 -3.29064488e-01
3.72355759e-01 3.68865430e-02 1.17704749e+00 2.27626666e-01
-2.93509603e-01 4.46087301e-01 -2.92025685e-01 -2.62757510e-01
-5.62661350e-01 -1.59380615e-01 -2.42727533e-01 -5.22697866e-01
3.82589102e-01 -7.08510637e-01 1.46345431e-02 3.19830120e-01
-1.01173973e+00 3.52751404e-01 2.75551677e-02 3.18443686e-01
2.19826221e-01 6.49031937e-01 7.88287669e-02 -6.13834620e-01
8.96216214e-01 -1.76312134e-01 -1.05907738e+00 5.64657032e-01
-3.43055218e-01 5.39202571e-01 8.63366425e-01 -4.09745455e-01
-1.22536087e+00 -2.86509722e-01 -2.23217010e-02 -2.80885231e-02
-9.65796560e-02 8.59756649e-01 2.50886649e-01 1.89934433e-01
5.39480329e-01 1.07250258e-01 -9.98952985e-02 8.33163708e-02
2.66414315e-01 -4.69098575e-02 4.42608029e-01 -1.48922300e+00
8.60875070e-01 4.10897523e-01 3.85694623e-01 -2.43439764e-01
-9.80133593e-01 -2.20194459e-01 -2.67730445e-01 -5.36263473e-02
5.45043766e-01 -7.25665033e-01 -1.41436493e+00 1.11616962e-01
-1.43130350e+00 -5.95398724e-01 -1.79330885e-01 1.28824160e-01
-6.49964929e-01 4.76868510e-01 -8.11113358e-01 -1.23713708e+00
-2.68777311e-01 -1.00164568e+00 8.07777226e-01 1.92582592e-01
-6.78725541e-01 -8.69054139e-01 4.26176041e-02 5.79162776e-01
2.13846087e-01 1.56897470e-01 1.44467664e+00 -5.55397689e-01
-8.46334755e-01 4.65550311e-02 -6.00986667e-02 -9.02494490e-02
-6.81494474e-01 -3.10620908e-02 -7.11962461e-01 2.11654305e-01
4.54974398e-02 -3.47677350e-01 1.30874649e-01 -1.31314144e-01
1.10170138e+00 -4.39512312e-01 -6.65065423e-02 -2.86766104e-02
1.19719863e+00 5.84769547e-02 8.15611303e-01 4.80703861e-01
-1.40517011e-01 6.06371641e-01 9.50334728e-01 3.51302356e-01
6.41088426e-01 4.69227731e-01 1.04561724e-01 5.73804855e-01
4.08684760e-01 -2.56970733e-01 4.43688005e-01 4.52700287e-01
-8.18737969e-02 2.97931999e-01 -1.47320938e+00 4.78870660e-01
-2.43985295e+00 -1.11529922e+00 -3.22196424e-01 1.52973485e+00
1.18026400e+00 2.65239209e-01 -6.35983720e-02 2.94750154e-01
6.99036047e-02 -8.68048239e-03 -3.72054994e-01 -8.55273962e-01
4.20016348e-01 5.20499885e-01 -2.11904213e-01 6.27166450e-01
-6.77216411e-01 1.23236203e+00 6.51021910e+00 6.21055722e-01
-8.01264822e-01 -9.20441225e-02 -4.92172949e-02 -6.20018877e-02
-2.45773077e-01 4.80565667e-01 -4.88435447e-01 -1.20902061e-01
1.16344690e+00 -5.46418488e-01 7.18989611e-01 9.20986593e-01
2.09235623e-01 -5.56122839e-01 -1.42412472e+00 5.55710196e-01
-2.84772098e-01 -1.20518816e+00 -1.43645823e-01 -5.13340056e-01
2.91547000e-01 -5.15014648e-01 -4.52088684e-01 6.81374729e-01
4.20879424e-01 -8.70474637e-01 7.61668384e-01 4.89614785e-01
-4.38254215e-02 -6.67261779e-01 7.75907755e-01 9.11844730e-01
-1.04699123e+00 -2.83602476e-01 -7.88439289e-02 -8.24364722e-01
3.38291675e-01 5.54669201e-01 -1.11226153e+00 6.99306548e-01
6.23565793e-01 7.52891526e-02 -5.78679621e-01 8.01057220e-01
-7.33155072e-01 3.32038939e-01 -5.51251709e-01 -2.02810615e-01
3.56517613e-01 -1.37783274e-01 2.93847084e-01 1.00151229e+00
-2.34657377e-01 6.08276129e-01 2.33180553e-01 1.20503819e+00
4.73890930e-01 -1.59726053e-01 -1.74165905e-01 -4.69173700e-01
7.45577887e-02 9.50230181e-01 -7.77291775e-01 -4.29153830e-01
-3.30264211e-01 4.59075749e-01 3.61760736e-01 2.31910110e-01
-1.23248065e+00 -9.05052423e-02 4.84285742e-01 1.61052376e-01
-1.40450135e-01 -8.12088788e-01 -2.66155303e-01 -1.29657435e+00
4.31715220e-01 -1.15083337e+00 7.99909592e-01 -1.11732972e+00
-7.86495984e-01 2.79352456e-01 5.30847073e-01 -5.18598676e-01
-6.10037744e-01 -5.17151654e-01 -6.27769172e-01 5.06483912e-01
-1.38119161e+00 -8.22548568e-01 -7.36010745e-02 5.75396061e-01
4.02630836e-01 5.08230388e-01 7.57054925e-01 8.29326734e-02
-4.16388482e-01 1.77341625e-02 -9.49782014e-01 -2.82757226e-02
2.64744639e-01 -1.00496602e+00 2.73317963e-01 1.16601837e+00
-9.61086750e-02 1.07699788e+00 7.70605326e-01 -3.55383694e-01
-2.07249022e+00 -8.31280947e-01 1.28889263e+00 -3.48070741e-01
1.20733190e+00 -1.77091062e-01 -8.82034779e-01 1.04133272e+00
2.23068893e-02 -1.64887935e-01 3.73631060e-01 2.11816296e-01
-5.52892804e-01 2.48633288e-02 -8.76537263e-01 9.03728426e-01
1.03355408e+00 -8.37143242e-01 -1.29288995e+00 6.73835278e-01
1.05293667e+00 -7.35899687e-01 -9.55696285e-01 2.79724956e-01
2.59794742e-01 -9.02801871e-01 1.01141286e+00 -1.04074669e+00
7.00750172e-01 -5.97938836e-01 1.18214592e-01 -5.11294842e-01
-2.30290741e-02 -7.07204938e-01 -4.20877635e-01 1.09144783e+00
1.95572138e-01 -6.82099700e-01 3.38339359e-01 1.39778066e+00
3.94826233e-01 -5.68687081e-01 -8.81300390e-01 -7.88031518e-01
-3.39320861e-03 -1.20013595e+00 5.60099661e-01 5.71210861e-01
7.47590303e-01 4.47685510e-01 2.43419975e-01 3.44202489e-01
2.42171437e-01 8.34505916e-01 5.98729193e-01 -1.02933419e+00
-5.23529172e-01 -2.70758748e-01 2.85423119e-02 -6.09970450e-01
6.28005981e-01 -9.78202820e-01 -9.83234569e-02 -1.68643975e+00
-1.73207652e-02 -2.56253988e-01 1.76277280e-01 9.69341516e-01
4.76824753e-02 -5.12865543e-01 3.16635758e-01 -1.78223416e-01
-1.20890677e+00 6.52218685e-02 1.20430827e+00 -1.92585468e-01
-6.29699603e-02 -3.17885261e-03 -3.37043583e-01 8.61334324e-01
4.77586150e-01 -4.18883383e-01 -4.23204482e-01 -3.99679244e-01
1.18801570e+00 7.52264798e-01 4.68557388e-01 -9.93475914e-01
7.69503653e-01 -9.41632032e-01 -3.93702716e-01 -4.30996329e-01
8.00078437e-02 -8.41026008e-01 1.73911184e-01 7.97640979e-01
-4.21665579e-01 1.24879554e-01 6.18356943e-01 3.67477648e-02
-4.54787970e-01 -3.54299098e-01 3.91838253e-01 -4.29507881e-01
-8.97227526e-01 -1.49193347e-01 -6.07364416e-01 1.94102436e-01
1.26244140e+00 4.74302977e-01 -3.02131891e-01 -3.32181394e-01
-6.67154968e-01 8.30104113e-01 5.17667420e-02 3.55551422e-01
3.74306262e-01 -8.02017391e-01 -2.40562573e-01 -2.79955685e-01
3.56405452e-02 1.85432777e-01 5.17881699e-02 1.20958745e+00
-7.42958307e-01 7.43927658e-01 -9.75400768e-03 -4.38819408e-01
-1.20163751e+00 1.02698851e+00 3.00567180e-01 -8.27741623e-01
-5.94272673e-01 5.64192832e-01 -2.71872073e-01 -4.73692805e-01
1.13089263e-01 -9.00578260e-01 6.70517311e-02 -1.58952504e-01
8.83313298e-01 3.85779560e-01 9.57733914e-02 4.07591403e-01
-8.48488033e-01 2.47558177e-01 9.36651155e-02 -3.07992637e-01
1.45383203e+00 3.09705138e-01 -9.15382087e-01 4.72307652e-01
3.80569130e-01 -2.74810731e-01 -4.31871444e-01 -3.79123360e-01
5.03384650e-01 -7.26438239e-02 -3.77569437e-01 -7.55255401e-01
-1.90758809e-01 7.52705872e-01 -3.68889123e-01 4.58007902e-01
9.99451995e-01 -6.76517114e-02 7.95175791e-01 9.89974320e-01
8.84513557e-01 -1.07526398e+00 -8.23958740e-02 1.04661405e+00
7.40508020e-01 -8.60805154e-01 1.46831512e-01 -5.67343414e-01
-3.22406411e-01 1.25510550e+00 8.07216406e-01 1.44021779e-01
2.51336601e-02 6.42296195e-01 -3.01327258e-01 -4.71043698e-02
-1.46992385e+00 -8.54868963e-02 -3.10821533e-01 -7.56066144e-02
1.48355663e-01 -1.42379537e-01 -5.48539937e-01 6.93015695e-01
-1.80990756e-01 6.60980403e-01 4.91465807e-01 1.44095659e+00
-2.07104206e-01 -1.22355473e+00 -8.75083089e-01 -2.23812237e-01
-3.77897322e-01 -1.01746492e-01 -3.27087283e-01 7.73389280e-01
-1.91822752e-01 1.10246897e+00 -4.58033718e-02 1.18514121e-01
2.45825782e-01 5.99018991e-01 8.47651064e-01 -4.98718530e-01
-8.02175820e-01 -5.41993856e-01 4.96461481e-01 -1.15573144e+00
-2.49155357e-01 -3.72828215e-01 -1.76445210e+00 -6.02892280e-01
1.86664179e-01 4.79170144e-01 3.72601032e-01 1.41100121e+00
1.05653621e-01 6.37133002e-01 -4.98590656e-02 -1.43363670e-01
-7.02894628e-01 -3.33116591e-01 2.06445213e-02 8.25185105e-02
-7.10574165e-02 -2.40550905e-01 -1.62586808e-01 2.95184813e-02] | [9.244367599487305, 7.314091682434082] |
93201fef-23ea-4bc5-b5c6-de23396c8341 | performance-analysis-of-empirical-open-2 | 2306.16542 | null | https://arxiv.org/abs/2306.16542v1 | https://arxiv.org/pdf/2306.16542v1.pdf | Performance Analysis of Empirical Open-Circuit Voltage Modeling in Lithium Ion Batteries, Part-1: Performance Measures | The open circuit voltage to the state of charge (OCVSOC) characteristic is crucial for battery management systems. Using the OCV-SOC curve, the SOC and the battery capacity can be estimated in real-time. Accurate SOC and capacity information are important to carry out the majority of battery management functionalities that ensure a safe, efficient, and reliable battery pack power system. Numerous approaches have been reported in the literature for improved SOC estimation and battery capacity estimation. These approaches focus on various estimation and filtering techniques to reduce the effect of measurement noise and uncertainties due to hysteresis and relaxation effects. Even though all the existing approaches to SOC estimation rely on the OCV-SOC characterization, little attention was paid to investigating the possibility of errors in the OCV-SOC characterization and the effect of uncertainty in the OCV-SOC curve on SOC and capacity estimates. In this paper, which is the first part of a series of three papers, the effect of OCV-SOC modeling error in the overall battery management system is discussed. The different sources of uncertainties in the OCV-SOC curve include cell-to-cell variation, temperature variation, aging drift, cycle rate effect, curve-fitting error, and measurement/estimation error. The proposed uncertainty models can be incorporated into battery management systems to improve their safety, performance, and reliability. | ['Balakumar Balasingam', 'James Nguyen', 'Prarthana Pillai'] | 2023-06-28 | null | null | null | null | ['management'] | ['miscellaneous'] | [-3.52977008e-01 -9.68690574e-01 -1.91739723e-01 -2.90386289e-01
-3.87079865e-01 -5.27012169e-01 3.07653487e-01 7.32257009e-01
-1.91502720e-01 1.38410306e+00 -3.07903856e-01 -3.05593848e-01
-5.58595546e-02 -7.47782171e-01 -6.50769532e-01 -7.70473301e-01
1.65627062e-01 1.07487716e-01 5.67152262e-01 -1.20931432e-01
4.18017298e-01 6.69136107e-01 -1.82191014e+00 -3.23895752e-01
1.17690480e+00 1.36835599e+00 7.57913366e-02 3.96395147e-01
-4.42495532e-02 1.76210001e-01 -6.58992827e-01 1.00475937e-01
-3.42919439e-01 -1.15794636e-01 2.12653756e-01 -6.52495444e-01
-4.30487752e-01 -2.42685422e-01 -1.42076969e-01 1.05181301e+00
5.02488136e-01 -1.86128870e-01 7.01503158e-01 -1.69703507e+00
1.86627463e-01 3.62356514e-01 9.29383337e-02 3.27062458e-01
3.32863510e-01 -6.18419163e-02 4.12621908e-02 -8.89672399e-01
2.27067508e-02 7.86182165e-01 5.04712403e-01 1.42763123e-01
-1.16886497e+00 -1.04082048e+00 -2.41706088e-01 3.90340924e-01
-1.83338904e+00 -3.90445709e-01 5.36672056e-01 -3.13039005e-01
1.04765964e+00 5.36389768e-01 1.02208781e+00 2.55836636e-01
1.17565751e+00 2.15176210e-01 1.43187594e+00 -3.50485444e-01
6.10343874e-01 5.83370030e-01 5.02486348e-01 -3.43270808e-01
1.03947771e+00 -3.35886590e-02 -3.17735702e-01 -2.55604118e-01
8.27260092e-02 -6.81252277e-04 -4.52610821e-01 -3.28814328e-01
-2.64847398e-01 2.55792350e-01 1.49809018e-01 1.79190055e-01
1.45210490e-01 3.61094713e-01 5.93618035e-01 -1.89411253e-01
-1.87304795e-01 5.35521023e-02 -2.48866826e-01 -2.78852493e-01
-9.44207609e-01 3.70366782e-01 8.78441334e-01 1.18928885e+00
4.41292107e-01 4.33676302e-01 -1.20325439e-01 2.16251582e-01
6.27520323e-01 1.00497079e+00 3.94030482e-01 -4.71509784e-01
6.18900023e-02 6.23327076e-01 6.36375248e-01 -2.53291249e-01
-2.46723384e-01 -1.09573990e-01 -4.55907881e-01 -4.42354269e-02
-9.23025329e-03 1.37620822e-01 -9.67455804e-01 8.97866070e-01
-8.67305994e-02 -3.01742673e-01 2.50460040e-02 4.54160273e-01
6.63310707e-01 5.61378360e-01 1.18610702e-01 -5.84813476e-01
1.09167123e+00 -1.64254263e-01 -1.38285220e+00 -7.22887814e-02
1.12369299e-01 -4.53231245e-01 5.94049692e-01 9.52450279e-03
-1.22767115e+00 -2.76796967e-01 -1.76209807e+00 1.39885634e-01
-7.07510471e-01 -8.85847658e-02 5.98491691e-02 9.20793295e-01
-6.26535416e-01 5.38361609e-01 -9.19094503e-01 -2.26238549e-01
-1.00648105e-01 6.01197779e-01 2.20925972e-01 5.46317101e-02
-1.46250677e+00 1.38804746e+00 1.97061911e-01 4.49740052e-01
-7.71714091e-01 -7.00723886e-01 -7.65662909e-01 -9.25157368e-02
2.38046162e-02 5.96133173e-02 1.01993454e+00 -2.04328492e-01
-1.17939293e+00 -8.99700895e-02 -6.47969663e-01 -5.18575728e-01
2.56518573e-01 -1.09724090e-01 -7.40016222e-01 -5.05358040e-01
-3.66103977e-01 -5.69865145e-02 3.25520635e-01 -1.25022352e+00
-2.59292156e-01 -3.57981354e-01 -6.25936747e-01 -2.71815300e-01
8.29870626e-02 -2.19688535e-01 -9.02619287e-02 -9.28992778e-02
2.41797213e-02 -1.00778401e+00 4.58935313e-02 -3.54062259e-01
8.56257677e-02 2.01258250e-02 8.92689109e-01 -6.19502425e-01
1.56874442e+00 -1.80308747e+00 -4.30372506e-01 4.42954719e-01
-6.78734183e-01 2.34721467e-01 6.71497881e-01 6.10252798e-01
3.44817609e-01 9.70638096e-02 -1.53922155e-01 1.52793497e-01
-9.11873206e-02 3.13249052e-01 -2.48691946e-01 7.72561610e-01
3.95071283e-02 6.94016099e-01 -5.79326272e-01 1.00049980e-01
2.99701750e-01 4.72664535e-01 2.33949974e-01 7.48757944e-02
4.17136885e-02 1.68050542e-01 6.33213203e-03 9.60820854e-01
9.41324770e-01 5.98431468e-01 2.76600689e-01 -4.68023539e-01
-4.90824074e-01 4.84741211e-01 -1.15285778e+00 6.72949672e-01
-2.70578414e-01 5.82323194e-01 5.78483455e-02 -4.47645277e-01
1.15479338e+00 1.12753436e-01 6.44618198e-02 -9.93585885e-01
4.84036803e-01 9.43872750e-01 1.34516343e-01 -1.68375835e-01
9.27218318e-01 -2.14108214e-01 -9.43883136e-02 -3.09620649e-01
-2.36368567e-01 -3.66905630e-01 1.72046512e-01 -6.96676597e-02
2.68999845e-01 -6.11651763e-02 2.59143077e-02 -9.21350777e-01
8.88775587e-01 -2.31259316e-01 6.66871071e-01 4.49732579e-02
-3.43551636e-01 -1.23373708e-02 5.64057887e-01 3.33119243e-01
-1.14857471e+00 -7.35228062e-01 -6.91994905e-01 9.31441262e-02
7.11343467e-01 -2.12387517e-01 -4.08728063e-01 1.68264091e-01
6.70401394e-01 1.11164784e+00 -2.51909077e-01 -5.30845106e-01
-3.12710583e-01 -7.44590104e-01 3.00629616e-01 1.10510373e+00
1.04170211e-01 -2.08593562e-01 -4.57662821e-01 3.23534846e-01
2.91160882e-01 -1.03980827e+00 -1.66593820e-01 5.81187904e-01
-1.16352117e+00 -7.82599032e-01 -2.21206024e-01 -1.78253666e-01
7.64889657e-01 5.90675659e-02 7.22853661e-01 1.27915874e-01
-9.34896842e-02 1.74671695e-01 1.37504771e-01 -9.95411694e-01
-4.31579977e-01 -3.89792800e-01 3.82348150e-01 -4.41104650e-01
4.15069252e-01 -1.65606197e-02 -6.64348125e-01 7.07749844e-01
-5.66971362e-01 -6.81161046e-01 2.83744950e-02 3.13898951e-01
8.47193062e-01 3.71096790e-01 1.01224506e+00 -2.29222327e-01
4.65337783e-01 -4.08378333e-01 -9.94064033e-01 7.09923580e-02
-1.34004676e+00 8.23448226e-02 4.00484204e-01 -2.16569722e-01
-8.83524776e-01 -9.77742299e-02 -1.12018742e-01 -2.71023899e-01
2.93372780e-01 3.37998003e-01 -7.32192039e-01 -2.42650539e-01
-1.44488797e-01 3.16455334e-01 3.22407752e-01 -1.45704955e-01
-5.10814428e-01 9.37935472e-01 5.91835439e-01 -2.82144994e-01
4.02574748e-01 2.69968182e-01 3.50681126e-01 -5.86460412e-01
-4.51348759e-02 -4.26759064e-01 -3.39569986e-01 -4.34794784e-01
3.97831142e-01 -1.32500398e+00 -8.90239835e-01 8.14049661e-01
-5.34690976e-01 -2.88800955e-01 2.72849828e-01 3.66846710e-01
1.46481052e-01 2.18171149e-01 -2.80297250e-01 -1.47140563e+00
-6.63320661e-01 -1.46995759e+00 4.53213781e-01 6.48579597e-01
-1.16574883e-01 -9.69863594e-01 -4.85662878e-01 3.55936512e-02
7.01545715e-01 1.66380867e-01 6.73940122e-01 -1.27824783e-01
-4.42277610e-01 -7.89691031e-01 6.34267107e-02 5.30840456e-01
-1.28190130e-01 2.33968511e-01 -7.82321155e-01 -8.53751421e-01
6.81123957e-02 1.68421164e-01 4.31294501e-01 4.10923362e-01
8.96152437e-01 3.55032116e-01 -5.68487883e-01 -5.87238232e-04
2.00283337e+00 8.52463067e-01 1.07332289e+00 2.26900399e-01
2.27688208e-01 2.83218399e-02 1.05449450e+00 4.02402133e-01
4.38651264e-01 4.63478625e-01 6.52195275e-01 5.82974374e-01
4.12171483e-01 -9.86677110e-02 7.10653484e-01 7.27282763e-01
2.63095438e-01 -4.65779513e-01 -6.90446556e-01 5.60122132e-01
-1.41962516e+00 -4.22669023e-01 -9.84969258e-01 2.94120693e+00
4.24722373e-01 5.38010597e-01 -5.26006781e-02 5.86951792e-01
7.09270120e-01 -4.23445851e-01 -6.24020100e-01 -7.81745791e-01
-8.56961757e-02 -5.94149902e-02 1.29569542e+00 4.06032056e-01
-3.87795895e-01 -8.34949315e-02 6.59665442e+00 4.88956720e-01
-1.36916828e+00 -8.54362249e-02 1.80517852e-01 -1.80497229e-01
-6.14601433e-01 3.13406408e-01 -1.37095582e+00 9.41527903e-01
1.47716916e+00 -2.77101189e-01 8.54444802e-02 3.92621368e-01
4.42378551e-01 -1.19326818e+00 -1.01146150e+00 7.42607653e-01
2.17705150e-03 -9.81436372e-01 -3.82328540e-01 7.17392415e-02
5.07978320e-01 -3.76119852e-01 -4.17930692e-01 4.43110436e-01
-6.99901819e-01 -7.47281790e-01 9.87337351e-01 8.92497897e-01
1.06146884e+00 -9.82981503e-01 1.47737777e+00 4.95521240e-02
-1.43014646e+00 -3.41518730e-01 -1.79004773e-01 -1.99347734e-01
3.91447902e-01 7.14505851e-01 -4.78512406e-01 5.26038527e-01
6.93912804e-01 1.05706982e-01 -5.86392522e-01 1.18692267e+00
3.92481945e-02 4.29732710e-01 -3.86205733e-01 -6.73508883e-01
-3.70425522e-01 -1.78531289e-01 1.50451049e-01 7.80953407e-01
5.31859994e-01 -1.87964849e-02 -4.53317761e-01 8.20829809e-01
1.80392444e-01 -3.50782603e-01 -1.25990152e-01 -1.19901396e-01
1.04632616e+00 9.22803104e-01 -5.49660385e-01 -2.55978972e-01
-2.85895497e-01 2.54155904e-01 -6.26619995e-01 2.21553370e-01
-9.31403220e-01 -6.95874572e-01 5.85172296e-01 8.32439303e-01
6.12101294e-02 -2.96398163e-01 -7.27408886e-01 -5.26411593e-01
1.41592950e-01 -2.15597838e-01 -5.04740961e-02 -7.55857587e-01
-6.79048717e-01 -1.21864974e-01 4.60446328e-01 -9.61507678e-01
2.43606716e-01 -6.32112682e-01 -9.09645319e-01 1.08473468e+00
-1.45379758e+00 -4.52439725e-01 -5.55996418e-01 4.92935255e-02
2.78373867e-01 1.72904536e-01 3.84756476e-01 5.02698839e-01
-6.70984626e-01 5.43605387e-01 8.22341084e-01 -6.77536666e-01
6.41962469e-01 -1.10226583e+00 -2.32753888e-01 7.08844543e-01
-1.37549305e+00 3.97300720e-01 1.13048744e+00 -1.04966998e+00
-2.19426847e+00 -8.31422508e-01 6.34543955e-01 -2.10928842e-01
6.68201745e-01 -5.84640205e-01 -1.19386756e+00 1.71209857e-01
-1.76039971e-02 -4.41913456e-02 3.18551272e-01 -6.10151052e-01
5.50176322e-01 -8.19835067e-01 -1.15251756e+00 2.57863492e-01
3.41528952e-02 -5.63731015e-01 -3.20936367e-02 -3.68453771e-01
1.63549766e-01 -6.28482521e-01 -1.38000369e+00 7.43263066e-01
9.56304491e-01 -4.45025295e-01 5.62637627e-01 3.41400325e-01
-5.96121907e-01 -7.51821458e-01 -2.23566204e-01 -8.22051823e-01
1.16663344e-01 -7.69045204e-02 -2.75611371e-01 1.65946054e+00
3.29834163e-01 -7.99049258e-01 2.61150539e-01 1.18234622e+00
-2.36997545e-01 -8.18368733e-01 -1.07284462e+00 -9.09785271e-01
-6.54065842e-03 -5.36633015e-01 7.38972604e-01 -1.28681570e-01
1.10592745e-01 -1.26734063e-01 -4.03518742e-03 2.52628177e-01
4.14694548e-01 -3.46662313e-01 5.21311820e-01 -1.15447640e+00
5.25555432e-01 1.07562892e-01 -4.76354718e-01 -8.82254690e-02
-1.46557748e-01 -2.68485546e-01 2.84462303e-01 -1.70651305e+00
2.55768239e-01 -4.93158668e-01 -5.35849094e-01 -3.18632722e-01
-2.01101854e-01 7.81655982e-02 1.99066311e-01 1.96702525e-01
-1.86447963e-01 6.09570444e-01 7.08077610e-01 -8.25851932e-02
-2.52605170e-01 9.86239314e-02 5.35663292e-02 3.02553087e-01
6.61513805e-01 -4.11010921e-01 -3.43243748e-01 2.31055737e-01
3.58249366e-01 1.06997199e-01 4.14683014e-01 -1.29950690e+00
2.88711220e-01 -5.87586500e-03 5.89419544e-01 -1.24925649e+00
2.50252932e-01 -1.10802400e+00 9.67585802e-01 7.12304771e-01
5.91266632e-01 8.43614936e-02 7.18098640e-01 6.63059533e-01
-9.09877047e-02 -4.75969404e-01 8.57363343e-01 3.54860127e-01
-5.30923367e-01 -2.37030089e-01 -9.42631483e-01 -6.07710361e-01
1.32394719e+00 -5.24881303e-01 -5.23794353e-01 -2.24605426e-01
-4.01590049e-01 6.06805205e-01 8.67189407e-01 2.97104418e-01
5.30573308e-01 -1.35943449e+00 2.29204837e-02 2.06076413e-01
1.87600508e-01 -8.63924697e-02 2.79348314e-01 1.16101551e+00
-4.09939528e-01 7.48852849e-01 -1.12801656e-01 -5.48429847e-01
-1.24957860e+00 4.74661589e-01 6.27001464e-01 1.73110947e-01
3.15777987e-01 2.24036574e-01 -7.90629864e-01 4.64963734e-01
1.53503701e-01 -4.85146731e-01 -5.62941097e-03 2.43584961e-01
2.93467462e-01 9.84570146e-01 4.85396475e-01 -6.52007759e-01
-8.15009236e-01 4.04151022e-01 4.01606232e-01 -3.71366590e-02
5.58198214e-01 -4.33334261e-01 -4.33454841e-01 1.11661243e+00
9.93026793e-01 -5.31283803e-02 -9.09599900e-01 4.65432197e-01
1.09914385e-01 -2.94666737e-01 2.74609238e-01 -7.25733697e-01
-5.76163769e-01 5.46015143e-01 9.49556410e-01 7.50817657e-02
1.04540062e+00 -5.20805418e-01 7.80703127e-01 -2.68357486e-01
5.01046717e-01 -1.53611648e+00 -3.86309087e-01 3.79348963e-01
9.24925506e-01 -8.46636117e-01 5.67246795e-01 -2.04007119e-01
-5.25579870e-01 1.12298942e+00 7.43737638e-01 7.95665905e-02
9.31977928e-01 7.66337097e-01 -3.95781398e-01 4.50124592e-01
-5.26371837e-01 2.68174350e-01 1.64581046e-01 2.41731718e-01
4.03091371e-01 8.87014270e-02 -9.28783000e-01 8.90472889e-01
4.08053190e-01 1.58243835e-01 7.43219852e-01 1.28445542e+00
-5.93000650e-01 -9.65712309e-01 -7.13753879e-01 6.41116977e-01
-3.10285717e-01 4.75702167e-01 7.82820135e-02 6.95100904e-01
1.38176873e-01 1.04272616e+00 3.42537075e-01 -2.65939742e-01
7.88930476e-01 2.85737604e-01 2.29013458e-01 9.62691754e-02
-3.28703374e-01 7.79837072e-02 3.54794919e-01 -2.17651621e-01
-1.99005872e-01 -6.82375073e-01 -1.75181401e+00 -3.32079411e-01
-1.01662242e+00 4.01937842e-01 1.44649422e+00 8.00796151e-01
1.56599790e-01 8.35558653e-01 5.02944112e-01 -8.00628841e-01
-2.89743841e-01 -9.36909735e-01 -9.49278951e-01 -1.23443633e-01
1.37314364e-01 -9.51000094e-01 -5.58054507e-01 -6.24569058e-01] | [6.30073881149292, 2.7510159015655518] |
b0cecc5b-3d35-4c05-98d7-b1b5fbfdc62b | from-inscription-to-semi-automatic-annotation | null | null | https://aclanthology.org/2022.lt4hala-1.16 | https://aclanthology.org/2022.lt4hala-1.16.pdf | From Inscription to Semi-automatic Annotation of Maya Hieroglyphic Texts | The Maya script is the only readable autochthonous writing system of the Americas and consists of more than 1000 word signs and syllables. It is only partially deciphered and is the subject of the project “Text Database and Dictionary of the Classic Maya” . Texts are recorded in TEI XML and on the basis of a digital sign and graph catalog, which are stored in the TextGrid virtual repository. Due to the state of decipherment, it is not possible to record hieroglyphic texts directly in phonemically transliterated values. The texts are therefore documented numerically using numeric sign codes based on Eric Thompson’s catalog of the Maya script. The workflow for converting numerical transliteration into textual form involves several steps, with variable solutions possible at each step. For this purpose, the authors have developed ALMAH “Annotator for the Linguistic Analysis of Maya Hieroglyphs”. The tool is a client application and allows semi-automatic generation of phonemic transliteration from numerical transliteration and enables multi-step linguistic annotation. Alternative readings can be entered, and two or more decipherment proposals can be processed in parallel. ALMAH is implemented in JAVA, is based on a graph-data model, and has a user-friendly interface. | ['Christian Prager', 'Cristina Vertan'] | null | null | null | null | lt4hala-lrec-2022-6 | ['decipherment', 'transliteration'] | ['natural-language-processing', 'natural-language-processing'] | [ 4.05137464e-02 -1.25312403e-01 2.69183517e-02 -5.80598488e-02
-3.15824449e-01 -1.03263986e+00 7.38347173e-01 7.76661187e-03
-2.97138631e-01 4.03545678e-01 4.11625355e-01 -7.34449029e-01
-2.54956931e-01 -6.00292146e-01 -5.23095131e-02 -1.19609468e-01
8.01285923e-01 9.78392303e-01 -1.10981045e-02 -5.43958724e-01
4.85937119e-01 5.85033894e-01 -1.57590210e+00 2.61080235e-01
7.94321001e-01 5.49279034e-01 1.86237127e-01 9.45981503e-01
-6.68673158e-01 8.13265443e-01 -6.72069311e-01 -6.46750033e-01
3.05017438e-02 -6.63005471e-01 -8.19530845e-01 -4.30250198e-01
3.90845299e-01 -3.22301298e-01 1.77840386e-02 7.72583246e-01
4.46885914e-01 -5.01148224e-01 6.39820158e-01 -7.30097413e-01
-5.18442750e-01 9.49916363e-01 1.60518855e-01 -4.10382956e-01
8.90630245e-01 8.89670551e-02 9.56781387e-01 -1.00684595e+00
1.02495527e+00 1.25258315e+00 6.70871794e-01 2.10422665e-01
-1.11576855e+00 -3.21871966e-01 -8.38643372e-01 2.04417884e-01
-1.38018119e+00 -4.62028474e-01 7.52792537e-01 -7.10326016e-01
8.21783483e-01 6.98131025e-01 1.03507948e+00 7.56478727e-01
-1.63643822e-01 4.52603847e-01 1.33847952e+00 -1.18090737e+00
7.11636767e-02 -6.33945093e-02 1.34447232e-01 8.42528224e-01
1.63928643e-01 -2.43089512e-01 -8.18237603e-01 8.51722260e-04
6.52110100e-01 -5.80058575e-01 8.14925414e-04 1.90859795e-01
-1.40507996e+00 5.99657118e-01 -5.94875813e-01 8.55024159e-01
-7.39307329e-02 -8.36184830e-04 6.47769868e-01 5.35170436e-01
-3.53277326e-02 4.01508361e-01 -8.94256085e-02 -9.82354462e-01
-1.17487538e+00 2.56239504e-01 1.13718688e+00 8.22651863e-01
5.09658575e-01 9.92759839e-02 2.68867463e-01 8.36521745e-01
5.91734827e-01 6.92310214e-01 6.09234989e-01 -7.77242362e-01
4.77486998e-01 8.05757403e-01 -4.68867719e-02 -9.14426982e-01
-2.35870451e-01 3.04042250e-01 -2.67847985e-01 5.11420548e-01
9.90897536e-01 1.11482672e-01 -7.04828203e-01 8.22706342e-01
1.29057467e-01 -8.77152562e-01 -1.09385669e-01 7.32405543e-01
1.00060987e+00 5.05190432e-01 -1.68169767e-01 -2.07827277e-02
1.46376801e+00 -1.99868187e-01 -1.22130334e+00 3.58806521e-01
8.46684158e-01 -1.49235213e+00 1.64892161e+00 5.51503062e-01
-1.35107744e+00 -1.53735965e-01 -1.20134807e+00 -6.79689229e-01
-6.86012268e-01 5.38960278e-01 3.74789864e-01 1.01214707e+00
-9.15349066e-01 4.06651825e-01 -7.44227886e-01 -5.39393961e-01
-1.06049143e-01 -2.40957662e-02 -3.52349952e-02 4.79740620e-01
-1.05039442e+00 1.18870425e+00 3.21467906e-01 2.89599776e-01
3.18838894e-01 -3.92633229e-01 -9.04302537e-01 -3.23893845e-01
5.96474968e-02 -2.60752738e-01 1.01580262e+00 -8.45047176e-01
-2.04831243e+00 1.42368770e+00 -4.72557098e-02 4.21884730e-02
1.01638448e+00 2.78945684e-01 -7.31810570e-01 1.44418150e-01
-1.24570496e-01 -7.40305055e-03 4.82533544e-01 -7.95919120e-01
-4.50828999e-01 -2.38799989e-01 -2.92900592e-01 3.46992947e-02
-5.77917546e-02 5.53015292e-01 -2.63856590e-01 -9.47277009e-01
7.65114129e-02 -5.38743675e-01 6.41488075e-01 6.94668442e-02
-1.41590357e-01 -9.43740755e-02 7.45983839e-01 -1.64065027e+00
1.75109899e+00 -1.86156201e+00 -1.69477999e-01 6.71574116e-01
-1.20198756e-01 2.28065804e-01 3.34073007e-01 7.43617713e-01
3.88288587e-01 1.16028719e-01 -2.68911690e-01 -1.80678427e-01
6.03609025e-01 4.46400225e-01 -2.08529964e-01 3.52503181e-01
-5.49845815e-01 8.94531846e-01 -8.41574013e-01 -7.84113348e-01
5.48805952e-01 4.76081491e-01 7.76160732e-02 -3.26523870e-01
-1.86497137e-01 2.06425354e-01 3.26281488e-02 9.40275013e-01
5.31796098e-01 2.45490685e-01 5.46672344e-01 -8.66556689e-02
-8.35945606e-01 5.18482208e-01 -1.37389910e+00 1.80327356e+00
-6.92637384e-01 1.03625643e+00 -5.11294119e-02 -5.22880971e-01
1.34254277e+00 3.12018126e-01 3.06864176e-03 -6.27412498e-01
3.15951198e-01 1.00106227e+00 -4.05734219e-02 -6.70583069e-01
1.16810942e+00 2.18262449e-01 -8.47211853e-02 7.82940686e-01
-9.81766060e-02 -6.36274993e-01 8.70567381e-01 9.21677276e-02
6.09220028e-01 5.98748684e-01 6.68351471e-01 -2.93239295e-01
6.88087165e-01 5.69296777e-02 8.02295282e-02 1.43929929e-01
5.27403831e-01 5.47384977e-01 6.31715596e-01 -4.56072360e-01
-1.39854646e+00 -9.03737962e-01 -4.49498504e-01 8.45075428e-01
-4.38240707e-01 -8.13171089e-01 -9.33200777e-01 1.10265650e-01
-3.71542484e-01 9.17796254e-01 -2.73743242e-01 6.16018414e-01
-6.97731614e-01 -6.02548160e-02 9.40663874e-01 7.12580681e-02
2.99253821e-01 -1.35975969e+00 -9.63048816e-01 2.63154745e-01
-2.49111310e-01 -8.37399125e-01 -3.22252899e-01 -4.70311612e-01
-2.94947028e-01 -1.10147512e+00 -6.69909716e-01 -8.74759257e-01
5.42238235e-01 -7.79557049e-01 6.57782614e-01 2.26447716e-01
-4.64268833e-01 3.67614865e-01 -5.74043870e-01 -1.94230556e-01
-1.16675329e+00 3.87563035e-02 -2.31012985e-01 -1.67962521e-01
2.78875977e-01 -4.45260495e-01 -1.51521992e-02 7.54456967e-02
-9.19817507e-01 2.09415972e-01 -7.01629966e-02 4.12582695e-01
3.29762816e-01 -6.15569472e-01 5.91928512e-02 -6.05885565e-01
7.59984374e-01 2.21384600e-01 -9.73588943e-01 4.57443178e-01
-5.74736774e-01 1.34245902e-02 9.41225529e-01 -7.10525215e-02
-8.36881816e-01 -1.14506714e-01 -3.55982006e-01 3.72800052e-01
-2.90840846e-02 7.28609741e-01 -1.73707351e-01 -1.39031574e-01
5.54882824e-01 2.88127691e-01 1.83795422e-01 -7.09619105e-01
4.94115591e-01 1.09335136e+00 9.30047631e-01 -4.47168231e-01
6.03476107e-01 2.43472293e-01 -1.31247407e-02 -1.07586348e+00
3.59269172e-01 -1.00638038e-02 -8.46199572e-01 -7.88530648e-01
5.94595969e-01 -2.00436160e-01 -1.03285587e+00 7.69542575e-01
-1.12079978e+00 -5.22978485e-01 -5.65061510e-01 4.14230078e-01
-7.06045687e-01 7.44223177e-01 -4.68832910e-01 -7.46350169e-01
-3.50833148e-01 -7.46654749e-01 5.96476853e-01 1.12361223e-01
-9.96227384e-01 -1.05928910e+00 2.86125809e-01 2.84925938e-01
2.27138311e-01 5.04752398e-01 1.05813313e+00 -2.88386941e-01
-2.47757006e-02 -4.59894985e-01 1.57139838e-01 1.69388592e-01
2.37222567e-01 7.27109492e-01 -4.63475674e-01 2.82837421e-01
-4.88595456e-01 -5.32021113e-02 1.59243748e-01 -1.93414554e-01
3.60213876e-01 -6.24617577e-01 3.33636612e-01 5.54233253e-01
1.29589081e+00 9.84326154e-02 9.02813196e-01 4.70681667e-01
5.32678843e-01 6.85611248e-01 3.67022216e-01 6.02894008e-01
5.18137872e-01 6.96947455e-01 -1.86393708e-01 4.66339111e-01
-5.78867555e-01 -4.87558097e-01 4.60501581e-01 1.37317586e+00
-3.76263797e-01 4.64540487e-03 -1.36419988e+00 5.43628812e-01
-1.65743780e+00 -1.01851606e+00 -9.51481760e-01 2.18691897e+00
8.78299236e-01 -1.13600038e-01 1.83242053e-01 7.67685711e-01
6.07953846e-01 -3.39200050e-02 2.09005460e-01 -1.02523971e+00
-4.86038595e-01 6.70758426e-01 5.39704263e-01 1.04010904e+00
-3.15305740e-01 1.19094682e+00 6.20272017e+00 7.25967348e-01
-1.09471989e+00 -9.93784964e-02 -2.05883607e-01 1.56740338e-01
-6.18487656e-01 2.89941698e-01 -5.66845059e-01 8.32861841e-01
7.05809653e-01 -3.98308218e-01 7.17425644e-01 2.56940842e-01
4.59877998e-01 -4.40173477e-01 -8.75510156e-01 1.14400780e+00
2.57627219e-01 -1.73012125e+00 1.58682659e-01 -5.74008040e-02
4.67564315e-01 -3.53394061e-01 -1.50656864e-01 -2.93204159e-01
2.37177163e-01 -8.82389128e-01 1.40492809e+00 8.80224764e-01
1.47597909e+00 -5.07229745e-01 3.18001896e-01 -5.07358052e-02
-1.05573308e+00 3.22442681e-01 5.38907275e-02 -2.91928470e-01
5.27082682e-01 2.06052080e-01 -6.90196931e-01 5.36738038e-01
1.12894803e-01 4.33987677e-01 -7.05054104e-01 8.07249010e-01
-5.08536518e-01 8.45087528e-01 -5.77561140e-01 -4.69550937e-01
-5.59156239e-02 -7.46638894e-01 6.30391181e-01 1.38443720e+00
6.80641651e-01 -1.50830045e-01 -3.67676169e-01 7.14638591e-01
2.78356463e-01 7.19391346e-01 -1.62788719e-01 -1.96297571e-01
5.95929384e-01 1.10771430e+00 -9.42831576e-01 -3.76199692e-01
-1.03192732e-01 8.86874318e-01 -2.31820241e-01 1.99160427e-01
-4.49173272e-01 -9.40580249e-01 -1.25815794e-01 4.33338076e-01
1.42187402e-01 -5.91801882e-01 -7.97701061e-01 -8.16769004e-01
2.88695991e-01 -9.94019151e-01 3.36665004e-01 -9.96604443e-01
-6.22873604e-01 2.46769771e-01 -1.51639834e-01 -1.07833028e+00
-6.96963131e-01 -8.00989151e-01 -4.74148870e-01 9.93649125e-01
-6.31578147e-01 -1.37817883e+00 -2.14939386e-01 5.16047835e-01
2.38140449e-02 -1.86747551e-01 1.20698726e+00 2.99515158e-01
-2.14611962e-02 5.67408323e-01 4.16298062e-01 4.19863611e-01
5.54194093e-01 -1.33584559e+00 4.19900626e-01 7.63341904e-01
2.21678853e-01 4.18383032e-01 8.13666999e-01 -7.60971487e-01
-1.11650145e+00 -2.45276868e-01 1.79680920e+00 -2.96571076e-01
1.10355377e+00 -5.23950219e-01 -6.15360320e-01 6.27542078e-01
2.64062911e-01 -7.06740439e-01 5.73728621e-01 -5.12612760e-01
-1.81387261e-01 8.28419477e-02 -9.91987348e-01 9.46130991e-01
7.98362732e-01 -8.22928727e-01 -7.59397089e-01 3.49088281e-01
-6.70132712e-02 -5.18737614e-01 -9.56141412e-01 -3.43415231e-01
8.97443593e-01 -7.47807980e-01 3.62953484e-01 1.95135295e-01
2.85277694e-01 -5.24965525e-01 1.11508213e-01 -8.64595115e-01
3.39267343e-01 -1.42929292e+00 6.45146891e-02 1.37350953e+00
4.85258877e-01 -8.35269928e-01 2.75503695e-01 4.45392758e-01
-1.86612010e-01 6.29353076e-02 -1.26413620e+00 -7.87351251e-01
-1.47461340e-01 -5.72361410e-01 6.92656755e-01 1.07515717e+00
7.28661358e-01 1.29410788e-01 -2.88713306e-01 -6.17352068e-01
4.41961437e-01 -1.24142796e-01 7.94007242e-01 -1.16302586e+00
-8.51141959e-02 -8.95308077e-01 -4.63845342e-01 -4.48211610e-01
-9.17614698e-02 -1.35854375e+00 -3.64605188e-01 -1.71700001e+00
-8.21674526e-01 -2.04225048e-01 8.03797662e-01 5.28958678e-01
6.98101819e-01 3.12312782e-01 3.09199363e-01 4.15903360e-01
2.17353538e-01 -6.27795160e-02 8.44948173e-01 1.60808221e-01
-4.81588632e-01 -4.26239371e-01 5.54316044e-02 8.64802003e-01
8.64761055e-01 -2.66315371e-01 2.84641683e-01 -2.73557901e-01
1.08908296e+00 -2.31326267e-01 3.58040601e-01 -8.52781653e-01
3.38199407e-01 8.35819021e-02 -5.28002530e-02 -1.06059206e+00
4.96900380e-02 -6.98307455e-01 4.97943878e-01 5.72747827e-01
-2.02322170e-01 2.13828415e-01 -2.16001049e-01 -3.91333848e-01
1.81266945e-02 -7.10026622e-01 5.97074270e-01 2.73937173e-02
-5.43746769e-01 -3.60903770e-01 -9.54366326e-01 -8.07910189e-02
6.08122289e-01 -8.61667573e-01 -2.38879472e-01 -3.29780787e-01
-5.41584194e-01 -2.78294712e-01 8.60576153e-01 1.99737549e-02
2.98272192e-01 -1.45832551e+00 -7.79096961e-01 1.26512334e-01
2.74851397e-02 -4.76328760e-01 -5.49859889e-02 7.17638910e-01
-1.79020023e+00 3.18605930e-01 -4.11676228e-01 -7.24319890e-02
-1.27237678e+00 1.00768536e-01 1.25209540e-01 -1.60366163e-01
-6.84115350e-01 -1.21248923e-01 -1.00268674e+00 -4.33448642e-01
-6.13264963e-02 -4.85272586e-01 -1.79319188e-01 3.95579934e-01
5.75353265e-01 1.03363454e+00 4.76738624e-02 -1.09673882e+00
-3.24687839e-01 9.10505295e-01 4.86394823e-01 -9.89509940e-01
9.75627303e-01 -2.15310544e-01 -4.82810140e-01 8.13333750e-01
9.75523174e-01 7.60143399e-01 -4.78518546e-01 1.74847439e-01
2.02359945e-01 -3.35589170e-01 -4.82742369e-01 -1.10307097e+00
-2.82242864e-01 5.79649270e-01 1.36279345e-01 1.58113703e-01
6.35076880e-01 -2.63961881e-01 6.08358741e-01 3.36936414e-01
-5.62943406e-02 -1.86868310e+00 -4.97461438e-01 8.94412279e-01
1.21177506e+00 -3.32244217e-01 -9.31561217e-02 -1.06704339e-01
-3.76018494e-01 1.73435378e+00 -2.36095950e-01 2.02856630e-01
2.95402318e-01 5.66903174e-01 5.07409394e-01 -1.75400987e-01
4.64915149e-02 -1.22808248e-01 2.00567245e-01 6.90814078e-01
5.13423264e-01 2.31715605e-01 -1.50569177e+00 4.38882947e-01
-1.20010448e+00 1.48708850e-01 7.59536803e-01 8.31617296e-01
-2.17976779e-01 -1.31170976e+00 -9.22592878e-01 2.41171475e-02
-3.58620435e-01 -3.31450760e-01 -8.02791536e-01 8.94877672e-01
3.19512039e-02 8.64893854e-01 1.06213614e-01 -1.23313934e-01
2.29754493e-01 3.57539207e-01 7.12547839e-01 -4.69384715e-02
-9.29756284e-01 -3.56863812e-02 5.79311311e-01 -1.83861852e-01
-1.50137886e-01 -8.71960044e-01 -1.46267629e+00 -7.36030400e-01
2.74513334e-01 -2.41060536e-02 1.16540456e+00 9.04219925e-01
-1.58007011e-01 -5.96216246e-02 5.27409576e-02 -4.59816635e-01
-1.28359795e-01 -9.17483985e-01 -5.91362298e-01 1.65743709e-01
-2.11664259e-01 9.54768434e-02 -3.89392287e-01 4.23462629e-01] | [10.413895606994629, 10.381119728088379] |
66c659d9-759d-40f3-bc01-e45bd273bf6c | earthquake-magnitude-prediction-in-hindukush | null | null | https://www.researchgate.net/publication/307951466_Earthquake_magnitude_prediction_in_Hindukush_region_using_machine_learning_techniques | https://www.researchgate.net/publication/307951466_Earthquake_magnitude_prediction_in_Hindukush_region_using_machine_learning_techniques | Earthquake magnitude prediction in Hindukush region using machine learning techniques | Earthquake magnitude prediction for Hindukush region has been carried out in this research using the temporal sequence of historic seismic activities in combination with the machine learning classifiers. Prediction has been made on the basis of mathematically calculated eight seismic indicators using the earthquake catalog of the region. These parameters are based on the well-known geophysical facts of Gutenberg–Richter’s inverse law, distribution of characteristic earthquake magnitudes and seismic quiescence. In this research, four machine learning techniques including pattern recognition neural network, recurrent neural network, random forest and linear programming boost ensemble classifier are separately applied to model relationships between calculated seismic parameters and future earthquake occurrences. The problem is formulated as a binary classification task and predictions are made for earthquakes of magnitude greater than or equal to 5.5 (\(M \ge\) 5.5), for the duration of 1 month. Furthermore, the analysis of earthquake prediction results is carried out for every machine learning classifier in terms of sensitivity, specificity, true and false predictive values. Accuracy is another performance measure considered for analyzing the results. Earthquake magnitude prediction for the Hindukush using these aforementioned techniques show significant and encouraging results, thus constituting a step forward toward the final robust prediction mechanism which is not available so far. | ['Francisco Martínez-Álvarez', 'Khawaja Asim', 'Abdul Basit', 'Talat Iqbal'] | 2016-09-08 | null | null | null | springer-2016-9 | ['earthquake-prediction'] | ['computer-vision'] | [ 2.32351869e-01 -7.68219978e-02 1.34508371e-01 -2.05922902e-01
-6.02763593e-01 -1.29573658e-01 4.90202874e-01 4.46978897e-01
-4.84091491e-01 9.81505156e-01 2.78233171e-01 -6.34087980e-01
-5.47894955e-01 -1.06707680e+00 -1.52845666e-01 -9.57226515e-01
-6.27321184e-01 4.51415002e-01 2.55310535e-01 -3.50751817e-01
7.33021796e-01 7.37466216e-01 -1.41014469e+00 2.41380021e-01
4.71108377e-01 8.87712896e-01 1.80274487e-01 8.29899371e-01
5.19927979e-01 1.15838361e+00 -4.68315035e-01 9.98372063e-02
-1.82750925e-01 -2.36599520e-01 -8.35918725e-01 -5.01589417e-01
-7.88742602e-01 -2.54378349e-01 -6.15412854e-02 1.75446063e-01
5.07005453e-01 2.39285916e-01 1.16778302e+00 -7.42456615e-01
-1.78531811e-01 6.97429597e-01 -4.52527136e-01 6.11714303e-01
4.76659268e-01 -2.11720541e-01 7.05774724e-01 -9.97919917e-01
9.35909003e-02 4.86618876e-01 1.01977122e+00 -2.71203279e-01
-7.98587918e-01 -4.93050367e-01 -5.64491630e-01 4.42284107e-01
-1.52171791e+00 8.93218666e-02 9.41840649e-01 -7.54540265e-01
1.33066082e+00 3.48520845e-01 7.24920750e-01 4.33792412e-01
5.68659425e-01 6.45306781e-02 9.43531513e-01 -7.33403087e-01
3.81153733e-01 -1.12300344e-01 3.95093113e-02 1.17128521e-01
1.80252478e-01 1.84838980e-01 -3.12122315e-01 -2.46979415e-01
3.41778725e-01 -4.89225835e-02 7.05412850e-02 7.40888774e-01
-9.39352870e-01 9.43279088e-01 3.15321386e-01 8.01815510e-01
-8.71969283e-01 -1.30705282e-01 4.90216523e-01 -3.96850111e-04
5.06314397e-01 4.76154923e-01 -5.40399909e-01 -1.28954753e-01
-1.15742338e+00 2.40978405e-01 7.22878873e-01 9.48067978e-02
5.01727879e-01 4.85397607e-01 4.63146538e-01 5.82519174e-01
3.00428569e-01 5.37399113e-01 6.33581340e-01 -4.91268963e-01
3.05702150e-01 4.62383777e-01 1.06215887e-01 -1.51976800e+00
-7.02533066e-01 -3.97920936e-01 -8.99370670e-01 2.80931522e-03
2.80830890e-01 -4.49216217e-01 -4.44768637e-01 1.16197729e+00
9.24742371e-02 1.38351858e-01 4.09075826e-01 4.68682706e-01
6.51009798e-01 1.00198102e+00 2.52017170e-01 -3.30745339e-01
1.14732206e+00 -1.28577262e-01 -3.80836427e-01 -6.07028790e-02
8.15691590e-01 -6.45546198e-01 4.53906596e-01 5.50029099e-01
-5.24262607e-01 -3.71604830e-01 -8.32442582e-01 6.75415337e-01
-3.12080175e-01 1.27850771e-01 6.20811880e-01 5.31725407e-01
-7.24829555e-01 6.21458411e-01 -8.90015662e-01 -8.18638578e-02
-1.08986452e-01 2.84267753e-01 -1.70771703e-01 3.91809434e-01
-1.43394375e+00 1.26340532e+00 6.83902681e-01 6.43339574e-01
-3.61278474e-01 -3.45171154e-01 -6.35422111e-01 5.35278693e-02
-3.57798755e-01 -1.86400265e-01 7.55510092e-01 -5.28327167e-01
-9.50661063e-01 6.39623582e-01 1.12676747e-01 -5.47580302e-01
2.45170966e-01 8.01441073e-02 -6.35140181e-01 3.87890451e-02
-1.83793774e-03 -3.73171866e-01 -5.94822913e-02 -9.39755857e-01
-6.94687366e-01 -3.91067505e-01 -5.21437466e-01 5.79246767e-02
-8.68525654e-02 1.78897813e-01 6.06284261e-01 -8.85927320e-01
4.80176628e-01 -6.23055339e-01 -4.19257045e-01 -1.12960482e+00
-1.46402538e-01 -1.51924446e-01 5.24442017e-01 -1.33283818e+00
1.55956614e+00 -1.83525729e+00 -3.04928631e-01 7.99452126e-01
-3.77135754e-01 -7.23543540e-02 6.71166718e-01 9.52097654e-01
-3.04139793e-01 -1.91256106e-01 -4.42738652e-01 5.81551313e-01
-6.16535723e-01 1.23922572e-01 -3.99580359e-01 3.15873265e-01
1.23963475e-01 7.27459714e-02 -6.31369114e-01 -3.88202727e-01
3.44259560e-01 2.97806770e-01 -2.84513324e-01 1.28337875e-01
4.11667436e-01 4.01358783e-01 -4.90744948e-01 5.81011474e-01
3.96948338e-01 2.07086548e-01 7.40852505e-02 1.39235690e-01
-4.58003789e-01 -4.91022244e-02 -1.06733859e+00 6.43779516e-01
-4.05330032e-01 3.82187515e-01 -7.46795475e-01 -1.54425514e+00
1.64267552e+00 7.73529053e-01 5.86054265e-01 -3.90249848e-01
1.22854702e-01 4.64125544e-01 -4.00950201e-02 -7.79443800e-01
4.15608585e-01 -3.77853036e-01 -2.54452288e-01 2.50850677e-01
-3.44675422e-01 -1.33874968e-01 1.88319296e-01 -2.89630949e-01
9.69237506e-01 -4.48103137e-02 6.53107524e-01 -2.80320138e-01
6.53008163e-01 2.10196242e-01 4.94374156e-01 6.28272891e-01
1.58245146e-01 3.93649906e-01 5.39155841e-01 -5.65263510e-01
-1.04970360e+00 -8.49364400e-01 -4.27822918e-01 7.77660668e-01
-4.27722722e-01 2.99459666e-01 -4.04327750e-01 1.68566361e-01
-3.07193130e-01 9.28796649e-01 -5.47344148e-01 -4.55823392e-02
-7.04477429e-01 -1.29619110e+00 5.81781149e-01 6.42103493e-01
3.82763118e-01 -1.25289893e+00 -1.06120455e+00 5.46294928e-01
-1.60882771e-01 -5.37916422e-01 6.53032899e-01 4.26698923e-01
-1.06301343e+00 -9.44615662e-01 -7.15253770e-01 -8.20367217e-01
3.52395236e-01 -3.35384578e-01 7.72142887e-01 1.34649441e-01
-1.25232399e-01 -4.50382605e-02 -5.38589239e-01 -5.53138673e-01
-4.20426041e-01 -5.57078682e-02 -1.67775750e-01 -1.54271260e-01
2.49358431e-01 -9.50664699e-01 -5.03186226e-01 1.80775329e-01
-6.18786514e-01 -1.54084668e-01 6.34620011e-01 7.77643800e-01
2.26706296e-01 4.57891762e-01 1.13775539e+00 -5.67437589e-01
2.69214720e-01 -1.09281635e+00 -3.10391098e-01 1.01808056e-01
-4.19964612e-01 -3.47349793e-01 5.66683948e-01 -2.22873449e-01
-1.27856624e+00 6.64173439e-02 -4.35686529e-01 4.80451643e-01
-3.05927038e-01 1.13688564e+00 3.70401293e-01 3.17063719e-01
7.13412821e-01 4.76982743e-01 -4.13386494e-01 -2.56840765e-01
-3.49110544e-01 8.95245016e-01 7.44895577e-01 -3.50664049e-01
3.05752516e-01 2.27684051e-01 1.27348363e-01 -9.75164473e-01
-3.90330851e-01 -5.64782441e-01 -6.64241493e-01 -8.50878000e-01
7.10560262e-01 -6.23282135e-01 -6.03285193e-01 8.17964315e-01
-9.40980256e-01 1.20453179e-01 2.35928833e-01 1.04723144e+00
-5.08256078e-01 1.34740502e-01 -4.87510592e-01 -1.52280629e+00
-4.94961470e-01 -7.35405147e-01 1.85161904e-01 1.67779759e-01
-4.57405537e-01 -1.08964694e+00 2.65608937e-01 2.42362469e-01
4.43785101e-01 9.45782781e-01 8.99498761e-01 -8.69251013e-01
1.40500769e-01 -6.52779698e-01 2.40555331e-01 2.00629950e-01
1.02494352e-01 3.32627505e-01 -6.69303715e-01 2.18485385e-01
3.21229130e-01 9.96559262e-02 6.49284601e-01 4.80186075e-01
6.81394696e-01 -4.30704623e-01 -7.59584978e-02 2.79364288e-02
1.65430009e+00 8.98662925e-01 6.73194051e-01 7.45856941e-01
1.34630412e-01 6.93692744e-01 6.06772780e-01 9.42271888e-01
3.50626141e-01 2.34736234e-01 2.20536038e-01 1.35154709e-01
4.82063949e-01 1.24381728e-01 5.18396161e-02 7.29526401e-01
-6.78921640e-01 4.38351370e-02 -1.67789519e+00 7.65704632e-01
-1.52838480e+00 -1.44337392e+00 -6.27339125e-01 2.17999601e+00
5.66520870e-01 3.77323985e-01 1.37611598e-01 1.05313373e+00
6.74220502e-01 3.14958319e-02 1.55064195e-01 -6.70140266e-01
-1.83703616e-01 8.68725702e-02 5.41886270e-01 2.43188113e-01
-9.30924058e-01 3.32309037e-01 6.09072113e+00 4.60726976e-01
-1.28166664e+00 -1.61985129e-01 8.28231514e-01 4.14308518e-01
-6.33562878e-02 1.39422685e-01 -3.47768396e-01 6.31119311e-01
1.20688128e+00 2.72655562e-02 3.27595212e-02 7.83345938e-01
5.31574845e-01 -7.86885202e-01 -3.39427233e-01 4.79404241e-01
-1.01115391e-01 -1.29560602e+00 -3.31634879e-01 -2.41239399e-01
6.51358485e-01 -8.01106468e-02 -1.29474357e-01 1.45569161e-01
9.63555351e-02 -7.95192719e-01 6.65105939e-01 9.89133060e-01
2.75411963e-01 -1.00731671e+00 1.23452306e+00 5.45912981e-01
-1.10306299e+00 -3.83322746e-01 8.95400122e-02 -5.54390788e-01
3.13531965e-01 8.94876957e-01 -1.09060216e+00 7.49955595e-01
9.01467085e-01 3.69706780e-01 -1.72739610e-01 9.37991083e-01
5.83977550e-02 1.09302223e+00 -6.21846139e-01 -1.49740383e-01
2.08329082e-01 -1.15423791e-01 1.48463815e-01 1.21144688e+00
6.79116845e-01 3.83721232e-01 -3.04826528e-01 2.16348678e-01
7.74218082e-01 4.01571423e-01 -5.52400529e-01 4.25084502e-01
6.10161960e-01 6.36995137e-01 -9.11865950e-01 -2.13486269e-01
-2.60192990e-01 8.38638842e-02 -9.31299701e-02 3.67480144e-02
-7.23053813e-01 -4.49290276e-01 -4.33781534e-01 1.92325458e-01
1.05099380e-01 -2.24832684e-01 -1.00560379e+00 -4.30603504e-01
-1.22363098e-01 -2.01414332e-01 6.59914613e-01 -7.50604451e-01
-9.67334032e-01 5.94215631e-01 2.86785990e-01 -1.19979286e+00
-6.72393799e-01 -2.74831802e-01 -1.23952699e+00 9.68047380e-01
-9.70175147e-01 -8.89165223e-01 3.60014737e-02 3.14772457e-01
1.84474573e-01 -2.62456924e-01 9.40653682e-01 2.14415833e-01
-5.47920406e-01 1.95256080e-02 4.23562169e-01 2.34468967e-01
-5.68709299e-02 -8.88206363e-01 -1.74695730e-01 6.98858917e-01
-5.62433422e-01 -3.84481624e-02 1.03812587e+00 -8.49216461e-01
-5.84967673e-01 -6.77477121e-01 1.47170889e+00 6.21642694e-02
8.18201125e-01 4.06126827e-01 -9.40531015e-01 4.71772254e-01
-2.80482411e-01 -4.19582188e-01 8.72385204e-01 -7.13422075e-02
2.65204638e-01 -9.95109379e-02 -1.10398436e+00 2.05145806e-01
1.28548974e-02 -1.21558331e-01 -8.77065003e-01 3.16846550e-01
-2.56174713e-01 -1.78778302e-02 -1.45873046e+00 5.76055706e-01
7.39039838e-01 -9.78118062e-01 9.99431312e-01 -2.13002160e-01
6.34560347e-01 -3.07480004e-02 -2.30469853e-01 -7.93340564e-01
-1.32908061e-01 4.18916382e-02 4.20869917e-01 1.19320035e+00
7.23702073e-01 -7.37859786e-01 7.72043228e-01 4.80854630e-01
-2.34127846e-02 -1.04125965e+00 -1.27210927e+00 -4.40103710e-01
1.24345712e-01 -6.05100513e-01 3.86100054e-01 9.18043017e-01
1.50110573e-01 1.73098501e-03 -5.36873221e-01 3.61328185e-01
4.36020464e-01 -8.86375830e-02 2.34712422e-01 -1.01795697e+00
-2.42780283e-01 -2.40856364e-01 -6.41620338e-01 8.68526325e-02
-2.92162716e-01 -5.03946900e-01 -1.65630445e-01 -1.42362738e+00
-1.39227092e-01 -6.88454866e-01 -3.79321247e-01 4.04788882e-01
-1.13353310e-02 1.44029826e-01 -1.55355826e-01 5.35951316e-01
5.20031095e-01 1.73355579e-01 5.10451853e-01 2.65960813e-01
-3.14385742e-01 5.47119439e-01 -8.08569342e-02 8.21266294e-01
1.25872946e+00 -6.28004611e-01 -1.77224979e-01 -3.77814211e-02
5.04021883e-01 8.58157992e-01 1.87480301e-01 -1.21678829e+00
6.42590970e-02 -4.47552085e-01 3.81077290e-01 -9.94355500e-01
1.40322700e-01 -6.20675921e-01 6.91694260e-01 8.58378351e-01
-2.02121764e-01 2.21443132e-01 -7.92317092e-02 4.02685881e-01
-4.16741699e-01 -5.75837016e-01 5.61564147e-01 -2.47717015e-02
-7.81245172e-01 -4.43395406e-01 -8.72574508e-01 -5.08919895e-01
1.34439957e+00 -5.57381988e-01 2.55193889e-01 -1.84967905e-01
-1.11442912e+00 -1.17294773e-01 -2.85195112e-01 -1.63911566e-01
5.90830564e-01 -9.64122236e-01 -1.12767673e+00 -1.13960266e-01
-3.51301990e-02 -4.04575646e-01 6.67624950e-01 9.82677221e-01
-1.10066330e+00 3.30973923e-01 -4.02316958e-01 -2.91417897e-01
-9.85307217e-01 3.08123410e-01 2.65710860e-01 -4.58067566e-01
-5.08160651e-01 6.34436011e-01 -5.17112494e-01 -9.43598822e-02
-2.95414567e-01 8.52332450e-03 -9.92683291e-01 2.45898724e-01
2.76810408e-01 7.82619238e-01 2.38828152e-01 -8.16115022e-01
-5.16125917e-01 5.00799775e-01 3.47997725e-01 -1.93063244e-01
1.67468655e+00 1.31111532e-01 -2.03812227e-01 6.97752059e-01
1.00969636e+00 -2.81314969e-01 -5.32239914e-01 8.04532245e-02
4.88593817e-01 -1.22478455e-01 1.93584785e-02 -7.40474224e-01
-7.71260977e-01 5.43446362e-01 4.22700942e-01 2.72143364e-01
1.39178932e+00 -1.16923586e-01 5.28754532e-01 2.77443588e-01
1.94023803e-01 -1.19168210e+00 -4.96761769e-01 4.07645106e-01
1.01019573e+00 -9.40182865e-01 7.59092271e-02 -7.68067092e-02
-5.53255975e-01 1.40970647e+00 -5.31923473e-02 -3.23754013e-01
1.04178119e+00 3.78603339e-01 -2.76041716e-01 1.18651070e-01
-5.63027501e-01 2.18506277e-01 -9.10074040e-02 4.36877936e-01
6.30108476e-01 3.11634719e-01 -8.78449082e-01 8.47263694e-01
-5.00415325e-01 1.32892355e-01 5.35241127e-01 1.18126905e+00
-7.67138243e-01 -4.14890081e-01 -8.74259710e-01 7.66842365e-01
-6.83388472e-01 -6.66384473e-02 2.16657430e-01 7.76023507e-01
4.85431552e-02 1.05735242e+00 1.14723183e-01 -4.65528250e-01
2.37043142e-01 -2.01143712e-01 -2.03877285e-01 -4.96264547e-02
-7.37764776e-01 -2.27005839e-01 3.46567482e-01 2.28305206e-01
-4.27225858e-01 -9.12124217e-01 -1.53950667e+00 -3.68110389e-01
-4.11218315e-01 4.74601150e-01 8.00294936e-01 1.18062341e+00
-4.81985122e-01 2.91229129e-01 1.04767096e+00 -9.79057729e-01
-3.28408241e-01 -1.18756986e+00 -6.63705230e-01 6.85198531e-02
-7.00344667e-02 -4.95943606e-01 -6.46726489e-01 1.92944050e-01] | [6.433723449707031, 3.001070022583008] |
daeabb89-5a9e-4e0e-8c65-16147b2f9380 | medical-image-registration-using-unsupervised | 2208.01825 | null | https://arxiv.org/abs/2208.01825v1 | https://arxiv.org/pdf/2208.01825v1.pdf | Medical image registration using unsupervised deep neural network: A scoping literature review | In medicine, image registration is vital in image-guided interventions and other clinical applications. However, it is a difficult subject to be addressed which by the advent of machine learning, there have been considerable progress in algorithmic performance has recently been achieved for medical image registration in this area. The implementation of deep neural networks provides an opportunity for some medical applications such as conducting image registration in less time with high accuracy, playing a key role in countering tumors during the operation. The current study presents a comprehensive scoping review on the state-of-the-art literature of medical image registration studies based on unsupervised deep neural networks is conducted, encompassing all the related studies published in this field to this date. Here, we have tried to summarize the latest developments and applications of unsupervised deep learning-based registration methods in the medical field. Fundamental and main concepts, techniques, statistical analysis from different viewpoints, novelties, and future directions are elaborately discussed and conveyed in the current comprehensive scoping review. Besides, this review hopes to help those active readers, who are riveted by this field, achieve deep insight into this exciting field. | ['Alireza Mehdizadeh', 'Reza Javidan', 'Hedieh Khorasani', 'Raouf Khayami', 'Mohammad Amin Mosleh Shirazi', 'Hamid Reza Boveiri', 'Meysam Tavakoli', 'Samaneh Abbasi'] | 2022-08-03 | null | null | null | null | ['medical-image-registration'] | ['medical'] | [ 4.10205752e-01 2.35946074e-01 -7.27243781e-01 -1.61827937e-01
-6.45998657e-01 7.68425921e-03 2.41355926e-01 5.71489990e-01
-7.67158270e-01 4.12435889e-01 2.53863305e-01 -2.10758850e-01
-5.36223412e-01 -6.00991011e-01 -1.94788173e-01 -1.13497734e+00
-3.79437625e-01 6.12651110e-01 -1.65600955e-01 -1.92875102e-01
2.06376314e-01 6.07321620e-01 -1.13514137e+00 1.12381302e-01
8.01963270e-01 7.85936594e-01 2.88864732e-01 2.20087767e-01
-5.36850058e-02 6.19921505e-01 -5.63540518e-01 -2.92809844e-01
4.27395254e-02 -6.17127120e-01 -8.80386829e-01 -1.92451701e-01
2.41242424e-01 -1.67010561e-01 -3.35329890e-01 9.45697248e-01
9.14934397e-01 9.66087654e-02 7.48272538e-01 -6.20302856e-01
-5.37708759e-01 3.97448301e-01 -5.44286489e-01 5.29549718e-01
-5.41859791e-02 -1.15490966e-01 2.83354580e-01 -5.13234437e-01
6.37490630e-01 4.62333471e-01 9.40334976e-01 5.75015008e-01
-6.68551922e-01 -5.66938043e-01 -3.78291816e-01 1.40045851e-01
-1.02893651e+00 -1.03606425e-01 6.86349154e-01 -4.72125351e-01
7.81666338e-01 2.62697667e-01 7.90908635e-01 7.00677156e-01
8.52258623e-01 4.81831998e-01 8.78201306e-01 -4.06469464e-01
-1.11100577e-01 -2.31686860e-01 5.96940964e-02 6.60393000e-01
3.16467345e-01 2.12827176e-01 -1.62835747e-01 4.86873314e-02
8.02854776e-01 1.22901700e-01 -2.82536805e-01 -2.36708060e-01
-1.33685768e+00 9.28330243e-01 6.42935991e-01 9.98009384e-01
-5.23714423e-01 -1.20602638e-01 7.80173540e-01 1.11931548e-01
6.12165332e-01 3.67141992e-01 -2.10470349e-01 1.19043447e-01
-9.72455204e-01 -2.56019711e-01 6.04015291e-01 3.86074185e-01
-2.51670908e-02 3.50762494e-02 -9.73443761e-02 8.57198894e-01
6.92921504e-02 2.51441211e-01 1.08449602e+00 -7.06587434e-01
1.95584875e-02 5.20221770e-01 -6.36497438e-01 -9.89653885e-01
-9.46003675e-01 -5.45135140e-01 -1.51395607e+00 2.23912567e-01
1.04528330e-01 -1.07859977e-01 -8.36546838e-01 1.11449039e+00
2.06974193e-01 -7.13776723e-02 -1.89772770e-02 9.11368191e-01
1.62899065e+00 1.06469810e-01 3.68796915e-01 -4.12716419e-01
1.80129004e+00 -9.46025491e-01 -1.18729246e+00 1.88231587e-01
7.40926087e-01 -1.00213230e+00 4.41229761e-01 9.22842175e-02
-1.06058192e+00 -4.14257795e-01 -9.68412280e-01 -4.54614200e-02
-3.74711871e-01 2.01991498e-01 8.98699284e-01 8.67798626e-01
-1.01107776e+00 5.23400128e-01 -1.35380447e+00 -7.33627737e-01
8.20267975e-01 9.35492277e-01 -6.85228705e-01 2.07681626e-01
-1.15571904e+00 1.31899011e+00 3.22481394e-01 3.66357088e-01
-4.45830554e-01 -8.03380847e-01 -8.67041469e-01 -6.06665790e-01
1.43215820e-01 -8.47698092e-01 9.84606624e-01 -7.83097386e-01
-1.34267664e+00 1.50282717e+00 -5.74724674e-02 -5.61480939e-01
3.36413324e-01 2.32075006e-01 -3.55193615e-01 2.85962194e-01
6.88769892e-02 4.83994246e-01 1.32542783e-02 -8.69969130e-01
-4.06177431e-01 -6.27602756e-01 -4.66262847e-01 2.94035822e-01
-2.79377580e-01 4.05193120e-01 -3.26308817e-01 -9.31312323e-01
1.20759413e-01 -9.16812360e-01 -5.94911933e-01 1.76695392e-01
-7.67818317e-02 -1.49147600e-01 4.93358701e-01 -7.22741127e-01
9.57848668e-01 -1.94051027e+00 -8.14712942e-02 1.29531398e-01
5.34792125e-01 5.12291551e-01 1.62881687e-01 3.19838673e-01
-3.23855817e-01 -3.50872576e-02 -5.76843262e-01 -2.22597986e-01
-7.52400696e-01 6.66089654e-02 3.69144738e-01 1.05833793e+00
-4.09950197e-01 1.34710705e+00 -8.95480633e-01 -5.51257789e-01
6.32520735e-01 6.51085138e-01 5.65211326e-02 2.06115320e-01
5.99089444e-01 9.86079574e-01 -4.05152351e-01 5.01160681e-01
5.87069571e-01 -1.75024986e-01 1.40144661e-01 -4.03794050e-01
-1.58687100e-01 2.47256756e-02 -5.41320384e-01 2.02865911e+00
-4.37762558e-01 7.83054709e-01 -3.39946747e-02 -1.41724038e+00
8.09196293e-01 7.16887712e-01 1.19900250e+00 -7.00669467e-01
5.38805962e-01 3.27086776e-01 3.17401528e-01 -7.37385869e-01
1.97468251e-01 -3.36014658e-01 2.69214630e-01 2.71630824e-01
7.34493881e-02 -7.06896782e-02 -8.37041810e-02 -2.86981970e-01
6.63689673e-01 -1.95431560e-01 7.62117863e-01 -1.68313414e-01
6.80919051e-01 1.52732015e-01 2.01164067e-01 6.86126769e-01
-5.13193250e-01 6.50434792e-01 5.57154529e-02 -7.77303457e-01
-6.45840406e-01 -7.61471748e-01 -5.56745112e-01 4.11509484e-01
2.50809547e-02 1.53361276e-01 -8.22470009e-01 -3.48346174e-01
-5.82351923e-01 -5.72459064e-02 -1.00914180e+00 -1.25133470e-01
-8.38815629e-01 -1.26331377e+00 6.14663899e-01 4.70740020e-01
5.33997297e-01 -1.44603467e+00 -7.03511298e-01 1.88708708e-01
-1.99986190e-01 -1.14626610e+00 -1.17915854e-01 1.32718652e-01
-1.41548610e+00 -1.24698281e+00 -1.38550234e+00 -1.39732027e+00
9.12073612e-01 1.68702438e-01 9.95276451e-01 4.44967717e-01
-5.89493752e-01 2.00490013e-01 -1.19361736e-01 -4.89588678e-01
-5.99418283e-01 4.54968512e-01 -9.83801857e-02 -5.13150275e-01
5.65293431e-01 -4.99133587e-01 -7.60260820e-01 1.14662677e-01
-1.08923519e+00 -1.76555380e-01 7.12649286e-01 9.85439241e-01
5.47984660e-01 -3.39096114e-02 3.68071467e-01 -1.02923012e+00
6.57467484e-01 -3.12932521e-01 -3.39413017e-01 9.44067612e-02
-5.51162541e-01 -3.30012888e-01 1.57770559e-01 -6.99471608e-02
-8.27480674e-01 -1.49763674e-01 -5.01097500e-01 1.54467538e-01
-4.70020056e-01 7.33126163e-01 4.03227031e-01 -6.42055035e-01
7.99725294e-01 2.78477371e-02 7.00084329e-01 -1.71820536e-01
4.64421064e-02 6.34709120e-01 5.80018580e-01 -2.26208940e-01
3.79496396e-01 8.37922812e-01 3.53987634e-01 -8.14189851e-01
-5.62211931e-01 -8.11502159e-01 -8.30935538e-01 -3.15686613e-01
1.23038089e+00 -5.27177155e-01 -6.48366451e-01 6.10940516e-01
-1.08772242e+00 -1.92092702e-01 -2.97490895e-01 8.86666954e-01
-5.84954262e-01 4.27556783e-01 -6.91830993e-01 -2.27341816e-01
-1.03810561e+00 -1.93410492e+00 7.70332396e-01 3.67857754e-01
-3.35437447e-01 -1.58238912e+00 3.83602977e-01 3.21425736e-01
7.19703734e-01 5.58894932e-01 7.19025850e-01 -9.33587968e-01
-7.08670169e-02 -5.08083761e-01 -8.86979606e-03 5.34047149e-02
6.10364139e-01 -2.78248608e-01 -7.70364583e-01 -1.77104115e-01
3.14370513e-01 9.53644887e-02 5.67204237e-01 9.53809083e-01
1.21535695e+00 1.60372928e-01 -5.33126175e-01 8.52333784e-01
1.27470481e+00 4.87229407e-01 7.95385301e-01 5.79629660e-01
5.60762107e-01 6.51160300e-01 3.86866540e-01 -1.18138663e-01
1.13959678e-01 7.09637284e-01 5.22604287e-01 -7.65148640e-01
-3.69490653e-01 5.42360306e-01 -4.50496793e-01 1.08518171e+00
-7.53792465e-01 1.65609524e-01 -1.05990255e+00 5.16589880e-01
-1.37062037e+00 -7.44933546e-01 -2.81941116e-01 2.18580604e+00
7.18431413e-01 -1.72513500e-01 -4.28628959e-02 4.71720323e-02
7.92341292e-01 1.56058952e-01 -4.51151542e-02 -1.49114206e-01
8.32592547e-02 6.04389608e-01 6.22319162e-01 2.22906619e-01
-1.37338364e+00 6.03441834e-01 6.84620094e+00 7.79075623e-01
-1.49225545e+00 2.54325420e-01 7.09459126e-01 3.16661060e-01
2.33165964e-01 -4.31837022e-01 -1.66905716e-01 1.41105413e-01
8.05861115e-01 1.54636502e-01 -7.96286240e-02 5.31736970e-01
3.40074062e-01 -3.01100940e-01 -5.75474322e-01 8.55019927e-01
3.10291737e-01 -1.69048393e+00 -2.76504189e-01 1.60866484e-01
7.14821935e-01 3.27753425e-01 3.61995637e-01 -2.55453974e-01
-3.35348845e-01 -1.09780049e+00 -3.43926668e-01 3.07580501e-01
1.03849816e+00 -4.96272177e-01 1.42991316e+00 -8.89587477e-02
-9.25411165e-01 4.16984469e-01 -1.42830729e-01 2.57393867e-01
2.76848763e-01 6.74518943e-01 -6.25014067e-01 1.00854671e+00
5.69158196e-01 7.99856961e-01 -1.22261882e-01 1.55212975e+00
-1.18920682e-02 3.27488810e-01 1.11348383e-01 2.26333022e-01
1.58230051e-01 -2.50866532e-01 4.71513778e-01 1.06848383e+00
1.04630798e-01 2.85082698e-01 -1.63627211e-02 3.48390251e-01
5.63013144e-02 4.58784819e-01 -5.33039570e-01 1.24159470e-01
-1.19930416e-01 1.39803934e+00 -1.32188392e+00 -1.02667533e-01
-5.19804955e-01 7.55531311e-01 -3.10541779e-01 9.52209253e-03
-6.18962109e-01 -3.49275291e-01 1.39720604e-01 1.43616498e-01
-4.37336773e-01 4.90185618e-03 -4.97952104e-01 -7.02822626e-01
-3.62750232e-01 -8.71099353e-01 6.50605023e-01 -4.14353967e-01
-1.03929889e+00 8.68473649e-01 1.75124764e-01 -1.34362423e+00
-6.76764175e-02 -6.01847172e-01 -5.23767233e-01 7.90730536e-01
-1.31711698e+00 -1.03763151e+00 -4.33980376e-01 3.24405313e-01
5.02189636e-01 -4.61739659e-01 1.23680961e+00 3.50669324e-01
-4.00643587e-01 5.12300611e-01 2.57034332e-01 5.30872762e-01
8.56640339e-01 -8.85123014e-01 8.73050392e-02 2.20963642e-01
-1.25630140e-01 8.17206860e-01 3.55239928e-01 -6.00523949e-01
-9.71964896e-01 -6.46517575e-01 8.06897461e-01 -2.02024221e-01
5.06204307e-01 4.58613425e-01 -7.16896594e-01 5.14865518e-01
6.38128519e-01 2.22757146e-01 1.15531063e+00 -2.24167243e-01
3.76701683e-01 -8.99646729e-02 -1.31053722e+00 4.60884988e-01
3.51412535e-01 -2.13795349e-01 -6.02633178e-01 4.37025189e-01
1.39995262e-01 -1.07520616e+00 -1.29186237e+00 8.05305719e-01
6.89589977e-01 -7.65028417e-01 1.15944016e+00 -4.91043180e-01
4.87587035e-01 8.92489310e-03 7.04311848e-01 -1.02385902e+00
-4.85446416e-02 -3.74172091e-01 3.83400291e-01 5.03556967e-01
3.82439457e-02 -7.14031279e-01 9.26690102e-01 3.65499437e-01
-4.47983801e-01 -1.11339009e+00 -1.11981452e+00 -3.38698685e-01
3.50873560e-01 -1.71337217e-01 1.77522749e-01 1.03692269e+00
6.74161315e-03 -4.04208601e-01 6.94973245e-02 -2.60236144e-01
5.53061783e-01 -3.78948241e-01 4.39945668e-01 -1.12980866e+00
4.14016783e-01 -7.37367153e-01 -7.79538929e-01 -4.70548540e-01
9.97199491e-02 -1.11605573e+00 -1.25699401e-01 -1.98074305e+00
3.57790798e-01 -3.04651439e-01 -3.69188339e-01 3.03648561e-01
-1.69705655e-02 5.82687616e-01 -2.18198642e-01 5.01226366e-01
-6.66713789e-02 4.71600406e-02 1.53311741e+00 -2.82380968e-01
-1.76877439e-01 2.45967448e-01 -7.27041304e-01 8.87673497e-01
9.52452481e-01 -6.41699612e-01 -9.15724859e-02 -2.19882026e-01
-1.05500318e-01 1.18901737e-01 8.72875899e-02 -1.03252089e+00
3.95854056e-01 2.07638279e-01 2.81195521e-01 -6.95966542e-01
-3.13015841e-02 -1.09143293e+00 1.22553937e-01 8.67160082e-01
-2.99774170e-01 3.29748243e-01 3.09470892e-01 -3.68376337e-02
-5.82535744e-01 -4.27238584e-01 8.62385154e-01 -3.42728913e-01
-5.21634221e-01 5.69135427e-01 -6.10042751e-01 -1.25574842e-01
1.09071147e+00 -3.91036242e-01 7.11515024e-02 -2.81777859e-01
-8.76839757e-01 -1.58736780e-01 3.80668491e-02 2.66480476e-01
5.50832689e-01 -1.27326274e+00 -6.00407004e-01 -1.42124772e-01
8.31150785e-02 1.54947117e-01 6.65454209e-01 1.70296180e+00
-9.56824541e-01 6.34531915e-01 -4.69220191e-01 -7.34594524e-01
-1.32475173e+00 5.09856164e-01 7.10727632e-01 -5.22106886e-01
-7.00737238e-01 5.43949008e-01 1.21210769e-01 -5.01912951e-01
2.63448119e-01 -1.39118239e-01 -8.30953479e-01 -3.07137612e-02
3.67393166e-01 3.29291493e-01 6.31420255e-01 -9.13723052e-01
-4.31215197e-01 9.31325853e-01 -1.13653034e-01 2.25355536e-01
1.28352678e+00 7.85487294e-02 -6.54808819e-01 2.13188052e-01
1.26529706e+00 -2.09386796e-01 -3.21047395e-01 -4.99942079e-02
-2.24027112e-01 1.22620717e-01 2.59059548e-01 -6.17578089e-01
-1.64216053e+00 8.74014497e-01 1.09656596e+00 -1.52807534e-01
1.33059537e+00 3.05360300e-03 7.28534043e-01 -1.02295205e-01
1.77646622e-01 -7.97734976e-01 -1.72556594e-01 3.09953958e-01
6.37467921e-01 -1.49997270e+00 3.35852802e-01 -4.88575101e-01
-3.81313920e-01 1.41306174e+00 2.16558695e-01 -1.26760006e-01
1.03223372e+00 5.11322260e-01 5.28305113e-01 -4.91605818e-01
3.83404702e-01 -4.25763428e-02 5.46332955e-01 7.69148827e-01
9.82479632e-01 -1.61078438e-01 -8.87608290e-01 3.73968422e-01
4.14764099e-02 1.63877383e-01 1.20585278e-01 8.07971179e-01
-1.00009166e-01 -1.39171779e+00 -4.41949457e-01 5.64126849e-01
-1.09449410e+00 -2.40946233e-01 1.32765118e-02 1.09938169e+00
1.01980656e-01 5.84565639e-01 -1.82712469e-02 1.43124297e-01
1.84279665e-01 -5.62503040e-01 5.09030879e-01 -3.86522800e-01
-9.94778812e-01 9.95492488e-02 -2.68488407e-01 -3.22519809e-01
-1.20528018e+00 -3.01139593e-01 -1.27134144e+00 5.65377884e-02
-3.32911640e-01 1.71319023e-01 9.62223172e-01 1.20681524e+00
-1.16029926e-01 8.50693345e-01 1.40045390e-01 -8.31239104e-01
1.57176062e-01 -8.42184544e-01 -2.62120754e-01 1.10825971e-01
1.97659045e-01 -5.70740998e-01 -2.73462236e-02 -6.00627773e-02] | [14.469731330871582, -2.5662286281585693] |
0a566926-eaea-4af3-80e8-9d337363c573 | simplifying-sparse-expert-recommendation-by | 2208.02438 | null | https://arxiv.org/abs/2208.02438v1 | https://arxiv.org/pdf/2208.02438v1.pdf | Simplifying Sparse Expert Recommendation by Revisiting Graph Diffusion | Community Question Answering (CQA) websites have become valuable knowledge repositories where individuals exchange information by asking and answering questions. With an ever-increasing number of questions and high migration of users in and out of communities, a key challenge is to design effective strategies for recommending experts for new questions. In this paper, we propose a simple graph-diffusion expert recommendation model for CQA, that can outperform state-of-the art deep learning representatives and collaborative models. Our proposed method learns users' expertise in the context of both semantic and temporal information to capture their changing interest and activity levels with time. Experiments on five real-world datasets from the Stack Exchange network demonstrate that our approach outperforms competitive baseline methods. Further, experiments on cold-start users (users with a limited historical record) show our model achieves an average of ~ 30% performance gain compared to the best baseline method. | ['Nino Antulov-Fantulin', 'Vaibhav Krishna'] | 2022-08-04 | null | null | null | null | ['community-question-answering', 'community-question-answering'] | ['miscellaneous', 'natural-language-processing'] | [-6.19338274e-01 -2.26951301e-01 -5.56499884e-02 -1.73997834e-01
-4.60915267e-01 -7.33476281e-01 6.01196408e-01 5.06437123e-01
-4.36465681e-01 3.60434949e-01 6.68339014e-01 -1.95041060e-01
-4.22174156e-01 -9.39964175e-01 -1.75044850e-01 1.10425517e-01
-1.93537459e-01 7.44525313e-01 7.48707056e-01 -5.76027572e-01
4.20125723e-01 -2.79880520e-02 -1.03242075e+00 4.14639622e-01
1.27320492e+00 8.17829549e-01 4.23181728e-02 6.76706016e-01
-5.22088110e-01 1.39878237e+00 -5.20925105e-01 -9.06848371e-01
6.92914203e-02 -1.99663535e-01 -1.39957488e+00 -2.25403115e-01
4.41864789e-01 -1.44741699e-01 -6.99770212e-01 6.71497047e-01
3.75424057e-01 5.36591291e-01 2.15651393e-01 -9.07710373e-01
-1.29729211e+00 5.97021937e-01 -1.53239638e-01 6.84949636e-01
5.98288953e-01 -4.35038432e-02 1.59569860e+00 -6.23222113e-01
6.49620175e-01 8.15339863e-01 8.87302220e-01 4.04389948e-01
-7.43408203e-01 -3.07230532e-01 6.69362128e-01 5.04746318e-01
-9.61095691e-01 1.24253519e-01 6.16307914e-01 -5.09073615e-01
1.10697770e+00 2.52152234e-01 9.76471126e-01 8.66843164e-01
-4.85453814e-01 9.24605489e-01 6.34762108e-01 -2.14740224e-02
3.82237673e-01 -2.23353673e-02 4.13790911e-01 7.49183118e-01
1.55176029e-01 -7.39431977e-01 -8.17007661e-01 -7.35541284e-01
3.57665151e-01 5.69591820e-01 -1.77726045e-01 -5.68963438e-02
-7.22439229e-01 1.01486242e+00 6.93925560e-01 7.37734973e-01
-6.00752950e-01 1.96201429e-01 -1.20747089e-01 5.11434972e-01
6.93735838e-01 8.08666527e-01 -4.82151687e-01 -1.65338829e-01
-6.22541368e-01 4.27184790e-01 1.38178289e+00 8.63104880e-01
6.02719247e-01 -6.80443406e-01 -4.77447808e-01 8.26881528e-01
6.82353973e-01 1.93987265e-01 2.74576813e-01 -1.29389548e+00
2.70132810e-01 1.18535018e+00 4.52902675e-01 -1.27371609e+00
-2.14294091e-01 -8.45550776e-01 -5.48809111e-01 -7.70020723e-01
4.97251868e-01 -3.42528224e-01 -2.44432464e-01 1.22509944e+00
4.42405879e-01 5.85189700e-01 -3.73004973e-01 7.24148750e-01
7.45847106e-01 6.98989689e-01 1.17094505e-05 -1.70986339e-01
1.13070214e+00 -1.30489886e+00 -5.93597531e-01 -2.05781102e-01
3.97922754e-01 -5.25833249e-01 8.97967160e-01 2.36383647e-01
-9.46459949e-01 -2.13728786e-01 -3.48166436e-01 -1.59655347e-01
-4.21713889e-01 -5.10961235e-01 7.88093030e-01 3.39167506e-01
-1.34199929e+00 6.63775802e-01 -4.26139176e-01 -7.38410234e-01
5.74527621e-01 1.50927857e-01 2.66043425e-01 -2.18008012e-01
-1.52116942e+00 3.39767069e-01 -3.71783614e-01 -4.07024324e-02
-7.35099375e-01 -9.85910773e-01 5.33401174e-03 3.71007115e-01
5.53496480e-01 -1.10188925e+00 1.60141838e+00 -6.34702444e-01
-1.22262585e+00 3.15506160e-01 -4.61167321e-02 -2.40753546e-01
3.11553478e-01 -5.66147029e-01 -6.51520014e-01 7.04331249e-02
3.15847136e-02 -6.79448843e-02 5.63749909e-01 -8.93882692e-01
-1.01426697e+00 -3.98225427e-01 5.58292568e-01 8.56419802e-02
-1.08187973e+00 7.42662475e-02 -8.12495410e-01 -2.03257442e-01
-3.71139884e-01 -6.17699564e-01 -5.10714650e-01 -1.16266437e-01
1.80800125e-01 -1.04670727e+00 6.26845896e-01 -8.04700255e-01
1.78934216e+00 -1.48692453e+00 6.51692078e-02 4.07286644e-01
7.11621165e-01 4.75028187e-01 -2.59669781e-01 9.59772408e-01
7.71313667e-01 3.48398834e-01 2.27918416e-01 -4.03503001e-01
-8.22755098e-02 8.57103020e-02 -7.46407285e-02 -5.07566370e-02
-2.71905750e-01 9.94023025e-01 -1.72805667e+00 -3.92493635e-01
-5.87799251e-01 4.04046535e-01 -8.01491320e-01 4.02100861e-01
-6.24518216e-01 4.40801412e-01 -1.06716704e+00 5.58737397e-01
2.05251463e-02 -1.26885104e+00 4.37147915e-01 3.13136935e-01
2.51020253e-01 3.82001758e-01 -6.39175951e-01 1.81130683e+00
-4.16694671e-01 5.05800724e-01 1.06024891e-01 -5.44492364e-01
6.74227536e-01 1.73447251e-01 6.14806116e-01 -8.16717267e-01
-1.42159075e-01 -4.00186554e-02 -2.18665257e-01 -6.96552098e-01
5.75603783e-01 8.91462088e-01 1.79664940e-01 8.45042706e-01
-2.79029701e-02 4.55438375e-01 2.92814136e-01 1.20010781e+00
1.61055636e+00 -4.78909820e-01 -1.92803144e-01 -2.87313879e-01
5.97783267e-01 -2.07193151e-01 3.94956172e-01 1.04139948e+00
-4.62470114e-01 1.57639325e-01 3.35940033e-01 -5.42205811e-01
-5.21980822e-01 -6.92182124e-01 7.54419386e-01 1.67211294e+00
-8.43191519e-02 -6.47172809e-01 -4.07586515e-01 -1.06780338e+00
2.62168527e-01 3.95832896e-01 -7.15364635e-01 1.59002513e-01
-6.33815706e-01 -4.00704563e-01 -5.37134595e-02 5.15350223e-01
5.88632226e-01 -9.12435234e-01 1.63738310e-01 4.07207519e-01
-6.94024920e-01 -8.77268493e-01 -9.23445344e-01 -9.84684169e-01
-8.19444358e-01 -1.25771868e+00 -8.55412364e-01 -6.68528497e-01
4.46204782e-01 6.95359111e-01 1.83746147e+00 8.82622659e-01
-4.97610494e-02 1.25258946e+00 -6.27289832e-01 -6.07964210e-03
1.56857342e-01 3.63944113e-01 -1.80917054e-01 3.97950411e-01
5.86056650e-01 -9.49756086e-01 -1.52075195e+00 3.46289158e-01
-6.71412349e-01 -6.50179029e-01 -1.51663851e-02 3.35621864e-01
1.02322246e-03 -1.47280648e-01 1.06602466e+00 -1.27003443e+00
1.30382442e+00 -1.10551584e+00 -4.13599998e-01 7.56863534e-01
-1.23189938e+00 -4.04042900e-01 3.73080671e-01 -2.96208799e-01
-1.27903712e+00 -4.91131544e-01 9.99531969e-02 6.98711798e-02
4.01014626e-01 7.30318904e-01 5.42840004e-01 3.19428705e-02
9.15673912e-01 -1.43051162e-01 -5.44231534e-01 -6.75385118e-01
8.18084002e-01 6.72339380e-01 1.68382078e-01 -5.22622168e-01
6.61096632e-01 5.17585516e-01 -6.91441059e-01 -4.93419051e-01
-1.32108927e+00 -1.42389894e+00 -3.16122293e-01 -6.34196460e-01
6.38483524e-01 -8.94551754e-01 -9.74723697e-01 4.89479870e-01
-1.09495175e+00 -1.90528736e-01 -1.59563616e-01 -1.14746280e-01
1.75966084e-01 5.99913716e-01 -9.44845617e-01 -8.34042907e-01
-8.31708312e-01 -3.23623329e-01 2.88535416e-01 7.79715478e-01
-5.29894009e-02 -1.62245858e+00 6.14120305e-01 1.23998511e+00
1.19778764e+00 -3.62012774e-01 4.90188479e-01 -9.99119401e-01
-1.08291364e+00 -2.13332206e-01 -2.76405156e-01 -2.37310864e-02
2.48666450e-01 -2.35769734e-01 -6.24488890e-01 -4.39043134e-01
-3.47998291e-01 -2.69440800e-01 9.70046759e-01 -1.17134720e-01
9.96897578e-01 -4.62749600e-01 -4.04544204e-01 -1.93349808e-01
1.05360162e+00 -2.83729494e-01 2.23223925e-01 -6.05346188e-02
7.40173519e-01 4.86120790e-01 9.88295674e-02 7.30088413e-01
1.13904607e+00 1.03564329e-01 5.34309149e-01 4.59700763e-01
-3.80528439e-03 -5.39832532e-01 4.72657979e-02 1.32782912e+00
-5.18551648e-01 -6.37474179e-01 -9.34708595e-01 9.92457151e-01
-2.27396607e+00 -1.11896932e+00 -3.82144779e-01 1.93470252e+00
6.67667985e-01 -2.61310697e-01 5.30993104e-01 -5.11752307e-01
4.62478369e-01 3.11060488e-01 -7.95422733e-01 2.58953482e-01
3.12343180e-01 2.91844420e-02 7.24013820e-02 5.19887865e-01
-5.88040709e-01 5.61464667e-01 6.14357185e+00 4.04895335e-01
-2.31594145e-01 4.93626535e-01 3.25766832e-01 3.67686935e-02
-6.43098712e-01 -6.85365885e-05 -5.88446081e-01 5.12740254e-01
1.10822272e+00 -3.75839055e-01 7.48598576e-01 5.66297054e-01
-2.16430917e-01 2.95478016e-01 -8.48353863e-01 5.34529030e-01
2.06557229e-01 -1.69023311e+00 -1.38331369e-01 -1.81410715e-01
1.39253962e+00 5.36150694e-01 -1.81978106e-01 6.92708611e-01
1.04302287e+00 -4.05986190e-01 -1.20000653e-01 1.13534856e+00
-3.09998661e-01 -2.95740247e-01 6.01144731e-01 5.01882434e-01
-1.10455430e+00 -5.79095125e-01 -1.79915667e-01 -1.25673428e-01
1.78027198e-01 6.37804449e-01 -6.49902880e-01 5.20378470e-01
1.10857749e+00 1.06971538e+00 -7.82951117e-01 1.27562165e+00
-2.44237646e-01 1.27830005e+00 -1.66054249e-01 -5.59676886e-01
2.81852692e-01 -1.74942747e-01 2.61837035e-01 1.05075634e+00
2.11812854e-01 3.33331436e-01 1.26347184e-01 7.07264602e-01
-7.63059437e-01 4.06515956e-01 -2.04407603e-01 -6.02627873e-01
7.31061459e-01 1.46467173e+00 -3.12215209e-01 -3.25515389e-01
-9.38690960e-01 1.10968292e+00 8.93249869e-01 6.24824643e-01
-3.35440785e-01 -1.10427491e-01 4.96962994e-01 4.52014118e-01
4.61461395e-01 -4.71891612e-02 4.86678034e-01 -1.21648252e+00
-8.31187591e-02 -6.09937072e-01 1.07179129e+00 -4.49893326e-01
-2.24176049e+00 3.21912616e-01 -6.24749243e-01 -5.33437431e-01
-8.77127573e-02 -8.64062756e-02 -1.03157699e+00 7.65678942e-01
-1.54331493e+00 -1.14295089e+00 -4.99375552e-01 8.24046910e-01
3.41592640e-01 -2.64475673e-01 5.08423090e-01 6.27923191e-01
-1.89328492e-01 3.57892364e-01 4.09200907e-01 3.30561161e-01
5.07964730e-01 -1.36824405e+00 6.43714488e-01 5.28451204e-01
4.02506739e-01 8.36456299e-01 2.06175923e-01 -7.14865327e-01
-1.42062497e+00 -9.79437888e-01 1.35873055e+00 -1.07339156e+00
1.01298583e+00 -1.69883266e-01 -1.13046217e+00 7.21475005e-01
7.22006798e-01 -6.37685582e-02 1.10232902e+00 9.51556385e-01
-4.98190939e-01 -2.50333279e-01 -8.90808284e-01 2.76181072e-01
1.19377160e+00 -8.14972103e-01 -5.14088154e-01 9.23362434e-01
9.50924456e-01 3.12466353e-01 -1.21450818e+00 -6.72850981e-02
3.23465765e-01 -8.67071807e-01 9.16497588e-01 -9.85898316e-01
-4.46976582e-03 -2.63914198e-01 3.91682297e-01 -1.21067572e+00
-9.12639201e-01 -9.86972809e-01 -8.96449864e-01 1.19368339e+00
4.65094358e-01 -6.68022990e-01 9.60064650e-01 9.73129034e-01
3.18910450e-01 -6.24619842e-01 -4.23559725e-01 -3.12675059e-01
-5.76253869e-02 1.48417622e-01 4.67010766e-01 1.34652579e+00
3.01147014e-01 6.45710588e-01 -3.19825351e-01 1.44254148e-01
3.98223460e-01 2.45177746e-01 3.75738800e-01 -1.89759076e+00
-4.80250478e-01 -4.30272579e-01 2.87542492e-01 -1.41172051e+00
-1.69954628e-01 -8.79126966e-01 -4.28213775e-01 -2.15808797e+00
4.56187397e-01 -4.12660420e-01 -8.58307600e-01 9.39694867e-02
-5.31735539e-01 -1.57591775e-01 -5.98844104e-02 4.01632577e-01
-1.69687963e+00 4.15962875e-01 1.30183589e+00 -1.85789004e-01
-2.83603251e-01 3.78946066e-01 -8.89255047e-01 4.68968093e-01
7.03109205e-01 -3.84867430e-01 -7.45168626e-01 -8.50971103e-01
1.36509252e+00 -1.21014202e-02 -1.57547165e-02 -6.54655039e-01
9.83309448e-01 -8.92462656e-02 -2.02641547e-01 -4.09066856e-01
4.25812863e-02 -6.18195891e-01 -2.48586480e-02 2.40350068e-01
-6.72146201e-01 1.17710933e-01 -6.16610885e-01 1.46032357e+00
1.36154462e-02 -1.83920816e-01 6.31323159e-02 -3.99136961e-01
-6.56715035e-01 7.69631624e-01 -3.44576687e-01 5.06424904e-01
5.01933634e-01 5.60224116e-01 -6.87831879e-01 -1.03405654e+00
-7.34841585e-01 1.06945598e+00 1.32472906e-02 8.78300011e-01
3.47495496e-01 -1.22276711e+00 -8.56355846e-01 -5.37082851e-01
1.63512155e-01 -1.31604493e-01 6.37597382e-01 5.68131030e-01
-1.79309204e-01 3.33348870e-01 4.71416444e-01 -1.50769725e-01
-8.31181347e-01 4.30537581e-01 4.36232388e-01 -7.25604713e-01
-1.57131448e-01 1.14182580e+00 -5.24411380e-01 -6.87831700e-01
2.71594584e-01 3.73184308e-02 -6.61374688e-01 9.45828035e-02
7.60247171e-01 8.56751323e-01 -2.71922983e-02 -5.55397831e-02
-1.09132968e-01 -3.14677283e-02 -4.27922130e-01 3.47309560e-01
1.39270937e+00 -3.10407549e-01 -2.84082532e-01 1.50033936e-01
8.45371246e-01 1.10658593e-01 -7.48013258e-01 -9.67431545e-01
3.14495176e-01 -4.63654637e-01 -3.23518887e-02 -1.10158587e+00
-1.26747167e+00 6.49678290e-01 3.19866687e-01 9.11183476e-01
7.31058657e-01 3.79749089e-01 1.09014463e+00 8.26534271e-01
1.37626857e-01 -1.25439537e+00 6.62115812e-01 5.83974183e-01
6.29743040e-01 -1.08162200e+00 -1.46513611e-01 1.03088096e-02
-2.88236380e-01 6.94667339e-01 8.53480518e-01 3.15728821e-02
1.06739604e+00 -6.61551654e-01 1.70364324e-02 -6.55049801e-01
-1.22837031e+00 -2.41570294e-01 3.91288638e-01 3.56333077e-01
4.67625082e-01 -1.59607768e-01 -2.76249677e-01 6.59502506e-01
3.78038913e-01 1.14930600e-01 2.75854588e-01 8.65688324e-01
-6.79387748e-01 -1.13133061e+00 3.77599418e-01 6.51180208e-01
-5.55367410e-01 -1.09009808e-02 -5.95155954e-01 8.93035680e-02
-2.63196856e-01 1.49712658e+00 -1.10075004e-01 -3.88231099e-01
3.38762105e-01 1.38609767e-01 -2.68359613e-02 -4.40320581e-01
-1.17792404e+00 -5.81254363e-01 2.73391843e-01 -7.13431299e-01
-6.33161247e-01 -7.02753246e-01 -9.44798172e-01 -4.03253645e-01
-3.56895179e-01 5.93919277e-01 2.04028845e-01 7.72434533e-01
9.33640718e-01 3.00400287e-01 6.91643775e-01 2.57293612e-01
-4.33481604e-01 -1.03377569e+00 -3.40120018e-01 6.57810152e-01
-1.03411283e-02 -4.66910005e-02 -3.45480531e-01 -2.37064168e-01] | [11.446191787719727, 7.925364017486572] |
31f012a8-4f7c-486c-a73e-b92ff2471957 | bn-drishti-bangla-document-recognition | 2306.09351 | null | https://arxiv.org/abs/2306.09351v1 | https://arxiv.org/pdf/2306.09351v1.pdf | BN-DRISHTI: Bangla Document Recognition through Instance-level Segmentation of Handwritten Text Images | Handwriting recognition remains challenging for some of the most spoken languages, like Bangla, due to the complexity of line and word segmentation brought by the curvilinear nature of writing and lack of quality datasets. This paper solves the segmentation problem by introducing a state-of-the-art method (BN-DRISHTI) that combines a deep learning-based object detection framework (YOLO) with Hough and Affine transformation for skew correction. However, training deep learning models requires a massive amount of data. Thus, we also present an extended version of the BN-HTRd dataset comprising 786 full-page handwritten Bangla document images, line and word-level annotation for segmentation, and corresponding ground truths for word recognition. Evaluation on the test portion of our dataset resulted in an F-score of 99.97% for line and 98% for word segmentation. For comparative analysis, we used three external Bangla handwritten datasets, namely BanglaWriting, WBSUBNdb_text, and ICDAR 2013, where our system outperformed by a significant margin, further justifying the performance of our approach on completely unseen samples. | ['Mohammad Khairul Islam', 'Md. Ataur Rahman', 'Nazifa Tabassum', 'Sheikh Mohammad Jubaer'] | 2023-05-31 | null | null | null | null | ['handwriting-recognition', 'handwritten-line-segmentation', 'handwritten-word-segmentation'] | ['computer-vision', 'computer-vision', 'computer-vision'] | [-1.00858800e-01 -2.70412862e-01 8.05286244e-02 -4.94324952e-01
-8.47608984e-01 -8.14788580e-01 4.63622808e-01 -1.96957797e-01
-7.11883426e-01 5.85275531e-01 -2.23151371e-01 -4.13122803e-01
3.99688073e-02 -6.71084046e-01 -5.27631283e-01 -7.93633342e-01
4.43010390e-01 8.10392082e-01 2.96814561e-01 -2.26090610e-01
5.95803499e-01 8.54325950e-01 -6.71070039e-01 4.26379025e-01
8.55216920e-01 7.83892453e-01 5.83965480e-02 8.33670974e-01
-2.36768529e-01 5.16907454e-01 -1.08842587e+00 -9.65430379e-01
3.04507285e-01 -5.17135382e-01 -8.37910891e-01 5.70744336e-01
7.79021025e-01 -7.18270361e-01 -5.90039313e-01 6.50365651e-01
8.21636140e-01 -2.72921801e-01 6.98977232e-01 -6.18050516e-01
-8.55810463e-01 6.91603541e-01 -8.95561278e-01 1.56907439e-02
-2.25532934e-01 8.02723169e-02 9.10168707e-01 -9.94303167e-01
6.84870481e-01 8.78035069e-01 9.91611838e-01 4.17003930e-01
-7.96246171e-01 -3.46456051e-01 -1.96712628e-01 1.89838544e-01
-1.24982381e+00 -2.13744923e-01 6.17827058e-01 -4.34814334e-01
1.07194018e+00 1.20525971e-01 5.25771201e-01 1.06419015e+00
-2.78543644e-02 1.34688354e+00 1.32150698e+00 -7.88606942e-01
1.06864490e-01 -1.26291931e-01 5.58505356e-01 7.00188935e-01
3.13270688e-01 -5.17608106e-01 -4.29501325e-01 5.38418770e-01
7.97326207e-01 -5.10822058e-01 -7.84854740e-02 7.44624510e-02
-1.01070797e+00 7.68692970e-01 -2.95034256e-02 3.16078395e-01
-1.17566235e-01 -3.09098482e-01 5.76245904e-01 7.26560550e-03
1.57482192e-01 2.85346746e-01 -3.43581706e-01 -2.88893014e-01
-1.41744411e+00 3.69331270e-01 8.70278537e-01 1.16235673e+00
2.93471366e-01 2.54543692e-01 -3.57151270e-01 1.18174767e+00
2.11369440e-01 7.22099423e-01 6.60634398e-01 -2.78657377e-01
7.48564899e-01 4.07878041e-01 -2.33815700e-01 -8.24713230e-01
-4.74739969e-01 -2.95897186e-01 -7.61774957e-01 2.05541514e-02
9.50622797e-01 -1.92239225e-01 -1.52100730e+00 7.62623489e-01
-2.40445346e-01 -8.36673439e-01 3.97107229e-02 1.16679418e+00
7.27012098e-01 6.82120264e-01 -5.00754654e-01 1.02808185e-01
1.30940759e+00 -1.31848538e+00 -8.90194237e-01 -2.06191316e-01
6.40816331e-01 -1.15820956e+00 1.03572273e+00 9.00680006e-01
-8.95727158e-01 -5.25503039e-01 -1.34903109e+00 -4.44432348e-01
-3.59475374e-01 8.14103782e-01 4.95720863e-01 9.98741329e-01
-6.93819702e-01 2.22939923e-01 -7.55839109e-01 -5.12619138e-01
7.64605105e-01 2.61130363e-01 -3.59045684e-01 -1.14745602e-01
-7.00224936e-01 9.45619762e-01 5.48969388e-01 5.78203738e-01
-5.75052679e-01 -1.90496296e-01 -4.96068090e-01 -4.34815884e-01
2.39341140e-01 3.69254321e-01 1.12493742e+00 -3.86239469e-01
-1.49049997e+00 1.22003472e+00 2.57954568e-01 -4.67010051e-01
1.22468793e+00 -5.32512426e-01 -3.37046146e-01 1.15591861e-01
-1.68948919e-01 5.40003240e-01 6.78149283e-01 -1.16727877e+00
-6.71018779e-01 -6.30357325e-01 -4.28370148e-01 -1.21078737e-01
-2.56203830e-01 -1.45422239e-02 -9.30768430e-01 -9.36055958e-01
2.88730532e-01 -6.04694724e-01 1.91112161e-01 -2.34114006e-01
-7.59170055e-01 -1.51430190e-01 1.15171993e+00 -1.44325280e+00
1.18491697e+00 -2.21175838e+00 -2.49116987e-01 2.88882971e-01
-1.35715589e-01 5.94286859e-01 -2.29315534e-01 3.64130974e-01
3.10410202e-01 -8.11992437e-02 -5.53872049e-01 -2.01021329e-01
8.41351524e-02 2.83535808e-01 -4.13339436e-01 6.71205699e-01
3.89689028e-01 9.68298435e-01 -3.81115824e-01 -6.36391819e-01
6.67640716e-02 2.57170349e-01 -6.73476458e-02 2.68261917e-02
-1.01100676e-01 -1.28812671e-01 6.59027025e-02 1.12585127e+00
7.97023594e-01 3.53956223e-01 8.37943330e-02 -2.47659445e-01
-2.43964359e-01 -1.73813626e-01 -1.24304318e+00 1.37832189e+00
-7.26046115e-02 1.08553660e+00 -1.88409880e-01 -7.24849701e-01
1.61174023e+00 -1.09153531e-01 -1.04023181e-01 -1.02679062e+00
2.45133311e-01 7.61700153e-01 1.83689907e-01 -5.80093622e-01
1.03192246e+00 3.19069475e-01 -5.62388264e-02 3.98265749e-01
1.86550155e-01 -4.39524114e-01 5.48293233e-01 -6.42050207e-02
5.04527986e-01 2.65375972e-01 -1.68549404e-01 -2.13571206e-01
3.31460774e-01 3.42023432e-01 2.70689905e-01 9.29348648e-01
-3.23808223e-01 1.19067085e+00 6.44064665e-01 -4.90487427e-01
-1.63245690e+00 -6.34772301e-01 -4.77599055e-01 8.07335973e-01
-1.86190069e-01 7.63921291e-02 -1.16256821e+00 -6.83898389e-01
-2.24438682e-02 5.64436078e-01 -5.50454557e-01 4.57326055e-01
-9.63587999e-01 -1.04518938e+00 1.18717015e+00 8.60642254e-01
1.01704764e+00 -1.25774276e+00 -3.89079243e-01 2.12892458e-01
1.55600414e-01 -1.30646014e+00 -4.91384625e-01 4.68569398e-01
-6.76836193e-01 -9.49833691e-01 -1.31771278e+00 -1.18984616e+00
5.78822315e-01 -2.28036374e-01 6.10649943e-01 -3.21071178e-01
-6.13540828e-01 -2.25418448e-01 -3.03762287e-01 -3.93267900e-01
-3.87222499e-01 3.11692327e-01 -2.91734010e-01 -1.45454675e-01
5.08499682e-01 4.50843841e-01 -9.57590863e-02 3.44069391e-01
-8.88921380e-01 -7.23112002e-02 8.50655079e-01 9.80114937e-01
4.69937444e-01 -7.98236802e-02 3.51252556e-01 -8.82024169e-01
7.70169675e-01 3.58829409e-01 -7.58769155e-01 5.16141772e-01
-3.27577710e-01 -2.87862360e-01 4.01988357e-01 -1.96104124e-01
-1.01587200e+00 -7.29857385e-02 -4.46685731e-01 2.15377495e-01
-2.27540016e-01 3.35028380e-01 -2.42276534e-01 7.11117014e-02
5.56482136e-01 4.66670692e-01 -1.23604305e-01 -5.59806049e-01
2.70212322e-01 1.44664037e+00 1.04791105e+00 -3.69750112e-01
3.95553857e-01 3.19044560e-01 -3.38768303e-01 -1.26763558e+00
-7.13713169e-01 -3.05889606e-01 -1.31079388e+00 -1.34704858e-01
8.17557871e-01 -4.33826894e-01 -4.32222992e-01 1.66698420e+00
-1.20870781e+00 -5.75835109e-01 -1.60638243e-01 3.15950841e-01
-4.35824662e-01 6.29556954e-01 -8.75125110e-01 -8.16221595e-01
-4.50162590e-01 -1.19430399e+00 9.26335216e-01 2.46712491e-01
-1.04011871e-01 -7.83471286e-01 -1.18386358e-01 7.35354304e-01
9.58869308e-02 1.65127680e-01 9.91694212e-01 -1.01089025e+00
-1.79485559e-01 -5.53019881e-01 -5.58615506e-01 8.13780844e-01
5.21032698e-03 3.12992543e-01 -1.08047378e+00 -2.02494800e-01
-3.36730361e-01 -6.03758156e-01 9.92459953e-01 1.02963150e-01
9.85347152e-01 -7.77482018e-02 3.24499577e-01 4.40012455e-01
1.19055068e+00 3.32528561e-01 8.23647082e-01 7.46170640e-01
9.52737987e-01 3.97722125e-01 6.64440930e-01 5.73324084e-01
2.25958914e-01 4.43956316e-01 -8.62779096e-02 -3.28178376e-01
-4.07962978e-01 -3.56915370e-02 1.14079103e-01 9.48129356e-01
-1.89322725e-01 -4.41784829e-01 -1.28718269e+00 6.76016867e-01
-1.48441315e+00 -4.23540115e-01 -6.35995209e-01 1.78748322e+00
1.12969923e+00 3.11949998e-01 2.96152949e-01 6.55426145e-01
6.82130098e-01 1.62800804e-01 -4.32349414e-01 -8.21984172e-01
-7.26686656e-01 1.70815825e-01 8.12215984e-01 3.43329757e-01
-1.40866232e+00 1.33075190e+00 6.03045177e+00 1.12486744e+00
-1.26304591e+00 -3.66333455e-01 8.92166615e-01 4.76729810e-01
2.76477426e-01 -6.71966553e-01 -1.16985607e+00 2.74995863e-01
4.93618459e-01 5.94448328e-01 1.53860316e-01 5.95021188e-01
-1.35681732e-02 -1.82819054e-01 -8.20810497e-01 9.60813165e-01
5.55268943e-01 -1.16630435e+00 1.06973462e-01 -4.28636186e-03
8.64816189e-01 -1.16362035e-01 1.54033443e-03 1.76858962e-01
-8.37151557e-02 -1.12935996e+00 9.96258676e-01 2.87137598e-01
8.62537146e-01 -7.29032695e-01 1.06490707e+00 1.78265944e-01
-2.80776232e-01 2.48771980e-01 -4.88530159e-01 2.45502576e-01
-7.46372566e-02 2.98274755e-01 -9.69554484e-01 3.75855029e-01
5.90119481e-01 3.86062682e-01 -8.46854687e-01 1.03914618e+00
-4.46853429e-01 9.17001724e-01 -2.73585767e-01 -1.91837624e-01
8.06104541e-01 -2.21988603e-01 4.69622836e-02 1.86211455e+00
8.55597630e-02 -1.74859539e-01 -2.10591510e-01 6.03065014e-01
-1.94532126e-01 1.57934397e-01 -6.49615948e-04 -3.32683146e-01
1.07709438e-01 1.29597127e+00 -1.16021919e+00 -1.13769449e-01
-2.70388275e-01 1.31823206e+00 6.15251064e-02 3.22565109e-01
-7.12488294e-01 -9.55019951e-01 -2.15825126e-01 -2.35834226e-01
6.88551784e-01 -5.93941867e-01 -8.04099798e-01 -8.04915309e-01
2.53043473e-01 -1.13458598e+00 1.43258750e-01 -3.39247257e-01
-1.24708188e+00 5.92988551e-01 -5.41774869e-01 -6.60126805e-01
1.17049374e-01 -1.27682948e+00 -3.46920192e-01 9.09544051e-01
-1.40412998e+00 -1.54715085e+00 -3.77604157e-01 3.30736846e-01
7.98680127e-01 -5.40290594e-01 5.94283581e-01 2.75627613e-01
-7.56348968e-01 1.04244149e+00 5.57113230e-01 1.05297112e+00
7.75732875e-01 -1.43812263e+00 7.82321811e-01 9.34194207e-01
3.09792429e-01 1.36050239e-01 3.33931893e-01 -5.61334193e-01
-1.37807322e+00 -9.10023689e-01 9.51743066e-01 -2.96393931e-01
6.36510372e-01 -7.00877190e-01 -1.02198148e+00 6.03473127e-01
1.42795831e-01 -1.67382896e-01 5.44089139e-01 -3.46310526e-01
-2.72770554e-01 4.51981463e-02 -1.20555699e+00 4.49112535e-01
4.63794321e-01 -3.18231285e-01 -6.91843390e-01 3.31919730e-01
-1.05155252e-01 -6.21661901e-01 -7.58845389e-01 1.63449094e-01
7.17963457e-01 -8.54457438e-01 3.85115057e-01 -4.49065506e-01
6.91349387e-01 -9.71548855e-02 -6.45925850e-02 -8.42845201e-01
-5.73047958e-02 -2.46988744e-01 9.01176035e-02 1.54170275e+00
4.88760173e-01 -2.71769553e-01 9.00000691e-01 3.41018915e-01
-2.54871428e-01 -7.89761066e-01 -6.69298291e-01 -9.33557749e-01
4.43420082e-01 -3.01035404e-01 2.60826021e-01 9.29814458e-01
-4.01511043e-01 -4.54986058e-02 -2.94881433e-01 -1.17590679e-02
5.92253327e-01 1.26201168e-01 6.82999611e-01 -6.08690798e-01
-1.46168619e-01 -5.91983736e-01 -5.16137362e-01 -1.12498724e+00
-1.64416328e-01 -6.60517335e-01 2.99548596e-01 -1.62668753e+00
1.47774205e-01 -2.18512267e-01 3.64647806e-01 7.54003167e-01
1.63154319e-01 8.36570323e-01 3.15254509e-01 3.94034386e-01
-3.35058391e-01 1.76172361e-01 1.30631185e+00 -4.20898467e-01
-5.87598532e-02 -2.09982947e-01 -1.23820014e-01 6.94179416e-01
7.46946514e-01 6.95191175e-02 3.39055568e-01 -8.71610761e-01
-2.47216836e-01 -3.72069985e-01 5.69677614e-02 -7.50358462e-01
2.10662678e-01 1.93585575e-01 9.42859590e-01 -1.23275197e+00
-5.53406477e-02 -3.68824244e-01 -8.61319065e-01 5.65387487e-01
-2.54363447e-01 -1.53233439e-01 4.34735268e-01 -1.45464420e-01
-2.86288947e-01 -6.09448612e-01 9.37385917e-01 1.75193012e-01
-9.78486180e-01 -1.57706067e-01 -5.21083891e-01 5.22567444e-02
1.00185812e+00 -7.95206964e-01 -5.22817969e-01 1.16573520e-01
-3.87198925e-01 2.32627522e-02 1.53256893e-01 4.53156441e-01
5.46983600e-01 -8.56528640e-01 -1.04636836e+00 3.93210560e-01
6.48929924e-02 2.00216368e-01 -8.34690854e-02 5.91471910e-01
-1.50873458e+00 6.77209020e-01 -5.16828895e-01 -6.27550483e-01
-1.10209084e+00 -6.40199184e-02 2.31895134e-01 -2.12501083e-02
-6.12609863e-01 9.83184814e-01 -4.71282750e-01 -5.15417099e-01
4.02022332e-01 -2.65810519e-01 8.53299350e-03 3.29205036e-01
3.02504689e-01 7.08831131e-01 5.54048002e-01 -6.06732428e-01
-3.39295536e-01 5.91132045e-01 -6.40650392e-01 -2.07222462e-01
1.34500790e+00 2.91332662e-01 -7.79568925e-02 3.21090430e-01
8.92165899e-01 8.01195651e-02 -1.22384608e+00 9.24207922e-03
3.25362861e-01 -5.18092394e-01 -7.58011267e-02 -1.30807233e+00
-1.03098345e+00 1.11189640e+00 5.17857850e-01 -7.36396015e-02
7.59019732e-01 -2.78361559e-01 8.08608294e-01 8.11734021e-01
7.83151463e-02 -1.70660925e+00 3.65956835e-02 9.21757281e-01
1.09886932e+00 -1.25681603e+00 3.51771154e-02 5.76390512e-02
-7.37448215e-01 1.71040380e+00 6.66898787e-01 -4.15410101e-02
-3.57506499e-02 5.28938413e-01 6.08143032e-01 8.53125304e-02
3.13345969e-01 1.27272516e-01 9.99733955e-02 5.06726205e-01
5.63556433e-01 -9.18482095e-02 -5.15576422e-01 6.81629837e-01
-4.73224878e-01 -2.46539012e-01 6.79912627e-01 1.09032595e+00
-2.59170741e-01 -9.26801503e-01 -6.38803661e-01 4.11446601e-01
-5.91329396e-01 8.54995623e-02 -8.18840504e-01 1.09143710e+00
-1.21626288e-01 4.96965408e-01 2.67330855e-01 2.51173433e-02
4.15108353e-01 -8.66326224e-03 6.60780489e-01 -1.67950198e-01
-5.48944652e-01 2.57475734e-01 -7.55454153e-02 1.38548836e-01
1.99497044e-02 -7.33914435e-01 -1.38740170e+00 -2.39564627e-01
-5.70219219e-01 -3.19182426e-01 9.65310574e-01 9.72499728e-01
-3.69373053e-01 4.62987602e-01 3.20945501e-01 -4.88343745e-01
-8.61644685e-01 -1.31180489e+00 -1.00696123e+00 2.77731508e-01
6.41350672e-02 2.14541182e-02 1.86218485e-01 3.30283046e-01] | [11.854950904846191, 2.575085163116455] |
3eda8985-f3c4-4577-9e2c-7acacef602d9 | how-do-languages-influence-each-other | 2305.13286 | null | https://arxiv.org/abs/2305.13286v1 | https://arxiv.org/pdf/2305.13286v1.pdf | How do languages influence each other? Studying cross-lingual data sharing during LLM fine-tuning | Multilingual large language models (MLLMs) are jointly trained on data from many different languages such that representation of individual languages can benefit from other languages' data. Impressive performance on zero-shot cross-lingual transfer shows that these models are capable of exploiting data from other languages. Yet, it remains unclear to what extent, and under which conditions, languages rely on each other's data. In this study, we use TracIn (Pruthi et al., 2020), a training data attribution (TDA) method, to retrieve the most influential training samples seen during multilingual fine-tuning for a particular test language. This allows us to analyse cross-lingual sharing mechanisms of MLLMs from a new perspective. While previous work studied cross-lingual sharing at the level of model parameters, we present the first approach to study cross-lingual sharing at the data level. We find that MLLMs rely on data from multiple languages from the early stages of fine-tuning and that this reliance gradually increases as fine-tuning progresses. We further study how different fine-tuning languages influence model performance on a given test language and find that they can both reinforce and complement the knowledge acquired from data of the test language itself. | ['Ekaterina Shutova', 'Dan Garrette', 'Rochelle Choenni'] | 2023-05-22 | null | null | null | null | ['zero-shot-cross-lingual-transfer', 'cross-lingual-transfer'] | ['natural-language-processing', 'natural-language-processing'] | [-3.74247879e-01 -1.19813956e-01 -4.75928426e-01 -3.88082355e-01
-1.08445442e+00 -8.40538561e-01 1.02322030e+00 1.31823927e-01
-7.58147120e-01 9.27237093e-01 3.65422130e-01 -2.50808328e-01
1.06372327e-01 -5.86451888e-01 -9.16263998e-01 -4.48580444e-01
2.03956366e-01 7.99510181e-01 2.66127288e-01 -5.96152127e-01
-1.42588332e-01 2.16178864e-01 -1.30547523e+00 4.34520602e-01
9.07466829e-01 1.43551659e-02 3.36040109e-01 4.26832259e-01
-2.54366279e-01 4.50637192e-01 -5.50932467e-01 -5.98825455e-01
2.31579617e-01 -4.68594909e-01 -7.15225875e-01 -3.58802348e-01
6.10348940e-01 2.23146200e-01 1.89414136e-02 9.80726898e-01
5.34452140e-01 -9.75584909e-02 7.04746902e-01 -9.35754001e-01
-8.52484047e-01 1.23889840e+00 -3.66317540e-01 2.36399040e-01
-9.54690948e-02 8.04821178e-02 8.55327368e-01 -9.68448937e-01
9.46626425e-01 1.45714307e+00 8.86715055e-01 5.58118761e-01
-1.55296016e+00 -9.07078087e-01 9.49724466e-02 -1.00331336e-01
-1.60393953e+00 -8.98083985e-01 4.51744080e-01 -7.00945616e-01
8.82691324e-01 -5.70836030e-02 2.94894695e-01 1.16589069e+00
1.35383070e-01 6.61408782e-01 1.65884280e+00 -7.69892991e-01
-2.60522872e-01 9.09885943e-01 9.23476145e-02 4.98816222e-01
1.33690909e-01 3.73653144e-01 -7.09136009e-01 -1.64912611e-01
3.21047664e-01 -6.07469797e-01 -1.08511768e-01 -4.23401743e-01
-1.12960458e+00 1.10049963e+00 4.27999049e-01 9.09070194e-01
7.85806254e-02 5.46904840e-02 4.22278881e-01 8.20882857e-01
7.40396559e-01 5.64337194e-01 -8.90237153e-01 3.68926600e-02
-1.07303584e+00 -1.16170436e-01 7.17061460e-01 7.03899145e-01
1.08036172e+00 -8.34220797e-02 6.45168424e-02 1.17933333e+00
4.30650920e-01 4.26040620e-01 8.97766948e-01 -4.16638136e-01
5.84823608e-01 3.34446460e-01 -3.99065226e-01 -3.15386444e-01
5.85027114e-02 -4.68835711e-01 -4.68663990e-01 2.90988773e-01
6.17846668e-01 -1.17237501e-01 -4.41414833e-01 2.22708774e+00
1.34131879e-01 -1.23746343e-01 2.79266596e-01 3.99759054e-01
2.26223096e-01 4.26514387e-01 3.72420907e-01 -1.08953454e-01
1.10440445e+00 -8.33833158e-01 -3.35104555e-01 -2.65581310e-01
1.27106726e+00 -1.14401376e+00 1.18519247e+00 9.81873795e-02
-8.44742239e-01 -7.69001663e-01 -8.99424851e-01 -1.05811141e-01
-7.95606673e-01 -1.06618643e-01 3.35193753e-01 7.28912413e-01
-1.20597792e+00 6.01699114e-01 -5.17651916e-01 -7.28451073e-01
2.47311696e-01 1.60523161e-01 -4.81717795e-01 -1.82114244e-01
-1.63247037e+00 1.33434951e+00 4.50898170e-01 -4.81551260e-01
-9.90395069e-01 -9.68321085e-01 -5.93345702e-01 -4.30929691e-01
6.11481220e-02 -4.93936747e-01 9.67328548e-01 -1.30788732e+00
-1.19858420e+00 1.28732872e+00 -7.80800506e-02 -3.47192854e-01
5.83004117e-01 -1.94855154e-01 -4.54259604e-01 -7.33282387e-01
2.73833454e-01 7.17072725e-01 6.41176283e-01 -1.32666564e+00
-4.54413593e-01 -3.37278336e-01 -1.41551688e-01 2.13359728e-01
-4.00106907e-01 3.30689013e-01 -3.69159192e-01 -6.51810825e-01
-6.44997299e-01 -1.09166169e+00 9.72628742e-02 -3.66795510e-01
-5.04996739e-02 -3.74761075e-01 4.65270877e-01 -6.46261930e-01
1.29309535e+00 -2.13957381e+00 2.72860825e-01 1.00869022e-01
-2.15409044e-02 4.89332527e-01 -4.00873452e-01 6.36110485e-01
-8.83930642e-03 4.55478519e-01 -1.91499680e-01 -4.76238698e-01
2.53464542e-02 3.29129547e-01 -2.49005601e-01 4.82513994e-01
7.42918625e-02 9.57399607e-01 -8.29458535e-01 -4.86593425e-01
3.40739861e-02 5.33748448e-01 -6.74742043e-01 4.06195484e-02
-3.78250591e-02 5.31856060e-01 -1.63179278e-01 2.73423195e-01
3.32616508e-01 4.72423108e-03 2.63194919e-01 -1.38272002e-01
-3.61772299e-01 3.20269376e-01 -7.75974333e-01 1.77986610e+00
-8.16246808e-01 7.04414010e-01 -3.27828750e-02 -5.53710699e-01
8.98617566e-01 4.08626705e-01 -1.36150837e-01 -7.73633599e-01
-4.62481044e-02 7.87864625e-01 4.07208711e-01 -7.12959915e-02
4.95708287e-01 -5.22683144e-01 -3.65264922e-01 8.93364072e-01
3.24401200e-01 -7.77973980e-02 1.38881370e-01 1.73041686e-01
4.90386367e-01 2.07059875e-01 4.72556233e-01 -7.00684190e-01
5.12227118e-01 5.27192056e-02 2.42437944e-01 6.08535171e-01
-2.17500299e-01 1.81886535e-02 -8.56096298e-02 -1.30159125e-01
-1.31637800e+00 -1.06231773e+00 -5.85661888e-01 1.93588519e+00
-4.35334742e-01 -2.99596220e-01 -6.48954988e-01 -6.54080153e-01
3.62400562e-01 9.82269704e-01 -7.99867392e-01 -3.57170582e-01
-5.12152970e-01 -6.87430322e-01 8.61963451e-01 1.56966954e-01
-2.89609227e-02 -9.75794137e-01 7.95732811e-02 2.21423745e-01
-3.43360044e-02 -7.30595529e-01 -6.27210259e-01 3.66781712e-01
-7.84573257e-01 -7.18180537e-01 -7.92536438e-01 -6.97680354e-01
5.46789110e-01 -2.04675142e-02 1.51597130e+00 -4.95954463e-03
1.92473065e-02 3.57809216e-01 -9.99604240e-02 -3.83695930e-01
-1.11609566e+00 5.42479515e-01 4.81129557e-01 -1.53167635e-01
6.48460329e-01 -4.55516845e-01 2.82662809e-01 4.13468748e-01
-6.39611781e-01 -4.27945890e-02 4.91167963e-01 5.47100186e-01
4.10550714e-01 -1.94664985e-01 6.42405748e-01 -1.31746840e+00
8.53290737e-01 -6.84692502e-01 -4.68784302e-01 6.48794413e-01
-6.54442430e-01 4.68042344e-01 3.53957713e-01 -5.33046663e-01
-1.04952133e+00 -3.57512891e-01 2.27148280e-01 -4.02170867e-01
5.30570224e-02 6.18363321e-01 7.47588798e-02 -1.55869976e-01
9.31267440e-01 1.07734017e-01 -2.11075678e-01 -8.09328020e-01
6.83466256e-01 7.39989579e-01 1.81546643e-01 -9.35410261e-01
8.34213257e-01 2.43420992e-02 -6.38798773e-01 -5.77400267e-01
-9.22707021e-01 -2.73630410e-01 -1.19536686e+00 -1.56161547e-01
6.99894369e-01 -1.34025478e+00 -3.09985932e-02 3.27303946e-01
-9.86338556e-01 -6.94654047e-01 -3.06606263e-01 5.23328185e-01
-3.44437659e-01 -1.29706860e-01 -4.64370608e-01 -4.31544513e-01
6.28599077e-02 -1.09044898e+00 7.42548108e-01 -2.03335434e-01
-4.65141892e-01 -1.61938000e+00 7.29888678e-01 3.14669579e-01
5.79869747e-01 -4.21548784e-01 1.00045562e+00 -9.67939258e-01
-3.63118619e-01 2.23957002e-01 6.83675893e-03 5.13044834e-01
1.99972913e-01 -5.48014902e-02 -1.22255027e+00 -7.27225840e-01
-1.91331729e-01 -7.36021399e-01 8.33922267e-01 -3.25281098e-02
3.19464386e-01 -6.12257794e-02 -3.03413451e-01 4.34854925e-01
1.50955558e+00 -4.00168419e-01 1.77899092e-01 4.37312275e-01
8.10761333e-01 9.03044701e-01 3.29209745e-01 -2.70105243e-01
4.61770028e-01 7.63663292e-01 -2.80088156e-01 -6.99856579e-02
-5.27214229e-01 -4.87550497e-01 8.22682202e-01 1.70833170e+00
5.45008965e-02 -1.93335954e-02 -1.05241883e+00 7.85818458e-01
-1.53044331e+00 -6.04714215e-01 6.75295144e-02 2.58308172e+00
1.53052115e+00 2.89446954e-02 -1.28088683e-01 -5.30880868e-01
7.21917450e-01 -8.58336017e-02 -4.20084476e-01 -5.56133211e-01
-4.86544698e-01 1.95861638e-01 7.48022854e-01 1.01880896e+00
-7.00458467e-01 1.44900012e+00 6.74613810e+00 1.05563021e+00
-1.06234658e+00 5.15598178e-01 3.41889769e-01 -9.02886242e-02
-6.84672773e-01 5.61315380e-02 -1.11654854e+00 3.19031656e-01
1.39303923e+00 -5.22459865e-01 5.60327530e-01 4.04087812e-01
-1.62671357e-01 -2.52266694e-02 -1.24938166e+00 3.36002290e-01
3.29260342e-02 -8.87434304e-01 1.75324202e-01 2.88452655e-01
1.12080967e+00 8.09071720e-01 4.59710285e-02 6.95585668e-01
1.04360795e+00 -9.52809155e-01 7.13595390e-01 6.39575243e-01
8.45142663e-01 -6.75338030e-01 3.80240500e-01 5.18377841e-01
-9.24659729e-01 4.10561383e-01 -4.83495325e-01 2.24050865e-01
5.03781885e-02 3.19206178e-01 -9.85693038e-01 3.80480170e-01
4.11611080e-01 6.91856325e-01 -1.03055406e+00 3.58865887e-01
-1.97873577e-01 8.58067870e-01 -1.58047199e-01 4.80991721e-01
2.30604187e-01 4.42671143e-02 3.77130836e-01 1.36495233e+00
2.98954874e-01 -6.89120591e-01 1.49147183e-01 7.72505820e-01
-2.52743095e-01 4.35696512e-01 -8.83947790e-01 -2.63595521e-01
4.72079247e-01 8.80373001e-01 -6.81835636e-02 -5.23512840e-01
-5.59749007e-01 7.74906337e-01 9.20338452e-01 2.55120635e-01
-5.22340059e-01 1.50797412e-01 6.94839537e-01 9.23886001e-02
3.89044844e-02 -2.71377385e-01 7.61056542e-02 -1.30256784e+00
-4.50150073e-01 -1.00994563e+00 3.79921347e-01 -7.22291827e-01
-1.62954307e+00 5.35581231e-01 -7.22076371e-02 -8.98864031e-01
-3.48966360e-01 -3.65309119e-01 -1.30117297e-01 1.39377630e+00
-1.64571905e+00 -1.44415522e+00 4.40121353e-01 9.31358397e-01
4.25694555e-01 -5.93594074e-01 9.36909616e-01 4.66059476e-01
-2.98917115e-01 9.01619971e-01 6.15237236e-01 1.11786626e-01
1.45572305e+00 -1.01543558e+00 2.34532967e-01 6.39268875e-01
6.28265202e-01 8.97433996e-01 5.12901425e-01 -6.89520061e-01
-9.67366040e-01 -1.08326805e+00 1.56064880e+00 -9.36865449e-01
1.07375407e+00 -3.51688892e-01 -1.28151131e+00 1.00107169e+00
5.25769472e-01 -2.37190336e-01 1.03339267e+00 6.47618830e-01
-7.92197049e-01 3.17532383e-02 -7.63104796e-01 5.06193876e-01
7.53072679e-01 -1.12931108e+00 -8.53352010e-01 1.92270324e-01
7.60540724e-01 2.70182103e-01 -1.16177154e+00 7.48918056e-02
2.75685221e-01 -7.80368209e-01 8.53191674e-01 -8.91461730e-01
2.41170689e-01 -1.17177755e-01 -4.45219874e-01 -1.83301163e+00
-2.95453638e-01 -3.19866985e-01 5.04171789e-01 1.66423309e+00
7.13667274e-01 -8.69952679e-01 -5.15847616e-02 2.27884173e-01
-1.95406768e-02 2.30676811e-02 -9.96583760e-01 -9.53700125e-01
8.79928470e-01 -6.01645887e-01 3.87213051e-01 1.69008791e+00
-2.43526340e-01 5.48782110e-01 -3.64222318e-01 -9.90655050e-02
7.35285401e-01 -1.89981520e-01 8.70332837e-01 -1.27232969e+00
-2.43265510e-01 -4.85866159e-01 2.63274722e-02 -3.55394483e-01
4.54468429e-01 -1.72095692e+00 -2.14132555e-02 -7.70766616e-01
3.50134790e-01 -9.06996250e-01 -7.21034884e-01 4.82411891e-01
-2.33086988e-01 4.55638915e-01 4.42310393e-01 6.17216170e-01
-3.67099673e-01 2.88586438e-01 1.16778266e+00 8.79279301e-02
-2.53648162e-02 -4.41224933e-01 -7.83346355e-01 6.96623921e-01
6.22710824e-01 -7.45924115e-01 -2.55775899e-01 -7.83434629e-01
3.35776627e-01 -4.80491936e-01 3.60622187e-04 -7.46770084e-01
1.82230875e-01 -5.57775721e-02 2.18721721e-02 -2.34653950e-02
1.85729228e-02 -5.94568193e-01 3.19798082e-01 4.18667227e-01
-5.61499178e-01 -5.47075346e-02 5.33424497e-01 2.38495231e-01
-2.74140894e-01 -3.13378751e-01 9.79174972e-01 -2.04326004e-01
-7.60706782e-01 7.81603828e-02 -3.94377589e-01 4.61900592e-01
8.03459227e-01 2.12781712e-01 -1.83597952e-01 1.51134983e-01
-7.85392344e-01 -1.12818457e-01 7.66342521e-01 8.55954289e-01
-3.79798621e-01 -1.74021208e+00 -1.26005352e+00 2.02751905e-01
6.08921587e-01 -7.23369718e-01 -7.56582990e-03 8.83744955e-01
1.17381893e-01 6.55057430e-01 -2.05743849e-01 -5.96087277e-01
-1.18640113e+00 4.52588379e-01 3.79056364e-01 -4.92725044e-01
1.83204822e-02 9.93953288e-01 5.86745203e-01 -1.12934935e+00
-4.06266332e-01 1.39137730e-01 -4.34836547e-04 5.38064182e-01
2.11135209e-01 -6.93417639e-02 -9.37720016e-02 -1.04139864e+00
-2.86149859e-01 7.08259821e-01 -3.91607642e-01 -3.75977635e-01
1.13906217e+00 -2.81877309e-01 -1.61371782e-01 1.16789591e+00
1.38520062e+00 5.07227182e-01 -8.36619198e-01 -9.31872427e-01
4.22114171e-02 -3.55086505e-01 5.35588153e-03 -1.05535936e+00
-7.54946113e-01 9.81582105e-01 4.48611856e-01 -1.77728981e-01
6.23733163e-01 3.76268089e-01 2.40723655e-01 4.37236130e-02
6.41987860e-01 -1.19426322e+00 -3.00306112e-01 7.13227272e-01
9.08368111e-01 -1.27091336e+00 -1.30889326e-01 1.18426949e-01
-6.16804361e-01 7.49392509e-01 4.59564656e-01 1.24334618e-01
6.08422458e-01 1.52522594e-01 3.35158169e-01 1.45206943e-01
-1.07481515e+00 -2.23250791e-01 4.79985297e-01 4.98929381e-01
1.04363275e+00 4.77081478e-01 -2.26225019e-01 3.09086740e-01
-4.64323014e-01 -8.17466378e-02 1.82139292e-01 6.38860941e-01
-4.40613747e-01 -1.63180947e+00 -4.25607949e-01 9.20083746e-02
-2.42985412e-01 -6.34254694e-01 -4.88689989e-01 1.02847862e+00
3.92893434e-01 4.52323258e-01 7.13445619e-02 -2.27008164e-01
1.93779189e-02 5.74481905e-01 6.11768544e-01 -9.49482501e-01
-9.38574791e-01 -9.07274485e-02 1.96615294e-01 -2.08661780e-01
-5.32941580e-01 -8.90928268e-01 -6.32713377e-01 -4.34812158e-01
-2.54468471e-01 2.55482793e-01 7.15144992e-01 9.59275126e-01
1.99501872e-01 4.19796288e-01 4.34733838e-01 -4.23723787e-01
-4.44699705e-01 -1.18100512e+00 -5.76921582e-01 4.83463585e-01
1.12171955e-01 -4.85331744e-01 -4.04319704e-01 2.26756066e-01] | [10.910711288452148, 9.982070922851562] |
59efb44a-0599-4a8d-8516-52a2201761f8 | a-survey-on-causal-discovery-theory-and | 2305.10032 | null | https://arxiv.org/abs/2305.10032v1 | https://arxiv.org/pdf/2305.10032v1.pdf | A Survey on Causal Discovery: Theory and Practice | Understanding the laws that govern a phenomenon is the core of scientific progress. This is especially true when the goal is to model the interplay between different aspects in a causal fashion. Indeed, causal inference itself is specifically designed to quantify the underlying relationships that connect a cause to its effect. Causal discovery is a branch of the broader field of causality in which causal graphs is recovered from data (whenever possible), enabling the identification and estimation of causal effects. In this paper, we explore recent advancements in a unified manner, provide a consistent overview of existing algorithms developed under different settings, report useful tools and data, present real-world applications to understand why and how these methods can be fruitfully exploited. | ['Fabio Stella', 'Alessio Zanga'] | 2023-05-17 | null | null | null | null | ['causal-inference', 'causal-discovery', 'causal-inference'] | ['knowledge-base', 'knowledge-base', 'miscellaneous'] | [ 4.20556724e-01 4.10777656e-03 -8.58801186e-01 -1.17333427e-01
2.07971409e-02 -7.42885530e-01 9.44435358e-01 4.34606493e-01
9.66134295e-02 9.16693568e-01 5.80436230e-01 -5.75529635e-01
-9.00305748e-01 -9.28674757e-01 -7.35672593e-01 -7.28880227e-01
-7.05745161e-01 2.62339443e-01 -3.83298192e-03 -1.45397499e-01
3.81008416e-01 4.42834139e-01 -1.14776397e+00 -1.43160686e-01
8.09138298e-01 2.10712075e-01 -1.13783769e-01 6.24359131e-01
6.23404160e-02 8.07843804e-01 -3.12159091e-01 -3.77764076e-01
-2.99223483e-01 -5.65388381e-01 -9.63087261e-01 -3.30255777e-01
3.17761675e-03 8.78212377e-02 -3.86043161e-01 8.80027294e-01
-1.60319343e-01 -3.73598814e-01 7.95165718e-01 -1.49326587e+00
-4.85236198e-01 8.50709736e-01 -6.95591509e-01 5.04494846e-01
6.41159296e-01 -1.59675002e-01 1.32183480e+00 -3.62899691e-01
5.58922052e-01 1.59631789e+00 2.36390218e-01 7.84883797e-02
-1.63921630e+00 -5.54748774e-01 2.33367816e-01 3.08085293e-01
-1.07058334e+00 -1.55061007e-01 5.12662590e-01 -6.46906197e-01
3.46212506e-01 5.06347895e-01 6.95398808e-01 1.21094465e+00
5.03328621e-01 5.40513694e-01 1.11313558e+00 -3.82618159e-01
2.05342859e-01 -5.43643832e-01 4.35462594e-01 5.39575636e-01
6.96913779e-01 5.36152899e-01 -1.02559471e+00 -5.05073309e-01
7.68851936e-01 -1.33297622e-01 -2.88074344e-01 -1.97884694e-01
-1.20808423e+00 9.90912497e-01 3.41640681e-01 4.84375000e-01
-2.67861575e-01 6.20955884e-01 1.41797945e-01 2.47254670e-01
2.81643420e-01 6.13867342e-01 -5.10636985e-01 -1.05969265e-01
-6.08763874e-01 5.34908652e-01 7.70509660e-01 5.89274824e-01
4.67201442e-01 -3.70835721e-01 -8.38746876e-02 1.81906596e-01
4.76194113e-01 3.84442121e-01 -2.97747940e-01 -8.39681268e-01
9.10788104e-02 6.25960529e-01 1.43531144e-01 -9.57012475e-01
-3.73621970e-01 -3.90397638e-01 -8.59010935e-01 -1.32776931e-01
5.78401923e-01 -9.05783549e-02 -4.72479582e-01 1.92541766e+00
4.66869831e-01 5.15491486e-01 -4.70727265e-01 6.56521916e-01
4.91496593e-01 4.21021044e-01 4.16605741e-01 -5.25113583e-01
1.41416407e+00 -4.40918580e-02 -1.00807512e+00 -1.44682199e-01
4.38637972e-01 -6.55631602e-01 8.35834265e-01 2.81795889e-01
-7.08404243e-01 1.20019931e-02 -6.98813558e-01 1.06805131e-01
-2.99410850e-01 -4.33599979e-01 1.21345091e+00 5.21333635e-01
-7.75186300e-01 6.84454620e-01 -7.14647353e-01 -3.98002923e-01
3.03752422e-01 2.23529980e-01 -8.64773318e-02 -1.16694666e-01
-1.54026079e+00 7.58463979e-01 1.14257775e-01 4.74730693e-02
-9.15943444e-01 -1.28542876e+00 -4.27190632e-01 -3.95381451e-02
6.48107231e-01 -1.06208062e+00 9.23356414e-01 -3.62272799e-01
-7.95516670e-01 5.11924982e-01 -4.70938593e-01 -2.27620810e-01
3.35772753e-01 -2.88904786e-01 -3.00142765e-01 -3.88920039e-01
1.41225278e-01 -4.94338386e-02 3.83361042e-01 -1.15113020e+00
-6.84720218e-01 -6.26815796e-01 5.00141606e-02 -1.93453386e-01
-6.52512982e-02 2.52984345e-01 -6.88415393e-02 -5.87018669e-01
-1.32092804e-01 -7.89152920e-01 -3.22501987e-01 -3.19686949e-01
-7.00637281e-01 -5.16502798e-01 6.82266116e-01 -2.49000311e-01
1.49204230e+00 -1.74286497e+00 5.34287453e-01 2.78324962e-01
6.75097883e-01 -3.29040915e-01 2.52646685e-01 1.00228536e+00
-3.26612622e-01 6.21734202e-01 -4.31298018e-01 2.06738278e-01
-1.60417452e-01 1.84212282e-01 -5.82210124e-01 6.75008476e-01
2.39594758e-01 1.01788354e+00 -1.20602608e+00 -2.92361706e-01
2.09477693e-01 2.07086697e-01 -2.87223160e-01 3.04283649e-01
-2.84913391e-01 6.63406432e-01 -8.13171566e-01 5.27375281e-01
1.30980074e-01 -3.75447899e-01 6.76720917e-01 1.62482485e-01
-4.46937025e-01 4.64798033e-01 -1.04315627e+00 1.15372097e+00
-1.20230019e-01 7.27603257e-01 -5.73305413e-02 -1.17008841e+00
5.25024414e-01 3.80863637e-01 5.81580818e-01 -4.12540793e-01
1.17610849e-01 -4.09614332e-02 3.21019024e-01 -7.25906849e-01
-5.77834211e-02 -2.29169950e-01 -8.66196230e-02 5.00728369e-01
-3.39074880e-01 1.04298419e-03 4.07351643e-01 3.75454754e-01
1.54692781e+00 -6.02596775e-02 9.63986218e-01 -5.11947572e-01
1.00477859e-01 1.28739133e-01 4.34180081e-01 9.07981098e-01
1.33895213e-02 -1.69674024e-01 1.01266611e+00 -3.33208948e-01
-7.88606346e-01 -1.21568358e+00 -4.12827194e-01 8.34484935e-01
6.46012649e-02 -4.75788444e-01 -3.40380371e-01 -4.41161186e-01
3.83240998e-01 7.72164941e-01 -1.17428172e+00 -9.59531441e-02
-4.87580389e-01 -1.21364939e+00 4.44549620e-01 2.72318542e-01
-9.33070555e-02 -7.03198075e-01 -4.03467238e-01 2.75293998e-02
-1.85895339e-01 -7.88116932e-01 1.66305408e-01 2.13738337e-01
-9.76553321e-01 -1.64766145e+00 2.21303524e-03 7.47908279e-02
3.51316333e-01 2.73664415e-01 1.33918953e+00 2.07822725e-01
-3.55631113e-01 1.78019181e-01 -2.07183287e-01 -4.44883466e-01
-4.64800090e-01 -2.15651050e-01 1.03799678e-01 -2.64009804e-01
2.51939327e-01 -8.96093547e-01 -3.06673497e-01 2.09248766e-01
-9.13761199e-01 -1.64458677e-01 5.26182055e-01 6.33448243e-01
9.45596993e-02 3.40305001e-01 5.52002609e-01 -1.11223900e+00
8.58911812e-01 -9.79748070e-01 -6.32009864e-01 3.51242483e-01
-8.72410238e-01 1.03266276e-01 3.44735742e-01 -9.39982831e-02
-9.45855260e-01 -3.06283683e-01 1.89343646e-01 1.47693425e-01
-2.26732060e-01 8.65240872e-01 -3.28781903e-01 3.78643095e-01
4.63889956e-01 -2.90922254e-01 -4.52641785e-01 -4.21191841e-01
7.54742563e-01 1.59792870e-01 3.73365879e-01 -7.72140741e-01
8.05079103e-01 5.83484948e-01 7.00031698e-01 -5.50156593e-01
-9.33667004e-01 -3.56292039e-01 -7.87064672e-01 -3.60642463e-01
5.57631016e-01 -3.82778674e-01 -1.10949600e+00 -1.02569431e-01
-1.15879905e+00 -2.53568560e-01 -3.80219780e-02 5.52576184e-01
-3.61027241e-01 -1.29970953e-01 -3.66959542e-01 -9.21491563e-01
2.34177828e-01 -8.36049974e-01 8.44693422e-01 1.70826674e-01
-6.71042323e-01 -1.53464413e+00 5.10201156e-01 9.30269510e-02
-1.43188402e-01 4.56172884e-01 1.32970357e+00 -3.55879605e-01
-5.28618276e-01 5.02463914e-02 -2.70232469e-01 -5.92725456e-01
4.86485332e-01 4.81759369e-01 -6.84832990e-01 2.03416422e-01
-2.89591879e-01 2.96825528e-01 7.07531393e-01 8.01495492e-01
9.19466078e-01 -2.31770113e-01 -8.74135375e-01 9.19018611e-02
1.29481232e+00 7.82593116e-02 5.29104710e-01 3.08504310e-02
7.20947385e-01 1.05055046e+00 5.56808770e-01 3.51636231e-01
3.01756382e-01 8.17472398e-01 5.97685337e-01 -1.34243503e-01
-1.50120053e-02 -4.39498752e-01 -2.03617290e-03 4.14214611e-01
-3.06880623e-01 -4.83090311e-01 -1.10738790e+00 5.40197670e-01
-2.00458336e+00 -1.11207771e+00 -1.16483748e+00 2.08783984e+00
1.06514215e+00 -1.54344356e-02 3.49920869e-01 1.61480919e-01
6.71020925e-01 -1.02233337e-02 -3.59639645e-01 -1.40100241e-01
1.77136902e-02 1.06969684e-01 6.57525837e-01 5.89788258e-01
-7.86286771e-01 8.13601315e-01 8.33881664e+00 5.40300012e-01
-7.82966912e-01 -6.26117736e-02 5.76328814e-01 7.24104494e-02
-5.79042196e-01 4.49999154e-01 -5.81860006e-01 3.01796436e-01
1.02115333e+00 -3.91610205e-01 3.93100381e-01 1.68981552e-01
1.02966368e+00 -2.92710066e-01 -1.42464364e+00 2.74055630e-01
-5.40306211e-01 -1.39635134e+00 -3.64399627e-02 4.88184363e-01
6.47476196e-01 -3.27738464e-01 -1.01251453e-01 -5.06598353e-01
1.02472866e+00 -1.23959470e+00 3.81115437e-01 6.42062306e-01
4.64119226e-01 -5.09030998e-01 3.58847529e-01 1.65420800e-01
-8.90591621e-01 -7.97677338e-02 -3.89141962e-02 -7.29759276e-01
3.76219124e-01 1.37999415e+00 -9.02137160e-01 8.20924282e-01
6.31471038e-01 9.80468333e-01 -1.93340242e-01 9.24629986e-01
-9.37955499e-01 1.19780076e+00 -1.36853024e-01 -7.45707676e-02
-2.31049463e-01 -2.33458489e-01 7.53573418e-01 1.10345852e+00
3.11115328e-02 1.49467066e-01 -3.39688152e-01 1.21896052e+00
1.50564061e-02 -1.43914193e-01 -8.08175027e-01 -2.64125019e-01
6.67506456e-01 9.88440514e-01 -7.57134378e-01 -2.82163247e-02
-1.93309143e-01 1.78330779e-01 2.83325642e-01 3.16936314e-01
-8.48956645e-01 2.98417091e-01 1.00258720e+00 2.29546472e-01
-3.99742126e-01 -4.23844188e-01 -7.10619152e-01 -8.88641238e-01
-4.69073534e-01 -6.02897644e-01 6.26531780e-01 -5.12923956e-01
-1.27369225e+00 -4.14396673e-01 3.49756420e-01 -5.69904625e-01
-1.79704264e-01 -4.11401868e-01 -7.70941019e-01 7.61954606e-01
-1.33077943e+00 -7.98917592e-01 -3.55221778e-02 2.89382011e-01
1.55696750e-01 5.54103017e-01 6.98036492e-01 7.75088519e-02
-8.53361070e-01 -1.66955590e-01 4.23396714e-02 -2.18764603e-01
5.59465587e-01 -1.45517206e+00 2.87245452e-01 9.00711060e-01
2.65531600e-01 1.09642708e+00 1.27520740e+00 -1.10671556e+00
-1.60462439e+00 -7.53728271e-01 9.79416549e-01 -7.28796721e-01
1.40648937e+00 -4.64899004e-01 -6.92243636e-01 7.22760499e-01
1.02572814e-01 -3.41707617e-01 7.62203515e-01 9.65144038e-01
-3.75192791e-01 7.45001808e-02 -5.34852087e-01 8.82593513e-01
1.43926764e+00 -1.06130503e-01 -6.27352834e-01 1.88424498e-01
6.33497000e-01 2.24162564e-01 -9.56873119e-01 4.22377080e-01
6.49326205e-01 -9.15573120e-01 9.11195815e-01 -9.10192072e-01
9.07596350e-01 -4.43734646e-01 7.79482648e-02 -1.37877595e+00
-4.39250857e-01 -7.29413033e-01 -1.60822436e-01 1.14447701e+00
3.39770347e-01 -5.91908336e-01 3.92415464e-01 4.84159738e-01
3.04295301e-01 -4.74686146e-01 -8.37876618e-01 -5.46855807e-01
1.18326843e-01 -5.62348247e-01 6.12185776e-01 1.43009555e+00
2.50666916e-01 5.61484337e-01 -4.28211272e-01 3.34599525e-01
8.91777515e-01 2.07420275e-01 7.05599546e-01 -1.62912273e+00
-8.64465162e-02 -6.85207069e-01 -2.63759434e-01 -6.91332817e-01
1.21989690e-01 -6.73295379e-01 -2.38307491e-01 -1.39402807e+00
5.53136766e-01 -5.65760016e-01 -1.55059248e-01 3.28357160e-01
-6.52797639e-01 -1.50138855e-01 -2.35941395e-01 2.44181395e-01
-8.52388740e-02 3.04168552e-01 1.12890077e+00 1.07462600e-01
-1.63506031e-01 1.51542770e-02 -9.54715312e-01 6.93694293e-01
6.53316021e-01 -8.08551490e-01 -4.55080152e-01 -2.57026535e-02
8.95132899e-01 1.57162428e-01 7.98261702e-01 -7.67848119e-02
1.89148381e-01 -9.37908053e-01 2.06889473e-02 -3.36095423e-01
-3.45926374e-01 -6.73158288e-01 4.65546638e-01 5.54754615e-01
-5.35284877e-01 -1.08304389e-01 -2.25721579e-02 8.09016645e-01
-6.65424019e-02 -1.59448832e-01 2.71165371e-01 1.73701420e-01
-3.88501018e-01 9.71496254e-02 -2.16206849e-01 -1.72846559e-02
9.53094184e-01 3.76744121e-01 -6.41977727e-01 -3.32073927e-01
-5.18634796e-01 2.43023410e-01 1.15018517e-01 4.87687916e-01
1.70675784e-01 -1.13248277e+00 -8.83037865e-01 -5.20068407e-01
9.62158442e-02 -3.76443326e-01 1.17308728e-01 1.12291563e+00
1.89338297e-01 5.75247109e-01 3.49927753e-01 -3.62749487e-01
-1.17073250e+00 7.10586965e-01 2.42075950e-01 -4.30487871e-01
-3.15595776e-01 5.93638599e-01 6.78982019e-01 1.45667866e-01
-3.12758207e-01 -6.84905574e-02 -2.81775445e-01 1.36750899e-02
6.09512866e-01 6.76933289e-01 -4.04680520e-01 -4.75611873e-02
-5.53264618e-01 2.79875517e-01 3.63390058e-01 -6.20458052e-02
1.41879976e+00 -1.86014876e-01 -7.23215103e-01 8.74738514e-01
7.31226265e-01 3.34674090e-01 -8.99584830e-01 1.00249603e-01
4.04401660e-01 -4.28708494e-01 1.30670384e-01 -8.74064326e-01
-8.09430361e-01 5.19549072e-01 -5.28397597e-02 8.14784765e-01
9.07853901e-01 5.57535470e-01 -4.53585871e-02 -2.66323626e-01
1.64863542e-01 -5.33092260e-01 -2.50391662e-01 7.48190060e-02
9.60194409e-01 -9.22308326e-01 4.91202384e-01 -1.03557515e+00
1.09428078e-01 1.00826919e+00 6.73545450e-02 -3.59676853e-02
8.80241632e-01 5.12076676e-01 -5.64764321e-01 -6.39692724e-01
-1.01143527e+00 -3.18815917e-01 3.31118077e-01 4.28922772e-01
8.81715953e-01 3.33867192e-01 -7.78615475e-01 3.12850147e-01
-1.57329783e-01 -3.51785608e-02 5.95146894e-01 5.75799644e-01
-1.26138911e-01 -1.37537146e+00 -5.12871981e-01 5.28457284e-01
-4.82225895e-01 -2.79722393e-01 -8.25609386e-01 1.01995659e+00
-1.11899036e-03 1.40433812e+00 -1.82353988e-01 -1.79631561e-01
1.93868324e-01 -2.62708664e-01 4.67080146e-01 -4.53407735e-01
-2.08942413e-01 1.13686040e-01 2.07860425e-01 -6.56720340e-01
-7.43643224e-01 -9.94440377e-01 -1.11599553e+00 -7.67148495e-01
-1.75845906e-01 9.90415812e-02 4.75818962e-01 1.27226472e+00
1.29817203e-01 8.84448051e-01 6.24258459e-01 3.41077857e-02
1.78574666e-01 -7.87616730e-01 -6.84495389e-01 9.96385813e-02
3.23079795e-01 -9.97650027e-01 -5.23779929e-01 2.64789194e-01] | [7.899211883544922, 5.38633394241333] |
3c421086-f757-4166-bd58-77a010a0a4ee | multi-scale-representation-learning-on-1 | 2204.02337 | null | https://arxiv.org/abs/2204.02337v1 | https://arxiv.org/pdf/2204.02337v1.pdf | Multi-Scale Representation Learning on Proteins | Proteins are fundamental biological entities mediating key roles in cellular function and disease. This paper introduces a multi-scale graph construction of a protein -- HoloProt -- connecting surface to structure and sequence. The surface captures coarser details of the protein, while sequence as primary component and structure -- comprising secondary and tertiary components -- capture finer details. Our graph encoder then learns a multi-scale representation by allowing each level to integrate the encoding from level(s) below with the graph at that level. We test the learned representation on different tasks, (i.) ligand binding affinity (regression), and (ii.) protein function prediction (classification). On the regression task, contrary to previous methods, our model performs consistently and reliably across different dataset splits, outperforming all baselines on most splits. On the classification task, it achieves a performance close to the top-performing model while using 10x fewer parameters. To improve the memory efficiency of our construction, we segment the multiplex protein surface manifold into molecular superpixels and substitute the surface with these superpixels at little to no performance loss. | ['Andreas Krause', 'Charlotte Bunne', 'Vignesh Ram Somnath'] | 2022-04-04 | multi-scale-representation-learning-on | http://proceedings.neurips.cc/paper/2021/hash/d494020ff8ec181ef98ed97ac3f25453-Abstract.html | http://proceedings.neurips.cc/paper/2021/file/d494020ff8ec181ef98ed97ac3f25453-Paper.pdf | neurips-2021-12 | ['protein-function-prediction'] | ['medical'] | [ 5.22198737e-01 4.09025162e-01 -3.13200235e-01 -3.51575047e-01
-8.47704232e-01 -5.49451888e-01 3.09107274e-01 5.87995112e-01
-3.54671687e-01 1.04866254e+00 2.34215766e-01 -3.91248375e-01
1.55132771e-01 -5.62041402e-01 -1.10495496e+00 -7.81220973e-01
-1.86769977e-01 5.08734286e-01 3.16542178e-01 -6.03304841e-02
1.89941779e-01 3.84505093e-01 -1.01159942e+00 5.51324248e-01
7.77619898e-01 7.56936550e-01 1.82301924e-01 8.04365516e-01
-1.60313442e-01 6.75291717e-01 -3.40687692e-01 -5.05817592e-01
6.14988878e-02 -5.03857017e-01 -1.04689729e+00 1.77152857e-01
6.91642761e-01 2.20987424e-01 -2.62607068e-01 8.89076114e-01
9.84755382e-02 -2.08788291e-01 5.92087388e-01 -8.09997678e-01
-8.46510053e-01 5.57393059e-02 -7.45193601e-01 1.44538566e-01
2.93411523e-01 1.55897230e-01 1.42158782e+00 -8.06669414e-01
1.20382786e+00 1.39062619e+00 8.35372031e-01 4.90322530e-01
-1.99176145e+00 -1.31219774e-01 2.51477301e-01 -5.66326566e-02
-1.22021091e+00 -2.88150251e-01 3.01256716e-01 -6.12881660e-01
1.48747420e+00 4.43712436e-02 4.78819638e-01 8.01198959e-01
6.68340981e-01 5.17109454e-01 8.36675644e-01 1.59223348e-01
4.16130014e-02 -3.62673223e-01 4.12467688e-01 1.09994495e+00
3.41233462e-01 -4.59587246e-01 -5.25146067e-01 -4.30463046e-01
7.40552604e-01 1.50772467e-01 -5.81664145e-01 -5.59684932e-01
-1.31627488e+00 7.59086728e-01 6.32970154e-01 2.07619622e-01
-4.75023836e-01 5.05546510e-01 2.38997176e-01 4.22282696e-01
5.49958825e-01 4.92103159e-01 -7.80442119e-01 1.46518871e-01
-7.26882875e-01 1.95231676e-01 9.23129618e-01 8.40476155e-01
1.02412617e+00 -4.13321435e-01 -7.60846585e-02 6.84788167e-01
1.93731174e-01 -4.81885374e-02 2.37700626e-01 -7.56345272e-01
4.02593195e-01 9.05081332e-01 1.14726365e-01 -6.32140100e-01
-6.96260452e-01 -3.30949098e-01 -6.52088821e-01 1.29447818e-01
5.67154109e-01 4.29673158e-02 -1.13097262e+00 1.71306610e+00
3.26827079e-01 1.53160945e-01 -2.03236684e-01 7.85715997e-01
7.06609845e-01 8.47788930e-01 3.99330527e-01 -1.82513744e-01
1.65765870e+00 -8.97091806e-01 -2.49367386e-01 -7.81731457e-02
7.99118161e-01 -4.01346147e-01 8.86683881e-01 1.52333423e-01
-1.01788700e+00 -4.52605516e-01 -1.19192910e+00 -6.87702954e-01
-3.36828381e-01 -2.06431895e-01 6.32625163e-01 1.76063463e-01
-1.28245008e+00 9.89565909e-01 -1.01706731e+00 -6.05895162e-01
3.82896602e-01 5.91033816e-01 -9.26501870e-01 6.19919263e-02
-9.81321812e-01 8.33825052e-01 8.45239088e-02 -5.19802749e-01
-7.38959849e-01 -1.05468130e+00 -9.03747022e-01 1.98986799e-01
-5.78289665e-02 -1.02217925e+00 8.56133461e-01 -7.05004394e-01
-1.09945512e+00 1.16495836e+00 -4.74122047e-01 -5.60023189e-01
1.96813151e-01 -2.82923486e-02 2.31479540e-01 1.88894883e-01
-2.54826508e-02 9.49323237e-01 4.43913043e-01 -1.03979707e+00
-4.97458667e-01 -5.49320221e-01 4.83056009e-02 2.63657033e-01
2.41437644e-01 -8.03249553e-02 -4.69655663e-01 -5.68469942e-01
1.87101871e-01 -8.80529821e-01 -2.42298841e-01 1.91460267e-01
-3.87122482e-01 -2.11112753e-01 4.71606106e-01 -8.25770676e-01
9.37773228e-01 -1.96058869e+00 7.51021743e-01 -1.31033406e-01
6.42475486e-01 -9.20311660e-02 -2.71881878e-01 6.52651489e-01
-4.00728852e-01 1.58042982e-01 -4.34432358e-01 -2.68433005e-01
-2.10474819e-01 2.43809626e-01 5.10387197e-02 7.42644429e-01
4.98621076e-01 1.33652985e+00 -8.36387753e-01 -1.97557718e-01
-1.15924530e-01 6.80165589e-01 -7.03041196e-01 8.91176462e-02
-5.44273555e-01 4.19403791e-01 -3.90728951e-01 5.37956715e-01
5.12206733e-01 -1.02309799e+00 6.74165428e-01 -4.84169304e-01
2.23527029e-01 4.97254193e-01 -6.41166389e-01 1.82769847e+00
1.84549969e-02 4.64467615e-01 5.09157598e-01 -1.02687252e+00
6.73176885e-01 1.48907065e-01 6.29199326e-01 -1.44741029e-01
-2.75618494e-01 4.39380016e-03 1.37831628e-01 -1.95561320e-01
3.49079928e-04 -1.29449755e-01 2.10918993e-01 1.71862096e-01
2.76329637e-01 2.12099805e-01 9.54995155e-02 3.09563667e-01
1.57578218e+00 4.30801392e-01 4.40490931e-01 -5.33681750e-01
4.83913213e-01 2.15533406e-01 6.42613888e-01 1.26264736e-01
-2.19302818e-01 5.09750903e-01 1.12188017e+00 -5.56733131e-01
-1.15778470e+00 -9.56246078e-01 2.01340951e-03 1.30751491e+00
1.67132989e-01 -5.97199380e-01 -1.01673114e+00 -6.87069595e-01
5.05179763e-01 3.39196362e-02 -7.22378433e-01 -9.20171514e-02
-6.03188634e-01 -1.06179214e+00 3.47981304e-01 4.44335461e-01
6.48542074e-03 -8.05894554e-01 -7.50660449e-02 3.52705002e-01
3.18934806e-02 -7.82043159e-01 -6.36127770e-01 6.00308061e-01
-9.72826302e-01 -1.44560158e+00 -5.85665762e-01 -1.02016628e+00
6.83637142e-01 1.37054771e-01 1.30318618e+00 1.67303339e-01
-5.28894186e-01 -9.93830785e-02 2.72413902e-02 -3.33405808e-02
-3.21274966e-01 3.27485092e-02 -2.55849987e-01 -7.24049136e-02
4.91002142e-01 -5.42053163e-01 -7.21449494e-01 1.00986317e-01
-6.83775485e-01 1.38233274e-01 5.28937638e-01 9.88016963e-01
8.65404189e-01 -5.17645955e-01 3.79487813e-01 -1.36020744e+00
5.69999635e-01 -4.26226407e-01 -4.88292783e-01 2.75400847e-01
-4.54594672e-01 4.31176096e-01 6.02636337e-01 7.94135258e-02
-4.85763967e-01 1.77832246e-01 3.38868089e-02 -1.72478750e-01
-1.85861707e-01 3.60067338e-01 -1.90703750e-01 -1.29855052e-01
6.88554227e-01 1.51273146e-01 2.24284545e-01 -6.46296620e-01
3.90374660e-01 3.66705239e-01 3.51214379e-01 -6.28073275e-01
4.11788076e-01 4.04485315e-01 2.19397679e-01 -8.55115473e-01
-5.26653886e-01 -5.10164201e-01 -8.12769353e-01 4.45840538e-01
1.29150450e+00 -9.55608189e-01 -7.73669362e-01 2.29194671e-01
-1.25300133e+00 -3.91427189e-01 5.91521189e-02 1.47745028e-01
-5.63621402e-01 4.78301674e-01 -1.24703026e+00 -1.31811902e-01
-1.15730487e-01 -1.25803864e+00 1.60189235e+00 -2.59256095e-01
-2.26345897e-01 -1.03110814e+00 1.73662424e-01 4.99070615e-01
8.25122744e-02 5.69870353e-01 1.33413792e+00 -4.91183132e-01
-6.59488261e-01 1.49825022e-01 -4.53393877e-01 -8.11143219e-03
2.34560624e-01 -1.78658530e-01 -8.86694908e-01 -6.01924896e-01
-3.45787108e-01 -3.85716438e-01 1.25176919e+00 3.23308796e-01
1.00937402e+00 -1.65754959e-01 -6.43607736e-01 7.54669011e-01
1.36890113e+00 7.91254565e-02 5.73460996e-01 -3.60272406e-03
8.40044200e-01 5.58461249e-01 2.33026028e-01 -1.22235054e-02
2.00686589e-01 5.46108544e-01 3.05100411e-01 -3.50993842e-01
-2.07585603e-01 -1.40304789e-01 4.89790022e-01 4.60569710e-01
-8.22434667e-03 -3.06938440e-01 -7.38265872e-01 2.60686487e-01
-1.95336068e+00 -8.18296015e-01 -7.16535896e-02 1.88224924e+00
9.17390823e-01 1.44361570e-01 1.32119030e-01 -3.32822591e-01
5.80144882e-01 2.89407909e-01 -9.50041771e-01 -3.44634265e-01
-1.39775515e-01 2.84305423e-01 6.14861131e-01 8.63487780e-01
-1.08096862e+00 1.00540066e+00 7.04168034e+00 4.57805574e-01
-9.87087667e-01 -3.09315920e-02 7.96398342e-01 -8.37152824e-02
-3.07443470e-01 1.46103110e-02 -6.32954478e-01 4.04666364e-01
9.96445775e-01 2.49682069e-01 5.31762719e-01 5.53311825e-01
1.53938243e-02 2.27591723e-01 -1.54975975e+00 7.38006055e-01
-2.08539858e-01 -1.68484139e+00 1.08922146e-01 2.91161120e-01
5.87926984e-01 4.26812708e-01 -2.27623612e-01 5.59859350e-02
4.85271066e-01 -1.22791564e+00 2.92817920e-01 3.48250240e-01
9.51908469e-01 -3.49347800e-01 4.18669045e-01 1.82264373e-01
-1.24124086e+00 3.54544699e-01 -4.28486496e-01 2.43558019e-01
-5.54339811e-02 4.61231172e-01 -7.20504880e-01 3.76985490e-01
3.84924561e-01 8.50337684e-01 -3.74164730e-01 5.16384006e-01
-1.37964413e-01 3.96090120e-01 3.32714953e-02 8.39292482e-02
2.14632466e-01 -4.79172945e-01 3.53058547e-01 1.48783970e+00
-2.95298159e-01 2.70532638e-01 3.18540782e-01 7.04090655e-01
-6.32038414e-01 1.08602881e-01 -4.24647808e-01 -3.74999583e-01
2.47121841e-01 1.20066750e+00 -7.07006991e-01 -4.46224064e-01
-7.36021996e-01 1.36661553e+00 8.92330825e-01 5.99330962e-01
-7.74364471e-01 -5.00904858e-01 1.34996867e+00 3.50039721e-01
4.23075467e-01 -2.29499906e-01 -1.18372098e-01 -1.12043691e+00
-1.35835424e-01 -8.01476061e-01 4.36531454e-01 -6.64683640e-01
-1.25894928e+00 2.13562042e-01 -5.24245381e-01 -5.59409082e-01
2.05653429e-01 -1.14439881e+00 -1.33647799e-01 1.05998802e+00
-1.29611337e+00 -1.07621980e+00 4.47817184e-02 1.12372756e-01
3.09846818e-01 2.10332330e-02 7.94999123e-01 1.91800162e-01
-5.84627211e-01 4.25621361e-01 2.50935704e-01 -6.99495301e-02
7.11870193e-01 -1.62774873e+00 7.86572874e-01 4.09279406e-01
-4.60658735e-03 1.03601599e+00 5.55094540e-01 -8.84482086e-01
-1.47514176e+00 -1.12544286e+00 9.77443278e-01 -6.87396526e-01
6.97378159e-01 -6.33181691e-01 -1.15319622e+00 8.01098645e-01
9.89409313e-02 6.11993074e-02 8.28760803e-01 1.93892956e-01
-6.12165332e-01 1.88352749e-01 -1.17231786e+00 4.40776020e-01
1.31138861e+00 -7.56719589e-01 -5.15236437e-01 5.62541664e-01
9.92916524e-01 -2.58805782e-01 -1.33035874e+00 2.60264844e-01
4.71266359e-01 -7.85900772e-01 1.03382123e+00 -1.25105131e+00
4.45392400e-01 -4.98546600e-01 -1.95921928e-01 -1.15045738e+00
-9.67684686e-01 -7.09165871e-01 -2.71938562e-01 4.28312987e-01
7.09829450e-01 -5.38916528e-01 1.16549897e+00 3.44915122e-01
-2.80119091e-01 -1.26818371e+00 -7.70600200e-01 -4.15034026e-01
2.13098034e-01 2.32131690e-01 4.63434964e-01 8.50454211e-01
2.08897397e-01 8.23506057e-01 2.70351525e-02 -9.11052078e-02
4.79493946e-01 2.53546208e-01 7.28061259e-01 -1.21439624e+00
-4.96063799e-01 -5.28169334e-01 -6.62477374e-01 -1.40885055e+00
1.04857441e-02 -1.20079672e+00 -2.89240450e-01 -1.79311311e+00
4.96391833e-01 2.41193667e-01 -4.83820587e-01 6.01326168e-01
-5.31273901e-01 2.61392266e-01 -1.00669496e-01 2.51849473e-01
-6.20348275e-01 2.02605680e-01 1.32459080e+00 -1.95663691e-01
4.70824875e-02 -6.38689160e-01 -9.25415993e-01 4.39534128e-01
4.39542919e-01 -2.61399269e-01 -2.60738611e-01 -3.54069412e-01
2.85621209e-04 6.92989752e-02 1.29611582e-01 -7.10617900e-01
-1.33207873e-01 -1.47752076e-01 5.34541488e-01 -1.68944195e-01
4.22766387e-01 -4.08681631e-01 2.00563073e-01 7.83824146e-01
-4.68000889e-01 1.99669167e-01 -2.15288643e-02 1.03539860e+00
2.27999985e-02 3.86360168e-01 1.13527727e+00 -1.97236612e-01
-3.53773534e-01 5.43066323e-01 -1.77332908e-01 1.71895344e-02
9.48574960e-01 -2.22630858e-01 -5.27327299e-01 -3.36114988e-02
-9.00838912e-01 2.34640703e-01 1.08702433e+00 1.73223853e-01
3.32573026e-01 -8.92120957e-01 -5.90676725e-01 1.65831298e-01
5.36919869e-02 -2.55424585e-02 -2.11523607e-01 7.88101673e-01
-9.68495786e-01 8.03758085e-01 -9.85774621e-02 -5.80568731e-01
-1.27617228e+00 6.32872880e-01 5.10601938e-01 -4.31372792e-01
-5.29634595e-01 1.16349816e+00 8.50221574e-01 -3.24005812e-01
3.22972722e-02 -5.50668538e-01 3.87718156e-02 -1.61502808e-01
3.60033482e-01 2.98084497e-01 1.50787726e-01 -6.51722312e-01
-4.31778222e-01 5.52585661e-01 -2.76330680e-01 3.09439331e-01
1.38625884e+00 1.18227363e-01 -4.57655400e-01 3.38297188e-01
1.49374485e+00 -2.60158956e-01 -1.50205743e+00 -1.06447682e-01
3.71616513e-01 4.48193215e-02 -4.00023311e-01 -7.03969538e-01
-6.65484786e-01 7.65573084e-01 2.55900800e-01 6.85749426e-02
4.98588353e-01 1.75813928e-01 9.07887578e-01 4.31738913e-01
4.81939167e-01 -8.03276241e-01 -9.90704373e-02 4.19685096e-01
6.63058579e-01 -1.13604891e+00 1.02494366e-01 -5.92478216e-01
-4.46583748e-01 8.50190580e-01 4.86522943e-01 -3.13216239e-01
4.85635430e-01 1.44236431e-01 -1.18787751e-01 -6.63239419e-01
-1.19659066e+00 3.41177359e-02 3.90911520e-01 4.08291489e-01
1.08441567e+00 1.65658787e-01 -3.95430952e-01 2.65590638e-01
1.96748510e-01 -2.41494641e-01 -2.31145509e-02 8.21001351e-01
-7.53158271e-01 -1.31445563e+00 2.56795324e-02 7.24150240e-01
-5.73546529e-01 -7.27419183e-02 -9.35238719e-01 5.52356601e-01
7.46527314e-02 7.58486271e-01 1.06139660e-01 -3.49248916e-01
3.13049644e-01 3.37371439e-01 5.95067620e-01 -9.21169758e-01
-5.80390573e-01 1.50304660e-01 2.00539276e-01 -1.04548287e+00
-2.02812299e-01 -5.67804456e-01 -1.60409355e+00 -3.14074516e-01
-1.33407623e-01 3.27502966e-01 4.95310009e-01 5.46199501e-01
7.85552859e-01 5.36052585e-01 2.41099969e-01 -7.13053107e-01
-5.48307061e-01 -8.23052943e-01 -7.11512268e-01 7.72952199e-01
4.05553102e-01 -5.41939080e-01 -5.52427471e-02 2.45023623e-01] | [4.849228858947754, 5.667567253112793] |
9225d7cb-0401-4fdb-9b1d-952f08e08c5a | spikecodec-an-end-to-end-learned-compression | 2306.14108 | null | https://arxiv.org/abs/2306.14108v1 | https://arxiv.org/pdf/2306.14108v1.pdf | SpikeCodec: An End-to-end Learned Compression Framework for Spiking Camera | Recently, the bio-inspired spike camera with continuous motion recording capability has attracted tremendous attention due to its ultra high temporal resolution imaging characteristic. Such imaging feature results in huge data storage and transmission burden compared to that of traditional camera, raising severe challenge and imminent necessity in compression for spike camera captured content. Existing lossy data compression methods could not be applied for compressing spike streams efficiently due to integrate-and-fire characteristic and binarized data structure. Considering the imaging principle and information fidelity of spike cameras, we introduce an effective and robust representation of spike streams. Based on this representation, we propose a novel learned spike compression framework using scene recovery, variational auto-encoder plus spike simulator. To our knowledge, it is the first data-trained model for efficient and robust spike stream compression. Extensive experimental results show that our method outperforms the conventional and learning-based codecs, contributing a strong baseline for learned spike data compression. | ['Wen Gao', 'Siwei Ma', 'Chuanmin Jia', 'Kexiang Feng'] | 2023-06-25 | null | null | null | null | ['data-compression'] | ['time-series'] | [ 7.80522883e-01 -5.56121647e-01 2.43226856e-01 -2.07442671e-01
-1.00721955e+00 -2.40400523e-01 4.24016416e-01 -6.66183457e-02
-5.01183510e-01 9.21314418e-01 3.48603725e-01 4.62979645e-01
-2.81086445e-01 -4.42629367e-01 -8.24933410e-01 -9.90729690e-01
5.72212562e-02 3.69758636e-01 3.20952445e-01 1.57474399e-01
4.63143587e-01 2.84344077e-01 -1.71888638e+00 5.77848911e-01
5.78415811e-01 1.22388828e+00 9.22707975e-01 4.62069452e-01
7.37444684e-02 9.15648580e-01 -2.10590526e-01 -2.20405042e-01
1.54375449e-01 -7.28304029e-01 -1.56844035e-01 -8.27097669e-02
8.17243457e-02 -7.00883508e-01 -1.12391889e+00 1.01068997e+00
6.29493892e-01 -2.19565004e-01 5.97722948e-01 -1.00864100e+00
-5.64806044e-01 7.35424817e-01 -2.94461042e-01 6.24840856e-01
2.28940830e-01 1.08474813e-01 5.32979786e-01 -7.10012853e-01
7.64838457e-01 4.93825823e-01 4.46044296e-01 7.70153821e-01
-1.21773946e+00 -6.67855024e-01 -3.22198063e-01 3.65112066e-01
-1.15525925e+00 -5.58932364e-01 8.08984756e-01 -1.40758514e-01
1.00789797e+00 1.18967555e-01 1.01043069e+00 1.28766191e+00
1.86693788e-01 8.27079237e-01 1.23399043e+00 2.23726645e-01
4.78799701e-01 -4.60629433e-01 -5.09818085e-02 1.92795619e-01
4.20136392e-01 1.18797325e-01 -1.04233611e+00 1.03240475e-01
9.81198370e-01 3.51403803e-01 -5.60602844e-01 3.60751599e-02
-1.33349085e+00 4.07404929e-01 2.18622074e-01 2.74985969e-01
-4.30824310e-01 5.76066375e-01 3.05664003e-01 1.80537000e-01
-2.77254954e-02 8.60648602e-03 5.47736771e-02 -6.42565668e-01
-1.50985050e+00 3.93234402e-01 3.23952347e-01 1.12524819e+00
2.33479306e-01 3.77525568e-01 -1.60038099e-01 5.62917352e-01
-4.02476527e-02 6.36824667e-01 7.98999131e-01 -1.25638473e+00
3.14438820e-01 2.09667265e-01 -2.72554457e-01 -7.74350584e-01
-7.54373744e-02 -5.52456677e-01 -1.19445932e+00 -4.98054326e-02
9.18831229e-02 4.51473862e-01 -6.20414138e-01 1.59934855e+00
-4.53589469e-01 8.68353188e-01 5.97518273e-02 8.86830688e-01
4.98486161e-01 7.14364529e-01 -4.63148594e-01 -7.18360603e-01
9.77223933e-01 -3.55911165e-01 -9.09906149e-01 -1.40337840e-01
5.50495321e-03 -3.78603607e-01 7.19427526e-01 5.93421876e-01
-1.54906476e+00 -9.21604559e-02 -1.16522229e+00 -1.95885643e-01
2.80749351e-01 -3.04813534e-01 5.37141919e-01 1.36949033e-01
-9.10280943e-01 8.30901802e-01 -1.05926085e+00 -7.68671408e-02
9.63528991e-01 1.88777253e-01 -3.68715674e-01 -4.54795629e-01
-4.92717803e-01 4.49100852e-01 3.66921723e-01 -2.98719972e-01
-1.17757416e+00 -6.22606874e-01 -3.72107059e-01 2.32347131e-01
-3.69335979e-01 -7.47102141e-01 9.88258660e-01 -9.66709554e-02
-1.20618355e+00 7.72774577e-01 -2.92801648e-01 -1.10198998e+00
1.79068074e-01 4.14925031e-02 -1.14667371e-01 6.89153671e-01
-1.17048785e-01 5.30539095e-01 9.23672915e-01 -1.11176610e+00
-2.11546525e-01 -5.65507889e-01 -5.99295735e-01 -2.68036444e-02
-3.83979261e-01 -9.92587879e-02 -1.70601562e-01 -7.69047201e-01
4.66722280e-01 -3.92328113e-01 1.27235442e-01 1.40276775e-01
1.38676852e-01 5.15039504e-01 8.12689126e-01 -5.27529359e-01
1.16638303e+00 -2.15426350e+00 1.59797207e-01 -4.09145981e-01
3.50950807e-01 1.76077321e-01 -9.21823457e-02 5.49999535e-01
8.53940845e-02 -3.65119725e-01 -8.99714530e-01 -5.26092470e-01
-3.07460040e-01 3.53959024e-01 -8.36883783e-01 5.86737096e-01
3.86631824e-02 9.96991813e-01 -7.35713184e-01 -3.86434108e-01
1.75706238e-01 7.18326151e-01 -8.02576244e-01 3.05950522e-01
9.66005549e-02 7.24377215e-01 -2.69805610e-01 5.99771619e-01
8.64787698e-01 -1.76415861e-01 -5.32507241e-01 -2.10603073e-01
-1.66162729e-01 1.75769880e-01 -7.16458559e-01 2.11324120e+00
1.10123709e-01 5.94805241e-01 -1.00647688e-01 -1.34942508e+00
9.59498942e-01 2.77288795e-01 8.12427700e-01 -1.06657732e+00
2.32830927e-01 4.08636093e-01 -3.00999552e-01 -4.78252679e-01
1.53554782e-01 -2.07643554e-01 1.08635336e-01 3.77697498e-01
1.95539206e-01 -3.44775319e-01 -2.78265439e-02 5.03162682e-01
1.61555636e+00 8.26453865e-02 -2.21488670e-01 1.57829612e-01
1.43614545e-01 -2.22333610e-01 5.66340744e-01 6.64307892e-01
-2.61458427e-01 1.31693041e+00 -1.44843489e-03 -7.27087632e-02
-1.36370063e+00 -1.23522806e+00 -4.43717599e-01 1.83267996e-01
2.85477430e-01 -4.20960546e-01 -9.58228946e-01 4.36963409e-01
-4.56894994e-01 4.87383425e-01 -9.51058790e-02 -3.88133824e-01
-6.45062387e-01 -7.24538207e-01 8.19864273e-01 3.14003408e-01
6.18651211e-01 -1.08948314e+00 -9.45666075e-01 5.50915778e-01
-5.33374250e-01 -1.53243542e+00 -1.95372000e-01 2.05581501e-01
-1.41791821e+00 -7.67658412e-01 -7.45758057e-01 -8.31406057e-01
2.76705503e-01 5.32424390e-01 7.34737694e-01 -2.07100525e-01
-7.52650440e-01 3.68851006e-01 -3.94172549e-01 -2.32276797e-01
9.78200510e-02 -5.67962945e-01 1.06665201e-01 -3.24961394e-02
2.81065375e-01 -1.28888714e+00 -9.22531128e-01 -2.41094545e-01
-1.29263914e+00 1.31446362e-01 5.87002218e-01 7.65550256e-01
1.03318465e+00 -7.84567930e-03 4.97277826e-01 -4.53630328e-01
4.62922931e-01 -4.64143753e-01 -4.25025165e-01 -1.76921889e-01
-3.38945448e-01 4.99870032e-02 7.17652202e-01 -1.04721189e-01
-7.24776566e-01 1.84449777e-01 -1.22198887e-01 -5.28056026e-01
-1.53611600e-01 3.44583452e-01 1.68861508e-01 -1.18406795e-01
3.50958318e-01 1.17065525e+00 -1.29562216e-02 -4.41840142e-01
-4.94008660e-02 4.10282612e-01 1.30849659e+00 -2.78213978e-01
7.68815637e-01 1.11016154e+00 1.79757290e-02 -7.37110496e-01
-3.49772722e-01 -4.52786297e-01 -4.20456350e-01 -2.20009699e-01
8.19286168e-01 -1.01744950e+00 -7.23181725e-01 8.09748769e-01
-1.36191869e+00 2.34893143e-01 -4.59070325e-01 7.77281761e-01
-1.22767985e+00 6.28138423e-01 -9.46667135e-01 -9.66091037e-01
-3.54641110e-01 -9.45374370e-01 1.12036014e+00 -6.19464405e-02
3.73488486e-01 -2.85670877e-01 2.39627853e-01 3.92127961e-01
6.59764290e-01 3.83811086e-01 5.43661177e-01 -1.11730851e-01
-1.17763698e+00 -2.79366791e-01 -3.11426938e-01 3.20031010e-02
-3.61574382e-01 -5.56626320e-01 -1.17065299e+00 -3.87322783e-01
8.82611096e-01 -6.42679036e-01 1.29556525e+00 7.20650196e-01
1.47967529e+00 -1.23535402e-01 -1.63166642e-01 1.24157262e+00
1.79824388e+00 7.01290891e-02 1.16295886e+00 1.31243378e-01
3.11492175e-01 1.97978824e-01 -3.24486918e-03 9.79284465e-01
1.42984256e-01 3.68176132e-01 6.59984708e-01 5.17292917e-01
-1.64391518e-01 -3.76121640e-01 3.57780606e-01 1.61117923e+00
-1.89256415e-01 -1.66710019e-01 -4.35225755e-01 5.08486807e-01
-1.77589607e+00 -1.56139290e+00 -1.92655057e-01 2.32454515e+00
6.83835268e-01 6.25145212e-02 -4.73447740e-02 3.65335405e-01
5.88053465e-01 -1.25335753e-02 -8.17329884e-01 1.63396567e-01
-6.83396459e-01 3.12656850e-01 4.15677279e-01 6.49806531e-03
-6.05179191e-01 5.23628116e-01 6.25109529e+00 7.87136972e-01
-7.77182519e-01 2.84824401e-01 1.91326320e-01 -6.67943120e-01
-3.42581123e-01 -1.20040260e-01 -6.51707590e-01 9.21540260e-01
1.23931742e+00 -4.00361329e-01 9.95181978e-01 3.08004081e-01
1.84680566e-01 3.80889587e-02 -8.69685590e-01 1.86923409e+00
2.86085218e-01 -1.58793890e+00 2.21746862e-01 2.50894159e-01
4.70241368e-01 2.51450121e-01 9.39916074e-02 -1.70605600e-01
-2.23237231e-01 -1.05468535e+00 7.78679669e-01 8.34644675e-01
8.72877419e-01 -6.17428780e-01 4.14971083e-01 6.50258303e-01
-1.15804780e+00 -2.94911712e-01 -1.01236069e+00 -1.81409717e-01
7.32095540e-01 8.03899288e-01 -7.35642063e-03 2.60608494e-01
7.51385510e-01 1.13072753e+00 -8.56293887e-02 1.49039662e+00
5.68185687e-01 6.32962286e-01 -3.42656195e-01 3.84478480e-01
1.69942342e-02 -2.93834925e-01 7.38512576e-01 1.01422930e+00
9.68184173e-01 5.77599227e-01 -3.04273039e-01 1.17919338e+00
-1.64840147e-01 -4.15891111e-01 -1.06550574e+00 -2.45938018e-01
5.77179909e-01 8.88795912e-01 -6.59308136e-01 -9.76023972e-02
-4.05933887e-01 1.15621996e+00 2.90243775e-01 7.66411424e-02
-8.44661772e-01 -9.10998136e-02 3.23557287e-01 2.85774261e-01
3.94460589e-01 -5.18983424e-01 -5.01890242e-01 -1.52457154e+00
2.61727452e-01 -4.16160375e-01 4.68942374e-02 -9.68249738e-01
-1.29431915e+00 7.49767959e-01 -1.36557460e-01 -1.66005576e+00
1.01507604e-01 -3.97695333e-01 -4.23565865e-01 2.91504204e-01
-1.58698130e+00 -9.07155752e-01 -4.75750685e-01 9.78226602e-01
5.39460480e-01 -3.54300112e-01 7.39170969e-01 4.45172101e-01
-2.20429689e-01 3.57239991e-01 6.43523633e-01 -2.82212794e-01
3.55232269e-01 -6.93002939e-01 1.90034628e-01 9.26120520e-01
2.85478115e-01 3.12579870e-01 6.52903020e-01 -5.24994731e-01
-1.85588849e+00 -8.99891734e-01 5.68188608e-01 -2.90167421e-01
5.66598415e-01 -6.00860000e-01 -1.14692521e+00 4.58248734e-01
2.31661648e-01 -1.13527551e-01 6.22438550e-01 -1.10682917e+00
-2.33332798e-01 -2.89633155e-01 -1.07275975e+00 2.48132050e-01
1.36130917e+00 -7.17667103e-01 -7.61274457e-01 4.21557993e-01
6.55355334e-01 7.98275229e-03 -6.54939532e-01 2.45546132e-01
4.38151956e-01 -1.23154092e+00 1.10207760e+00 1.95978638e-02
7.89558887e-01 -1.30235016e-01 -4.84437376e-01 -8.94279301e-01
-4.70298797e-01 -6.19938314e-01 -3.26730728e-01 1.12582886e+00
-1.76958129e-01 -3.22573692e-01 7.55965769e-01 3.13362330e-01
-5.35732031e-01 -5.74643254e-01 -1.33176756e+00 -9.27995741e-01
-2.10508749e-01 -6.20390296e-01 3.18992913e-01 4.62518692e-01
2.24108711e-01 -1.60673559e-01 -6.65803134e-01 -1.05556346e-01
1.25123501e+00 4.33067754e-02 1.29437700e-01 -1.09589994e+00
-5.40722668e-01 -3.27049017e-01 -8.39076161e-01 -1.17665207e+00
-1.40192896e-01 -1.15149486e+00 1.43750459e-01 -1.41238976e+00
6.14703834e-01 -6.26304075e-02 -4.13793057e-01 -1.47681430e-01
3.65363449e-01 7.29332089e-01 2.58136809e-01 8.30322981e-01
-6.65596604e-01 9.08030272e-01 1.10785413e+00 -3.64945792e-02
3.46790224e-01 -5.60738564e-01 -4.30191904e-01 3.34862500e-01
5.66092849e-01 -6.72414124e-01 -4.73461330e-01 -7.58225739e-01
2.30550140e-01 3.73105347e-01 5.30026734e-01 -1.71729946e+00
8.42402756e-01 2.30101928e-01 2.14902997e-01 -6.45926654e-01
6.99198127e-01 -8.60011816e-01 3.72040808e-01 4.94591087e-01
-3.61712664e-01 5.99380136e-02 2.38043871e-02 9.54786003e-01
-5.46451330e-01 -2.73422807e-01 9.57745910e-01 -2.73979843e-01
-3.58936787e-01 7.05142081e-01 -5.23262501e-01 3.35386693e-01
8.40127230e-01 -7.26304233e-01 -4.15183008e-01 -3.68922025e-01
-3.66014272e-01 -3.29683065e-01 5.10172009e-01 -7.29241222e-02
1.38470066e+00 -1.42262149e+00 -8.73132169e-01 5.23098171e-01
7.82969818e-02 -1.12730689e-01 4.15498465e-01 8.40360582e-01
-6.34474993e-01 4.60222095e-01 -7.81147242e-01 -6.30286992e-01
-7.72363484e-01 6.72395349e-01 -2.47992262e-01 1.45374164e-01
-1.19425821e+00 8.12372446e-01 2.15031594e-01 4.26401526e-01
2.23011255e-01 -2.34889537e-01 7.70701319e-02 -3.35974365e-01
9.67626393e-01 2.79766619e-01 -3.56998369e-02 -4.85964924e-01
-1.59086764e-01 7.99284637e-01 -1.32703586e-02 -3.13240677e-01
1.85860932e+00 -3.74330819e-01 -1.85100392e-01 6.21932328e-01
1.24737585e+00 -5.89823008e-01 -1.60419512e+00 -1.09014034e-01
-3.32375288e-01 -6.56630099e-01 1.32996142e-01 -3.54856104e-01
-1.04950762e+00 1.30036426e+00 6.12098157e-01 -5.04789501e-02
1.40897346e+00 -1.09850571e-01 1.47779608e+00 2.68115193e-01
9.69552457e-01 -9.21877265e-01 1.86134577e-01 4.16496307e-01
8.37886870e-01 -9.12191391e-01 -2.85101831e-01 -1.10181481e-01
-5.42385578e-01 9.88853276e-01 1.66706175e-01 -6.35474980e-01
3.90111506e-01 8.08877230e-01 -5.10730505e-01 -2.33326301e-01
-1.02366269e+00 1.85967475e-01 -3.29067290e-01 8.70280385e-01
8.94823447e-02 -2.72841662e-01 -2.64132828e-01 6.03692710e-01
-1.97465166e-01 4.90899026e-01 6.13123119e-01 8.67729127e-01
-8.11958849e-01 -8.12931716e-01 -1.56751424e-01 7.53173172e-01
-4.38225150e-01 -2.82303512e-01 2.44349614e-02 -2.04892922e-02
-2.23814607e-01 6.34511411e-01 1.71336427e-01 -4.17687446e-01
7.94080494e-04 -2.08849404e-02 8.89735758e-01 -2.29338095e-01
-1.22694492e-01 2.08157748e-01 -7.02978313e-01 -7.37934768e-01
-5.26625097e-01 -7.19314873e-01 -1.46191859e+00 -3.39337915e-01
-2.04537623e-02 -2.96016961e-01 6.44612849e-01 8.08777928e-01
5.55828154e-01 3.92046690e-01 4.64101523e-01 -9.30648863e-01
-3.96938294e-01 -6.17693484e-01 -8.97336721e-01 7.82180071e-01
3.36945385e-01 -4.57562625e-01 -3.94211590e-01 4.67616528e-01] | [8.830697059631348, -1.2198576927185059] |
4b7e8da4-1297-48a6-89fd-8f30aa279be3 | deep-multimodal-subspace-clustering-networks | 1804.06498 | null | http://arxiv.org/abs/1804.06498v3 | http://arxiv.org/pdf/1804.06498v3.pdf | Deep Multimodal Subspace Clustering Networks | We present convolutional neural network (CNN) based approaches for
unsupervised multimodal subspace clustering. The proposed framework consists of
three main stages - multimodal encoder, self-expressive layer, and multimodal
decoder. The encoder takes multimodal data as input and fuses them to a latent
space representation. The self-expressive layer is responsible for enforcing
the self-expressiveness property and acquiring an affinity matrix corresponding
to the data points. The decoder reconstructs the original input data. The
network uses the distance between the decoder's reconstruction and the original
input in its training. We investigate early, late and intermediate fusion
techniques and propose three different encoders corresponding to them for
spatial fusion. The self-expressive layers and multimodal decoders are
essentially the same for different spatial fusion-based approaches. In addition
to various spatial fusion-based methods, an affinity fusion-based network is
also proposed in which the self-expressive layer corresponding to different
modalities is enforced to be the same. Extensive experiments on three datasets
show that the proposed methods significantly outperform the state-of-the-art
multimodal subspace clustering methods. | ['Vishal M. Patel', 'Mahdi Abavisani'] | 2018-04-17 | null | null | null | null | ['multi-view-subspace-clustering', 'multiview-learning', 'multi-modal-subspace-clustering'] | ['computer-vision', 'computer-vision', 'computer-vision'] | [-8.59250082e-04 -5.22546060e-02 -1.19139351e-01 -5.04757822e-01
-9.87612247e-01 -5.73044538e-01 6.30108178e-01 -9.65162367e-02
-4.84099984e-01 3.89022827e-01 6.34659886e-01 2.21320242e-01
-7.78077543e-02 -4.60396379e-01 -8.74909222e-01 -1.10685372e+00
3.35577995e-01 5.64906061e-01 3.90358232e-02 5.91467209e-02
2.35314757e-01 2.97978252e-01 -1.51482081e+00 7.97583520e-01
8.70854676e-01 1.09851825e+00 1.56448558e-01 8.41680944e-01
-1.93167999e-01 6.60819471e-01 -1.12701841e-01 -3.56077522e-01
1.62313387e-04 -3.81494582e-01 -8.27352941e-01 2.63662368e-01
2.35696703e-01 -3.28736931e-01 -5.92037320e-01 1.22849441e+00
4.42767471e-01 1.94880709e-01 8.41550946e-01 -1.35695171e+00
-6.79210365e-01 5.27808607e-01 -4.99636471e-01 -4.14401323e-01
5.04540145e-01 -2.91954190e-01 6.75290704e-01 -1.06198907e+00
6.56799316e-01 1.22502387e+00 4.15364295e-01 5.45145035e-01
-1.10589504e+00 -3.27834815e-01 -5.77267185e-02 3.29445340e-02
-1.51258564e+00 -6.89941883e-01 1.01694787e+00 -5.42974949e-01
4.62010980e-01 1.58852667e-01 4.62709218e-01 1.12470865e+00
-5.02097420e-02 1.10485685e+00 7.45076835e-01 -3.43740851e-01
3.63552302e-01 1.20493129e-01 1.53168023e-01 7.42517531e-01
-5.18179774e-01 -1.97282657e-01 -7.25097537e-01 -1.77171737e-01
6.56411648e-01 1.71945482e-01 -2.14128941e-01 -7.63708174e-01
-1.50407588e+00 6.74343169e-01 4.47751015e-01 4.48655665e-01
-4.22757536e-01 3.66916955e-02 2.98011959e-01 4.84461859e-02
-2.09301282e-02 -2.15106428e-01 -2.53929365e-02 -2.21550260e-02
-1.22156441e+00 -1.68420225e-01 6.05788887e-01 1.08066463e+00
9.25221562e-01 -1.22216366e-01 -1.81481957e-01 8.10852170e-01
5.97896159e-01 2.98768014e-01 6.11445725e-01 -1.17490709e+00
6.42297804e-01 9.54264402e-01 -1.38421692e-02 -8.30659568e-01
-2.41694078e-01 -1.68706194e-01 -1.18069196e+00 -1.06698282e-01
1.22301541e-01 -1.65750533e-01 -9.02582645e-01 1.91972125e+00
1.24044806e-01 3.16870660e-01 4.80784655e-01 1.07165313e+00
8.88431430e-01 7.74441898e-01 -4.48731482e-02 -4.99145016e-02
9.68167543e-01 -1.00963652e+00 -7.39375353e-01 2.95953006e-01
5.09512365e-01 -7.58019865e-01 7.00933933e-01 4.64078225e-02
-1.23781025e+00 -6.85444176e-01 -1.00691271e+00 -2.11530045e-01
-5.48329473e-01 6.21380687e-01 2.26334497e-01 1.95645645e-01
-1.33638310e+00 3.47643644e-01 -8.58248651e-01 -6.31942511e-01
1.30169645e-01 6.17213070e-01 -8.90668154e-01 1.92040861e-01
-9.02391911e-01 5.42833447e-01 6.90572977e-01 1.52708873e-01
-9.19469893e-01 -9.49618146e-02 -1.08465815e+00 4.79047634e-02
-4.58721966e-02 -8.81370962e-01 8.61841738e-01 -1.35012114e+00
-1.51417255e+00 7.68526852e-01 -4.96745557e-01 -1.55509580e-02
1.71805367e-01 3.90909798e-03 -5.40209055e-01 3.12333137e-01
1.24247357e-01 1.08065927e+00 1.03190577e+00 -1.87129128e+00
-6.59719169e-01 -4.43065792e-01 -2.15449542e-01 6.24913394e-01
-5.04916072e-01 -1.85381353e-01 -1.07449889e+00 -3.87993485e-01
5.60795128e-01 -9.60652769e-01 1.16345651e-01 -2.02159315e-01
-7.20844805e-01 7.80713260e-02 1.08248568e+00 -5.84093928e-01
1.26618946e+00 -2.49078822e+00 1.07449436e+00 4.15114641e-01
2.99023092e-01 -1.04409866e-01 -2.37919420e-01 5.77216744e-01
-2.01594815e-01 -3.03400964e-01 -5.49391031e-01 -7.93915868e-01
3.10389906e-01 3.70292157e-01 -1.79837823e-01 4.60294873e-01
-3.09616551e-02 8.82323205e-01 -6.88797832e-01 -7.22372770e-01
4.24472034e-01 8.51105571e-01 -3.31947863e-01 5.02136409e-01
1.45548910e-01 6.06851697e-01 -5.69637977e-02 7.72156894e-01
7.77923286e-01 -1.29483402e-01 2.41819277e-01 -7.33924568e-01
-2.24833950e-01 -2.58371204e-01 -1.35661936e+00 2.38515210e+00
3.61221991e-02 4.37859982e-01 4.99387801e-01 -8.84782434e-01
4.66661960e-01 5.33413649e-01 5.13223529e-01 -5.46590924e-01
2.04329088e-01 2.24369392e-01 -3.04346800e-01 -5.45521557e-01
7.66177833e-01 -1.55802861e-01 -2.62214482e-01 2.85971820e-01
5.49758255e-01 6.06010973e-01 3.82804908e-02 5.71930051e-01
5.46187878e-01 3.41455519e-01 -2.16192260e-01 -1.70399323e-01
9.02109981e-01 -2.50997365e-01 4.00177211e-01 3.16401809e-01
-6.07787408e-02 7.01863527e-01 4.78778630e-01 -8.49985033e-02
-1.04213691e+00 -1.32281935e+00 1.56271681e-01 1.10726523e+00
5.62833488e-01 -4.36274558e-01 -9.92365777e-01 -6.23839319e-01
-1.78002909e-01 6.17704630e-01 -8.55104923e-01 -2.57579863e-01
-3.11381936e-01 -1.59615099e-01 7.00801790e-01 7.35602081e-01
7.38954604e-01 -6.05459869e-01 -2.42924616e-01 -1.48889378e-01
-6.44728243e-01 -8.92633379e-01 -4.90681618e-01 1.02120735e-01
-8.26883972e-01 -7.94889092e-01 -9.07154500e-01 -9.45817530e-01
9.26589191e-01 1.91013515e-01 5.43930531e-01 -2.24678025e-01
4.68473077e-01 6.53371453e-01 -2.19739661e-01 3.07059795e-01
-3.66337031e-01 1.05330713e-01 7.97462389e-02 7.09739804e-01
2.82512337e-01 -4.92905378e-01 -6.01338863e-01 2.28864953e-01
-1.31901693e+00 2.31939510e-01 6.02994442e-01 8.47331464e-01
7.19854057e-01 -4.41452302e-02 1.27158716e-01 -5.13895273e-01
4.98852730e-01 -6.17162049e-01 -1.32547453e-01 4.87488419e-01
3.00035104e-02 4.32376713e-01 5.37513733e-01 -4.06680852e-01
-1.17965341e+00 6.06943369e-01 -5.84702231e-02 -9.13810194e-01
-5.78881085e-01 4.97167170e-01 -5.24857581e-01 2.06767738e-01
2.01289281e-01 4.77325261e-01 -1.02413138e-02 -5.64001560e-01
6.51286364e-01 8.70016515e-01 9.68815386e-01 -6.58813179e-01
6.27876937e-01 6.09894931e-01 -2.10613921e-01 -6.45311832e-01
-4.37362462e-01 -5.87281346e-01 -1.29086530e+00 -3.89952481e-01
1.34603429e+00 -9.36193943e-01 -6.44256055e-01 4.42780465e-01
-1.18314493e+00 1.47642210e-01 -2.85829902e-01 5.17161846e-01
-6.40033305e-01 5.61921418e-01 -6.25510156e-01 -7.24576533e-01
-1.97089761e-02 -1.45119369e+00 1.65538621e+00 2.09106311e-01
3.38994749e-02 -1.11994720e+00 1.78443238e-01 4.06863242e-01
5.88577799e-02 1.67256773e-01 9.70026731e-01 -4.35823113e-01
-3.94912452e-01 -2.46077403e-02 -2.17482790e-01 1.21043950e-01
-7.62839988e-02 -1.65993012e-02 -1.18851364e+00 -2.95435071e-01
-1.58308208e-01 -3.51414412e-01 1.16152096e+00 2.71580130e-01
7.24351227e-01 -1.95632800e-01 -3.41283739e-01 6.72564209e-01
1.36036253e+00 2.14020193e-01 7.41459191e-01 -3.15859728e-02
9.95288551e-01 5.96502781e-01 2.09534720e-01 1.15638211e-01
6.28066480e-01 5.14858723e-01 6.10643148e-01 -3.39198858e-01
1.11522622e-01 -2.92789727e-01 6.23031199e-01 1.27653527e+00
3.35976668e-02 -1.90650105e-01 -7.55555332e-01 4.67703789e-01
-2.42897201e+00 -1.01266086e+00 -2.27968972e-02 2.27972579e+00
4.66076344e-01 -4.19567645e-01 1.51636720e-01 1.10010229e-01
7.43793905e-01 3.94555628e-02 -4.37032849e-01 -2.05439851e-01
-4.96306956e-01 -4.96708483e-01 2.55536940e-02 6.33495450e-01
-1.43145311e+00 9.50520396e-01 5.57200384e+00 6.26066267e-01
-1.07535839e+00 1.67114630e-01 1.83535099e-01 1.19905405e-01
-4.56667095e-01 8.68506134e-02 -2.90999681e-01 3.34582925e-01
7.04586804e-01 4.22226936e-01 5.72827816e-01 4.04112875e-01
-2.71730810e-01 -2.23153651e-01 -1.35306275e+00 1.23384404e+00
3.22048932e-01 -1.25921345e+00 3.09041798e-01 3.11451536e-02
6.99227929e-01 -1.40058935e-01 2.78007925e-01 -3.92284105e-03
-3.14169154e-02 -8.82676423e-01 9.19550359e-01 1.23074770e+00
9.02118623e-01 -9.50010657e-01 9.64158654e-01 4.36943620e-01
-1.48003864e+00 -6.41988888e-02 -1.59913629e-01 4.63071436e-01
1.38356581e-01 7.30370358e-02 -1.94093093e-01 1.01684678e+00
6.03157461e-01 8.51317704e-01 -6.09872341e-01 4.49535131e-01
2.24909812e-01 2.32322305e-01 -2.10563198e-01 6.15518928e-01
4.67823535e-01 -4.73760188e-01 5.41279256e-01 1.33201265e+00
4.29140210e-01 -1.45352602e-01 3.32323983e-02 7.85816908e-01
-9.31775495e-02 -9.16719511e-02 -7.80995965e-01 -2.01669335e-01
2.67259359e-01 1.28441381e+00 -4.91564035e-01 -4.73707139e-01
-4.45941359e-01 1.59653807e+00 3.80798221e-01 8.20750535e-01
-7.73284972e-01 -1.65925235e-01 2.63664812e-01 -1.64524466e-01
1.69764549e-01 -2.86565095e-01 -1.76943913e-01 -1.39698505e+00
-1.85581073e-01 -6.22109175e-01 5.46552300e-01 -9.76387322e-01
-1.20031941e+00 6.59504473e-01 8.62797629e-03 -1.40583742e+00
-2.39553317e-01 -4.21734363e-01 -4.78435308e-01 8.48563254e-01
-1.10372055e+00 -1.60898793e+00 -2.97336191e-01 1.19447410e+00
2.14635640e-01 -6.13301754e-01 9.50876474e-01 4.51957196e-01
-6.72677338e-01 5.41499972e-01 3.48690808e-01 2.06068709e-01
7.34171033e-01 -1.24802089e+00 -5.85364044e-01 7.37845838e-01
6.97647035e-02 8.98039162e-01 1.89363316e-01 -5.75355172e-01
-1.79973590e+00 -9.32101190e-01 7.57256567e-01 -1.81164071e-01
4.47332442e-01 -2.27440819e-01 -8.45893979e-01 6.73234403e-01
7.69856691e-01 -4.60333794e-01 9.33848441e-01 -1.21753722e-01
-4.09481168e-01 -1.27433807e-01 -9.10204172e-01 5.37015259e-01
5.83657324e-01 -9.25624967e-01 -5.43052793e-01 -3.43821608e-02
4.10471380e-01 -2.62015343e-01 -9.15918291e-01 3.44525278e-01
6.05535448e-01 -1.11545801e+00 8.27876747e-01 -4.26212758e-01
4.59275484e-01 -7.00181067e-01 -7.71019101e-01 -1.12176943e+00
-4.75713789e-01 -1.77680865e-01 -3.35780770e-01 1.23711169e+00
2.78028160e-01 -1.29448831e-01 6.36795163e-01 5.05121112e-01
-2.63551444e-01 -4.92760897e-01 -1.08421922e+00 -3.09087753e-01
-1.45771012e-01 -2.71199107e-01 5.08313119e-01 1.13316894e+00
9.88680348e-02 6.25743449e-01 -4.14409995e-01 4.77082759e-01
7.38327920e-01 2.61454731e-01 6.26264632e-01 -9.39583957e-01
-6.36873208e-03 -4.47812766e-01 -3.25327605e-01 -1.06231260e+00
3.68034303e-01 -1.08159375e+00 -7.76171014e-02 -1.76380289e+00
5.73199332e-01 2.23511904e-01 -5.44583857e-01 5.01060307e-01
2.19704166e-01 1.42876878e-01 1.42775565e-01 4.89526898e-01
-1.02859259e+00 8.11302423e-01 8.85832369e-01 -2.76102334e-01
-3.54972124e-01 -3.13727319e-01 -3.81482720e-01 7.16578305e-01
6.09295905e-01 -6.04943894e-02 -5.88959813e-01 -5.47919571e-01
-1.96361858e-02 3.10652554e-01 5.64197361e-01 -1.11588037e+00
8.63118231e-01 1.82459265e-01 7.10133433e-01 -1.06980860e+00
6.84935868e-01 -1.33151269e+00 4.90041226e-01 1.05656952e-01
-5.38486421e-01 1.51136041e-01 1.23522915e-01 5.50572991e-01
-5.91174901e-01 -6.85727829e-03 7.07147539e-01 1.79164186e-01
-5.38527131e-01 1.11441106e-01 -3.18185389e-01 -5.81957877e-01
8.80332708e-01 -4.29317057e-01 -1.20552398e-01 -7.16856360e-01
-1.07740414e+00 1.93264544e-01 6.12610519e-01 3.21639806e-01
9.90413308e-01 -1.79956365e+00 -4.37247127e-01 4.59156245e-01
2.57926315e-01 -1.75278276e-01 5.33645332e-01 1.18931067e+00
-2.86495566e-01 3.55968505e-01 -2.03231320e-01 -8.99257362e-01
-1.14561880e+00 6.88708246e-01 2.98815608e-01 7.11806938e-02
-1.87232301e-01 4.76264894e-01 3.75596732e-01 -6.91996396e-01
5.33955336e-01 6.55401349e-02 -2.20948130e-01 1.27425984e-01
9.35179889e-02 4.47295547e-01 -3.47948015e-01 -1.45710862e+00
-3.78733307e-01 6.93246484e-01 1.76099241e-01 -6.31774783e-01
1.01650620e+00 -3.25933665e-01 -4.71463442e-01 7.85976112e-01
1.43412423e+00 -1.60846800e-01 -1.21147776e+00 -2.80458480e-01
-1.38905138e-01 1.01054152e-02 -3.97974290e-02 -6.39184594e-01
-1.06016922e+00 1.14473045e+00 7.58965075e-01 -1.63587287e-01
1.53868544e+00 8.86393432e-03 6.80778801e-01 2.69127518e-01
3.45002189e-02 -1.35195470e+00 2.55433246e-02 5.71376920e-01
7.93657005e-01 -1.12513828e+00 -3.53492022e-01 -1.52703291e-02
-9.04950619e-01 1.11007583e+00 4.95421946e-01 2.08342090e-01
6.80694163e-01 1.70075536e-01 -8.90311524e-02 -2.66561925e-01
-4.47491646e-01 -1.90495878e-01 7.69400120e-01 4.00257945e-01
3.41738582e-01 5.64645045e-02 7.30853006e-02 5.98043084e-01
2.78827816e-01 -2.20637187e-01 -2.15659454e-01 7.95934737e-01
-2.91468024e-01 -1.00014639e+00 -5.37914515e-01 1.15884781e-01
2.39022225e-01 4.56759967e-02 -7.83565819e-01 5.98799109e-01
3.57861966e-01 8.82660449e-01 1.71334833e-01 -9.08153415e-01
1.41430832e-02 3.87524486e-01 3.61072809e-01 -1.49269000e-01
-4.85707253e-01 7.67859161e-01 -1.34014666e-01 -6.45218909e-01
-8.70563090e-01 -4.85314757e-01 -1.60341287e+00 -2.71345586e-01
-3.83467227e-02 3.24658781e-01 6.39859319e-01 8.37711036e-01
5.95924377e-01 3.45011473e-01 6.12894714e-01 -9.94009495e-01
3.99538800e-02 -8.08722198e-01 -5.46153605e-01 5.89330852e-01
2.82052547e-01 -4.99478966e-01 7.19750449e-02 4.83845383e-01] | [13.128105163574219, 5.003395080566406] |
550c41e0-836e-4fe8-891b-38ecc7c4416a | acnlp-at-semeval-2020-task-6-a-supervised | null | null | https://aclanthology.org/2020.semeval-1.58 | https://aclanthology.org/2020.semeval-1.58.pdf | ACNLP at SemEval-2020 Task 6: A Supervised Approach for Definition Extraction | We describe our contribution to two of the subtasks of SemEval 2020 Task 6, DeftEval: Extracting term-definition pairs in free text. The system for Subtask 1: Sentence Classification is based on a transformer architecture where we use transfer learning to fine-tune a pretrained model on the downstream task, and the one for Subtask 3: Relation Classification uses a Random Forest classifier with handcrafted dedicated features. Our systems respectively achieve 0.830 and 0.994 F1-scores on the official test set, and we believe that the insights derived from our study are potentially relevant to help advance the research on definition extraction. | ['Mhamed Hajaiej', 'Mathis Linger', 'Pirashanth Ratnamogan', 'Fabien Caspani'] | 2020-12-01 | null | null | null | semeval-2020 | ['definition-extraction'] | ['natural-language-processing'] | [ 4.83597130e-01 6.87374353e-01 -4.02178228e-01 -4.98912126e-01
-1.15892899e+00 -7.56730795e-01 8.17755878e-01 3.42049658e-01
-5.78488410e-01 1.22910833e+00 2.85212040e-01 -7.45664597e-01
-2.63552129e-01 -6.66469574e-01 -5.76590180e-01 -2.70167470e-01
-1.44846305e-01 5.62562525e-01 1.90002263e-01 -7.08856523e-01
-7.72472918e-02 8.87175575e-02 -9.32661057e-01 7.43285477e-01
7.79362440e-01 1.28043354e+00 -2.01097518e-01 7.50862777e-01
7.92134404e-02 9.38614488e-01 -7.05257714e-01 -8.78108442e-01
9.96274054e-02 -1.87924623e-01 -1.68046236e+00 -5.10723829e-01
3.10691029e-01 1.33001551e-01 -1.31456733e-01 8.40903580e-01
1.63596764e-01 1.40056148e-01 7.76368499e-01 -9.22838151e-01
-5.61373711e-01 1.08600736e+00 -1.85564846e-01 4.44572717e-01
5.98068655e-01 -5.22578359e-02 1.56990957e+00 -7.59295762e-01
7.68412173e-01 7.34794497e-01 8.81071866e-01 5.91882408e-01
-1.31016636e+00 -5.29763043e-01 5.02332486e-03 3.80207866e-01
-1.22069407e+00 -7.01531172e-01 2.03380451e-01 -3.81752819e-01
1.92532468e+00 5.22005200e-01 3.87902886e-01 1.18899345e+00
3.66143405e-01 7.00137854e-01 9.56090808e-01 -7.36334622e-01
-1.02564596e-01 1.42495036e-01 6.88503861e-01 6.20876551e-01
1.89075932e-01 5.61292581e-02 -5.53233027e-01 -8.72273967e-02
2.03507543e-01 -7.69016683e-01 -4.21446472e-01 2.51624018e-01
-1.08392668e+00 7.95227289e-01 2.94682682e-01 6.13620698e-01
-2.55271584e-01 -8.59298706e-02 6.24135137e-01 8.48272204e-01
7.43693709e-01 9.41330969e-01 -1.29303551e+00 -2.37826705e-01
-8.97797585e-01 2.54897445e-01 1.19342661e+00 1.20622694e+00
6.43993437e-01 -6.60743713e-01 -5.47391176e-01 8.70205224e-01
7.55823180e-02 9.55093727e-02 4.77755427e-01 -4.99737084e-01
8.97220433e-01 4.32490408e-01 -1.04318596e-01 -3.25094640e-01
-5.85052133e-01 -5.61875284e-01 -4.10967231e-01 -5.13270676e-01
3.20801109e-01 -6.54809535e-01 -9.31468129e-01 1.54219663e+00
-1.00715786e-01 -1.14157975e-01 2.29601920e-01 4.25573945e-01
1.24757719e+00 8.62384066e-02 3.21272314e-01 -2.52371371e-01
1.58700764e+00 -1.12147653e+00 -6.87002182e-01 -3.60087663e-01
1.08161461e+00 -7.99452841e-01 6.10306442e-01 7.63447136e-02
-9.38619912e-01 -2.63765633e-01 -1.17641580e+00 -1.74491808e-01
-5.48422515e-01 5.32225426e-03 7.81069160e-01 4.01655227e-01
-9.64367867e-01 5.29431462e-01 -5.84011436e-01 -2.62546152e-01
3.31017137e-01 3.72208267e-01 -5.26878297e-01 2.64386296e-01
-2.00666118e+00 1.32844341e+00 5.62195480e-01 4.67716046e-02
-4.34802204e-01 -7.87530720e-01 -8.67506385e-01 8.98432359e-02
4.17270124e-01 -9.01924014e-01 1.63837445e+00 -6.09943986e-01
-1.22899187e+00 1.15302145e+00 -8.69077295e-02 -8.10716331e-01
1.06819220e-01 -4.08926040e-01 -5.92141211e-01 -6.26185685e-02
4.21008050e-01 2.09783763e-02 2.17156455e-01 -5.94751060e-01
-8.36700022e-01 -1.18848830e-01 1.60638034e-01 1.02613345e-01
-1.53053313e-01 3.01285803e-01 -1.54911861e-01 -4.60313588e-01
-4.02304977e-01 -5.82600951e-01 -5.61433844e-02 -9.58142042e-01
-6.00840867e-01 -6.85757220e-01 6.74071983e-02 -9.08139646e-01
1.32482302e+00 -1.37843466e+00 1.14911430e-01 4.86606210e-02
3.11964244e-01 4.12437618e-01 5.23009486e-02 5.53371668e-01
-3.33114624e-01 2.26043239e-01 -1.37067184e-01 -1.41287014e-01
-1.20244883e-02 -3.16377610e-01 -2.30937362e-01 -1.07278898e-01
5.17294586e-01 1.26961601e+00 -1.17537498e+00 -4.38869178e-01
-1.37656063e-01 -3.37534435e-02 7.32959732e-02 2.34243974e-01
-2.29754478e-01 -6.35802820e-02 -5.37642419e-01 2.29210973e-01
3.45049679e-01 -1.68495297e-01 3.88744026e-01 -1.76876545e-01
1.96376011e-01 1.31864655e+00 -3.94520432e-01 1.58770752e+00
-5.85926175e-01 7.30766296e-01 -9.23162326e-02 -1.11964405e+00
6.57508194e-01 6.40355766e-01 1.47642925e-01 -3.05871189e-01
3.06329057e-02 1.28641844e-01 2.20502913e-01 -6.83605850e-01
4.28455442e-01 -3.36193204e-01 -2.76916057e-01 3.80684912e-01
4.79972452e-01 -1.20174192e-01 3.41605783e-01 3.85428429e-01
1.63154399e+00 1.07921809e-01 4.76071805e-01 -3.94759119e-01
6.58811510e-01 -1.17541486e-02 3.66302013e-01 6.42578781e-01
-1.17103778e-01 3.22055697e-01 6.73624158e-01 -3.76667887e-01
-4.31608409e-01 -1.05209398e+00 -4.00515825e-01 1.16737735e+00
-5.14819026e-01 -8.21405232e-01 -6.80127621e-01 -1.33119011e+00
-2.33388841e-02 1.01898623e+00 -7.86978900e-01 -6.96287528e-02
-5.29473603e-01 -8.85469437e-01 8.81486714e-01 5.92973590e-01
5.22951365e-01 -1.10050368e+00 -1.33114457e-01 2.52094567e-01
-4.82508868e-01 -1.35150003e+00 -2.90123314e-01 6.24424398e-01
-3.98555368e-01 -1.23992193e+00 -2.14559317e-01 -1.02697635e+00
1.77061826e-01 -3.44082147e-01 1.65469253e+00 5.99922333e-03
6.12469576e-02 -2.86229700e-03 -5.56767583e-01 -3.83294761e-01
-3.04802567e-01 8.80077779e-01 -2.21105844e-01 -4.02582824e-01
1.11675107e+00 -4.53560591e-01 -1.83841020e-01 -2.03988045e-01
-2.58945793e-01 1.13047637e-01 5.12404025e-01 9.89687681e-01
6.35950491e-02 -2.80088902e-01 8.65132451e-01 -1.40310025e+00
8.73085141e-01 -4.60397810e-01 -1.97179630e-01 5.59735894e-01
-1.05976212e+00 1.55594170e-01 5.24530828e-01 4.99526039e-02
-1.14180839e+00 -1.17772721e-01 -3.86232793e-01 4.52385694e-01
-3.07200737e-02 7.08260357e-01 -2.09379733e-01 3.34162682e-01
8.41735244e-01 1.58201512e-02 -3.99556398e-01 -3.99233401e-01
3.81117076e-01 9.74308550e-01 4.30792838e-01 -5.96346319e-01
7.38056839e-01 -3.24500531e-01 -1.75018966e-01 -5.28462410e-01
-1.51504767e+00 -4.59443003e-01 -8.20633233e-01 1.59528226e-01
8.80029202e-01 -7.27915525e-01 -3.81128728e-01 1.78108782e-01
-1.20793271e+00 -5.39979041e-01 -2.23608106e-01 2.31987938e-01
-3.57571602e-01 2.74946354e-02 -9.11629498e-01 -5.88256955e-01
-9.76918340e-01 -5.78373253e-01 9.39063013e-01 1.76614262e-02
-6.80393398e-01 -1.30649209e+00 2.17765331e-01 7.20369518e-01
3.64217401e-01 7.93576837e-02 9.07044113e-01 -1.17049623e+00
-1.29309073e-01 -3.06462467e-01 -2.00650096e-01 3.66997391e-01
1.86853841e-01 -2.81264186e-01 -1.05342031e+00 3.47882062e-02
-1.68176353e-01 -6.39759243e-01 1.20551586e+00 1.05418168e-01
9.58248198e-01 -2.35057533e-01 -5.67709386e-01 3.37595582e-01
9.12783980e-01 -2.39037409e-01 4.01827186e-01 4.87795860e-01
4.64087874e-01 5.75747728e-01 7.35419571e-01 -1.44313812e-01
6.80951715e-01 7.11379468e-01 -1.88343301e-01 -4.53125918e-03
-2.22222418e-01 -2.70057619e-01 6.65758476e-02 6.23493731e-01
-2.52312601e-01 -6.13676244e-03 -1.15935981e+00 6.92189693e-01
-1.95206094e+00 -1.01747906e+00 -2.14109734e-01 1.79665959e+00
1.66235375e+00 5.54496884e-01 -6.62847981e-02 1.00499421e-01
4.49002743e-01 1.51663885e-01 1.91742238e-02 -4.78218406e-01
-1.20710805e-02 9.16898489e-01 4.04991329e-01 8.83537114e-01
-1.46196508e+00 1.26117218e+00 6.61183405e+00 8.76601398e-01
-4.98042196e-01 1.99629769e-01 5.75630307e-01 8.53395760e-02
-3.09472233e-01 4.18780223e-02 -1.07244563e+00 1.88185111e-01
1.48005724e+00 -2.42443591e-01 1.02764256e-01 3.65017503e-01
-1.96096435e-01 4.66738604e-02 -1.47878969e+00 2.93190181e-01
-6.65321574e-02 -1.27030122e+00 -2.49022037e-01 -1.37842178e-01
5.01202226e-01 2.41137549e-01 -3.94990891e-01 9.06617820e-01
4.80702996e-01 -1.35379219e+00 1.50719166e-01 5.41611850e-01
9.61948454e-01 -5.97007513e-01 1.21926677e+00 3.28032255e-01
-1.05048907e+00 4.63959068e-01 5.84242679e-02 -4.27134335e-01
1.09172128e-01 8.72477531e-01 -1.31386030e+00 6.49988949e-01
2.46840462e-01 6.94843948e-01 -6.25009239e-01 6.09671533e-01
-9.81144667e-01 1.04745007e+00 -4.46252897e-02 -3.93111259e-01
-1.17667489e-01 2.64846116e-01 3.91092002e-01 1.64135838e+00
-5.16984403e-01 1.59534708e-01 -1.44161597e-01 5.96632421e-01
-3.86660516e-01 1.44290701e-01 -3.25832099e-01 6.78061843e-02
5.42792082e-01 1.42883682e+00 -2.34573081e-01 -5.49104035e-01
-4.05727446e-01 8.12390208e-01 8.46447945e-01 3.04107159e-01
-6.69513166e-01 -8.14747691e-01 4.07102257e-01 -2.51689613e-01
3.31039041e-01 3.94494943e-02 -7.33722985e-01 -1.51183927e+00
4.66131717e-02 -7.72832096e-01 4.82282996e-01 -3.51658821e-01
-1.50689566e+00 9.40584481e-01 -2.42120642e-02 -4.43084896e-01
-6.75256133e-01 -6.74526572e-01 -4.96909797e-01 1.26643157e+00
-1.43743634e+00 -1.37236547e+00 1.91286787e-01 2.68104851e-01
3.75206500e-01 -2.12398782e-01 1.35951388e+00 2.29354843e-01
-4.90242690e-01 1.07462335e+00 -3.84781599e-01 6.81305587e-01
7.66029477e-01 -1.73699307e+00 9.03724670e-01 5.91925144e-01
3.29851389e-01 9.08256650e-01 6.00576937e-01 -5.56228995e-01
-8.15679729e-01 -1.16345811e+00 1.80901694e+00 -9.12003219e-01
9.31201458e-01 -4.39836234e-01 -5.75340688e-01 1.13064694e+00
3.78293842e-01 -1.89939871e-01 1.05380344e+00 1.09439111e+00
-3.67997706e-01 6.88061342e-02 -9.94689465e-01 1.77205533e-01
1.12457919e+00 -6.96024358e-01 -1.21319556e+00 6.45357788e-01
8.23353291e-01 -5.55689633e-01 -1.41244578e+00 4.79453802e-01
6.91272199e-01 -4.05459970e-01 6.60982311e-01 -1.08919442e+00
7.00315535e-01 2.48481870e-01 -7.82635901e-03 -1.35059869e+00
-3.53057832e-01 -5.46695828e-01 -3.44275475e-01 1.37106025e+00
1.44959950e+00 -6.49075270e-01 6.32427871e-01 6.72148764e-01
1.03203207e-01 -1.10058618e+00 -9.20979738e-01 -7.07460523e-01
5.50798774e-01 -5.35833359e-01 4.70188379e-01 9.23576534e-01
4.83966947e-01 1.42118335e+00 1.77032277e-01 -2.27580726e-01
3.23598385e-01 1.92367002e-01 4.88287061e-01 -1.07061923e+00
-5.68039119e-01 -2.65105188e-01 -1.71258837e-01 -1.13361216e+00
5.62632561e-01 -1.31201518e+00 -4.69250605e-02 -1.46673381e+00
3.81253690e-01 -2.73318559e-01 -4.41022068e-01 8.07685494e-01
-5.88860273e-01 1.32344693e-01 -1.83410615e-01 -9.27965790e-02
-6.26281381e-01 3.95074666e-01 7.80321598e-01 -1.61584228e-01
1.39135243e-02 4.22550887e-01 -1.05867279e+00 2.98751593e-01
6.60255909e-01 -3.45827579e-01 -1.48325369e-01 -4.07077968e-01
1.65912077e-01 -2.38456324e-01 -1.06473923e-01 -3.01139593e-01
-1.15026899e-01 2.42176726e-02 2.43095562e-01 -3.09604317e-01
1.56472042e-01 -3.13664913e-01 -5.07794976e-01 2.85883367e-01
-6.90787017e-01 -2.93832928e-01 1.34920314e-01 1.67478189e-01
-8.60304907e-02 -3.81659776e-01 2.37415791e-01 -2.41817180e-02
-3.60275954e-01 -3.41629013e-02 -3.46541107e-01 5.29545248e-01
7.53944874e-01 4.91402805e-01 -4.55911845e-01 -2.91736811e-01
-8.47926795e-01 4.63267773e-01 -1.09962791e-01 3.74115735e-01
3.10121596e-01 -1.07979476e+00 -1.09502029e+00 1.66606121e-02
1.76822975e-01 1.90531183e-02 -5.58149457e-01 7.25516856e-01
-2.35136643e-01 8.97500575e-01 3.94197106e-01 -1.54319614e-01
-1.36816001e+00 1.54125109e-01 5.06734610e-01 -9.24503863e-01
-4.41227138e-01 1.43935990e+00 -2.18867511e-01 -5.38894773e-01
-4.82290089e-02 -3.22124988e-01 -2.93530524e-01 -8.65122080e-02
6.79535806e-01 1.48497403e-01 5.46737254e-01 -2.92055190e-01
-7.28322744e-01 -1.46195956e-03 -6.18147790e-01 -1.65749773e-01
1.54190600e+00 1.98779181e-01 -3.10404539e-01 3.93883556e-01
1.30052638e+00 -1.21630374e-02 -3.70557368e-01 -4.47785974e-01
6.54645920e-01 1.69819444e-01 1.73513949e-01 -1.42271852e+00
-5.43208838e-01 4.33992594e-01 -8.39350596e-02 4.40410942e-01
1.13237703e+00 1.62981004e-01 9.79367733e-01 5.72285771e-01
3.23868304e-01 -9.99883831e-01 -6.03435695e-01 1.04074013e+00
9.54746664e-01 -1.23306370e+00 5.74978925e-02 -7.36589432e-01
-3.80924672e-01 1.14098752e+00 4.40091699e-01 -9.08368360e-03
6.83263719e-01 5.57706356e-01 -4.07490954e-02 -4.23157245e-01
-1.34741282e+00 -3.84781182e-01 5.94326079e-01 6.36045694e-01
1.11096692e+00 2.84173638e-01 -8.70295644e-01 1.15426207e+00
-6.66393161e-01 -1.84656251e-02 1.04474306e-01 6.55627370e-01
-1.59157485e-01 -1.41124833e+00 2.24448070e-01 9.01176214e-01
-7.01015413e-01 -7.31713116e-01 -7.70358801e-01 6.37244701e-01
1.19892463e-01 1.24347913e+00 -3.44976068e-01 -7.42371738e-01
3.85637492e-01 4.86024320e-01 7.67400146e-01 -1.20507586e+00
-9.97527599e-01 -5.20714819e-01 1.19263661e+00 -3.30635428e-01
-2.41594791e-01 -7.80099809e-01 -1.10897720e+00 1.22133285e-01
-6.24961019e-01 5.56321681e-01 1.74821749e-01 1.37009168e+00
1.83965892e-01 6.89639986e-01 3.99482816e-01 -1.30153567e-01
-8.33470106e-01 -1.68825459e+00 -2.58812279e-01 2.19642386e-01
3.34547907e-01 -4.32458013e-01 -3.86628151e-01 -3.49210761e-02] | [9.693426132202148, 8.932903289794922] |
19302f70-3e14-442f-859a-74a15575cf62 | ace-net-fine-level-face-alignment-through | 2012.01461 | null | https://arxiv.org/abs/2012.01461v2 | https://arxiv.org/pdf/2012.01461v2.pdf | ACE-Net: Fine-Level Face Alignment through Anchors and Contours Estimation | We propose a novel facial Anchors and Contours Estimation framework, ACE-Net, for fine-level face alignment tasks. ACE-Net predicts facial anchors and contours that are richer than traditional facial landmarks while overcoming ambiguities and inconsistencies in their definitions. We introduce a weakly supervised loss enabling ACE-Net to learn from existing facial landmarks datasets without the need for reannotation. Instead, synthetic data, from which GT contours can be easily obtained, is used during training to bridge the density gap between landmarks and true facial contours. We evaluate the face alignment accuracy of ACE-Net with respect to the HELEN dataset which has 194 annotated facial landmarks, while it is trained with only 68 or 36 landmarks from the 300-W dataset. We show that ACE-Net generated contours are better than contours interpolated straight from the 68 GT landmarks and ACE-Net also outperforms models trained only with full supervision from GT landmarks-based contours. | ['Amir Tamrakar', 'Jihua Huang'] | 2020-12-02 | null | null | null | null | ['face-alignment'] | ['computer-vision'] | [-4.57280055e-02 8.87157381e-01 -2.49904260e-01 -8.96752298e-01
-1.06962514e+00 -4.08052295e-01 6.02118850e-01 -4.34460104e-01
-2.50823885e-01 6.32631719e-01 3.57659370e-01 2.75065213e-01
2.57364780e-01 -5.12054443e-01 -6.89676821e-01 -3.03743899e-01
-5.74276932e-02 7.03271687e-01 -1.99888870e-01 -2.77328283e-01
-2.53934026e-01 8.26540232e-01 -1.30684841e+00 -1.36505112e-01
5.03667772e-01 1.25780785e+00 -5.90372205e-01 2.94933558e-01
-1.85848817e-01 2.14225098e-01 -3.28639537e-01 -9.59010243e-01
7.60388911e-01 -3.95652652e-01 -4.78484452e-01 1.16615459e-01
1.49949265e+00 -4.28589702e-01 -6.60430416e-02 8.90837312e-01
5.63806057e-01 -9.67951939e-02 7.68052518e-01 -1.44219351e+00
-5.62279880e-01 2.23649114e-01 -9.37702477e-01 -3.17404717e-01
2.07456410e-01 -3.32356483e-01 1.01041031e+00 -1.45644259e+00
9.51868057e-01 1.53140700e+00 1.40842664e+00 1.00774944e+00
-1.30109441e+00 -8.80268633e-01 6.77362233e-02 -2.87366360e-01
-1.68058896e+00 -1.14477408e+00 9.09922600e-01 -1.98474124e-01
4.18681085e-01 -1.05259351e-01 6.02176607e-01 9.25351202e-01
-2.45202273e-01 4.13134515e-01 6.56178713e-01 -3.83911341e-01
-7.43241329e-03 -8.60831439e-02 -4.02132004e-01 1.19829857e+00
-1.69741422e-01 2.04577133e-01 -8.21084619e-01 -1.53256461e-01
1.07029462e+00 -5.86149812e-01 -9.52187404e-02 -4.23128635e-01
-7.94866800e-01 6.48236334e-01 6.95612073e-01 -1.54087646e-02
-2.18533382e-01 2.23933235e-01 1.77287217e-02 6.12253770e-02
9.06216443e-01 1.93867721e-02 -3.77683461e-01 1.35128304e-01
-1.44982779e+00 1.76813275e-01 5.85938156e-01 1.26254869e+00
1.14647198e+00 2.69281149e-01 -3.83143872e-02 1.01340926e+00
4.04237241e-01 5.34209549e-01 -7.71367028e-02 -1.34276152e+00
2.23529607e-01 2.90768206e-01 -1.10659217e-02 -9.79275346e-01
-3.43212277e-01 -7.62280375e-02 -6.75045192e-01 4.94695425e-01
7.64821231e-01 -3.33493888e-01 -1.31354070e+00 1.87677526e+00
5.03339946e-01 7.14542508e-01 -2.48327732e-01 6.48745418e-01
8.84963512e-01 1.57974228e-01 6.29880577e-02 1.70645311e-01
1.07140911e+00 -8.72835934e-01 -5.39321542e-01 -2.78578639e-01
3.87699753e-01 -7.89021373e-01 8.00790429e-01 -5.59708476e-02
-1.16674030e+00 -5.51895618e-01 -8.46381664e-01 -4.02987748e-01
-1.54640153e-01 1.70988694e-01 5.77533484e-01 6.47870898e-01
-1.69311655e+00 5.70451260e-01 -8.06583285e-01 -1.32969648e-01
1.10049951e+00 5.28301775e-01 -8.83016586e-01 1.90278694e-01
-6.70235932e-01 8.92524362e-01 -8.04853216e-02 5.90894103e-01
-5.80489278e-01 -1.29409051e+00 -1.36421645e+00 -1.60970390e-01
2.12857090e-02 -3.99796844e-01 1.03245950e+00 -1.37204969e+00
-1.55994785e+00 1.19472098e+00 -4.97948021e-01 -2.16211483e-01
7.79464543e-01 -3.30684008e-04 -1.20700695e-01 2.69083440e-01
2.06082329e-01 1.59761834e+00 8.82556796e-01 -1.40899396e+00
-3.96721452e-01 -3.98945063e-01 -3.22829455e-01 -8.92956778e-02
-4.12683822e-02 -9.30233821e-02 -4.28057075e-01 -6.48041248e-01
2.47323383e-02 -8.40627849e-01 -1.18970416e-01 8.62498283e-01
-5.57559192e-01 -2.59041995e-01 8.81466746e-01 -8.53710055e-01
4.48566228e-01 -1.92854798e+00 -1.25098974e-01 7.67873466e-01
1.97087914e-01 1.00017391e-01 -5.77771902e-01 -2.37904787e-01
-1.70666337e-01 1.24271683e-01 -2.46357784e-01 -1.14230895e+00
1.27991587e-01 1.81062520e-01 -1.79325849e-01 5.69557130e-01
4.04641628e-01 1.04201412e+00 -5.30190349e-01 -8.21992218e-01
-1.43947704e-02 8.42441857e-01 -7.30637193e-01 9.84705463e-02
-2.04948917e-01 4.76760328e-01 -1.12913430e-01 1.15216660e+00
9.60259497e-01 1.13505617e-01 -1.88642833e-02 -4.21869189e-01
1.91595301e-01 -2.77131557e-01 -7.80665576e-01 1.81436324e+00
-3.83669704e-01 6.65427327e-01 2.68081397e-01 -4.93776500e-01
1.29584253e+00 3.43212813e-01 6.29540563e-01 -6.71037138e-01
7.47673726e-03 1.38746142e-01 -5.55335939e-01 1.75502658e-01
1.80438519e-01 -2.70178974e-01 2.37951219e-01 1.53324619e-01
5.09472430e-01 -3.07124078e-01 -1.77645341e-01 -1.66355059e-01
5.08237600e-01 6.01610124e-01 4.19327356e-02 -3.08326304e-01
4.52498972e-01 -2.60970980e-01 8.64259303e-01 4.59117666e-02
-2.58353174e-01 9.89476383e-01 4.71056044e-01 -7.21225679e-01
-1.23746955e+00 -1.31238329e+00 -4.25207734e-01 9.76267338e-01
-2.48206601e-01 -4.79082406e-01 -1.03023493e+00 -9.16909814e-01
1.92131206e-01 1.94250360e-01 -9.72564042e-01 2.43817091e-01
-6.30364716e-01 -1.83281586e-01 8.71321619e-01 6.17446959e-01
5.13275802e-01 -9.49226618e-01 1.57686442e-01 -8.16222951e-02
-1.06634879e-02 -1.12543833e+00 -1.10235798e+00 -4.03650045e-01
-5.78315318e-01 -1.07688701e+00 -7.40712821e-01 -8.97933483e-01
1.27359784e+00 -4.82758433e-01 1.28981400e+00 3.42020333e-01
-2.99154639e-01 3.74337345e-01 1.92562044e-01 -4.42151189e-01
-9.94011983e-02 1.68406907e-02 -8.21563080e-02 1.42932579e-01
5.38533032e-01 -6.65467322e-01 -5.15394568e-01 4.60335791e-01
-2.50333130e-01 -9.09654126e-02 1.34258270e-01 8.64890814e-01
7.49998987e-01 -6.75976098e-01 7.55260289e-01 -8.66017342e-01
-1.13354204e-02 -1.63100317e-01 -7.95342565e-01 2.64791399e-01
-3.10195953e-01 -8.99239108e-02 1.72377959e-01 -1.14064127e-01
-1.13070881e+00 4.25599754e-01 -5.81718624e-01 -6.71153486e-01
-1.52288333e-01 -2.33762190e-02 -1.46609202e-01 -7.10841835e-01
5.33214867e-01 -3.58252525e-01 3.33925903e-01 -4.23078507e-01
5.41940928e-01 2.53532380e-01 8.75971377e-01 -1.00342536e+00
1.03299570e+00 5.95335126e-01 2.65620649e-01 -7.81912863e-01
-1.05698371e+00 -7.41235614e-02 -1.19845355e+00 -2.24870905e-01
7.88162708e-01 -9.26676452e-01 -6.45458817e-01 2.63649851e-01
-1.22245049e+00 -4.98723149e-01 -4.52322364e-01 -5.03339991e-02
-7.56743729e-01 1.42069086e-01 -5.85232615e-01 -4.96250033e-01
-3.32799941e-01 -8.79608631e-01 1.57899404e+00 3.38774204e-01
-4.45911199e-01 -1.27262652e+00 -3.06973737e-02 1.18826710e-01
2.31566608e-01 6.95986211e-01 4.66009438e-01 -3.14017892e-01
-3.37521344e-01 -1.11224495e-01 -2.50274420e-01 3.63863140e-01
4.34160382e-01 3.07979345e-01 -1.29811263e+00 -2.58033127e-01
-6.00816369e-01 -5.92244983e-01 5.20856559e-01 3.00773531e-01
1.10915077e+00 -4.09291267e-01 -2.28757218e-01 1.16292822e+00
1.13749194e+00 -1.39529049e-01 6.16971076e-01 3.73047329e-02
7.47176409e-01 8.65495920e-01 2.83697933e-01 1.76609382e-01
4.80072796e-01 7.11438239e-01 2.65552282e-01 -5.60320020e-01
-3.31322610e-01 -7.48848200e-01 1.32482052e-01 2.78553039e-01
-3.04836333e-01 4.45509642e-01 -7.77225316e-01 4.95502502e-01
-1.37096572e+00 -7.65384793e-01 4.15118068e-01 1.87229514e+00
1.12310660e+00 -2.08927825e-01 1.84757024e-01 -4.28166509e-01
6.32159591e-01 1.40095383e-01 -4.49965924e-01 -1.50330186e-01
-1.28421411e-01 7.62727618e-01 3.75946015e-01 1.02034283e+00
-1.10138059e+00 1.35367048e+00 7.35099697e+00 7.39070952e-01
-9.71859455e-01 -7.48650283e-02 1.12085736e+00 2.02547340e-03
-3.43863875e-01 -2.14990184e-01 -1.08985305e+00 1.72140300e-01
6.06307328e-01 -3.74460556e-02 1.46577626e-01 1.01668084e+00
-9.97731835e-02 3.47087234e-01 -1.29378414e+00 9.61469173e-01
2.61395365e-01 -1.57229745e+00 1.23904750e-01 5.33278994e-02
8.19499016e-01 -1.56043664e-01 1.54695079e-01 -8.53612125e-02
4.62825239e-01 -1.49386108e+00 7.59723306e-01 5.77300131e-01
1.33011615e+00 -8.84354770e-01 5.71222723e-01 -1.76099300e-01
-1.28032935e+00 5.84793925e-01 -4.91459519e-01 3.82554948e-01
2.02785373e-01 2.73450822e-01 -1.03977346e+00 3.82689565e-01
5.01942039e-01 7.54035294e-01 -5.82761467e-01 6.95116162e-01
-2.05704123e-01 3.41851294e-01 -6.92016780e-01 7.14297891e-01
1.87326536e-01 -5.02948463e-01 9.02440995e-02 1.06704938e+00
4.73454297e-01 -3.64270527e-03 2.20605016e-01 9.66558993e-01
-5.26174963e-01 4.49181274e-02 -7.04045951e-01 3.89273584e-01
5.78246057e-01 1.50184619e+00 -6.10219717e-01 -2.83369608e-02
-3.82037103e-01 7.98543215e-01 5.52415431e-01 4.38036591e-01
-6.55864298e-01 -5.24056144e-02 1.04973090e+00 2.79748976e-01
4.28408943e-02 -1.31813377e-01 -1.94368020e-01 -8.13519478e-01
-6.28636628e-02 -5.53692579e-01 1.49204269e-01 -7.84773529e-01
-1.50075650e+00 8.98113072e-01 -1.54943364e-02 -8.22914183e-01
-4.11509186e-01 -5.37525117e-01 -7.33127475e-01 1.00783741e+00
-1.73644197e+00 -1.83848226e+00 -3.56436282e-01 6.85497999e-01
1.96806654e-01 -2.08795756e-01 8.24767649e-01 2.55355150e-01
-2.59210169e-01 1.24068439e+00 -5.58558047e-01 7.77144670e-01
9.05562937e-01 -1.19241798e+00 7.31319726e-01 4.92789894e-01
4.33575630e-01 5.08710444e-01 2.02864662e-01 -6.11447573e-01
-6.17539585e-01 -1.18343234e+00 9.74470913e-01 -5.90619922e-01
3.08987081e-01 -6.70334160e-01 -7.26574123e-01 1.26493633e+00
4.87561487e-02 6.94211841e-01 6.77128315e-01 7.61990622e-02
-5.77815413e-01 -2.57244021e-01 -1.58649337e+00 6.96790457e-01
1.30096483e+00 -5.92050374e-01 -3.85943264e-01 1.05278879e-01
3.21572602e-01 -6.39765978e-01 -1.01266253e+00 3.55590820e-01
7.58772492e-01 -7.89512038e-01 1.21855378e+00 -4.99450684e-01
1.07884429e-01 -1.75647259e-01 -2.63865869e-02 -1.31220472e+00
6.87445104e-02 -7.52643466e-01 4.39690620e-01 1.46160960e+00
5.29322326e-01 -5.68386614e-01 1.49224424e+00 7.61192918e-01
-1.37235343e-01 -6.54345214e-01 -1.36466527e+00 -2.93700963e-01
2.02246517e-01 -1.29996344e-01 9.31671083e-01 1.03887296e+00
-3.25159371e-01 -1.16208263e-01 -2.75504529e-01 9.00472607e-03
9.84275222e-01 -3.94653976e-01 9.66427147e-01 -1.34518027e+00
4.22059596e-01 -3.40976894e-01 -5.10679126e-01 -8.80881369e-01
9.61493552e-01 -9.62177634e-01 7.13851750e-02 -1.08158994e+00
-3.99238229e-01 -7.65017092e-01 2.88813502e-01 1.06081259e+00
1.51842564e-01 1.11352623e+00 1.17626987e-01 -2.44355515e-01
-2.08883341e-02 7.07879841e-01 1.34485245e+00 -1.32853031e-01
2.13284828e-02 -1.25979170e-01 -6.03376567e-01 1.23919821e+00
4.94429022e-01 -1.49718076e-01 -3.34740251e-01 -2.19392389e-01
-3.81410010e-02 3.60119119e-02 3.09827745e-01 -7.43390143e-01
3.76019627e-01 -6.47700299e-03 1.01345873e+00 -4.03983235e-01
6.37940347e-01 -8.17782402e-01 1.42928809e-01 -3.31062019e-01
-2.25829989e-01 1.21070474e-01 2.95714259e-01 1.72741637e-01
-1.88396707e-01 6.41621798e-02 1.03092301e+00 -1.88080827e-03
-5.40442944e-01 8.50014567e-01 4.55231786e-01 1.04155980e-01
8.34751844e-01 -4.94193017e-01 -1.31870687e-01 -4.80060875e-01
-1.19710290e+00 2.40804076e-01 6.23842299e-01 2.96088547e-01
7.56334305e-01 -1.77194560e+00 -7.78571844e-01 5.72706580e-01
-1.26975358e-01 3.42003554e-01 -1.00999363e-01 4.37796384e-01
-7.20997453e-01 4.86466587e-02 -3.89683455e-01 -6.02113307e-01
-1.28745472e+00 -2.80349739e-02 7.85483599e-01 3.41875046e-01
-6.18777990e-01 1.06411934e+00 2.36311764e-01 -7.70670354e-01
2.61037469e-01 -9.14183483e-02 4.52733552e-03 1.17922977e-01
4.60341632e-01 1.22167587e-01 3.24107967e-02 -1.21766412e+00
-3.62254977e-01 1.21548319e+00 1.13989606e-01 -1.91810697e-01
1.26016498e+00 1.16478642e-02 -2.27774143e-01 -1.02321751e-01
1.20685148e+00 3.54393661e-01 -1.79646778e+00 -1.93481162e-01
1.86528608e-01 -5.55073619e-01 -2.04916596e-01 -6.18224204e-01
-1.46127617e+00 6.78063512e-01 4.37294960e-01 -7.45484054e-01
9.22656894e-01 -7.41766766e-02 6.32125020e-01 3.40131335e-02
4.85184699e-01 -8.91853154e-01 9.10460874e-02 2.02951863e-01
1.01655042e+00 -1.18418324e+00 -1.85866207e-01 -7.17660010e-01
-4.55460340e-01 1.22025919e+00 9.39149022e-01 -2.85575420e-01
9.81200933e-01 6.13224328e-01 4.67746764e-01 -1.63358986e-01
-3.20161134e-01 -1.20815463e-01 5.88307679e-01 1.02049422e+00
3.93204629e-01 -3.15226972e-01 3.48133385e-01 1.38986200e-01
-7.51446605e-01 -2.81295497e-02 8.77412334e-02 4.85428602e-01
-5.71707040e-02 -1.23797405e+00 -2.58526593e-01 2.70301700e-01
-5.06617486e-01 -7.25402310e-02 -4.28342640e-01 9.77080822e-01
2.91098535e-01 4.47696805e-01 5.61353564e-01 -1.25376303e-02
2.44564176e-01 2.38237724e-01 6.76325262e-01 -4.69655931e-01
-1.62221104e-01 1.46913275e-01 1.82653442e-01 -7.43277133e-01
-5.58405817e-01 -5.44963658e-01 -1.23231590e+00 -4.80009645e-01
-9.04482752e-02 2.77800381e-01 4.45490927e-01 7.50895083e-01
3.02424252e-01 1.47871692e-02 5.43867052e-01 -1.27635074e+00
-1.25352800e-01 -8.03676486e-01 -5.30705452e-01 2.80330062e-01
6.22367084e-01 -7.34528065e-01 -3.19671214e-01 3.41831177e-01] | [13.39801025390625, 0.2623896300792694] |
b8fcae79-53ee-4c5c-97b5-d8b2ace6d958 | motion-attentive-transition-for-zero-shot | 2003.04253 | null | https://arxiv.org/abs/2003.04253v3 | https://arxiv.org/pdf/2003.04253v3.pdf | Motion-Attentive Transition for Zero-Shot Video Object Segmentation | In this paper, we present a novel Motion-Attentive Transition Network (MATNet) for zero-shot video object segmentation, which provides a new way of leveraging motion information to reinforce spatio-temporal object representation. An asymmetric attention block, called Motion-Attentive Transition (MAT), is designed within a two-stream encoder, which transforms appearance features into motion-attentive representations at each convolutional stage. In this way, the encoder becomes deeply interleaved, allowing for closely hierarchical interactions between object motion and appearance. This is superior to the typical two-stream architecture, which treats motion and appearance separately in each stream and often suffers from overfitting to appearance information. Additionally, a bridge network is proposed to obtain a compact, discriminative and scale-sensitive representation for multi-level encoder features, which is further fed into a decoder to achieve segmentation results. Extensive experiments on three challenging public benchmarks (i.e. DAVIS-16, FBMS and Youtube-Objects) show that our model achieves compelling performance against the state-of-the-arts. | ['Tianfei Zhou', 'Yazhou Yao', 'Shunzhou Wang', 'Jianwu Li', 'Yi Zhou', 'Ling Shao'] | 2020-03-09 | null | null | null | null | ['unsupervised-video-object-segmentation'] | ['computer-vision'] | [ 2.33002141e-01 -1.36687979e-01 -4.87065226e-01 -4.64824021e-01
-5.25680721e-01 -2.22349405e-01 3.74685168e-01 -3.15310180e-01
-4.47158694e-01 3.78402710e-01 2.72504926e-01 6.24695383e-02
4.60564196e-01 -6.01402938e-01 -9.29498494e-01 -6.45898998e-01
-1.24838009e-01 -1.68339871e-02 8.80506516e-01 1.10242201e-03
-4.40688990e-02 1.31353498e-01 -1.12604141e+00 4.95882541e-01
7.94343591e-01 1.13540637e+00 5.49133241e-01 8.33749592e-01
-1.79414809e-01 1.36493981e+00 -8.16839710e-02 -2.37709999e-01
9.28875133e-02 -5.38541973e-01 -1.05282009e+00 3.81550819e-01
5.59155107e-01 -9.10339475e-01 -9.02480304e-01 9.43688095e-01
-7.99882691e-04 3.00085783e-01 5.21614015e-01 -1.18865871e+00
-9.99677718e-01 5.79713523e-01 -8.14901650e-01 5.90399206e-01
1.67730257e-01 6.31789505e-01 1.01088345e+00 -8.32499385e-01
7.19683528e-01 1.53787827e+00 4.29136992e-01 6.61855161e-01
-1.05373549e+00 -3.16495329e-01 7.10530996e-01 4.62466359e-01
-1.00276995e+00 -3.41654360e-01 6.62279427e-01 -4.22358960e-01
8.04570556e-01 -2.94697955e-02 8.48577619e-01 1.11019468e+00
9.43427309e-02 1.54314637e+00 3.00098896e-01 1.02187507e-01
3.17598991e-02 -5.26000977e-01 -1.23316295e-01 7.87176490e-01
-9.53964144e-02 -5.01279980e-02 -3.89749497e-01 3.10080171e-01
1.21732247e+00 4.26975578e-01 -2.49601066e-01 -3.79225403e-01
-1.18158960e+00 5.65464795e-01 1.03798366e+00 2.91190118e-01
-4.71766710e-01 7.48351038e-01 4.39897358e-01 -6.12052530e-02
4.87024575e-01 -3.05442624e-02 -3.92172515e-01 -2.30062917e-01
-9.86342013e-01 5.19622527e-02 1.13815464e-01 1.06703424e+00
7.84103274e-01 3.43244344e-01 -7.09669232e-01 7.36794353e-01
5.62728643e-01 2.61917174e-01 4.35445249e-01 -1.18182611e+00
3.92035395e-01 5.95658779e-01 -4.26331200e-02 -6.94047093e-01
-3.58959846e-02 -1.93088755e-01 -9.53589797e-01 -1.81502193e-01
1.64538249e-01 -3.10650542e-02 -1.47864282e+00 1.66848755e+00
3.96514118e-01 8.62672627e-01 -1.55793214e-02 1.26107132e+00
1.11573327e+00 1.01993740e+00 5.59672594e-01 -5.93498088e-02
1.26220393e+00 -1.74361110e+00 -7.57600725e-01 -4.29274619e-01
3.14868033e-01 -4.62426335e-01 9.67247128e-01 -3.97607684e-02
-1.48081636e+00 -8.93770695e-01 -8.57190013e-01 -4.82984960e-01
-4.15168330e-02 -1.72169879e-01 8.51026654e-01 -8.71850103e-02
-8.60079467e-01 6.45047665e-01 -1.15529621e+00 -1.71563044e-01
8.52494836e-01 2.51958221e-01 -8.72378424e-02 -3.23549747e-01
-1.06899440e+00 4.02862489e-01 2.85409749e-01 2.25201458e-01
-1.27431703e+00 -6.75671935e-01 -1.15745807e+00 1.92353874e-01
2.60831058e-01 -8.70422482e-01 1.44312727e+00 -1.29871047e+00
-1.64336729e+00 5.01975119e-01 -2.93969303e-01 -4.41617757e-01
5.33216596e-01 -3.20678920e-01 -1.78753868e-01 6.18949294e-01
1.14366524e-01 1.38468814e+00 1.04738140e+00 -1.12459230e+00
-8.23340178e-01 -6.33657258e-03 5.20056263e-02 1.48216173e-01
-2.77317852e-01 1.33866638e-01 -1.24853575e+00 -9.14164841e-01
-3.83061618e-02 -6.55058265e-01 -5.30433953e-01 2.38619104e-01
-2.91257828e-01 -1.70813933e-01 1.19091117e+00 -5.51578462e-01
1.12494266e+00 -2.23140740e+00 3.87972951e-01 -3.29872847e-01
2.32206389e-01 4.99654293e-01 -4.22428459e-01 -2.42808890e-02
3.43420692e-02 -1.95131432e-02 -5.40799737e-01 -5.93801796e-01
-2.73165762e-01 4.36920166e-01 -1.00987643e-01 4.08873320e-01
6.84325457e-01 1.62755179e+00 -1.14723384e+00 -6.16370857e-01
3.15586746e-01 4.04520422e-01 -7.85087287e-01 4.22353417e-01
-5.35071909e-01 4.28824484e-01 -5.43733597e-01 8.84510696e-01
6.28497839e-01 -5.31171024e-01 -2.06253678e-01 -4.49228249e-02
-2.70459056e-02 3.00383586e-02 -6.44962490e-01 2.07346249e+00
-1.65771976e-01 6.88081682e-01 -1.13592334e-01 -8.33870411e-01
5.94241023e-01 1.24333113e-01 6.63356602e-01 -6.94868624e-01
2.93861032e-01 -2.15602189e-01 -2.14564830e-01 -7.49679685e-01
6.13768637e-01 2.18510360e-01 7.95325115e-02 1.00777201e-01
2.74399698e-01 1.66560933e-01 1.64936692e-01 3.93389285e-01
9.40221608e-01 4.19630408e-01 -1.23848967e-01 -6.59065172e-02
4.78656620e-01 -1.71421677e-01 9.34778333e-01 4.05410409e-01
-4.78909314e-01 1.03782403e+00 2.56396383e-01 -5.02683163e-01
-1.03296149e+00 -1.17353797e+00 2.24690527e-01 1.08931589e+00
6.61846399e-01 -2.79151469e-01 -5.47495246e-01 -7.41477728e-01
-3.24596800e-02 3.51978153e-01 -7.76884198e-01 -4.46166426e-01
-7.25598872e-01 -4.02674198e-01 9.61509421e-02 1.09181988e+00
8.12808037e-01 -1.23116016e+00 -6.01863205e-01 3.85287106e-01
-3.74818027e-01 -1.33589005e+00 -9.82162178e-01 -1.30472809e-01
-9.81280684e-01 -7.94498384e-01 -1.06540453e+00 -1.10368359e+00
5.75010240e-01 5.56485534e-01 9.15989935e-01 1.55229509e-01
-2.95891672e-01 2.68029988e-01 -5.97785294e-01 9.81323421e-02
-4.03404497e-02 1.23298980e-01 -5.06517410e-01 2.57646620e-01
1.72877789e-01 -4.88959104e-01 -1.10908282e+00 1.68758869e-01
-1.20682514e+00 3.01977187e-01 8.20702672e-01 7.75080919e-01
5.46196103e-01 -6.12019300e-01 5.31104028e-01 -6.98035598e-01
-1.41506448e-01 -7.21233010e-01 -2.65205443e-01 2.22429320e-01
-1.34220282e-02 -1.13262460e-01 5.40871918e-01 -5.85099161e-01
-1.21036434e+00 3.94307762e-01 -1.92225859e-01 -9.21470344e-01
-4.64078933e-02 1.36954740e-01 -5.60018569e-02 -4.55115102e-02
7.33091682e-02 4.47301090e-01 -1.16659902e-01 -5.08147359e-01
6.56164825e-01 4.23587143e-01 9.63611245e-01 -3.18745375e-01
6.61670625e-01 6.18380547e-01 -3.73188108e-01 -5.93931139e-01
-8.64561737e-01 -6.86578810e-01 -7.67652452e-01 -2.64716983e-01
1.34457159e+00 -1.06748474e+00 -4.82566953e-01 7.41423428e-01
-1.25890744e+00 -7.51344323e-01 -4.73461449e-01 4.21450883e-01
-7.71034837e-01 4.50911611e-01 -1.09542489e+00 -5.01992941e-01
-3.03798020e-01 -1.19021809e+00 1.26154971e+00 6.25135779e-01
9.02636945e-02 -8.50822687e-01 -1.31899461e-01 2.20245987e-01
3.36014420e-01 1.31162122e-01 5.78220308e-01 -3.20814252e-02
-1.01359236e+00 1.38674840e-01 -5.35155892e-01 3.04015189e-01
8.66120532e-02 2.96006739e-01 -7.43692815e-01 -2.92129159e-01
-3.89210463e-01 -5.04736245e-01 1.38290763e+00 6.21890783e-01
1.43503356e+00 -2.45603353e-01 -3.24384481e-01 1.03841293e+00
1.20957637e+00 1.94974780e-01 1.00591063e+00 2.24386528e-01
1.12910116e+00 1.31412476e-01 6.74315155e-01 3.92043382e-01
6.00777984e-01 6.25739038e-01 6.40782297e-01 -5.37636518e-01
-3.51236701e-01 -3.51025254e-01 4.00790125e-01 7.99464583e-01
-1.24634176e-01 -1.69569507e-01 -4.64566171e-01 7.02853441e-01
-2.12943506e+00 -1.01502705e+00 -2.19777927e-01 1.68468297e+00
7.54990995e-01 -1.86807029e-02 3.21255922e-01 -3.71659636e-01
6.14358604e-01 5.42406023e-01 -7.39721417e-01 -7.93893039e-02
2.16240901e-02 -7.94235393e-02 3.62523645e-01 3.34201932e-01
-1.33797300e+00 1.34286118e+00 6.27729940e+00 6.11889660e-01
-1.07917976e+00 5.82295358e-02 8.92144501e-01 -8.95453766e-02
-1.71151698e-01 -3.13620195e-02 -5.09552360e-01 7.21653044e-01
5.03797233e-01 2.82607321e-02 5.42896166e-02 9.35445547e-01
1.57542139e-01 -1.66140553e-02 -9.56958711e-01 7.90936708e-01
-8.04007947e-02 -1.44271076e+00 1.38398692e-01 -2.23361820e-01
1.00513458e+00 2.03586400e-01 1.32354632e-01 4.37593222e-01
3.52260917e-01 -8.37726355e-01 9.20231283e-01 5.84255695e-01
7.52234757e-01 -7.62514591e-01 4.41713989e-01 -2.57925130e-02
-1.74164176e+00 -1.62991047e-01 -4.76700246e-01 1.98326766e-01
5.44179797e-01 1.83644667e-01 -1.52178749e-01 5.88229179e-01
8.53127003e-01 1.33403468e+00 -4.70692217e-01 1.27990353e+00
-1.97683573e-01 5.68414330e-01 3.41998115e-02 2.46191233e-01
7.86459088e-01 8.28508195e-03 3.31283987e-01 1.36029935e+00
4.09515649e-02 3.40623975e-01 3.19468588e-01 9.23196316e-01
-1.37724176e-01 -2.56078035e-01 -2.21259817e-01 -2.34648008e-02
7.63316676e-02 1.33800507e+00 -7.97541559e-01 -7.64073968e-01
-6.69326305e-01 1.40515935e+00 3.84924799e-01 6.15593612e-01
-1.05316341e+00 -4.00575325e-02 8.09757888e-01 -2.82911900e-02
9.40361559e-01 -3.15195918e-01 1.43481717e-01 -1.29979849e+00
-2.50690073e-01 -5.27849376e-01 4.37310725e-01 -7.31244683e-01
-1.21535420e+00 5.12305379e-01 -1.01526976e-01 -1.36563540e+00
-7.72447735e-02 -3.22895557e-01 -8.46939445e-01 5.18803179e-01
-1.66109872e+00 -1.24428904e+00 -4.78258491e-01 5.69830120e-01
1.07598352e+00 2.78657615e-01 8.15785751e-02 4.84438658e-01
-8.76840353e-01 4.99859273e-01 -9.95208323e-02 2.99710661e-01
5.05318284e-01 -1.09627044e+00 7.34525800e-01 1.06148148e+00
6.15146644e-02 2.46101633e-01 2.06814334e-01 -6.88181758e-01
-1.36297369e+00 -1.64713717e+00 3.79128844e-01 -1.14708818e-01
5.21585345e-01 -1.90039858e-01 -1.19684780e+00 7.30348110e-01
3.88453096e-01 6.10297322e-01 3.08061033e-01 -4.45502996e-01
-1.09562904e-01 8.83416831e-02 -7.18558669e-01 6.54682219e-01
1.38300896e+00 -2.68188238e-01 -6.16303921e-01 7.43388832e-02
1.11649573e+00 -4.17555302e-01 -6.40699089e-01 4.72042859e-01
4.15759414e-01 -7.60787249e-01 1.09433746e+00 -9.02545333e-01
9.85184908e-01 -3.44880760e-01 5.84865827e-03 -1.05736184e+00
-7.06718564e-01 -7.36981690e-01 -5.37287295e-01 1.14664447e+00
7.23093376e-02 3.06097046e-02 8.79391849e-01 3.62810433e-01
-4.37887758e-01 -1.09399390e+00 -5.38918555e-01 -6.94458485e-01
-6.20447099e-02 -2.37186834e-01 3.93972218e-01 8.40754747e-01
-2.68789828e-01 3.39370608e-01 -5.15045226e-01 -8.61072913e-02
4.46530670e-01 1.03211232e-01 6.94004476e-01 -6.04765832e-01
-4.48706627e-01 -5.33142686e-01 -5.21886230e-01 -1.86926997e+00
4.33209762e-02 -7.49963105e-01 4.33980614e-01 -1.70710647e+00
4.73651499e-01 -1.43099234e-01 -4.21974063e-01 4.13905799e-01
-7.31769145e-01 4.23224032e-01 3.83673459e-01 3.04184943e-01
-1.12422752e+00 1.13038814e+00 1.88708377e+00 -2.68198788e-01
-2.32006699e-01 -2.48149410e-01 -4.21848804e-01 7.98464417e-01
3.46927911e-01 -3.69358301e-01 -4.61658686e-01 -6.52736366e-01
-4.98312026e-01 1.38028234e-01 3.65302593e-01 -9.42790627e-01
2.95467347e-01 -3.68107796e-01 6.12237692e-01 -7.87126124e-01
4.06150997e-01 -5.26067555e-01 -2.61894345e-01 5.19818485e-01
-3.59436274e-01 -4.88782413e-02 8.83912519e-02 8.99078071e-01
-4.43937361e-01 1.06656529e-01 9.28779542e-01 -7.28859007e-02
-1.33362043e+00 1.08724105e+00 -3.53364080e-01 1.42885193e-01
1.21457434e+00 -2.53104240e-01 -2.06806764e-01 -3.07440102e-01
-6.83444560e-01 7.35690236e-01 3.70800793e-01 6.33469522e-01
8.04984987e-01 -1.58327305e+00 -6.96152925e-01 -5.47525361e-02
1.49630504e-02 5.30337393e-01 7.05831528e-01 9.14771020e-01
-5.67946017e-01 1.12807147e-01 -3.85566115e-01 -9.55472946e-01
-1.00747097e+00 6.92473531e-01 2.04008296e-01 1.36007620e-02
-9.50231791e-01 1.27841485e+00 7.56982863e-01 3.58372688e-01
2.81108856e-01 -6.85345292e-01 -7.34364390e-02 -2.64637858e-01
6.75488353e-01 8.49184170e-02 -5.72357416e-01 -6.57591760e-01
-2.75622278e-01 5.11068523e-01 -2.91481286e-01 7.40592182e-02
1.31034839e+00 -2.86502093e-01 7.81504512e-02 4.20113653e-01
1.42438400e+00 -6.12746119e-01 -2.22930121e+00 -3.87236983e-01
-2.12156564e-01 -7.28731215e-01 -9.36797410e-02 -1.60430700e-01
-1.59830379e+00 9.67619419e-01 2.38816485e-01 -5.52098714e-02
1.22477293e+00 1.21352047e-01 1.26434982e+00 1.00471549e-01
-8.18428919e-02 -1.00922501e+00 6.92971706e-01 4.96146917e-01
7.52929091e-01 -1.25354302e+00 -1.79605097e-01 -5.71363509e-01
-8.95614564e-01 9.37834859e-01 1.11224091e+00 -3.99005830e-01
3.95030022e-01 1.03322059e-01 -1.70571595e-01 1.77345667e-02
-9.21720088e-01 -4.40236509e-01 3.91620159e-01 4.41561997e-01
3.31673831e-01 -2.08446115e-01 6.81533366e-02 5.27662396e-01
4.62498009e-01 1.10306174e-01 2.23670468e-01 9.07744944e-01
-4.98050839e-01 -8.18404913e-01 7.46557340e-02 3.37509871e-01
-3.35445672e-01 -4.36117500e-02 -3.03668082e-01 6.23455226e-01
1.91804320e-01 5.57044029e-01 4.13322121e-01 -3.20397943e-01
9.78351757e-02 -3.37204039e-01 4.65116411e-01 -6.83145046e-01
-4.43044305e-01 2.83904046e-01 -1.62469909e-01 -8.41212809e-01
-5.48125386e-01 -5.97364128e-01 -1.46623361e+00 -1.05554871e-01
-4.45615966e-04 -1.79824024e-01 5.36089391e-02 8.76477420e-01
3.45971525e-01 1.08732092e+00 6.40744686e-01 -1.14778852e+00
-1.66290388e-01 -8.69254589e-01 -3.33537221e-01 5.22062600e-01
5.02886832e-01 -6.66567981e-01 3.85569371e-02 3.85499120e-01] | [9.25650405883789, -0.10987801104784012] |
f7796008-ea6e-4358-a327-de6f8c6a96b8 | enhancing-local-feature-learning-using | 2207.01174 | null | https://arxiv.org/abs/2207.01174v1 | https://arxiv.org/pdf/2207.01174v1.pdf | Enhancing Local Feature Learning Using Diffusion for 3D Point Cloud Understanding | Learning point clouds is challenging due to the lack of connectivity information, i.e., edges. Although existing edge-aware methods can improve the performance by modeling edges, how edges contribute to the improvement is unclear. In this study, we propose a method that automatically learns to enhance/suppress edges while keeping the its working mechanism clear. First, we theoretically figure out how edge enhancement/suppression works. Second, we experimentally verify the edge enhancement/suppression behavior. Third, we empirically show that this behavior improves performance. In general, we observe that the proposed method achieves competitive performance in point cloud classification and segmentation tasks. | ['Masashi Matsuoka', 'Qiong Chang', 'Takayuki Shinohara', 'Kyoung-Sook Kim', 'Weimin WANG', 'Xin Liu', 'Haoyi Xiu'] | 2022-07-04 | null | null | null | null | ['point-cloud-classification'] | ['computer-vision'] | [ 6.72488380e-03 1.14234842e-01 -2.31949449e-01 -1.99300319e-01
-3.45291018e-01 -3.46421421e-01 4.20475528e-02 2.72846729e-01
-1.41154662e-01 4.70079482e-01 -1.54654205e-01 -1.47482634e-01
-1.31411701e-01 -9.82182801e-01 -1.06866193e+00 -4.62547779e-01
-2.90606171e-01 2.37429872e-01 5.68765461e-01 3.57363820e-02
2.81895995e-01 7.51970232e-01 -1.45336807e+00 5.60985096e-02
1.18786407e+00 8.40095580e-01 1.39076442e-01 4.23155576e-01
-2.34517202e-01 3.08221519e-01 -2.46037617e-01 -1.68849036e-01
5.23386061e-01 5.40225580e-03 -7.17348397e-01 3.17710817e-01
6.72223508e-01 -3.52449089e-01 -1.96949303e-01 1.16000521e+00
1.48381442e-02 -7.34540960e-03 5.84530056e-01 -1.50228131e+00
-2.69830376e-01 4.49050426e-01 -1.04337740e+00 1.11123472e-01
-1.59592599e-01 -2.37152409e-02 1.08803403e+00 -9.49270010e-01
7.27166533e-01 8.99429381e-01 8.55875373e-01 2.14239299e-01
-1.00947154e+00 -7.06997395e-01 4.82445002e-01 6.92766579e-03
-1.67346990e+00 -1.03948802e-01 1.11043167e+00 -3.50086123e-01
6.31016433e-01 1.77983105e-01 8.23745489e-01 3.48381847e-01
5.91665274e-03 1.05738139e+00 1.02453971e+00 -4.32825297e-01
9.61175337e-02 6.79670423e-02 4.52409178e-01 7.30327725e-01
5.22106886e-01 1.99715234e-03 -4.34229016e-01 -3.48148542e-03
8.62168431e-01 6.18789159e-02 -3.97822827e-01 -6.91635549e-01
-7.68939078e-01 6.84306204e-01 7.89910197e-01 1.97509140e-01
-3.49999577e-01 3.00061941e-01 1.32386819e-01 9.56986248e-02
5.84682524e-01 2.79373139e-01 -5.47886252e-01 6.18769154e-02
-1.14076161e+00 3.87134515e-02 5.67227781e-01 1.28815269e+00
1.10676265e+00 -1.70384005e-01 1.86711505e-01 7.32997537e-01
1.29277691e-01 1.72672018e-01 -2.70482630e-01 -1.02897727e+00
3.07601869e-01 8.03342283e-01 -3.49869095e-02 -1.15242350e+00
-4.87102747e-01 -6.84239209e-01 -7.23618746e-01 3.58836323e-01
4.07519162e-01 -6.45327643e-02 -9.02761877e-01 1.57082903e+00
3.75669628e-01 6.88826323e-01 -2.51462519e-01 7.88680911e-01
7.04563737e-01 3.53399217e-01 -1.24416739e-01 -8.76555368e-02
1.06253159e+00 -1.16946518e+00 -5.32759666e-01 -2.92253554e-01
6.20165050e-01 -5.21354914e-01 9.49112773e-01 2.50500500e-01
-1.01693249e+00 -4.55916047e-01 -9.13702428e-01 1.18794620e-01
-1.61388069e-01 1.66591376e-01 1.04856324e+00 5.75424075e-01
-1.08100832e+00 7.88108528e-01 -1.12819147e+00 -1.67382956e-01
5.92876852e-01 4.90360975e-01 -3.81589085e-01 -4.71246578e-02
-7.43153453e-01 4.57470059e-01 3.73991638e-01 -1.79208040e-01
-1.92588136e-01 -8.63812923e-01 -7.84467638e-01 3.58763814e-01
5.21338642e-01 -8.57902884e-01 1.13339579e+00 -9.10137951e-01
-9.56897974e-01 5.85279703e-01 -4.80187625e-01 -4.60079879e-01
5.42787552e-01 -2.74645865e-01 1.16093993e-01 3.08586419e-01
-5.48129156e-02 1.06808054e+00 6.52065337e-01 -1.84369707e+00
-9.51214135e-01 -4.49102044e-01 2.10415035e-01 9.32670087e-02
-6.26278222e-01 -5.43050945e-01 -9.41288114e-01 -6.36035800e-01
4.66845781e-01 -1.19724929e+00 -3.92360777e-01 1.25051349e-01
-3.73018682e-01 -7.49270618e-02 1.06201148e+00 -2.36928344e-01
1.09044683e+00 -2.13295841e+00 -4.14867610e-01 6.24424100e-01
6.07011855e-01 -6.99624047e-02 2.74470914e-02 2.00243145e-01
-2.35418472e-02 4.33399707e-01 -3.14111620e-01 -3.97584409e-01
-1.94603637e-01 1.04930431e-01 -1.76815584e-01 3.83587271e-01
1.38385922e-01 9.72863913e-01 -7.75698781e-01 -5.96516073e-01
4.62572873e-01 5.98680079e-01 -8.35840404e-01 -3.20351213e-01
2.42865365e-02 2.38664120e-01 -4.97583061e-01 7.24347293e-01
1.10933125e+00 -4.99689758e-01 -7.47945765e-03 -3.62742215e-01
-5.48887663e-02 1.13189630e-01 -1.51251912e+00 1.30267727e+00
-3.32547516e-01 8.41590345e-01 2.01411441e-01 -8.67088199e-01
7.88927138e-01 -5.76631092e-02 9.20043707e-01 -3.54235530e-01
2.94989021e-03 1.18502200e-01 -6.71227425e-02 -6.25747293e-02
6.27785683e-01 2.67025903e-02 5.37230968e-01 2.34293059e-01
-3.17528695e-01 -5.73551804e-02 -7.46377110e-02 2.57609338e-01
1.09201419e+00 -2.44761091e-02 1.01410978e-01 -3.23620915e-01
3.31895240e-02 1.94135934e-01 5.62216282e-01 8.65565062e-01
-1.83350012e-01 7.73005426e-01 2.42924288e-01 -6.62379619e-03
-8.09110105e-01 -8.72424662e-01 -2.32997403e-01 7.71221697e-01
8.52987170e-01 -6.34371698e-01 -8.88254225e-01 -7.07420945e-01
1.36155024e-01 5.41115105e-01 -4.37842548e-01 -1.50510624e-01
-4.78625804e-01 -6.11596227e-01 2.32191652e-01 9.69548821e-01
4.92353499e-01 -6.19919896e-01 -2.34705850e-01 1.95810385e-02
-1.85938943e-02 -1.21388245e+00 -5.11525154e-01 -1.40885757e-02
-1.43912590e+00 -1.07482636e+00 -3.56800884e-01 -7.96974599e-01
9.96018171e-01 1.02529669e+00 1.28882515e+00 5.57454705e-01
3.65361795e-02 4.34682131e-01 -2.97300875e-01 -4.66582745e-01
1.52359396e-01 2.60381281e-01 -2.48905182e-01 -4.33055848e-01
3.17283273e-01 -7.01778829e-01 -7.06080675e-01 2.38045558e-01
-7.94713616e-01 2.30880588e-01 6.21403575e-01 5.66344380e-01
9.06371057e-01 4.42130774e-01 3.15272957e-01 -1.13042068e+00
4.27072644e-01 -9.40671787e-02 -6.34983242e-01 9.41518620e-02
-8.26079369e-01 8.88419002e-02 2.20488355e-01 -2.09565729e-01
-8.49677503e-01 4.48723644e-01 -2.51654238e-01 -5.38433135e-01
-2.62762427e-01 3.25059295e-01 -1.42622992e-01 -5.46775281e-01
3.16558123e-01 -1.05624601e-01 -3.75029325e-01 -3.90587151e-01
4.46178585e-01 4.40881163e-01 2.68043965e-01 -6.30674779e-01
8.73960793e-01 8.57882559e-01 -3.14959814e-03 -8.88684928e-01
-7.29186893e-01 -7.77331293e-01 -6.01958156e-01 -3.75622064e-01
5.46280205e-01 -8.29880416e-01 -7.49827862e-01 1.62461951e-01
-1.12974942e+00 -3.03406864e-01 -1.12546660e-01 5.05996346e-01
-4.30489391e-01 5.10350108e-01 -7.97773659e-01 -8.32031667e-01
-2.37772986e-01 -1.11721468e+00 1.22137749e+00 3.19346189e-01
3.72399949e-02 -1.10644698e+00 -2.52355099e-01 1.47402108e-01
5.87819070e-02 9.78544950e-02 6.14905775e-01 -3.09168845e-01
-9.68554199e-01 -1.39339203e-02 -4.36022818e-01 1.90849155e-01
2.75228262e-01 1.81670845e-01 -9.04585361e-01 -2.40216240e-01
-6.28969744e-02 3.49572688e-01 1.16615331e+00 6.62222505e-01
1.48661423e+00 -2.74600610e-02 -8.17881584e-01 6.37406230e-01
1.42065048e+00 -1.59551233e-01 6.81003928e-01 2.67015636e-01
9.34731066e-01 4.96089131e-01 6.03850782e-01 1.82150066e-01
4.90876436e-01 5.12635469e-01 5.38504004e-01 -5.03280282e-01
-2.56428361e-01 -2.31289372e-01 -1.91376209e-01 6.20039105e-01
-1.74136370e-01 -1.28847510e-01 -1.09276974e+00 6.06765509e-01
-1.90686548e+00 -5.44642091e-01 -6.01309061e-01 2.05027628e+00
5.39272666e-01 6.12797856e-01 -1.58987623e-02 1.31759807e-01
9.08073008e-01 -1.24723494e-01 -4.57095444e-01 1.93331882e-01
3.20742056e-02 1.94462731e-01 9.65726674e-01 5.95512152e-01
-1.11040747e+00 1.12881947e+00 6.85984278e+00 7.12634742e-01
-1.18420279e+00 -1.15676068e-01 5.47065437e-01 1.68812320e-01
-3.30702394e-01 -1.27585111e-02 -7.11377859e-01 1.88901275e-01
1.07592158e-01 -8.05980936e-02 1.31082937e-01 1.27004910e+00
1.12805188e-01 5.41550368e-02 -1.07170773e+00 9.78631079e-01
-7.53234848e-02 -1.40964389e+00 8.52881407e-04 1.12392604e-01
8.78597438e-01 7.32971951e-02 -1.89800337e-01 1.56551927e-01
1.53380796e-01 -7.72457540e-01 5.36838353e-01 2.51749486e-01
3.77211839e-01 -8.05443585e-01 5.77957988e-01 4.99882340e-01
-1.41142428e+00 2.68877953e-01 -2.59781301e-01 -9.19262171e-02
1.16492406e-01 9.42888916e-01 -6.83852494e-01 6.07731938e-01
6.69395387e-01 6.61778927e-01 -5.86434245e-01 1.45406687e+00
-2.59506971e-01 7.60260761e-01 -5.92265844e-01 4.66591358e-01
9.82875973e-02 -4.51089174e-01 6.17293358e-01 1.22326493e+00
2.92535275e-01 1.70261562e-01 2.68424660e-01 9.43440080e-01
-3.42321843e-01 8.15908462e-02 -7.58120179e-01 1.83511704e-01
6.22880518e-01 1.23310220e+00 -1.22926772e+00 -3.05637658e-01
-7.28120625e-01 7.32398629e-01 3.68621528e-01 3.71272385e-01
-8.40310991e-01 -1.69841915e-01 8.22730064e-01 3.27527523e-01
3.77895325e-01 -4.93106782e-01 -8.18728149e-01 -1.01795554e+00
2.74774402e-01 -4.37067121e-01 7.04032704e-02 -6.95272982e-01
-1.32889915e+00 3.78143251e-01 -2.64770016e-02 -1.19032633e+00
3.96280408e-01 -5.03016651e-01 -7.21241355e-01 3.20702285e-01
-1.73758125e+00 -1.00431168e+00 -7.44516730e-01 1.49402350e-01
5.21638453e-01 5.72814703e-01 1.15346193e-01 5.56492269e-01
-3.24831963e-01 6.00011587e-01 -1.27305552e-01 2.45979071e-01
5.94047606e-01 -1.29275846e+00 4.56828743e-01 8.42338979e-01
3.53841126e-01 8.23121965e-01 5.24272561e-01 -8.96046162e-01
-1.07667613e+00 -1.19044042e+00 4.27789479e-01 -1.94348291e-01
3.83738250e-01 -2.91455597e-01 -1.26983595e+00 7.67477930e-01
-7.14935362e-02 1.25108123e-01 3.39814812e-01 3.08795720e-01
-1.85182452e-01 1.22717075e-01 -1.10610425e+00 6.94241345e-01
1.35604525e+00 -2.56439209e-01 -4.12008971e-01 3.29491138e-01
7.67604649e-01 -5.09821117e-01 -7.07278311e-01 8.20639253e-01
3.17179918e-01 -9.74350572e-01 1.13685930e+00 -4.68928844e-01
3.00328553e-01 -3.71953696e-01 1.81270808e-01 -1.27652860e+00
-2.83099085e-01 -1.27222821e-01 -2.22851649e-01 1.21495008e+00
4.18577880e-01 -6.41632259e-01 1.32505429e+00 5.86041868e-01
-3.27753663e-01 -8.16031575e-01 -5.91038406e-01 -1.01431251e+00
1.14569210e-01 -6.56371832e-01 5.78461766e-01 1.13505244e+00
-2.56365269e-01 2.24241719e-01 -8.43902081e-02 4.73572969e-01
6.01459324e-01 2.92747021e-01 7.73746014e-01 -1.38921249e+00
-9.93633270e-02 -5.88324845e-01 -3.56312007e-01 -1.60407126e+00
1.76814094e-01 -9.71005082e-01 -5.70236333e-02 -1.85503590e+00
2.65918165e-01 -8.71563673e-01 -2.28899658e-01 4.27758187e-01
-5.76726079e-01 2.22453311e-01 2.14592651e-01 2.90989876e-01
-6.84984446e-01 4.19150144e-01 1.34957063e+00 -1.25515461e-01
-3.93579632e-01 1.39193442e-02 -6.85994148e-01 9.49053347e-01
1.10342455e+00 -5.16045928e-01 -3.38007867e-01 -5.48927426e-01
1.25460505e-01 -3.49572212e-01 3.65709007e-01 -8.88878524e-01
3.00586224e-01 -1.70837030e-01 2.48844072e-01 -7.04827070e-01
2.37733021e-01 -1.01270688e+00 -8.70826002e-03 3.15716177e-01
5.46364784e-02 -6.96999952e-02 4.01866794e-01 6.33155763e-01
-2.97490001e-01 -2.91185200e-01 7.22222030e-01 3.19344737e-02
-8.07012677e-01 4.17897552e-01 1.09046055e-02 -1.13207564e-01
1.09648800e+00 -4.96545613e-01 -1.84914172e-01 -3.58758479e-01
-5.15044451e-01 4.84286368e-01 8.60339224e-01 2.83381552e-01
5.76023340e-01 -1.31431448e+00 -4.67312813e-01 -2.40622293e-02
4.65035029e-02 2.25661680e-01 6.96575195e-02 9.42948341e-01
-5.27935505e-01 1.75517276e-01 6.82622269e-02 -8.39519978e-01
-1.53908002e+00 5.82030118e-01 2.89601743e-01 8.57656822e-03
-8.52870345e-01 8.12200963e-01 4.83210653e-01 -2.54508555e-01
3.01453978e-01 -5.00546396e-01 -5.95482253e-02 -2.24961907e-01
1.27843380e-01 4.57847953e-01 2.69676566e-01 -3.58321577e-01
-4.55330938e-01 8.05689156e-01 -5.57344183e-02 2.03056276e-01
1.33958077e+00 -5.33858351e-02 5.57202622e-02 8.61704424e-02
9.16801870e-01 3.80934745e-01 -1.28730369e+00 -1.84812900e-02
1.69148892e-01 -6.89613283e-01 3.70596528e-01 -1.33170903e-01
-1.57607830e+00 7.67625749e-01 4.50699896e-01 3.49547803e-01
1.21188879e+00 9.20714289e-02 1.01328528e+00 3.56287211e-01
4.18297321e-01 -1.10597670e+00 -1.22848861e-01 3.84216815e-01
4.83944595e-01 -1.44910645e+00 8.74436125e-02 -1.41924751e+00
-4.44131285e-01 8.24769020e-01 8.11202288e-01 -2.02796578e-01
8.93583477e-01 6.13171041e-01 -1.28605828e-01 -3.95068467e-01
-4.93691713e-01 -5.21137178e-01 3.67526650e-01 5.29613674e-01
3.99866074e-01 -1.15032960e-03 -3.53171945e-01 3.64054710e-01
-2.69448072e-01 -1.03430152e-01 3.19237649e-01 8.96210372e-01
-6.52650893e-01 -1.08299434e+00 -4.28138971e-01 5.81342518e-01
-9.54006985e-02 -1.05178550e-01 -6.94613516e-01 1.03821290e+00
1.26215219e-01 8.85729730e-01 2.68116295e-01 -3.37375373e-01
4.75455493e-01 -2.43004426e-01 4.74987984e-01 -5.38973987e-01
-3.24868351e-01 5.26640303e-02 -1.86732158e-01 -3.90899718e-01
-2.93761998e-01 -3.98437172e-01 -1.76286173e+00 -2.48714074e-01
-8.99671495e-01 1.85932368e-01 6.48170412e-01 8.04466546e-01
5.92185259e-01 7.18071759e-01 4.55943286e-01 -5.55455387e-01
-5.69935068e-02 -5.81032693e-01 -4.58481252e-01 5.35065293e-01
2.10541755e-01 -9.10009861e-01 -4.02091980e-01 9.72574018e-03] | [8.002401351928711, -3.218546152114868] |
4738fe9a-4806-4171-b9c6-6246365c0fe6 | disentangling-object-motion-and-occlusion-for | 2203.15174 | null | https://arxiv.org/abs/2203.15174v2 | https://arxiv.org/pdf/2203.15174v2.pdf | Disentangling Object Motion and Occlusion for Unsupervised Multi-frame Monocular Depth | Conventional self-supervised monocular depth prediction methods are based on a static environment assumption, which leads to accuracy degradation in dynamic scenes due to the mismatch and occlusion problems introduced by object motions. Existing dynamic-object-focused methods only partially solved the mismatch problem at the training loss level. In this paper, we accordingly propose a novel multi-frame monocular depth prediction method to solve these problems at both the prediction and supervision loss levels. Our method, called DynamicDepth, is a new framework trained via a self-supervised cycle consistent learning scheme. A Dynamic Object Motion Disentanglement (DOMD) module is proposed to disentangle object motions to solve the mismatch problem. Moreover, novel occlusion-aware Cost Volume and Re-projection Loss are designed to alleviate the occlusion effects of object motions. Extensive analyses and experiments on the Cityscapes and KITTI datasets show that our method significantly outperforms the state-of-the-art monocular depth prediction methods, especially in the areas of dynamic objects. Code is available at https://github.com/AutoAILab/DynamicDepth | ['Bing Li', 'YingLi Tian', 'HaiYan Wang', 'Longlong Jing', 'Liang Yang', 'Ziyue Feng'] | 2022-03-29 | null | null | null | null | ['motion-disentanglement'] | ['computer-vision'] | [-1.23593852e-01 -2.36230373e-01 -4.94870931e-01 -4.11694437e-01
-4.89502460e-01 -2.30359957e-01 5.26079118e-01 -5.40435374e-01
-9.14077908e-02 6.35653436e-01 3.01214606e-01 2.21465588e-01
1.56511039e-01 -6.29574180e-01 -4.94084328e-01 -8.06834161e-01
5.02816677e-01 3.87109607e-01 6.22317076e-01 1.47506922e-01
1.43273905e-01 3.90436113e-01 -1.57491279e+00 2.24281624e-01
1.08805704e+00 7.80729830e-01 5.05654156e-01 4.14815545e-01
6.66069388e-02 1.06400084e+00 3.25670615e-02 -1.60166875e-01
5.95865130e-01 -2.64583528e-01 -6.07234180e-01 3.66782844e-01
8.27829778e-01 -8.18985701e-01 -9.40705299e-01 8.50887954e-01
6.14176452e-01 7.53196701e-02 2.90336400e-01 -1.43784487e+00
-3.47596318e-01 -9.70252976e-03 -7.46800065e-01 2.83808261e-01
3.40689659e-01 3.27940315e-01 9.60162342e-01 -1.13161778e+00
7.94037819e-01 1.35106540e+00 3.40508610e-01 6.34631157e-01
-1.11060965e+00 -6.41482651e-01 5.03397703e-01 5.89656413e-01
-1.18033552e+00 -3.98808062e-01 1.06499362e+00 -6.17490888e-01
7.60031044e-01 5.78826182e-02 7.71054149e-01 8.92746270e-01
-6.32766113e-02 1.25028694e+00 8.45219731e-01 -2.17781186e-01
-6.26233360e-03 -4.91318814e-02 -9.74869356e-02 6.57975912e-01
2.68635601e-01 4.27361578e-01 -6.11847758e-01 1.90927729e-01
9.24707174e-01 7.08937570e-02 -4.67242539e-01 -1.05647850e+00
-1.11645007e+00 6.19683325e-01 5.37376940e-01 -4.53279447e-03
5.39346971e-02 1.50736526e-01 2.86305100e-01 -4.98920493e-02
7.36071885e-01 -6.85347766e-02 -6.14191115e-01 -1.50638940e-02
-6.76683605e-01 4.82091278e-01 3.35323989e-01 9.74060833e-01
8.64009559e-01 1.54637173e-01 -1.50278881e-01 7.18772471e-01
5.15375316e-01 3.97586077e-01 4.19393748e-01 -1.11141288e+00
8.44029188e-01 7.19628692e-01 2.96750814e-01 -1.08986449e+00
-4.30080920e-01 -3.55799794e-01 -6.53830469e-01 2.15210587e-01
4.55133110e-01 5.37884869e-02 -5.93109488e-01 1.59968662e+00
9.87975121e-01 4.96830225e-01 -8.56363624e-02 1.24865663e+00
1.03030717e+00 5.37421882e-01 -1.01217873e-01 -2.11829871e-01
8.29182684e-01 -1.67924607e+00 -6.88259006e-01 -4.71240282e-01
7.88056016e-01 -7.46789336e-01 1.17191291e+00 2.59034187e-01
-1.10121906e+00 -6.38315976e-01 -9.18906271e-01 -4.37910169e-01
2.22953752e-01 3.32243085e-01 6.61066294e-01 4.43776101e-01
-6.68123126e-01 3.57675016e-01 -9.50634241e-01 -4.68417347e-05
4.45071548e-01 2.32200194e-02 -1.88154027e-01 -4.31106150e-01
-7.91726470e-01 5.91986120e-01 2.62541652e-01 1.16994448e-01
-7.03140259e-01 -8.56739640e-01 -1.09449291e+00 -2.19272733e-01
3.64198774e-01 -8.69466722e-01 1.25573945e+00 -9.38797832e-01
-1.42815614e+00 8.69536042e-01 -2.80352384e-01 -1.19256258e-01
1.04334450e+00 -5.77016473e-01 -5.06674014e-02 1.29956394e-01
2.60499537e-01 8.72762203e-01 5.60040116e-01 -1.31604898e+00
-7.65322864e-01 -5.23781717e-01 6.90264627e-02 7.29481161e-01
-1.20833494e-01 -3.87254357e-01 -7.70964861e-01 -7.28410304e-01
5.03069758e-01 -1.03191507e+00 -1.52687192e-01 5.55762947e-01
-1.84214771e-01 -1.33562600e-02 1.09961915e+00 -5.54597676e-01
1.10364807e+00 -1.97344649e+00 3.27961206e-01 -5.01863480e-01
2.31751204e-01 1.89759389e-01 -7.18753338e-02 6.32585213e-02
3.67897227e-02 -5.97431779e-01 -1.14634849e-01 -7.58889258e-01
-3.20935607e-01 2.06874087e-01 -6.98922724e-02 7.20463514e-01
-1.41289547e-01 8.14574242e-01 -1.14127612e+00 -5.71129858e-01
6.08948171e-01 3.58405441e-01 -7.31235147e-01 3.46580446e-01
-3.46708417e-01 7.93238997e-01 -3.55185866e-01 7.87026882e-01
1.09820557e+00 -2.74265945e-01 -2.93102991e-02 -2.74960667e-01
-1.60905361e-01 2.48762831e-01 -1.26952684e+00 2.02573895e+00
-3.55586708e-01 6.82592928e-01 -1.57294065e-01 -7.68546462e-01
7.18737483e-01 9.57985148e-02 7.26236522e-01 -7.21862674e-01
-1.14244660e-02 3.22055370e-01 -2.63396859e-01 -5.85913777e-01
4.85156417e-01 4.68492545e-02 4.12854433e-01 7.09389672e-02
-3.78275245e-01 -2.83028603e-01 -2.91414391e-02 3.28889117e-02
7.61167109e-01 6.54942155e-01 2.13167429e-01 -4.64568660e-02
6.81188643e-01 -4.04021889e-02 1.06342053e+00 9.90628749e-02
-6.29716933e-01 9.56826866e-01 1.83580846e-01 -7.08655715e-01
-1.07222378e+00 -9.12583649e-01 -2.33619556e-01 4.69177932e-01
8.56151521e-01 -1.62845477e-01 -2.89258122e-01 -6.88206017e-01
1.38592899e-01 3.28184247e-01 -3.79213750e-01 3.61440703e-02
-6.95469201e-01 -6.62137985e-01 -5.53172939e-02 6.54231966e-01
9.16494966e-01 -7.93803930e-01 -5.81776917e-01 1.61951438e-01
-6.33558929e-01 -1.59451854e+00 -6.03293121e-01 -3.29312921e-01
-1.10720253e+00 -1.11156905e+00 -8.16915452e-01 -7.26733088e-01
4.17466134e-01 8.71541083e-01 9.08198595e-01 -1.18355848e-01
-1.79639012e-01 1.32545501e-01 -2.58130908e-01 -1.92051791e-02
-3.52817937e-03 5.94360866e-02 -1.47745982e-01 4.36554290e-02
2.69014508e-01 -6.24062955e-01 -1.15352488e+00 5.74145794e-01
-6.36185229e-01 6.33947492e-01 2.59995639e-01 7.55349636e-01
6.58689737e-01 -1.54265657e-01 2.07163453e-01 -5.93339503e-01
-4.96254176e-01 -4.40742642e-01 -8.53745699e-01 -7.51212686e-02
-5.71823835e-01 -1.70239493e-01 2.15763196e-01 -4.69184279e-01
-1.36052668e+00 3.28627676e-01 7.62506947e-02 -6.93304300e-01
-1.08129289e-02 -7.03697950e-02 -6.38154447e-01 -1.84652179e-01
4.58269864e-01 2.14898840e-01 -3.13988984e-01 -5.22508323e-01
1.18443981e-01 2.92091638e-01 2.80104041e-01 -2.33369857e-01
8.11744869e-01 9.53384519e-01 1.07022345e-01 -5.45736670e-01
-1.04443634e+00 -7.79018641e-01 -6.92145467e-01 -3.89425397e-01
7.86357284e-01 -1.48231769e+00 -2.32803598e-01 8.96430969e-01
-1.20331538e+00 -5.90793490e-01 -1.26176134e-01 6.64439857e-01
-7.67193496e-01 6.57791972e-01 -5.56869149e-01 -5.90969026e-01
-1.21805258e-01 -1.04059494e+00 1.25598323e+00 7.05762580e-02
2.95561373e-01 -8.57022524e-01 2.41682827e-01 7.60368764e-01
-3.09889261e-02 3.35699797e-01 4.60296154e-01 2.62500495e-01
-1.24764311e+00 1.50212884e-01 -4.76326734e-01 2.58584470e-01
1.46490470e-01 -2.29307577e-01 -9.49726403e-01 -3.23957503e-01
-3.79369929e-02 -2.84046233e-01 8.46312881e-01 4.73940402e-01
1.08288646e+00 -2.50124514e-01 -3.06041479e-01 9.52071309e-01
1.58381939e+00 1.27982438e-01 7.66428947e-01 5.84298253e-01
1.30937278e+00 6.31218493e-01 9.79176283e-01 5.97257614e-01
7.43142068e-01 1.27585137e+00 5.78623950e-01 -2.82964446e-02
-4.62626964e-01 -3.99292856e-01 1.81947306e-01 5.87841213e-01
2.16613680e-01 -2.97743797e-01 -8.11352372e-01 6.04269862e-01
-2.15031409e+00 -8.53850245e-01 -4.98978823e-01 1.92047191e+00
6.44703507e-01 2.94289328e-02 3.63610052e-02 2.15866625e-01
3.74060690e-01 4.29351956e-01 -8.11797559e-01 2.39470690e-01
-3.50574136e-01 -5.95503092e-01 3.77216697e-01 6.96007788e-01
-1.17835963e+00 1.07765758e+00 4.98489618e+00 8.10798824e-01
-9.72713709e-01 3.67486775e-01 6.37619555e-01 -3.73191893e-01
-3.55496138e-01 1.63769141e-01 -9.52488482e-01 5.16735554e-01
8.55205953e-02 1.22795194e-01 -7.66468048e-03 1.10349166e+00
4.38267469e-01 -4.18920845e-01 -1.00450253e+00 1.42200148e+00
1.00104861e-01 -1.21964753e+00 -6.01354875e-02 4.91941199e-02
1.24135935e+00 1.11540064e-01 -5.09345755e-02 -5.73392659e-02
-5.76969124e-02 -4.92839128e-01 8.83267760e-01 2.96497047e-01
5.89342535e-01 -3.82581323e-01 4.34060395e-01 3.58349293e-01
-1.57234180e+00 -2.37853304e-01 -3.68573040e-01 -2.27262273e-01
4.33062106e-01 7.91303217e-01 -2.19138339e-01 6.16984665e-01
7.12524712e-01 1.29704225e+00 -5.00540853e-01 1.18066764e+00
-3.37607145e-01 1.15307733e-01 -8.27865750e-02 4.97230381e-01
1.32289469e-01 -1.66534871e-01 6.82552695e-01 6.88750803e-01
8.84705633e-02 2.97792524e-01 1.73869297e-01 6.43134058e-01
2.15911910e-01 5.24217151e-02 -4.37073916e-01 5.64548850e-01
2.96151251e-01 9.77336705e-01 -4.10478473e-01 -2.68815160e-01
-7.08419621e-01 1.11410820e+00 4.46138650e-01 3.55057508e-01
-8.38583708e-01 2.44200379e-01 8.35875511e-01 3.64930481e-01
2.77944177e-01 -3.59810621e-01 -5.25090456e-01 -1.74772191e+00
3.97268772e-01 -5.47019482e-01 2.35349193e-01 -9.53529358e-01
-9.08739269e-01 2.85925329e-01 9.34378058e-02 -1.76491702e+00
1.67643018e-02 -2.99425751e-01 -3.71469557e-01 4.11512345e-01
-1.91540623e+00 -1.13566554e+00 -9.44970787e-01 5.52177250e-01
1.03027534e+00 9.66718271e-02 3.02075949e-02 6.51457012e-01
-6.50609314e-01 4.03940171e-01 1.84866451e-02 -2.76335716e-01
5.49884439e-01 -9.00412202e-01 1.73054963e-01 9.18641329e-01
-5.10349795e-02 -3.08186561e-01 5.51799655e-01 -5.65256715e-01
-1.06907415e+00 -1.25718307e+00 9.72757161e-01 -4.57464337e-01
2.09001943e-01 -2.69363225e-01 -8.41677547e-01 5.97973645e-01
-3.57778519e-01 3.44153881e-01 1.89176366e-01 -5.25201142e-01
-2.52447277e-01 -2.45036617e-01 -9.55412865e-01 6.51529968e-01
1.62422121e+00 -2.22969934e-01 -1.17419966e-01 4.14192021e-01
9.55044866e-01 -7.55250692e-01 -4.17931348e-01 7.27637649e-01
6.52851403e-01 -1.25847840e+00 1.12207294e+00 -4.89849672e-02
7.35551298e-01 -3.71620387e-01 -3.04132670e-01 -7.18793452e-01
-1.94885150e-01 -3.10318023e-01 -5.75428426e-01 9.31726396e-01
-1.74452409e-01 -4.68766958e-01 1.20996130e+00 5.60066998e-01
-2.08911106e-01 -8.20651889e-01 -1.07050920e+00 -9.15044785e-01
-9.48341936e-02 -2.97135442e-01 3.34499806e-01 8.68163526e-01
-4.13708448e-01 2.20825523e-01 -5.69723427e-01 1.84295908e-01
8.19795787e-01 3.16416830e-01 1.21390092e+00 -9.30377781e-01
-4.29318607e-01 -2.51298159e-01 -5.95965385e-01 -1.69588399e+00
1.04992419e-01 -4.75930512e-01 -1.17644183e-01 -1.36824238e+00
2.72531420e-01 -6.71697140e-01 2.68157095e-01 7.17975423e-02
-2.94507682e-01 2.78630644e-01 2.24067599e-01 5.90370595e-01
-4.39045101e-01 1.19022048e+00 1.75694871e+00 3.25738341e-02
-4.19346780e-01 1.71820004e-03 -1.27272859e-01 8.41999769e-01
6.73518598e-01 -2.99912393e-01 -7.78802693e-01 -7.85655141e-01
-2.45986413e-02 1.99178055e-01 5.55442393e-01 -1.14215100e+00
2.56442726e-02 -4.26902831e-01 2.25283071e-01 -1.03636825e+00
6.40462279e-01 -7.52030730e-01 1.55347422e-01 5.66688240e-01
5.11291549e-02 -7.44573548e-02 9.41361412e-02 7.66736269e-01
-2.34825566e-01 1.75552264e-01 1.01607454e+00 2.19413061e-02
-9.03476655e-01 8.05159450e-01 1.76618829e-01 1.42617092e-01
1.13917470e+00 -4.05516922e-01 -2.46430680e-01 -3.14961195e-01
-4.88127261e-01 4.01270628e-01 5.71547627e-01 5.94188333e-01
7.06542253e-01 -1.54011059e+00 -6.72160566e-01 1.96447000e-01
2.71479994e-01 5.58722138e-01 6.51172817e-01 8.46562922e-01
-7.99174428e-01 4.94098157e-01 -2.64538005e-02 -8.14479530e-01
-1.30843735e+00 4.21380728e-01 4.59552765e-01 -3.27996284e-01
-8.14139307e-01 8.35023642e-01 7.88619697e-01 -3.81339222e-01
2.97079802e-01 -3.50164950e-01 2.12587509e-02 -2.17379272e-01
3.22685629e-01 6.54550254e-01 -6.15851432e-02 -7.38855124e-01
-3.54437143e-01 9.17677104e-01 7.53282532e-02 -1.06521674e-01
1.10750246e+00 -6.09619439e-01 3.12753975e-01 3.06387931e-01
1.23747313e+00 -2.26527795e-01 -1.96501493e+00 -5.53150713e-01
-4.71492499e-01 -1.13835299e+00 5.00850156e-02 -3.89320314e-01
-1.26840293e+00 8.11538398e-01 6.97309494e-01 -5.37796497e-01
1.19582331e+00 -7.53138959e-02 1.01603580e+00 1.34433033e-02
4.79322821e-01 -8.70431423e-01 5.54307342e-01 3.44721884e-01
8.46756756e-01 -1.55915987e+00 2.22151712e-01 -9.02194858e-01
-5.80749094e-01 7.59614825e-01 1.32030475e+00 -1.41068816e-01
7.65346885e-01 -1.93565845e-01 -3.85499299e-02 3.73842008e-02
-7.23155022e-01 -3.36982384e-02 2.52368748e-01 4.69770759e-01
7.13670999e-02 -1.67465806e-01 -4.62564170e-01 4.55296561e-02
6.56609908e-02 -4.19724509e-02 3.62269312e-01 9.42753017e-01
-2.71480292e-01 -9.97972727e-01 -2.41049439e-01 -6.58025369e-02
5.55923916e-02 8.93504545e-02 7.24450350e-02 7.80827284e-01
3.30784887e-01 7.70901799e-01 5.16706593e-02 -2.42919549e-01
2.78454781e-01 -4.86970037e-01 6.59939349e-01 -5.53230822e-01
3.07371635e-02 1.69638798e-01 -8.04359168e-02 -8.97572875e-01
-7.43196666e-01 -9.14866745e-01 -1.02763093e+00 -2.73718983e-01
-3.97931486e-01 -5.18235207e-01 3.72903079e-01 8.32391322e-01
2.74706721e-01 1.67884752e-01 8.89428496e-01 -1.19469202e+00
-3.33251327e-01 -6.61947310e-01 -4.95501041e-01 4.84782666e-01
4.30861741e-01 -9.71545398e-01 -4.87308204e-01 1.07740080e-02] | [8.66174030303955, -2.2848546504974365] |
a169790a-c0cb-4690-a133-51d1f8c9ae76 | unsupervised-domain-adaptation-without-source | 2010.12427 | null | https://arxiv.org/abs/2010.12427v5 | https://arxiv.org/pdf/2010.12427v5.pdf | Casting a BAIT for Offline and Online Source-free Domain Adaptation | We address the source-free domain adaptation (SFDA) problem, where only the source model is available during adaptation to the target domain. We consider two settings: the offline setting where all target data can be visited multiple times (epochs) to arrive at a prediction for each target sample, and the online setting where the target data needs to be directly classified upon arrival. Inspired by diverse classifier based domain adaptation methods, in this paper we introduce a second classifier, but with another classifier head fixed. When adapting to the target domain, the additional classifier initialized from source classifier is expected to find misclassified features. Next, when updating the feature extractor, those features will be pushed towards the right side of the source decision boundary, thus achieving source-free domain adaptation. Experimental results show that the proposed method achieves competitive results for offline SFDA on several benchmark datasets compared with existing DA and SFDA methods, and our method surpasses by a large margin other SFDA methods under online source-free domain adaptation setting. | ['Shangling Jui', 'Luis Herranz', 'Joost Van de Weijer', 'Yaxing Wang', 'Shiqi Yang'] | 2020-10-23 | null | null | null | null | ['source-free-domain-adaptation'] | ['computer-vision'] | [ 3.18632096e-01 -1.73375830e-01 -4.18496221e-01 -4.17857856e-01
-7.90257812e-01 -7.09320664e-01 2.61211336e-01 2.73295432e-01
-3.71978015e-01 9.62433577e-01 -4.44901884e-01 -1.42035827e-01
3.05509157e-02 -7.18436241e-01 -6.25073969e-01 -1.01800740e+00
-1.13042600e-01 9.91283596e-01 6.22869253e-01 -3.98938395e-02
-1.85235720e-02 4.23708439e-01 -1.35857379e+00 1.14777915e-01
8.23834062e-01 1.54418683e+00 2.46164456e-01 5.07632852e-01
-1.46852180e-01 3.98109019e-01 -7.42038727e-01 -2.19375148e-01
5.43940902e-01 -3.30263138e-01 -7.50859320e-01 2.25583792e-01
1.18883662e-01 -3.46741974e-01 -1.41222239e-01 8.57945800e-01
3.91450465e-01 4.06283349e-01 5.29581368e-01 -1.47256184e+00
-5.59793830e-01 2.54007101e-01 -3.12106162e-01 5.52185893e-01
1.65818438e-01 -1.47880195e-02 6.84861541e-01 -9.99402881e-01
5.06149352e-01 8.86020720e-01 3.65569443e-01 7.51644254e-01
-1.03499341e+00 -7.81286716e-01 8.00303161e-01 2.27220371e-01
-1.08569205e+00 -4.78314966e-01 7.53423393e-01 -1.88200399e-01
6.63805544e-01 -1.61733136e-01 2.15156034e-01 1.21255267e+00
-2.68960714e-01 1.01168334e+00 1.01762021e+00 -2.77728230e-01
7.58248389e-01 4.01392430e-01 4.05387908e-01 1.67353779e-01
2.81205177e-01 1.42805308e-01 -5.47397196e-01 -5.31886160e-01
4.93073016e-01 1.38020560e-01 -9.85592529e-02 -7.93908775e-01
-1.05781484e+00 6.98807955e-01 2.19528854e-01 -1.45046249e-01
-5.51533997e-01 -5.88935018e-01 5.42468786e-01 8.58788192e-01
4.25260246e-01 7.62644261e-02 -1.02607465e+00 -1.09251201e-01
-5.28576791e-01 1.98731735e-01 9.38069344e-01 1.28850067e+00
8.21150601e-01 -1.19251862e-01 -1.08541258e-01 9.89540339e-01
-2.12268278e-01 5.76058865e-01 8.59627306e-01 -4.26128626e-01
7.21650600e-01 7.41039991e-01 6.16990983e-01 -3.29628199e-01
-2.16156334e-01 -5.91772735e-01 -6.90948784e-01 -5.30343913e-02
6.49919450e-01 -2.02943459e-01 -8.14914167e-01 1.60416329e+00
8.68671119e-01 4.20613736e-01 3.46232951e-01 9.49288130e-01
2.29079530e-01 5.83908439e-01 -1.43757658e-02 -4.80915725e-01
9.01210487e-01 -1.10050213e+00 -2.62793362e-01 -4.07413244e-01
5.39502025e-01 -4.15771753e-01 1.04051721e+00 7.65945017e-01
-7.17502475e-01 -6.45707726e-01 -1.21883333e+00 4.57746625e-01
-4.17809576e-01 1.73236251e-01 1.85657606e-01 4.65855956e-01
-4.31345791e-01 4.08239901e-01 -7.58792758e-01 -3.19485545e-01
4.92750734e-01 4.56919640e-01 -2.97943056e-01 -3.91501188e-01
-1.18552184e+00 5.07701993e-01 5.47090411e-01 -2.90107489e-01
-1.10517251e+00 -5.88715613e-01 -5.69530785e-01 -1.15301087e-01
6.15676284e-01 -4.43004638e-01 1.60489500e+00 -1.42480588e+00
-1.60530198e+00 5.74977398e-01 -4.83299553e-01 -5.41705787e-01
5.79240739e-01 -1.13309607e-01 -7.74019718e-01 -7.98047036e-02
-7.37155089e-04 1.67708471e-01 1.06134999e+00 -1.10404539e+00
-1.33176422e+00 -4.73555654e-01 -1.31600844e-02 4.19738114e-01
-6.38290644e-01 -3.55940372e-01 -3.65253568e-01 -4.49557900e-01
3.05262655e-02 -8.08998346e-01 2.67680697e-02 -6.91400319e-02
2.50567514e-02 -3.34316164e-01 1.14431715e+00 -3.38693649e-01
1.36115396e+00 -2.29077291e+00 -4.21015024e-02 3.27019334e-01
5.21272793e-02 4.74635690e-01 -1.79683894e-01 1.74973875e-01
3.28754932e-02 -3.50217044e-01 -3.38417262e-01 -2.44117513e-01
-2.13992298e-01 3.82118970e-01 -4.59668577e-01 3.27224433e-01
1.49348453e-01 3.96393180e-01 -1.05819261e+00 -1.36973232e-01
-4.75003421e-02 -1.92394570e-01 -4.84056860e-01 3.99744362e-01
-1.83656827e-01 3.48782867e-01 -7.18305588e-01 6.84470296e-01
9.20074046e-01 -3.78588796e-01 2.32113630e-01 3.98337692e-01
3.62950832e-01 1.87685445e-01 -1.51293027e+00 1.38816464e+00
-6.46599889e-01 5.29258624e-02 2.53245197e-02 -1.29439211e+00
1.16320562e+00 7.62476921e-02 4.69213575e-01 -6.94081068e-01
-2.49447182e-01 3.13676685e-01 -2.16504708e-02 -3.19416001e-02
2.21307859e-01 1.96804330e-01 -2.19126746e-01 2.39982277e-01
1.18734255e-01 5.77827930e-01 -3.04128677e-02 -9.83958393e-02
1.12882113e+00 5.46291694e-02 5.96135318e-01 -1.40037134e-01
8.99773657e-01 2.33982340e-01 8.48295033e-01 8.31141949e-01
-7.09595263e-01 3.76210630e-01 1.98368087e-01 -7.61440754e-01
-8.89367223e-01 -1.18355072e+00 2.30424330e-02 1.49261558e+00
4.55699235e-01 1.92547679e-01 -5.50463617e-01 -1.33914471e+00
3.53637695e-01 4.37678725e-01 -6.10150993e-01 -4.00290728e-01
-5.09321153e-01 -4.67544347e-01 7.51259038e-03 4.69022363e-01
7.19555974e-01 -6.95361316e-01 -3.38737190e-01 5.62901139e-01
-6.71066642e-02 -9.30038691e-01 -6.33531690e-01 5.69919765e-01
-9.40175593e-01 -1.10245657e+00 -7.59557784e-01 -8.44513834e-01
7.27084696e-01 3.86078477e-01 7.38435686e-01 -3.82937253e-01
2.54164159e-01 4.23030257e-02 -5.08240163e-01 -3.80367458e-01
-4.48223174e-01 4.75342512e-01 3.00754815e-01 4.14045990e-01
7.01429844e-01 -3.99238080e-01 -5.78818440e-01 8.02719951e-01
-7.96837926e-01 -5.64012110e-01 5.30824065e-01 1.03621042e+00
7.80798912e-01 3.93643528e-02 1.25309110e+00 -1.05700517e+00
4.57642406e-01 -1.04397237e+00 -6.13595665e-01 2.52248555e-01
-8.67535710e-01 -2.42008697e-02 1.38304830e+00 -1.00362158e+00
-9.78071928e-01 2.56995678e-01 3.01260531e-01 -6.74545407e-01
-3.12340826e-01 2.09208459e-01 -5.54929674e-01 1.68643713e-01
8.13958228e-01 5.23263752e-01 -2.58023962e-02 -6.74769104e-01
-2.15655103e-01 9.91067231e-01 5.00552416e-01 -5.59986889e-01
8.61409426e-01 2.13234484e-01 -3.86274755e-01 -3.50791425e-01
-8.58566642e-01 -9.02303696e-01 -8.21459949e-01 1.00531809e-01
7.29407668e-02 -1.00967658e+00 -2.22354680e-01 6.23148322e-01
-6.29548073e-01 -4.32399273e-01 -3.96789908e-01 2.17942312e-01
-3.17392141e-01 1.99041083e-01 -5.73081076e-02 -7.10974276e-01
-1.90098047e-01 -8.97998691e-01 8.71921003e-01 4.11462188e-01
8.80569443e-02 -9.73722219e-01 -6.66679665e-02 -6.25823066e-02
3.08283389e-01 6.28589094e-02 8.24549556e-01 -1.57456803e+00
-8.91665146e-02 -3.44528943e-01 1.01809882e-01 4.75645483e-01
4.45261151e-01 -6.12817407e-01 -9.15506959e-01 -6.40877128e-01
-9.42053795e-02 -3.28865498e-01 5.10713398e-01 4.12229970e-02
1.24703908e+00 -4.09578413e-01 -5.87113619e-01 3.92698854e-01
1.12805915e+00 6.11159444e-01 1.10576078e-01 5.22938728e-01
1.27628192e-01 -3.98523845e-02 1.28453708e+00 8.38108122e-01
3.78585190e-01 6.93718433e-01 2.81587064e-01 1.53825402e-01
8.26751534e-03 -2.89135009e-01 5.79843760e-01 4.21255410e-01
6.39107704e-01 -3.74698013e-01 -8.86950970e-01 7.80047476e-01
-1.90327775e+00 -5.23052633e-01 2.45986834e-01 2.86848974e+00
8.60621691e-01 2.99358010e-01 6.51396394e-01 1.30379483e-01
8.29206645e-01 -2.65992463e-01 -1.42091393e+00 -3.55165511e-01
-3.68510857e-02 1.54280469e-01 5.75354397e-01 2.69683093e-01
-1.39324045e+00 8.50953877e-01 5.93964005e+00 7.73213625e-01
-1.17050135e+00 2.36079037e-01 5.44711411e-01 -1.25339970e-01
2.77458310e-01 -1.08713165e-01 -1.09805381e+00 7.52306104e-01
9.73639905e-01 -3.29158664e-01 6.35292828e-01 1.40773511e+00
-1.99903145e-01 -1.31045297e-01 -1.34948969e+00 8.39716792e-01
-1.66949481e-01 -7.20808506e-01 -1.28874660e-01 -1.12352774e-01
7.22207487e-01 -6.65987134e-02 3.69198956e-02 7.04242170e-01
4.96382326e-01 -3.08955908e-01 3.35104883e-01 1.43865481e-01
8.19747269e-01 -9.03885424e-01 5.50394356e-01 7.24757791e-01
-1.21396589e+00 -5.90192854e-01 -5.07658005e-01 9.67462137e-02
-4.24291253e-01 5.89353561e-01 -1.03182983e+00 5.96315384e-01
6.44335866e-01 6.99027002e-01 -4.43248212e-01 1.15032840e+00
2.06370026e-01 9.09567654e-01 -4.26340044e-01 6.78719506e-02
4.26635556e-02 7.29353428e-02 6.46390975e-01 8.60246480e-01
3.22444230e-01 1.26767743e-04 5.93605816e-01 1.42178655e-01
3.00867041e-03 1.72179475e-01 -4.39304918e-01 3.79314095e-01
1.02907360e+00 1.08099389e+00 -3.53123516e-01 -6.05560422e-01
-4.24223363e-01 1.42776072e+00 3.99950534e-01 3.87194782e-01
-8.54426920e-01 -4.68306154e-01 7.31574655e-01 5.07359207e-03
5.45621634e-01 3.93156707e-02 -8.77362117e-02 -1.22090101e+00
2.94007868e-01 -8.95955443e-01 9.88632619e-01 -2.63776511e-01
-1.64819872e+00 5.25089979e-01 -5.92470765e-02 -1.65697348e+00
-2.15613320e-01 -5.33150434e-01 -3.17140132e-01 8.25722992e-01
-1.75209415e+00 -7.80020595e-01 -3.33444893e-01 9.66517150e-01
7.87841380e-01 -3.79014462e-01 8.58501911e-01 3.37504119e-01
-5.59761882e-01 1.14535010e+00 6.90026104e-01 1.63698453e-03
1.03972530e+00 -1.22359729e+00 3.06857079e-01 6.76869333e-01
-2.52585113e-01 2.03072220e-01 1.81024313e-01 -6.06111169e-01
-1.36998594e+00 -1.53950000e+00 6.25304043e-01 -3.10293585e-01
4.58736986e-01 -4.33438420e-01 -1.10599017e+00 6.75076365e-01
-5.33114791e-01 3.60354513e-01 6.32817566e-01 -1.07006401e-01
-3.86802256e-01 -5.93087316e-01 -1.63861966e+00 1.66223720e-01
1.04987466e+00 1.42204417e-02 -5.06881833e-01 3.18940133e-01
5.41433036e-01 -4.91991878e-01 -8.40186715e-01 1.95341051e-01
2.22329006e-01 -4.94803518e-01 7.49007523e-01 -8.39706540e-01
-1.16278790e-01 -2.96107858e-01 -1.54486358e-01 -1.53572118e+00
-3.31327051e-01 -6.45195186e-01 -5.11147499e-01 1.16260326e+00
4.18838829e-01 -1.12116015e+00 7.21128762e-01 4.19898391e-01
5.38316816e-02 -5.53237915e-01 -1.20979500e+00 -1.35623956e+00
2.62856603e-01 4.14062887e-02 9.26549971e-01 8.98443699e-01
-1.84257075e-01 1.48873881e-01 -2.19147846e-01 5.71502268e-01
4.76433396e-01 3.64902318e-01 9.47966814e-01 -1.33348048e+00
-3.26961130e-01 -9.19446275e-02 -9.55525339e-02 -1.50532627e+00
4.43778634e-02 -7.92436302e-01 -1.99857056e-02 -8.58328998e-01
1.88027155e-02 -8.38851869e-01 -7.43274808e-01 6.21802986e-01
-4.03408527e-01 -2.72978634e-01 -1.16934277e-01 4.61012632e-01
-7.83581197e-01 3.10268760e-01 8.84888887e-01 8.07667598e-02
-5.46642482e-01 5.74026346e-01 -7.07648575e-01 3.58467102e-01
8.34934711e-01 -5.40987849e-01 -6.02027297e-01 -6.36087209e-02
-4.96024013e-01 1.22937046e-01 -1.87222622e-02 -1.03431547e+00
1.96548939e-01 -5.17920136e-01 6.70205534e-01 -3.65931809e-01
9.12028253e-02 -1.16818964e+00 -2.12880090e-01 3.95914972e-01
-3.98644805e-01 -1.14497282e-01 1.14986613e-01 9.08509493e-01
-1.70969397e-01 -1.65864937e-02 1.06498647e+00 2.03834131e-01
-9.85751390e-01 5.94203770e-01 -2.63450116e-01 2.24616423e-01
1.35805416e+00 -4.17247713e-01 -2.60151058e-01 -1.69920877e-01
-9.77247417e-01 5.98995507e-01 5.01651466e-01 5.60019195e-01
4.79809940e-01 -1.38645589e+00 -3.96476537e-01 3.92095864e-01
4.54213142e-01 2.47812450e-01 2.21922528e-02 5.36662161e-01
2.07783356e-01 2.77493238e-01 3.41468640e-02 -6.76172793e-01
-9.76586342e-01 1.02504766e+00 4.24200118e-01 -4.23740208e-01
-3.98973942e-01 6.81315720e-01 1.98229164e-01 -6.13409340e-01
3.93247008e-01 -3.06352407e-01 -2.94725690e-02 -1.31821871e-01
5.19696891e-01 4.42229211e-01 2.75750399e-01 -1.82671279e-01
-4.88455057e-01 7.80223235e-02 -3.56639624e-01 2.62973398e-01
1.07431757e+00 -3.89785558e-01 5.51719666e-01 4.78377879e-01
1.14761531e+00 -1.89425439e-01 -1.57268155e+00 -8.54418874e-01
1.37963623e-01 -5.80909014e-01 -2.88483441e-01 -1.31792831e+00
-9.31533813e-01 6.05603576e-01 8.69084299e-01 9.28139687e-02
1.59664381e+00 -2.09552914e-01 9.37623084e-01 5.17384529e-01
5.71015060e-01 -1.43512332e+00 4.13200110e-02 5.53928196e-01
5.03803432e-01 -1.38405669e+00 -3.33274215e-01 -7.76990429e-02
-1.00449753e+00 1.08870399e+00 1.04512239e+00 -1.40165299e-01
5.86256564e-01 7.10231438e-02 -1.78681076e-01 5.47288716e-01
-1.07465565e+00 -4.89344150e-02 7.21183955e-04 9.41220939e-01
-3.97552699e-01 1.82530209e-01 2.28686854e-01 1.03117990e+00
1.50369510e-01 3.93655449e-02 3.00937474e-01 9.82392967e-01
-6.38412118e-01 -1.38013017e+00 -2.97195226e-01 4.99874592e-01
-2.00759992e-01 2.51337171e-01 -5.44806182e-01 6.97480500e-01
7.44183809e-02 8.96780908e-01 1.64677203e-01 -2.09320202e-01
4.37065929e-01 5.00718296e-01 1.50502607e-01 -6.11139715e-01
-4.25092012e-01 -1.09984942e-01 -1.46648269e-02 -3.52724165e-01
1.63954288e-01 -8.54317069e-01 -1.28141880e+00 2.14334745e-02
-2.70793468e-01 3.29666346e-01 3.79698455e-01 7.98596203e-01
7.16229022e-01 3.27949855e-03 1.31442904e+00 -3.44218552e-01
-1.04391587e+00 -7.92849362e-01 -5.15002608e-01 3.74196649e-01
6.50225222e-01 -8.18847299e-01 -3.73569548e-01 1.50470972e-01] | [10.360246658325195, 3.145914316177368] |
aac4c998-c36b-44bb-9b95-fda819ce1ad7 | concise-answers-to-complex-questions | 2305.19271 | null | https://arxiv.org/abs/2305.19271v1 | https://arxiv.org/pdf/2305.19271v1.pdf | Concise Answers to Complex Questions: Summarization of Long-form Answers | Long-form question answering systems provide rich information by presenting paragraph-level answers, often containing optional background or auxiliary information. While such comprehensive answers are helpful, not all information is required to answer the question (e.g. users with domain knowledge do not need an explanation of background). Can we provide a concise version of the answer by summarizing it, while still addressing the question? We conduct a user study on summarized answers generated from state-of-the-art models and our newly proposed extract-and-decontextualize approach. We find a large proportion of long-form answers (over 90%) in the ELI5 domain can be adequately summarized by at least one system, while complex and implicit answers are challenging to compress. We observe that decontextualization improves the quality of the extractive summary, exemplifying its potential in the summarization task. To promote future work, we provide an extractive summarization dataset covering 1K long-form answers and our user study annotations. Together, we present the first study on summarizing long-form answers, taking a step forward for QA agents that can provide answers at multiple granularities. | ['Eunsol Choi', 'Fangyuan Xu', 'Abhilash Potluri'] | 2023-05-30 | null | null | null | null | ['long-form-question-answering', 'extractive-summarization'] | ['natural-language-processing', 'natural-language-processing'] | [ 2.23093376e-01 7.99661815e-01 -1.07180163e-01 -3.27946007e-01
-1.60367012e+00 -9.50952411e-01 5.50086260e-01 8.47405910e-01
-4.55881238e-01 1.03312564e+00 1.07318044e+00 -4.03849214e-01
-2.73296982e-01 -5.76678813e-01 -5.92475414e-01 9.48825106e-02
6.18294418e-01 9.33276772e-01 4.51059550e-01 -7.83962667e-01
4.30855453e-01 -1.90122813e-01 -1.35242069e+00 1.07730913e+00
1.67968702e+00 5.04899263e-01 3.32146287e-01 1.09025550e+00
-7.12651789e-01 1.10171473e+00 -1.31548274e+00 -8.59796643e-01
-3.84082586e-01 -6.78872764e-01 -1.40623355e+00 -1.55104041e-01
1.04656589e+00 -4.89050150e-01 -2.74233371e-01 6.77307367e-01
6.05838001e-01 -3.01009081e-02 7.48803616e-01 -7.38788307e-01
-8.47291231e-01 7.78795660e-01 -3.79263831e-04 3.70735586e-01
1.20753074e+00 1.75702378e-01 1.34445179e+00 -6.04731500e-01
7.30064631e-01 1.14674830e+00 4.22880828e-01 6.51175857e-01
-1.16846228e+00 1.97380856e-01 8.09359774e-02 3.30356538e-01
-5.70357084e-01 -5.77887714e-01 4.98359293e-01 -1.42686561e-01
1.44148529e+00 9.09345329e-01 2.69794077e-01 7.09378362e-01
-8.18763301e-02 1.12205172e+00 6.78387105e-01 -4.22936827e-01
-1.73437148e-02 7.51655102e-02 7.48981953e-01 5.07366478e-01
3.77157360e-01 -8.24268818e-01 -5.89745998e-01 -3.01147550e-01
-1.13906138e-01 -3.35115403e-01 -4.45513666e-01 2.12756783e-01
-1.14511156e+00 7.86938012e-01 2.04237431e-01 1.07596107e-01
-6.34763479e-01 -1.51105061e-01 4.90090460e-01 6.37015998e-01
3.81758362e-01 1.06474292e+00 -6.48077786e-01 -3.01625103e-01
-1.08064377e+00 1.04475796e+00 1.52988791e+00 9.87095475e-01
7.07429171e-01 -4.31964129e-01 -8.90904486e-01 7.49149561e-01
-2.63856620e-01 5.28356791e-01 3.74205947e-01 -1.53559101e+00
1.01040661e+00 8.40964317e-01 5.16239822e-01 -8.98667157e-01
-2.69537210e-01 -6.36974216e-01 -5.03276706e-01 -6.41206741e-01
4.74422127e-01 -2.66671330e-01 -3.48978996e-01 1.33990014e+00
7.94155449e-02 -7.33700871e-01 3.61671776e-01 5.96090198e-01
1.75221765e+00 9.10683870e-01 -7.83070773e-02 -3.17453444e-01
1.60577023e+00 -1.30954421e+00 -1.05967593e+00 -5.08219540e-01
7.98684776e-01 -8.28731656e-01 1.37467086e+00 1.49568439e-01
-1.50441337e+00 -5.08992791e-01 -7.91580975e-01 -7.64613450e-01
-1.41634615e-02 1.07815839e-01 2.51576841e-01 3.46824378e-01
-1.04451215e+00 3.46790671e-01 -2.43729353e-01 -4.17930037e-01
9.40985382e-02 -1.41848087e-01 -2.62734741e-01 -3.77312720e-01
-1.12452793e+00 1.17471731e+00 1.69045314e-01 -5.52847445e-01
-1.60323769e-01 -9.67824757e-01 -9.51252401e-01 3.07014436e-01
6.38963282e-01 -1.42148459e+00 1.96029031e+00 -2.01054931e-01
-1.04892278e+00 5.90700746e-01 -7.79136479e-01 -6.15498364e-01
3.35012317e-01 -4.95949328e-01 -1.39549196e-01 6.49324477e-01
4.42543566e-01 6.53105557e-01 3.54051322e-01 -1.08819139e+00
-7.21794963e-01 -2.24332795e-01 6.83628321e-01 3.48156393e-01
-1.26226217e-01 3.26814130e-02 -3.81367505e-01 -4.47647631e-01
-9.87396762e-02 -1.81037188e-01 -9.27558020e-02 -5.84262490e-01
-4.57329482e-01 -6.62127495e-01 5.58213711e-01 -1.42407012e+00
1.87867594e+00 -1.46759665e+00 1.49602100e-01 -5.12891352e-01
4.95374054e-01 2.64998138e-01 -3.29108268e-01 1.10478914e+00
3.55983377e-01 2.00412005e-01 -4.97326195e-01 -4.09416676e-01
4.04361367e-01 2.68810213e-01 -7.38114059e-01 -3.37313801e-01
2.99428970e-01 1.42891467e+00 -1.01047182e+00 -6.32388532e-01
-3.82641196e-01 -8.20760950e-02 -8.76940191e-01 3.80927056e-01
-8.18805516e-01 2.02077195e-01 -3.73638958e-01 4.44969147e-01
2.74232537e-01 -4.37713176e-01 -3.56339335e-01 -2.98887521e-01
1.90943629e-01 1.02599454e+00 -7.41203189e-01 1.60876405e+00
-4.12745386e-01 5.67284405e-01 -3.85213806e-03 -6.44974828e-01
4.25797850e-01 3.20475757e-01 4.79984172e-02 -8.93208325e-01
-3.34181786e-01 2.95094937e-01 -1.75363138e-01 -9.95414555e-01
1.12623954e+00 1.39737412e-01 -3.42335165e-01 6.40391350e-01
1.58570036e-01 -6.85154200e-01 1.06142247e+00 1.03900778e+00
1.29414594e+00 -1.97575733e-01 4.85610247e-01 3.43323573e-02
4.91550684e-01 5.13614833e-01 -2.81259120e-02 1.19272327e+00
3.17395002e-01 5.97464383e-01 7.78798938e-01 3.93479317e-03
-8.55038345e-01 -6.90980315e-01 2.84525186e-01 1.07328284e+00
-1.66329473e-01 -1.07992947e+00 -8.97950172e-01 -1.00430083e+00
-3.00961193e-02 1.23905158e+00 -2.63673335e-01 9.88129824e-02
-7.86151707e-01 -1.69841811e-01 5.10356605e-01 4.14743304e-01
4.06745106e-01 -1.01430988e+00 -7.84880817e-01 3.72316331e-01
-1.08558857e+00 -9.58112240e-01 -6.14583254e-01 -5.58944494e-02
-9.07158673e-01 -9.98940885e-01 -8.29295516e-01 -5.85628152e-01
4.54435050e-01 2.41561517e-01 1.89704049e+00 4.44833606e-01
4.21936810e-01 7.70169973e-01 -5.35601676e-01 -3.48593831e-01
-6.59187973e-01 5.01230061e-01 -6.26610696e-01 -7.36107409e-01
2.68783540e-01 -4.01896805e-01 -5.74763834e-01 -9.46611632e-03
-1.00138485e+00 1.07585736e-01 5.73506236e-01 5.77253699e-01
2.51523316e-01 -2.97298104e-01 1.09962845e+00 -1.21662474e+00
1.38830411e+00 -4.81101871e-01 1.00590602e-01 5.35095751e-01
-2.33062446e-01 4.70787793e-01 6.02331877e-01 -6.20469898e-02
-1.21890712e+00 -5.51411569e-01 -6.43180907e-01 6.60307288e-01
-5.40598519e-02 9.63471532e-01 -6.66966140e-02 6.46307349e-01
9.74862814e-01 1.15557320e-01 -2.06111029e-01 -6.75273955e-01
9.56406534e-01 6.73044920e-01 8.70450914e-01 -7.60132790e-01
6.19627535e-01 3.29224989e-02 -5.68719208e-01 -7.81419933e-01
-1.40239966e+00 -7.69328594e-01 -2.65049040e-01 -3.60730886e-02
6.27397239e-01 -5.07049799e-01 -4.93775934e-01 -1.80631131e-01
-1.63478756e+00 -2.55957007e-01 -5.86788833e-01 -2.49364644e-01
-5.95353484e-01 8.16057563e-01 -6.94355786e-01 -5.20574927e-01
-7.00326443e-01 -6.90285325e-01 1.04265559e+00 3.56952965e-01
-9.26625550e-01 -8.09305966e-01 6.95180148e-02 1.04481459e+00
5.58509529e-01 -1.53646469e-02 1.43744206e+00 -1.07049620e+00
-4.15772408e-01 -1.91768065e-01 -1.43673066e-02 1.72786057e-01
3.90460566e-02 -3.43174458e-01 -6.15679860e-01 -1.72890630e-02
1.89820081e-01 -5.77930212e-01 1.10901117e+00 8.51546451e-02
8.68881941e-01 -1.09716165e+00 1.08985431e-01 -3.26197445e-01
7.81310380e-01 -2.91176260e-01 5.85588336e-01 1.37166008e-01
2.79816329e-01 1.07997191e+00 4.99518573e-01 1.67205930e-01
9.99922752e-01 4.22340840e-01 2.06558585e-01 2.09261566e-01
-4.53068137e-01 -4.36211973e-01 1.33605719e-01 1.09030592e+00
3.22007835e-01 -3.94091189e-01 -6.13905728e-01 7.44879425e-01
-1.86744225e+00 -1.20007432e+00 -4.25688177e-01 1.71705270e+00
1.30886531e+00 4.09813784e-02 1.88372582e-01 1.55063912e-01
1.43290907e-01 2.02662408e-01 -4.47968721e-01 -4.04973567e-01
-2.92700201e-01 4.47020531e-01 -1.80586666e-01 8.85637879e-01
-5.40971339e-01 6.76709414e-01 6.31579399e+00 8.58690202e-01
-3.98245156e-01 -4.59914915e-02 2.76601821e-01 9.27965119e-02
-9.73568261e-01 9.69185308e-02 -7.54516602e-01 2.22059622e-01
9.99500751e-01 -5.08137703e-01 8.72165188e-02 6.10644996e-01
5.96395880e-02 -4.23534364e-01 -1.20736098e+00 6.45981252e-01
3.51339728e-01 -1.68897140e+00 4.41838354e-01 -4.22758311e-01
6.70922935e-01 -3.98581475e-01 -1.08484365e-01 8.04538548e-01
7.21370354e-02 -9.63111699e-01 6.74031913e-01 6.86925769e-01
4.83766764e-01 -6.07876718e-01 9.51146424e-01 8.86976004e-01
-7.39435673e-01 -5.32560889e-03 -2.54197627e-01 -2.74313778e-01
4.93526548e-01 4.86404359e-01 -6.44120872e-01 8.49561095e-01
3.78906459e-01 3.70534062e-01 -1.14050865e+00 1.08514345e+00
-4.72517729e-01 5.52645028e-01 -1.85773835e-01 -3.69229108e-01
1.80581406e-01 6.30995333e-02 5.98139584e-01 1.26781976e+00
2.66234547e-01 6.26545429e-01 -1.33111747e-03 6.93409443e-01
-2.63781250e-01 2.10428625e-01 -2.10684106e-01 -1.55825317e-01
3.90806496e-01 1.09636724e+00 -1.90235883e-01 -7.62534857e-01
-3.79989892e-01 9.88113642e-01 3.52685422e-01 4.64714825e-01
-2.74133593e-01 -6.38815701e-01 2.39832681e-02 2.12672696e-01
2.67032403e-02 -1.69139370e-01 -3.98456544e-01 -1.37259793e+00
4.80433732e-01 -1.59592700e+00 8.39334488e-01 -1.13371110e+00
-1.20447326e+00 5.24916232e-01 1.15496881e-01 -7.27839649e-01
-7.12170064e-01 -1.54605180e-01 -5.75378358e-01 8.83712053e-01
-1.59148991e+00 -8.37940156e-01 -2.78458953e-01 1.72435641e-01
9.78581727e-01 2.05573320e-01 7.48058617e-01 3.30707133e-01
-9.11648348e-02 2.92197973e-01 -2.72918731e-01 -2.39652842e-01
8.82171512e-01 -1.51809633e+00 3.04302514e-01 8.31738830e-01
3.19260955e-01 1.06206548e+00 1.30605662e+00 -7.95627832e-01
-1.38324475e+00 -6.93881452e-01 1.58324647e+00 -1.14572310e+00
4.04429764e-01 1.13793507e-01 -1.23348248e+00 5.31871498e-01
9.08525765e-01 -8.88594389e-01 7.20132709e-01 1.07188739e-01
-2.29997396e-01 1.00372545e-01 -1.06104231e+00 6.22308135e-01
7.01440275e-01 -5.39330184e-01 -1.68617404e+00 4.67365116e-01
1.32764792e+00 -6.11661792e-01 -4.47144240e-01 3.83613318e-01
9.95060652e-02 -8.70596886e-01 8.01596284e-01 -7.81776547e-01
7.92110562e-01 -3.75594437e-01 -8.65249112e-02 -1.42635942e+00
1.56666245e-02 -6.87813938e-01 -8.29978883e-01 1.29237044e+00
7.54889190e-01 -3.08848381e-01 6.97817147e-01 7.30074406e-01
-4.22419071e-01 -8.01259279e-01 -6.22868538e-01 -4.09013271e-01
8.69343206e-02 -3.40600759e-01 5.50722063e-01 3.36866200e-01
4.62403327e-01 8.81622434e-01 -4.39301766e-02 -1.78201139e-01
2.97658503e-01 2.08830848e-01 9.86785591e-01 -9.82409596e-01
-3.43925804e-01 -4.52759832e-01 4.73201901e-01 -1.85137355e+00
5.36156222e-02 -8.91423583e-01 -3.25602353e-01 -2.72032070e+00
3.53200704e-01 3.70807052e-01 5.37311018e-01 1.50615558e-01
-5.75017810e-01 -6.16896525e-02 9.43143144e-02 1.58992603e-01
-1.12504148e+00 5.67742944e-01 1.37651861e+00 -2.48614550e-01
-6.06663972e-02 9.81708094e-02 -1.43054199e+00 6.41830504e-01
4.73653823e-01 -2.07420349e-01 -5.33967018e-01 -6.74360514e-01
7.36889899e-01 6.20541394e-01 2.52004892e-01 -7.17707574e-01
6.10748768e-01 -5.16380705e-02 -3.17202695e-02 -1.18047714e+00
1.76731259e-01 -3.95944476e-01 -4.34923291e-01 2.96023875e-01
-6.58215404e-01 3.67864370e-01 1.41051918e-01 3.04558396e-01
-4.81940955e-01 -5.64693093e-01 1.55052066e-01 -3.37954015e-01
-4.05488372e-01 -5.60327768e-02 -4.87604380e-01 7.32653379e-01
1.16643056e-01 -4.81778830e-02 -9.67421293e-01 -1.07168281e+00
-4.24982846e-01 6.71158075e-01 1.64449111e-01 2.81498004e-02
7.26663053e-01 -1.07014489e+00 -1.12402403e+00 -4.28737193e-01
3.40322256e-01 -9.14793462e-03 4.53835458e-01 6.48028314e-01
-5.42888105e-01 9.37896848e-01 1.83090761e-01 -2.64667690e-01
-1.24969351e+00 4.03620213e-01 3.76224816e-02 -6.07725918e-01
-6.66085899e-01 6.54096484e-01 -7.17607066e-02 -5.78879774e-01
2.44450599e-01 -7.89068043e-01 -6.66388690e-01 5.85157096e-01
9.06074286e-01 4.54354405e-01 3.10183316e-01 -1.15189388e-01
4.44295630e-02 2.93635339e-01 -1.18533589e-01 -2.87895560e-01
1.29888678e+00 -5.04948497e-01 -4.03656662e-01 1.42268613e-01
8.71650279e-01 5.51049054e-01 -7.78460383e-01 -4.60536540e-01
3.03657919e-01 -1.24824814e-01 -6.13263011e-01 -1.20799971e+00
-1.81756124e-01 7.79490411e-01 -5.15748024e-01 6.09096825e-01
1.03717387e+00 5.36816537e-01 1.22164476e+00 9.92308021e-01
1.10971555e-01 -9.49198663e-01 4.93519068e-01 9.49679375e-01
1.43577003e+00 -1.05739772e+00 -5.84718846e-02 -2.85299420e-01
-8.03879499e-01 1.00826645e+00 5.26860356e-01 3.26615691e-01
-2.52930254e-01 -8.94443691e-02 6.45655915e-02 -5.41554570e-01
-1.07824183e+00 -3.14477116e-01 6.79400265e-01 5.17407298e-01
5.23954272e-01 -2.90869832e-01 -5.33173800e-01 1.08785427e+00
-8.74125540e-01 -5.18454313e-01 8.20664108e-01 8.16203654e-01
-9.89413261e-01 -8.59848857e-01 -1.83533922e-01 6.92916572e-01
-5.75479090e-01 -3.34334731e-01 -6.85822845e-01 4.42241669e-01
-5.14832616e-01 1.66092491e+00 -3.24910104e-01 2.15811506e-01
7.19532430e-01 3.40340495e-01 5.42479396e-01 -1.03468275e+00
-8.46448481e-01 -4.43718612e-01 8.27958763e-01 -3.83213818e-01
-2.07707480e-01 -3.16834450e-01 -1.16534448e+00 -3.13723773e-01
2.84422506e-02 7.66663373e-01 2.68810511e-01 1.14152491e+00
4.48346913e-01 6.77850485e-01 -1.36204928e-01 -2.26756290e-01
-7.03677058e-01 -1.26531982e+00 2.03044206e-01 3.69636089e-01
6.64556503e-01 8.84217918e-02 -3.40653628e-01 1.70423388e-01] | [11.601612091064453, 8.186164855957031] |
6de79ec9-4838-4c73-bfbc-4d987b8f1929 | negation-scope-delimitation-in-clinical-text | null | null | https://aclanthology.org/W13-5635 | https://aclanthology.org/W13-5635.pdf | Negation Scope Delimitation in Clinical Text Using Three Approaches: NegEx, PyConTextNLP and SynNeg | null | ['Sumithra Velupillai', 'Hideyuki Tanushi', 'Hercules Dalianis', 'Martin Duneld', 'Maria Skeppstedt', 'Maria Kvist'] | 2013-05-01 | null | null | null | ws-2013-5 | ['negation-detection'] | ['natural-language-processing'] | [-8.63703638e-02 1.71006292e-01 -6.22772932e-01 -4.08054382e-01
-8.41685571e-03 -9.08429027e-01 6.55310392e-01 -6.53472245e-01
-2.85945535e-01 1.06888819e+00 -4.63127941e-02 -1.01159286e+00
-3.91567826e-01 -9.63214397e-01 -4.95059669e-01 -6.31337762e-01
-9.79754329e-01 7.25764990e-01 3.30370307e-01 -6.93831444e-01
7.03166842e-01 7.88774848e-01 -1.68942046e+00 7.18545914e-01
7.04417467e-01 8.52217197e-01 2.49141872e-01 1.14950800e+00
-1.95044339e-01 1.55633950e+00 -7.48382092e-01 -5.46825826e-01
3.13719302e-01 -1.23176083e-01 -7.22945035e-01 -1.01074085e-01
9.28529128e-02 -8.59008506e-02 -2.09758401e-01 9.22211111e-01
5.37373662e-01 4.49454933e-02 1.08379531e+00 -1.42548037e+00
-5.91619551e-01 6.10313773e-01 -4.01565880e-02 1.21627934e-01
1.03678203e+00 -5.39447069e-01 1.19919395e+00 -1.13026452e+00
7.20913768e-01 1.26888943e+00 8.66221786e-01 5.44149756e-01
-1.22286928e+00 -1.94712028e-01 -3.26822817e-01 -9.51717794e-02
-1.46558487e+00 -3.25250506e-01 4.25783843e-02 -2.08119690e-01
1.66093647e+00 1.26596653e+00 1.20609856e+00 1.01401424e+00
1.26658809e+00 8.34431887e-01 1.04267764e+00 -5.13792276e-01
3.35295945e-01 3.66983831e-01 1.54683650e-01 6.33519173e-01
8.40953708e-01 5.26628852e-01 -7.06372619e-01 -9.13127720e-01
9.33553874e-01 -2.94925272e-01 1.71355158e-01 -5.05680561e-01
-9.05919552e-01 6.91228509e-01 1.78732842e-01 3.83959889e-01
-1.39880210e-01 9.89067405e-02 1.26390755e-01 5.30987144e-01
-2.58292928e-02 6.47037446e-01 -9.11868811e-01 -1.33165747e-01
-8.71728659e-01 5.10332465e-01 1.25398111e+00 1.52653182e+00
1.24482810e-01 2.94908643e-01 -9.34252143e-02 3.17179203e-01
8.92314315e-01 1.01808000e+00 4.28362608e-01 -1.36146402e+00
-6.87414408e-02 1.72361732e-01 5.01781464e-01 -8.52631688e-01
-6.33224547e-01 -9.64177120e-03 -8.93263519e-01 4.49267089e-01
3.49161088e-01 4.57367361e-01 -8.02827001e-01 5.07305264e-01
4.33481112e-02 -2.34125629e-01 4.53833073e-01 5.55570945e-02
4.99930978e-01 3.76208365e-01 -1.34477139e-01 -5.73289394e-01
1.06082785e+00 -1.36716676e+00 -1.35299087e+00 2.33215362e-01
9.05734658e-01 -1.07320261e+00 4.35900748e-01 5.33875942e-01
-1.55548143e+00 -1.37560293e-01 -1.08699942e+00 1.78573877e-01
-7.27255583e-01 -3.14239264e-01 8.57801437e-01 1.43120694e+00
-1.60129595e+00 9.73287821e-01 -4.91727620e-01 6.59165755e-02
1.20568443e-02 8.24621081e-01 -2.64718989e-03 4.62812334e-01
-1.33193445e+00 1.08501506e+00 2.22979754e-01 -1.21242590e-01
-1.65216476e-01 -2.13068098e-01 -8.23704481e-01 -5.63443303e-01
-4.78693932e-01 -5.29636025e-01 1.44139910e+00 -2.59346128e-01
-1.65295815e+00 9.71794367e-01 -1.42069459e-01 -1.97814897e-01
6.14786744e-01 -1.28011424e-02 -8.31891418e-01 2.42498964e-01
-1.89849049e-01 5.76383233e-01 9.28263724e-01 -1.35132408e+00
-7.59897232e-01 -1.67359829e-01 -1.23336017e-01 2.66287565e-01
-1.25510961e-01 1.89734384e-01 2.11616129e-01 -1.12999000e-01
3.27147305e-01 -7.20919967e-01 -2.53068686e-01 -5.32041907e-01
-1.46512717e-01 -7.10518599e-01 7.70373225e-01 -4.51523662e-01
1.83705616e+00 -1.67618537e+00 -2.06720144e-01 3.98590982e-01
3.57815564e-01 -1.24705513e-03 2.40583986e-01 1.08380008e+00
-2.76906848e-01 7.32199550e-01 4.11965609e-01 -1.10722095e-01
2.24991128e-01 4.85861301e-01 -4.16602850e-01 3.05609167e-01
-1.29282743e-01 1.15307164e+00 -1.16605783e+00 -5.23096442e-01
4.11106765e-01 9.42391157e-02 -4.58732933e-01 4.79237735e-01
2.99364805e-01 1.47170946e-01 -3.56553018e-01 1.39399457e+00
1.15709066e+00 -1.31984919e-01 1.45911396e-01 5.30878425e-01
-3.79135728e-01 3.55090618e-01 -6.96863770e-01 1.05554795e+00
7.08333924e-02 5.00173986e-01 1.02364108e-01 -7.94621468e-01
3.33247900e-01 8.61540735e-01 4.37155962e-01 -9.67555881e-01
-2.26398129e-02 6.92409754e-01 1.12803578e-01 -6.07703328e-01
7.58228302e-01 3.94563079e-02 -3.64872098e-01 6.40070081e-01
-2.37588286e-01 -6.59476995e-01 9.64643434e-02 3.08100313e-01
6.22585893e-01 -5.18246442e-02 5.62923312e-01 -1.05613089e+00
7.86340594e-01 -1.82965681e-01 -1.81299388e-01 1.03415680e+00
-3.09923887e-01 3.19085121e-01 2.99841821e-01 -6.65102363e-01
-6.45341039e-01 -1.12307119e+00 -4.89381433e-01 1.30636716e+00
3.24267983e-01 -4.39044595e-01 -9.54439282e-01 -2.49762803e-01
1.77620783e-01 6.89606130e-01 -5.90509653e-01 3.84124845e-01
-5.03739953e-01 -8.32535863e-01 7.39044368e-01 3.45434904e-01
-5.07752821e-02 -1.33414865e+00 -6.58416986e-01 1.25490099e-01
-2.20292807e-01 -6.63697243e-01 -6.23428151e-02 4.48765576e-01
-1.35989368e+00 -5.18594682e-01 -6.66252747e-02 -8.20914626e-01
5.87345481e-01 2.46782884e-01 1.27047324e+00 5.39230824e-01
-2.31483161e-01 4.26904231e-01 -1.21292919e-01 -4.95818377e-01
-4.59671497e-01 -8.00336525e-02 5.28869390e-01 -5.87835789e-01
5.19427478e-01 -2.50617653e-01 -7.29350567e-01 5.37953973e-01
-6.88540697e-01 1.62748516e-01 1.79803044e-01 1.04410267e+00
1.35816500e-01 -9.34035778e-02 1.22507080e-01 -6.38007045e-01
8.72274399e-01 -1.69219792e-01 -3.78732830e-01 5.77745810e-02
-6.77108407e-01 -3.74140263e-01 3.21430594e-01 -3.25342178e-01
-1.01981449e+00 -4.87835288e-01 -9.82677937e-02 2.45538145e-01
1.11353043e-02 -1.46784872e-01 6.47139177e-02 -5.24923325e-01
8.02199244e-01 9.25758183e-02 1.99174434e-02 -6.80815242e-03
3.01039815e-01 7.09525108e-01 -6.82967342e-03 -6.68678164e-01
8.44880998e-01 4.91470337e-01 7.98524171e-02 -9.57177758e-01
-1.52186140e-01 -2.60129690e-01 -9.51962709e-01 -6.54426932e-01
6.56643391e-01 -6.78531289e-01 -9.10833478e-01 3.91110867e-01
-9.38691139e-01 -3.38627815e-01 -3.91645581e-01 4.25431967e-01
-1.01278400e+00 2.75717527e-02 -3.90154392e-01 -1.27895141e+00
-5.10977268e-01 -1.02017939e+00 9.43384409e-01 5.30070923e-02
-5.10597289e-01 -1.26927447e+00 5.87685481e-02 2.71537274e-01
1.81734428e-01 -1.73075795e-01 6.90226793e-01 -2.38256708e-01
-4.24233019e-01 -1.53791070e-01 2.34436691e-02 -1.39755070e-01
1.70832314e-02 4.95917559e-01 -9.81751978e-01 -5.31145096e-01
6.65065646e-02 -1.92070693e-01 -1.08835101e-01 6.52520418e-01
5.91872573e-01 -2.29931593e-01 -8.56000841e-01 5.40386558e-01
1.38545322e+00 3.85070026e-01 5.32770038e-01 7.28214979e-01
1.41836226e-01 5.53460240e-01 9.17806149e-01 4.63203549e-01
1.30579369e-02 3.28798652e-01 2.40537539e-01 1.49327129e-01
1.11720070e-01 -1.54819340e-01 3.77893507e-01 1.16112018e+00
-8.18235934e-01 -2.69281328e-01 -5.07867396e-01 4.42987174e-01
-1.72482407e+00 -1.40330648e+00 -4.32368398e-01 6.90478683e-01
6.25676990e-01 1.56016424e-01 -1.48347050e-01 3.35214496e-01
4.99015123e-01 -2.03574806e-01 -1.19133167e-01 -1.06291151e+00
-1.43546045e-01 3.15233678e-01 7.37729073e-01 1.00061214e+00
-7.20721722e-01 1.03317809e+00 1.29781246e+01 1.02230716e+00
2.21112028e-01 1.03134915e-01 5.16071796e-01 3.48020852e-01
-4.36954498e-01 -4.56139445e-02 -1.04416132e+00 2.72933897e-02
1.38140702e+00 -4.30666685e-01 6.85999811e-01 5.44219851e-01
3.44648361e-01 -4.23268199e-01 -1.26188684e+00 5.26221812e-01
9.73738134e-02 -1.40886843e+00 -2.83300440e-04 6.85225725e-01
7.73699820e-01 -5.08050561e-01 6.22419357e-01 3.24184299e-01
6.09259963e-01 -1.14389277e+00 8.60300779e-01 2.53660440e-01
1.03040910e+00 -6.05088234e-01 5.67372203e-01 1.68872893e-01
-1.14389896e+00 -2.20873043e-01 -8.77727985e-01 -1.00755692e+00
3.93533185e-02 -1.81779593e-01 -4.29956943e-01 3.48861217e-01
9.58353162e-01 2.99398601e-01 -3.93658698e-01 9.95779395e-01
-4.78694476e-02 1.04875881e-02 -2.71853864e-01 -4.48467314e-01
4.83122796e-01 -3.54241252e-01 4.66730654e-01 1.00164843e+00
2.48499006e-01 3.51035744e-01 -9.84472036e-02 4.01770771e-01
5.45058846e-01 3.29446048e-02 -1.19659424e+00 -1.78908288e-01
2.83276141e-01 9.16795909e-01 -4.83487815e-01 -4.22520459e-01
-2.00212970e-01 8.62069130e-01 -3.55488248e-02 5.01107454e-01
-6.11489356e-01 -4.35615242e-01 9.72222984e-01 -1.27327025e-01
-1.14700586e-01 -3.48497719e-01 -6.23769283e-01 -7.30352640e-01
-5.89872956e-01 -4.54965204e-01 5.93606755e-02 -5.53365827e-01
-1.39813089e+00 5.79277515e-01 -2.27688253e-02 -1.40553558e+00
-6.99901402e-01 -1.27676582e+00 -4.76714373e-01 4.92853165e-01
-1.11898029e+00 -1.10984349e+00 2.50124663e-01 4.52870727e-01
1.64141744e-01 -5.34416080e-01 1.39563632e+00 3.57715860e-02
1.00637585e-01 9.24474537e-01 6.69434488e-01 -7.37814724e-01
5.56605101e-01 -1.27867436e+00 5.68737745e-01 -1.37897313e-01
-4.31265175e-01 9.05828118e-01 6.28349900e-01 -5.39804697e-01
-1.41196322e+00 -3.66917729e-01 1.08350635e+00 -9.83769417e-01
6.55218959e-01 -3.86345625e-01 4.23767231e-02 7.88592756e-01
7.15902448e-01 -6.18741751e-01 8.21781039e-01 -1.83753878e-01
1.80774391e-01 5.75296998e-01 -1.39248300e+00 6.12354755e-01
1.66275799e+00 -4.63594139e-01 -6.25784039e-01 7.60327101e-01
8.13696027e-01 -6.94087505e-01 -1.30082703e+00 3.34633321e-01
8.65424156e-01 -8.75409484e-01 1.61978090e+00 -1.32660246e+00
-4.43697497e-02 2.81152606e-01 -2.61993498e-01 -9.32519078e-01
-5.96193194e-01 -1.23518765e+00 -5.33532679e-01 -5.83747849e-02
5.96577883e-01 -1.13057327e+00 3.42365682e-01 8.74560475e-01
-2.82833427e-01 -6.42737269e-01 -1.06996536e+00 -1.32016802e+00
-3.35779637e-02 -1.45572275e-01 4.90409225e-01 7.63798356e-01
6.83744550e-01 1.09839931e-01 -6.36873543e-02 -1.00294888e-01
5.33176839e-01 9.55312885e-03 4.41501856e-01 -1.34294486e+00
3.86843324e-01 -5.75816095e-01 -3.07655483e-01 -9.45992947e-01
-8.85957032e-02 -8.12076271e-01 -6.53862000e-01 -1.28511906e+00
-8.32044985e-03 -1.91056758e-01 -1.11109078e-01 -1.65725678e-01
3.67937148e-01 2.13746816e-01 1.20859891e-02 1.03788137e-01
-3.71160030e-01 6.18435517e-02 1.29639816e+00 6.91750320e-05
-1.62315920e-01 4.85058486e-01 -4.81304944e-01 7.84440815e-01
8.58408585e-02 -2.99253196e-01 -6.78878546e-01 6.11881316e-02
6.69384480e-01 4.61409837e-02 3.24159935e-02 -7.42885649e-01
5.37211418e-01 -3.75702560e-01 4.78586555e-01 -1.32223868e+00
1.30741090e-01 -9.61415648e-01 6.95283338e-02 9.40189242e-01
2.74610907e-01 1.20424610e-02 8.66204947e-02 5.39500564e-02
-1.44871444e-01 -5.70943117e-01 9.21121240e-01 -4.07591403e-01
-4.92852688e-01 -5.20386267e-03 -1.03226590e+00 8.97834301e-02
9.92593169e-01 -7.84614205e-01 -3.59281451e-01 -4.20183957e-01
-8.29068601e-01 -1.95836127e-02 6.50830388e-01 3.03609259e-02
7.30431557e-01 -1.51530886e+00 -2.30721906e-01 7.18729138e-01
-3.22939813e-01 -3.74400020e-01 -1.70157343e-01 6.58265352e-01
-1.32361674e+00 1.02442718e+00 -5.41665435e-01 -4.55340147e-01
-1.14228773e+00 4.78126436e-01 4.28307921e-01 -2.41845414e-01
-2.15481281e-01 1.11954463e+00 2.71224789e-02 -8.23025763e-01
1.85185194e-01 -9.21545625e-02 -7.54407048e-01 3.55081353e-03
6.88606799e-01 1.05194807e+00 -2.91290224e-01 -6.04341030e-01
-4.56784427e-01 6.65885091e-01 2.32151806e-01 -2.87484169e-01
9.17833567e-01 -2.19243199e-01 -9.89108324e-01 4.28274393e-01
8.38715494e-01 -1.36269778e-01 -3.69319022e-02 4.16855574e-01
1.25943512e-01 -8.02164078e-01 -4.32406247e-01 -3.55811834e-01
-1.85641110e-01 5.54822803e-01 5.34874737e-01 8.99602413e-01
8.57008278e-01 -3.02566767e-01 8.18335712e-01 9.66778398e-01
5.72402716e-01 -1.68019545e+00 -2.13140488e-01 6.89524531e-01
9.12339568e-01 -9.28350806e-01 5.44190466e-01 -7.27165341e-01
-4.14997995e-01 1.32979155e+00 4.68304873e-01 -1.55325383e-01
1.27306652e+00 5.74917436e-01 1.14069022e-02 -3.70670199e-01
-9.44949508e-01 1.12705544e-01 3.75366658e-01 1.11147714e+00
5.06513238e-01 5.07374525e-01 -9.66500878e-01 3.21953118e-01
-7.28706717e-01 -2.34555230e-01 4.90474731e-01 1.41972518e+00
-6.43810987e-01 -1.20391917e+00 -7.23931909e-01 4.61561680e-01
-5.49773455e-01 -1.16372630e-01 -5.28106689e-01 8.46754074e-01
-5.76629937e-02 1.49448860e+00 -1.97535474e-03 -5.03491640e-01
4.11356747e-01 1.41089618e-01 7.21762300e-01 -1.23501487e-01
-9.04846430e-01 4.15413082e-01 3.84890139e-01 -1.23056793e+00
-8.58632207e-01 -1.05834293e+00 -1.40667629e+00 -1.19437599e+00
-5.12782812e-01 1.89310342e-01 3.83317530e-01 3.90289724e-01
-2.06836104e-01 2.85260603e-02 9.80917513e-01 -1.07949340e+00
-3.90341938e-01 -9.39418614e-01 -1.00026262e+00 -7.84516707e-02
2.89751232e-01 -8.00943017e-01 -7.83523321e-01 2.83909619e-01] | [-7.395808696746826, 3.7034409046173096] |
cad5ef65-ddf8-4fc4-ba26-d85fb9ae74de | formal-verification-of-piece-wise-linear-feed | 1705.01320 | null | http://arxiv.org/abs/1705.01320v3 | http://arxiv.org/pdf/1705.01320v3.pdf | Formal Verification of Piece-Wise Linear Feed-Forward Neural Networks | We present an approach for the verification of feed-forward neural networks
in which all nodes have a piece-wise linear activation function. Such networks
are often used in deep learning and have been shown to be hard to verify for
modern satisfiability modulo theory (SMT) and integer linear programming (ILP)
solvers.
The starting point of our approach is the addition of a global linear
approximation of the overall network behavior to the verification problem that
helps with SMT-like reasoning over the network behavior. We present a
specialized verification algorithm that employs this approximation in a search
process in which it infers additional node phases for the non-linear nodes in
the network from partial node phase assignments, similar to unit propagation in
classical SAT solving. We also show how to infer additional conflict clauses
and safe node fixtures from the results of the analysis steps performed during
the search. The resulting approach is evaluated on collision avoidance and
handwritten digit recognition case studies. | ['Ruediger Ehlers'] | 2017-05-03 | null | null | null | null | ['handwritten-digit-recognition'] | ['computer-vision'] | [ 4.55174387e-01 8.47346544e-01 -1.64846718e-01 -5.87474346e-01
-2.46090263e-01 -6.45668983e-01 1.86009318e-01 2.39950851e-01
-1.82006344e-01 9.83688295e-01 -6.77142859e-01 -1.04774261e+00
-5.30117691e-01 -9.33627665e-01 -1.43599021e+00 -5.90191007e-01
-5.21316588e-01 7.38237679e-01 3.76440287e-01 -2.29621485e-01
5.08256927e-02 7.28007674e-01 -1.25902033e+00 4.42459524e-01
2.06115171e-01 1.20362639e+00 -3.48484755e-01 7.26422668e-01
-1.33612990e-01 1.02617371e+00 -4.80905771e-01 -2.37816349e-01
5.16621232e-01 -2.98536688e-01 -1.21822989e+00 -3.30874622e-01
6.42810166e-01 -4.11988884e-01 -2.25161672e-01 1.30556166e+00
-1.50269017e-01 -1.42084554e-01 1.93574697e-01 -2.06029105e+00
2.70231992e-01 1.50943983e+00 -2.26217240e-01 7.51119629e-02
-1.68758571e-01 -4.64772470e-02 1.20803666e+00 -5.90163842e-02
5.81913590e-01 1.28595936e+00 7.69439757e-01 6.08252048e-01
-1.23013866e+00 -7.87778199e-01 3.34317982e-01 4.44223881e-01
-1.37987959e+00 -3.16131920e-01 6.49777591e-01 -8.73054415e-02
1.56695545e+00 3.72466773e-01 7.84020185e-01 4.34538811e-01
3.48635554e-01 5.86109042e-01 6.02592647e-01 -5.07883787e-01
6.55865252e-01 3.13788429e-02 4.04578328e-01 1.06192756e+00
5.72439790e-01 2.65122175e-01 -3.37175071e-01 1.81480616e-01
6.78649426e-01 -4.90324408e-01 -6.32188469e-02 -3.79771024e-01
-7.90229976e-01 7.55198658e-01 7.21791744e-01 2.08754107e-01
-8.27479828e-03 9.10331428e-01 7.45621145e-01 4.63816375e-01
-2.93973684e-01 4.91283804e-01 -7.41654336e-01 3.98093015e-01
-1.04320467e+00 4.41162556e-01 1.22892058e+00 1.00724745e+00
6.19706810e-01 2.42511123e-01 1.72189415e-01 -3.55496854e-01
4.10738289e-01 2.05835328e-01 -1.58268645e-01 -1.06245458e+00
4.71946180e-01 5.49231410e-01 -3.53035808e-01 -9.13627207e-01
-4.25765067e-01 -5.44075370e-01 -8.80382061e-01 6.52596354e-01
5.18705308e-01 -4.27273899e-01 -7.78007448e-01 1.92119002e+00
2.31976762e-01 4.01515901e-01 7.00075030e-02 6.80462062e-01
2.71184802e-01 6.53364599e-01 -4.96473372e-01 -2.57498801e-01
1.07762361e+00 -6.99326217e-01 -4.56907243e-01 -2.94714034e-01
1.00199008e+00 2.27336541e-01 -5.00198156e-02 8.70651960e-01
-1.50810266e+00 -1.10780649e-01 -1.59171164e+00 1.80594847e-01
-1.93427473e-01 -3.92431915e-01 9.11438882e-01 5.03940821e-01
-1.32520866e+00 9.30425346e-01 -9.97252524e-01 4.62726951e-01
5.77546299e-01 1.23586369e+00 -4.17182714e-01 -2.38897074e-02
-1.10569870e+00 9.56646144e-01 8.24437797e-01 6.86220229e-01
-8.99728358e-01 -8.42338443e-01 -1.08050489e+00 4.96845961e-01
6.03932202e-01 -3.09187621e-01 1.57080722e+00 -1.11010504e+00
-1.14814794e+00 5.02408385e-01 -1.97779417e-01 -1.07974708e+00
3.29698265e-01 6.65379763e-01 2.18044415e-01 1.60375252e-01
-5.92267692e-01 8.39095533e-01 3.86773407e-01 -9.87889230e-01
-4.29541141e-01 -3.58941019e-01 4.22603637e-01 -3.81978035e-01
2.92154223e-01 -1.48710147e-01 -6.03013439e-03 1.16494656e-01
3.45369607e-01 -9.84876335e-01 -4.51652169e-01 4.12789434e-01
-7.87332535e-01 -5.62831722e-02 5.98064661e-01 -3.10840786e-01
7.58034170e-01 -1.66614234e+00 2.60178387e-01 9.55386639e-01
4.51716512e-01 7.66878277e-02 3.07743018e-03 -5.44646271e-02
-4.65131164e-01 1.40031844e-01 -1.03741493e-02 -2.15201885e-01
3.48981112e-01 6.26724303e-01 -3.54676992e-01 6.48965001e-01
3.69026959e-01 1.05834424e+00 -6.24072373e-01 -5.54169774e-01
-1.90324560e-02 1.11578278e-01 -7.48104036e-01 -4.22610402e-01
-7.77442873e-01 -3.42814207e-01 -1.70747384e-01 2.64421403e-01
8.56435001e-01 -1.94385514e-01 5.81807017e-01 -1.44850314e-01
5.84910661e-02 4.75838631e-01 -1.46089482e+00 1.19023418e+00
-1.74385235e-01 1.10831797e+00 3.87947410e-01 -1.36963701e+00
4.62780029e-01 2.14583457e-01 1.11193903e-01 -6.62505388e-01
8.08925748e-01 1.09394334e-01 4.24201310e-01 7.95289278e-02
9.59061086e-02 -2.50026315e-01 -2.08228286e-02 4.93436366e-01
1.20668925e-01 -3.31053650e-03 1.44112691e-01 1.90585703e-01
1.24348474e+00 -2.24525094e-01 4.51407321e-02 -4.18639690e-01
5.92284024e-01 3.00160766e-01 5.27631462e-01 7.52862930e-01
1.32337704e-01 1.87103084e-04 1.29689360e+00 -4.52116013e-01
-9.77868795e-01 -3.59716117e-01 7.84907937e-02 2.61905015e-01
-2.19989836e-01 -3.04017007e-01 -1.00872588e+00 -4.60766017e-01
-3.61008823e-01 5.85217476e-01 -8.07428241e-01 -3.55335504e-01
-7.38894165e-01 7.05685765e-02 9.88064408e-01 9.08549905e-01
2.31708631e-01 -7.93316245e-01 -9.10558939e-01 1.08699000e-03
3.53604704e-01 -1.13275862e+00 2.96997845e-01 1.20658481e+00
-9.62313533e-01 -1.18274677e+00 1.98991686e-01 -9.02538180e-01
1.05695379e+00 -4.09293443e-01 6.57395124e-01 5.36298513e-01
-1.91109374e-01 -4.36593205e-01 2.81961799e-01 -3.34751844e-01
-6.24566257e-01 -4.87178843e-03 4.39428650e-02 -6.01743519e-01
6.44601211e-02 -3.92039388e-01 2.78314680e-01 1.31524593e-01
-8.93048167e-01 2.18277186e-01 2.61469334e-01 8.26598167e-01
5.91231287e-01 8.92436385e-01 -2.10498482e-01 -8.17884982e-01
1.61904290e-01 -2.66971439e-02 -1.60262012e+00 5.06955743e-01
-5.47371507e-01 7.14499652e-01 9.14669812e-01 -5.29116571e-01
-2.26841867e-01 3.95801932e-01 8.05182755e-02 -6.60163522e-01
7.68365934e-02 7.88632214e-01 -3.94035190e-01 -6.77021265e-01
5.42096794e-01 -1.97312608e-01 1.23734042e-01 2.19133198e-01
-1.73065543e-01 -2.42707089e-01 6.02521300e-01 -6.81676984e-01
1.16443408e+00 1.40776029e-02 1.03301632e+00 -5.16866595e-02
-4.77348477e-01 3.57812494e-01 -4.44422126e-01 -8.71245787e-02
3.28436852e-01 -3.77879381e-01 -1.72099161e+00 4.54982698e-01
-1.29749656e+00 -6.85186863e-01 -4.00253862e-01 8.91676843e-02
-2.69715577e-01 -5.52416891e-02 -5.38886189e-01 -1.08145046e+00
-1.10550046e-01 -1.38514376e+00 6.55830681e-01 9.72220600e-02
-1.38635769e-01 -9.57922101e-01 -3.27896118e-01 -2.49270499e-01
1.49852425e-01 3.13256443e-01 1.29643404e+00 -8.60583901e-01
-9.18712616e-01 -3.53454828e-01 -2.05954984e-01 2.11729497e-01
-7.11563587e-01 1.57954395e-01 -9.68953133e-01 -1.19553901e-01
-6.95156083e-02 -1.59665421e-01 5.32713532e-01 5.14386296e-01
1.19960582e+00 -8.38150740e-01 -3.88810426e-01 4.45830077e-01
1.68790138e+00 2.70683300e-02 5.59511244e-01 2.42799044e-01
3.31977636e-01 6.04412973e-01 2.51673222e-01 1.71498179e-01
-1.27762690e-01 4.94422466e-01 1.02737796e+00 2.57824004e-01
3.96806270e-01 -9.24217999e-02 2.05155715e-01 -8.22675675e-02
2.80621558e-01 -1.86795652e-01 -1.03876519e+00 2.47501984e-01
-1.81434250e+00 -9.27572727e-01 -3.43542099e-01 1.93385327e+00
8.64985704e-01 7.60775685e-01 -1.42302945e-01 9.15649712e-01
3.77584070e-01 -5.89358807e-01 -6.78787529e-01 -8.86966705e-01
2.23538756e-01 6.44403815e-01 1.00761414e+00 1.08547986e+00
-5.78296959e-01 6.91708684e-01 6.37148905e+00 4.49895561e-01
-1.04212284e+00 -2.54685551e-01 4.59094018e-01 -1.66116819e-01
-2.79010862e-01 1.26497999e-01 -8.54630232e-01 -2.86022455e-01
1.10271382e+00 2.79843975e-02 7.36886263e-01 7.46148884e-01
-2.88571209e-01 -1.96585968e-01 -1.75269496e+00 3.52950364e-01
-1.32699087e-01 -1.56675327e+00 -4.33601856e-01 2.20795721e-01
2.52446085e-01 -1.61287352e-01 -9.55690145e-02 1.89450964e-01
4.43845570e-01 -1.30065739e+00 1.10147059e+00 -1.54319063e-01
6.92485809e-01 -1.08065116e+00 8.72843504e-01 5.33019722e-01
-8.69921029e-01 -3.00077200e-01 -1.95073873e-01 -2.83857137e-01
-2.20911279e-02 2.97470480e-01 -1.22069371e+00 2.72440791e-01
2.52055526e-01 1.27139434e-01 -2.12888762e-01 1.09395850e+00
-4.16638076e-01 2.32469872e-01 -8.35156977e-01 -1.97106734e-01
4.30503607e-01 1.59334823e-01 2.99499720e-01 7.94310570e-01
-9.61576402e-02 -5.93864918e-02 -4.98838067e-01 1.18264997e+00
-1.35660879e-02 -8.62751007e-01 -3.41385692e-01 -6.67075068e-02
3.38008136e-01 9.65563118e-01 -1.01145375e+00 -8.58553201e-02
9.01410356e-02 4.61043984e-01 2.18222722e-01 8.46024156e-02
-1.06198704e+00 -2.50326365e-01 5.14751554e-01 -9.75712761e-02
5.39670706e-01 -1.67166114e-01 -6.33235037e-01 -6.22894704e-01
1.99647799e-01 -9.33954656e-01 6.44065514e-02 -7.41566598e-01
-4.69381779e-01 5.23396313e-01 3.90853167e-01 -4.83142257e-01
-5.81744075e-01 -1.15310979e+00 -5.80913544e-01 9.20543194e-01
-1.42604280e+00 -6.45648479e-01 -5.55478595e-03 5.78762650e-01
-8.16064030e-02 1.69020638e-01 8.15459490e-01 9.33510587e-02
-8.22831035e-01 9.86981094e-01 -6.57174706e-01 2.13498399e-01
-4.66422200e-01 -1.02945626e+00 4.46460009e-01 1.16734731e+00
-7.57168382e-02 8.12315106e-01 9.81186807e-01 -4.04116064e-01
-1.99398696e+00 -8.28528047e-01 1.03040922e+00 2.41546631e-01
7.92635143e-01 -5.63534915e-01 -6.14834905e-01 1.12649322e+00
6.36712834e-02 1.11030862e-01 8.68528560e-02 9.95932594e-02
-5.01064062e-01 -1.41775116e-01 -1.30525649e+00 3.26983899e-01
8.45835686e-01 -2.89646089e-01 -1.82134703e-01 3.44995797e-01
6.44037545e-01 -1.00752091e+00 -3.58582675e-01 2.12709457e-01
5.52202046e-01 -3.41583014e-01 8.03377509e-01 -8.76683414e-01
6.00878656e-01 -4.52833593e-01 -1.30266890e-01 -8.05671513e-01
-1.39202639e-01 -5.96890509e-01 -5.78376234e-01 5.75678289e-01
7.82866299e-01 -4.48471069e-01 1.13753963e+00 1.04415905e+00
-2.95294195e-01 -9.55379009e-01 -1.22908056e+00 -7.44497895e-01
4.24617119e-02 -6.37457013e-01 4.76282865e-01 6.78442240e-01
3.11415493e-01 2.74107724e-01 1.41549826e-01 5.06311059e-01
5.11862755e-01 -1.12176068e-01 1.46562159e-01 -1.40100276e+00
-7.25887060e-01 -7.28680134e-01 -6.75403774e-01 -5.65019011e-01
6.70352995e-01 -1.03519166e+00 2.53781974e-01 -1.30605149e+00
-1.93471476e-01 -4.97016668e-01 3.18536572e-02 1.20969093e+00
9.07454252e-01 -1.64682150e-03 -6.97088316e-02 -6.27282619e-01
-4.03637201e-01 -3.12563419e-01 8.68463695e-01 -5.05911946e-01
1.64478794e-01 7.71188363e-02 -5.91952980e-01 3.84603053e-01
6.06654763e-01 -6.35103762e-01 -4.51704234e-01 -3.91706228e-01
1.09253383e+00 3.13244969e-01 6.13013148e-01 -1.06367886e+00
7.83174217e-01 -1.82752460e-01 1.04250327e-01 -5.00069499e-01
2.38032460e-01 -1.42540312e+00 4.26216036e-01 1.05801213e+00
-6.52521968e-01 9.44982544e-02 7.31916308e-01 -3.08879111e-02
5.61455684e-03 -6.76983237e-01 6.05794907e-01 9.16269198e-02
-3.93624812e-01 1.67925611e-01 -2.57750064e-01 -3.57888669e-01
1.11101484e+00 -1.92064375e-01 -3.86213183e-01 -2.44485959e-01
-5.50258517e-01 5.38257778e-01 -2.80504376e-02 -1.86539903e-01
6.45557404e-01 -9.12451208e-01 -2.22687513e-01 2.93216318e-01
-4.14197177e-01 4.92548257e-01 5.10583669e-02 1.03856730e+00
-8.63932610e-01 7.16818810e-01 -4.18900192e-01 -3.34544361e-01
-1.53789508e+00 7.09511995e-01 1.01194322e+00 -3.33160728e-01
-3.17753673e-01 1.06007099e+00 -4.04272079e-01 -5.32174781e-02
8.01396787e-01 -1.10278881e+00 1.77381963e-01 -3.00209314e-01
6.04665875e-01 5.94720952e-02 5.03036022e-01 -1.54095501e-01
-6.29895151e-01 -2.21995693e-02 -5.22042140e-02 -2.31779963e-01
1.42551482e+00 5.46160579e-01 -4.42965806e-01 3.91248204e-02
1.09211683e+00 -6.35185301e-01 -8.50900829e-01 -9.71751735e-02
-2.76228875e-01 3.48710954e-01 4.13814038e-01 -6.92083597e-01
-1.48405468e+00 9.05924141e-01 1.36621371e-01 1.85390696e-01
1.25330913e+00 -2.64254779e-01 4.18680817e-01 1.16906679e+00
5.34804583e-01 -7.59945989e-01 -7.79177725e-01 8.14500570e-01
2.85704345e-01 -5.63405216e-01 8.55167434e-02 -3.93964589e-01
1.32100448e-01 1.38641524e+00 6.79644406e-01 -2.25740761e-01
3.47709864e-01 1.36303425e+00 -7.05013037e-01 -5.01996614e-02
-1.00533473e+00 3.96491796e-01 4.13601473e-02 2.27596059e-01
-2.16457129e-01 -1.95138127e-01 1.83844507e-01 7.09996760e-01
-3.98547739e-01 1.15705922e-01 5.29750049e-01 1.02276969e+00
-3.93287212e-01 -1.12143314e+00 -2.27505118e-01 1.51935220e-01
-3.26677889e-01 -1.76434010e-01 -5.31616271e-01 9.48804736e-01
5.59254102e-02 4.74321872e-01 5.85145988e-02 -5.67730844e-01
4.34210300e-02 9.44117382e-02 1.11863649e+00 -2.47437760e-01
-8.11132014e-01 -4.50968951e-01 3.17185163e-01 -7.40522504e-01
-1.43856676e-02 -5.53484738e-01 -1.86398220e+00 -5.71797848e-01
-6.51855707e-01 1.20866641e-01 9.26538885e-01 1.27202857e+00
-1.35640904e-01 9.02605653e-01 1.90159410e-01 -7.58509219e-01
-5.21477878e-01 -3.09673041e-01 -2.89594203e-01 -6.26969457e-01
7.89929926e-01 -2.68438429e-01 -3.49546760e-01 -3.54790956e-01] | [8.72257137298584, 6.931527614593506] |
411e0dca-d224-4154-9296-94814d534e3b | fast-graph-sampling-for-short-video | 2110.11420 | null | https://arxiv.org/abs/2110.11420v2 | https://arxiv.org/pdf/2110.11420v2.pdf | Fast Graph Sampling for Short Video Summarization using Gershgorin Disc Alignment | We study the problem of efficiently summarizing a short video into several keyframes, leveraging recent progress in fast graph sampling. Specifically, we first construct a similarity path graph (SPG) $\mathcal{G}$, represented by graph Laplacian matrix $\mathbf{L}$, where the similarities between adjacent frames are encoded as positive edge weights. We show that maximizing the smallest eigenvalue $\lambda_{\min}(\mathbf{B})$ of a coefficient matrix $\mathbf{B} = \text{diag}(\mathbf{a}) + \mu \mathbf{L}$, where $\mathbf{a}$ is the binary keyframe selection vector, is equivalent to minimizing a worst-case signal reconstruction error. We prove that, after partitioning $\mathcal{G}$ into $Q$ sub-graphs $\{\mathcal{G}^q\}^Q_{q=1}$, the smallest Gershgorin circle theorem (GCT) lower bound of $Q$ corresponding coefficient matrices -- $\min_q \lambda^-_{\min}(\mathbf{B}^q)$ -- is a lower bound for $\lambda_{\min}(\mathbf{B})$. This inspires a fast graph sampling algorithm to iteratively partition $\mathcal{G}$ into $Q$ sub-graphs using $Q$ samples (keyframes), while maximizing $\lambda^-_{\min}(\mathbf{B}^q)$ for each sub-graph $\mathcal{G}^q$. Experimental results show that our algorithm achieves comparable video summarization performance as state-of-the-art methods, at a substantially reduced complexity. | ['Chia-Wen Lin', 'Gene Cheung', 'Sadid Sahami'] | 2021-10-21 | null | null | null | null | ['graph-sampling'] | ['graphs'] | [ 4.15721178e-01 4.65277284e-01 -8.12589601e-02 1.26579953e-02
-9.61072624e-01 -7.19919503e-01 -3.74591023e-01 2.56217390e-01
-2.17811465e-01 5.71664631e-01 -2.62009710e-01 -7.23264217e-01
-6.04516685e-01 -1.01216257e+00 -8.86845112e-01 -6.24576151e-01
-1.04987776e+00 -1.31533578e-01 9.30317864e-02 -2.80296147e-01
1.75213441e-01 2.59857893e-01 -1.33142006e+00 -2.06297323e-01
8.91693532e-01 1.33286703e+00 1.95172504e-01 1.02485335e+00
1.11237608e-01 6.66387260e-01 -4.76769209e-01 -8.04613888e-01
6.41509950e-01 -1.12819731e+00 -6.95037663e-01 2.02056170e-01
7.39804924e-01 -8.48714635e-02 -7.68504679e-01 1.67086542e+00
1.06590174e-01 2.77826071e-01 7.10225165e-01 -9.68802691e-01
-3.76662135e-01 4.87696439e-01 -1.02859557e+00 3.64542842e-01
5.79349935e-01 1.20735385e-01 1.37038803e+00 -5.23545206e-01
5.79344213e-01 9.00560856e-01 6.29440486e-01 -1.94126163e-02
-1.32178450e+00 -1.16260111e+00 3.87136549e-01 1.41316711e-03
-1.79963684e+00 -1.68358997e-01 5.34563959e-01 -3.44097048e-01
7.39995718e-01 6.53413773e-01 8.47347915e-01 -1.65646300e-01
1.84969559e-01 4.46664721e-01 8.04050446e-01 -5.02452672e-01
8.84228125e-02 -3.62495661e-01 1.73502326e-01 1.52532101e+00
4.21715409e-01 -2.96552807e-01 -3.94058019e-01 2.26000324e-02
1.08153450e+00 -2.78894603e-01 -6.37810886e-01 1.23680532e-01
-1.07966101e+00 8.55465949e-01 7.51766562e-02 7.23882392e-03
-4.70463820e-02 7.40673482e-01 3.43013443e-02 6.45740688e-01
8.37570652e-02 1.54894575e-01 -4.32536490e-02 -4.46203649e-02
-1.18630970e+00 1.55173257e-01 6.96807683e-01 1.67122650e+00
1.35241127e+00 3.59101385e-01 -6.98358193e-02 6.51898086e-01
7.38445893e-02 8.28775406e-01 -3.66868824e-01 -1.62516117e+00
7.67266512e-01 5.42677104e-01 -1.18443750e-01 -1.23976243e+00
-3.07309985e-01 -3.11901778e-01 -1.17987609e+00 -4.26199147e-03
6.33215010e-01 -3.56368899e-01 -4.01554674e-01 1.96521652e+00
-1.39892057e-01 1.72252581e-02 -5.73131144e-01 5.93579471e-01
5.09502649e-01 8.34486306e-01 -1.87538147e-01 -1.02268445e+00
1.36174488e+00 -4.38588887e-01 -2.42170736e-01 -3.22189592e-02
4.50930446e-01 -8.64474118e-01 1.12508690e+00 5.79756558e-01
-1.51725388e+00 -4.93092686e-01 -1.08181870e+00 5.03736079e-01
3.83773535e-01 -2.16190979e-01 5.18552125e-01 1.10907125e+00
-1.28604519e+00 5.21769285e-01 -3.15033019e-01 -1.23859316e-01
1.69632956e-01 5.68735778e-01 -3.22593123e-01 -2.15377808e-01
-7.25014269e-01 -9.85191166e-02 7.00201690e-02 -2.09153056e-01
-7.28878975e-01 -5.80915630e-01 -7.93870211e-01 8.77760872e-02
7.08842456e-01 -3.67103010e-01 5.64023435e-01 -6.85297430e-01
-9.47288454e-01 9.46275234e-01 -1.48924068e-01 -2.11559594e-01
1.21411264e-01 5.08753002e-01 -3.80538583e-01 6.44479334e-01
1.99813128e-01 3.45223665e-01 9.90422547e-01 -1.12093723e+00
-7.54489958e-01 -5.19845307e-01 3.48569423e-01 8.44095740e-03
-1.63371578e-01 1.23793088e-01 -8.44800234e-01 -1.00817204e+00
7.19913244e-01 -1.02469885e+00 -2.69711465e-01 -3.63909245e-01
-5.04053116e-01 2.07585290e-01 9.84446034e-02 -5.51483572e-01
2.10605264e+00 -2.16739702e+00 1.78010181e-01 9.52732027e-01
7.50403881e-01 -7.61267170e-02 1.22984082e-01 5.69662035e-01
-4.41472232e-02 1.50708854e-01 -1.84614107e-01 2.51872480e-01
-1.36067018e-01 -1.09757401e-01 5.91171235e-02 5.87441623e-01
-6.17830873e-01 3.30383807e-01 -9.34004784e-01 -3.22587788e-01
-6.11366443e-02 -1.19267451e-02 -6.40523493e-01 -1.59853891e-01
-1.74150720e-01 1.35179266e-01 -4.78300810e-01 5.82425594e-01
8.00035954e-01 -3.98201615e-01 5.43774545e-01 -2.93273062e-01
9.21622142e-02 -3.13745856e-01 -1.82929409e+00 1.55409563e+00
2.74820536e-01 4.53320831e-01 6.89485371e-01 -1.35168040e+00
7.73608267e-01 -6.68122470e-02 1.04715276e+00 -6.87888622e-01
3.03417295e-01 1.04754746e-01 -2.35083655e-01 -1.27149761e-01
3.31886560e-01 -2.16552019e-01 -2.82109141e-01 6.94403350e-01
-2.92515848e-02 -3.71565849e-01 8.61945689e-01 7.30333745e-01
1.58540571e+00 -1.42586663e-01 6.95649385e-02 -5.99448025e-01
6.68479502e-01 -2.17634872e-01 6.30093098e-01 1.03000891e+00
-3.02358955e-01 4.65369731e-01 1.03036678e+00 4.92972285e-02
-8.60334396e-01 -1.13633692e+00 2.68443555e-01 1.13331401e+00
4.84317571e-01 -1.07277429e+00 -1.31153536e+00 -4.39486742e-01
-1.67307273e-01 4.23719138e-01 -3.26695889e-01 -2.13190857e-02
-7.30622292e-01 -6.68076813e-01 5.42670965e-01 2.82254100e-01
5.19262731e-01 -7.03385651e-01 -5.66668451e-01 3.02213281e-01
-1.85254931e-01 -8.07974637e-01 -1.03649175e+00 -5.86443394e-03
-8.04151833e-01 -1.24797916e+00 -6.81567371e-01 -5.73759913e-01
9.22161520e-01 5.33234596e-01 1.06443560e+00 2.78577685e-01
-4.27913100e-01 7.20010757e-01 -4.29612517e-01 -3.11561916e-02
-5.08241840e-02 -4.08258438e-01 -4.61765118e-02 8.75981525e-02
-4.58092205e-02 -8.28230083e-01 -1.14432359e+00 3.14169079e-01
-9.42442954e-01 -6.28850982e-02 2.51610130e-01 6.24492943e-01
1.08591497e+00 3.18594426e-01 1.73024938e-01 -6.82723701e-01
3.99257272e-01 -1.99594232e-03 -8.38666260e-01 4.87102233e-02
-6.55472696e-01 -1.54254690e-01 8.31044495e-01 -8.00206587e-02
-2.15114906e-01 -1.83871388e-01 7.41421953e-02 -6.02369845e-01
3.79607767e-01 4.98311847e-01 4.10120748e-02 -2.08575502e-01
5.95012724e-01 4.63431120e-01 -1.69850230e-01 -1.12287104e-01
6.13144398e-01 1.15630649e-01 5.32839239e-01 -6.86088324e-01
7.28042424e-01 4.51466799e-01 4.94081587e-01 -9.14373398e-01
-4.24586415e-01 -2.59241968e-01 -8.77185836e-02 -4.03586298e-01
7.36136973e-01 -8.95179451e-01 -1.34733200e+00 -2.00498700e-01
-3.67486835e-01 -2.80962467e-01 -5.30748785e-01 5.46938002e-01
-8.65464628e-01 8.90660942e-01 -5.95467389e-01 -9.23935711e-01
-3.19102079e-01 -1.24324167e+00 6.61664248e-01 1.09395981e-01
-2.23544076e-01 -5.56841850e-01 -5.21784008e-01 4.62136537e-01
-5.80624118e-03 2.15707108e-01 1.17568600e+00 5.71952350e-02
-9.99038935e-01 -3.21777314e-01 -6.58174694e-01 5.69006085e-01
-1.10029548e-01 -2.23804727e-01 -1.46548852e-01 -7.55718052e-01
-2.62654394e-01 2.33280882e-01 6.53047562e-01 5.19009292e-01
1.15857351e+00 -6.49479032e-01 -2.83880740e-01 5.94012082e-01
1.54169023e+00 3.41286868e-01 6.18755400e-01 -2.62867481e-01
4.77205276e-01 -3.54236215e-02 3.54252726e-01 8.46166313e-01
3.34750086e-01 3.81048054e-01 3.55053306e-01 3.02201569e-01
-1.63530767e-01 5.42775281e-02 4.79315132e-01 7.98842251e-01
-2.68697143e-01 -2.38912746e-01 -5.69515049e-01 5.09559393e-01
-1.56350756e+00 -9.40159857e-01 -2.91740805e-01 2.62594080e+00
7.00956404e-01 3.03173184e-01 3.00021142e-01 3.32181633e-01
6.44242227e-01 2.99602002e-01 -1.10244185e-01 -2.70953208e-01
-1.14762932e-01 1.06589937e+00 8.27033699e-01 5.51218867e-01
-7.90942430e-01 7.66914368e-01 4.16205168e+00 1.33661139e+00
-6.94118023e-01 -1.51584908e-01 6.42732799e-01 -3.19963336e-01
-3.90780538e-01 3.82327408e-01 -6.04591668e-01 6.86367035e-01
7.39803076e-01 -2.77854860e-01 6.78482473e-01 5.11980295e-01
2.07377031e-01 -5.65743685e-01 -9.50429261e-01 1.25383651e+00
1.94766119e-01 -1.39201248e+00 -8.79776105e-02 1.24404646e-01
7.74827182e-01 -5.65668106e-01 8.26802775e-02 9.81402397e-02
4.66876894e-01 -8.77503097e-01 8.82856905e-01 -8.46604537e-03
1.52959192e+00 -8.74420643e-01 -1.05328463e-01 1.65567011e-01
-2.04898047e+00 -3.02247614e-01 -2.20898986e-01 8.39927793e-02
2.33028963e-01 6.56538904e-01 -1.19494259e-01 8.44930351e-01
9.97347653e-01 1.53232038e-01 -4.33592796e-02 7.76451290e-01
8.72445926e-02 6.01120770e-01 -3.04527193e-01 3.91646884e-02
2.25493267e-01 -8.71037900e-01 7.72934914e-01 1.27290010e+00
6.24257088e-01 9.11033154e-01 6.33831561e-01 4.84081864e-01
-3.75787586e-01 4.87140507e-01 -3.77709836e-01 -4.85981964e-02
3.88348728e-01 8.19222212e-01 -1.14767599e+00 -4.45319563e-01
-4.06165361e-01 8.32678974e-01 4.35055792e-02 5.36301374e-01
-8.49090815e-01 -1.11052692e+00 6.69677675e-01 6.06035590e-01
2.80360430e-01 -5.70043385e-01 -2.41956830e-01 -9.68375504e-01
6.82015046e-02 -9.76550281e-01 7.58170784e-01 -4.84224468e-01
-7.36758649e-01 3.05363178e-01 9.12720487e-02 -1.24819648e+00
7.07660615e-02 -2.95657694e-01 1.24143651e-02 8.19714546e-01
-7.80013382e-01 -6.07430398e-01 -2.82705903e-01 1.02448595e+00
1.08267508e-01 -3.25996548e-01 6.14528835e-01 3.57035518e-01
-3.40877652e-01 9.02793109e-01 1.60797417e-01 1.98348746e-01
2.59568006e-01 -8.34962368e-01 -1.82912752e-01 9.00880158e-01
1.42816111e-01 6.13495886e-01 6.13060176e-01 -6.44592226e-01
-1.90129435e+00 -8.45613182e-01 4.80132431e-01 2.16914952e-01
5.51783979e-01 -2.19086051e-01 -4.53752607e-01 8.77319872e-01
1.78933099e-01 -8.42690840e-02 7.46554554e-01 -4.26246494e-01
-3.41573387e-01 -4.96264488e-01 -1.22011077e+00 8.10909331e-01
1.34812689e+00 -5.32372773e-01 2.60689288e-01 3.86819839e-01
4.96947765e-01 -4.36989635e-01 -1.10313404e+00 4.10572678e-01
4.01837140e-01 -1.29575658e+00 8.78932238e-01 -1.47474989e-01
-7.20833242e-02 -3.60785484e-01 -5.96825838e-01 -6.36932671e-01
-5.27144074e-01 -1.44131625e+00 -8.03105161e-02 6.64753795e-01
4.00970489e-01 -3.00302237e-01 1.15054905e+00 3.87174249e-01
-1.81489944e-01 -5.70620298e-01 -1.18863189e+00 -5.95441759e-01
5.80455251e-02 -6.77046478e-01 1.02480389e-01 7.39880860e-01
2.91543752e-01 -6.92887418e-03 -2.66083717e-01 -4.80336361e-02
7.79083252e-01 6.63226619e-02 6.76710427e-01 -7.60936439e-01
-7.59131968e-01 -6.83651268e-01 -4.13488537e-01 -1.38752413e+00
-3.36867124e-01 -9.29466903e-01 -4.49104995e-01 -1.44302845e+00
3.06364387e-01 -7.54783809e-01 -2.02366069e-01 2.69055128e-01
7.47967288e-02 4.46472406e-01 3.94611984e-01 -1.42552540e-01
-6.81487739e-01 -2.74009258e-02 1.31053877e+00 -2.44752035e-01
-1.33885741e-01 -1.63142547e-01 -8.27589214e-01 5.74810386e-01
1.88122347e-01 -1.83522061e-01 -6.09166026e-01 -2.96413332e-01
7.97735035e-01 9.27567661e-01 8.45713317e-02 -1.01345980e+00
1.71276465e-01 -2.54294455e-01 -1.50678838e-02 -3.71134281e-01
3.80581141e-01 -5.66833436e-01 3.54219109e-01 2.78719395e-01
4.32702936e-02 3.30900937e-01 -2.03937888e-02 6.64145708e-01
6.50359765e-02 -9.04229581e-02 8.77248108e-01 -4.18973833e-01
-3.53371233e-01 5.16896665e-01 -3.53362232e-01 1.88832596e-01
1.17086852e+00 -6.94570541e-01 3.26450951e-02 -9.00872529e-01
-7.36104786e-01 2.06274077e-01 2.57919163e-01 -3.02296609e-01
4.71873522e-01 -1.14398825e+00 -6.97629452e-01 3.72496456e-01
-8.78310576e-02 9.25624594e-02 5.18493712e-01 1.09084833e+00
-7.05322742e-01 6.94160238e-02 1.85589299e-01 -5.59593797e-01
-1.09316933e+00 2.80596584e-01 1.91784844e-01 -2.95879960e-01
-3.77012640e-01 1.54908180e+00 6.05777800e-02 3.64793211e-01
3.38540465e-01 -1.80059671e-01 4.67537165e-01 5.87097816e-02
1.68272138e-01 8.31604421e-01 -2.45958362e-02 -7.90691376e-01
-1.46006912e-01 6.72575653e-01 1.95299506e-01 -2.34459922e-01
9.23743784e-01 -4.78381395e-01 -4.91327137e-01 -1.79263905e-01
1.31654227e+00 2.66897112e-01 -1.02025175e+00 -6.98402226e-02
-5.10778964e-01 -7.08530426e-01 -4.84415293e-01 -3.50491732e-01
-1.25832403e+00 6.06396079e-01 6.08643532e-01 6.05978489e-01
1.39154780e+00 -1.07991628e-01 6.93033874e-01 4.58750464e-02
9.62834597e-01 -1.47605336e+00 4.01325792e-01 4.77747053e-01
4.19837445e-01 -6.80135548e-01 2.94218808e-01 -5.87049067e-01
-2.96087980e-01 1.00006115e+00 8.91600028e-02 -3.31670344e-01
9.21433151e-01 -6.05940372e-02 -3.35404515e-01 -2.89806634e-01
-1.07704801e-02 -1.08210713e-01 1.84950903e-01 8.31851438e-02
5.32757699e-01 1.91725150e-01 -7.09476769e-01 7.44004667e-01
-3.87082487e-01 -3.22820127e-01 5.01550019e-01 9.78274524e-01
-6.37670994e-01 -1.18733895e+00 -3.16640556e-01 8.44722211e-01
-5.37060261e-01 -8.75261948e-02 1.82806298e-01 6.68733120e-01
3.55540127e-01 1.23691332e+00 -1.53047577e-01 -4.04754251e-01
3.93949360e-01 -7.21996874e-02 1.00044954e+00 -4.18323129e-01
-4.48141009e-01 6.50579095e-01 6.66948482e-02 -6.58113241e-01
-1.36565343e-01 -5.45800507e-01 -1.64876831e+00 -7.58926690e-01
-7.30562210e-03 3.16851169e-01 1.11191690e-01 5.13131380e-01
2.35024974e-01 2.82373220e-01 6.67369545e-01 -4.70405430e-01
-2.34579206e-01 -5.38462102e-01 -1.20918179e+00 5.88497102e-01
-1.81695804e-01 -3.34774882e-01 -2.16127038e-01 2.77572006e-01] | [6.5999627113342285, 4.765214920043945] |
308b2af1-e415-46ed-9fc8-cd455e9a29d1 | low-resource-named-entity-recognition-via | null | null | https://aclanthology.org/W18-6125 | https://aclanthology.org/W18-6125.pdf | Low-resource named entity recognition via multi-source projection: Not quite there yet? | Projecting linguistic annotations through word alignments is one of the most prevalent approaches to cross-lingual transfer learning. Conventional wisdom suggests that annotation projection {``}just works{''} regardless of the task at hand. We carefully consider multi-source projection for named entity recognition. Our experiment with 17 languages shows that to detect named entities in true low-resource languages, annotation projection may not be the right way to move forward. On a more positive note, we also uncover the conditions that do favor named entity projection from multiple sources. We argue these are infeasible under noisy low-resource constraints. | ["{\\v{Z}}eljko Agi{\\'c}", 'S{\\o}ren Harrison', 'Jan Vium Enghoff'] | 2018-11-01 | null | null | null | ws-2018-11 | ['low-resource-named-entity-recognition'] | ['natural-language-processing'] | [-5.72781935e-02 2.87304699e-01 -6.15067542e-01 -5.92188179e-01
-1.27116168e+00 -8.85458171e-01 4.58054662e-01 -2.36320153e-01
-8.33171725e-01 1.09022737e+00 6.10630810e-01 -7.56303549e-01
3.38528454e-01 -4.99831110e-01 -6.77425921e-01 -3.36808443e-01
3.71543407e-01 6.18779480e-01 8.48193914e-02 -2.86540926e-01
-2.59229019e-02 1.98453173e-01 -7.61825204e-01 5.84729373e-01
7.45558918e-01 2.00562373e-01 1.57329850e-02 1.21614322e-01
-7.60319352e-01 6.42685413e-01 -5.30858576e-01 -1.02364337e+00
4.40249681e-01 -4.00128126e-01 -1.11286831e+00 -2.99284250e-01
4.51296151e-01 1.25729172e-02 2.33357206e-01 1.24433899e+00
3.99170130e-01 -1.00450002e-01 4.74063903e-01 -8.40192497e-01
-9.98882771e-01 1.13141143e+00 -5.79243422e-01 2.62418121e-01
1.78256929e-01 -1.28448054e-01 1.42254436e+00 -1.31960976e+00
1.11097598e+00 1.04984426e+00 8.61854732e-01 6.06416762e-01
-1.25398016e+00 -7.76165247e-01 3.63204449e-01 -1.38749555e-01
-1.34651446e+00 -5.80318093e-01 4.80859637e-01 -5.09737790e-01
1.03201056e+00 1.08443588e-01 1.12924106e-01 1.28961647e+00
-2.88334429e-01 7.77429223e-01 1.36140668e+00 -7.61850417e-01
-2.14526311e-01 4.73188847e-01 3.08295012e-01 3.90040576e-01
7.14815676e-01 -2.57829100e-01 -7.91378558e-01 -3.87980133e-01
3.08595061e-01 -6.30157471e-01 -3.12392622e-01 -2.88890698e-03
-1.24199641e+00 8.46029341e-01 -5.47093004e-02 4.32316929e-01
-1.62681922e-01 4.01869491e-02 4.84968930e-01 2.62439191e-01
6.00333750e-01 7.19847381e-01 -1.03674221e+00 -1.45946056e-01
-6.31183743e-01 -1.77132219e-01 8.77721488e-01 1.21784604e+00
7.81443179e-01 1.29762337e-01 1.73222095e-01 8.78437221e-01
3.83458227e-01 6.22842848e-01 3.32952738e-01 -7.94232011e-01
9.19823289e-01 4.41622257e-01 3.54328692e-01 -4.17364508e-01
-2.00765446e-01 -1.56972244e-01 -1.88187838e-01 4.77061700e-03
9.25631285e-01 -6.48450613e-01 -5.82042038e-01 1.96327996e+00
2.42397457e-01 -2.98984945e-01 1.90303847e-01 7.68956721e-01
5.15398622e-01 4.08956289e-01 5.29653549e-01 -2.14075595e-01
1.54877794e+00 -6.33748710e-01 -7.03339517e-01 -6.42309606e-01
1.10091639e+00 -1.02894115e+00 1.36742783e+00 -4.46594395e-02
-8.09426248e-01 -2.09281981e-01 -7.14390993e-01 -2.07520008e-01
-4.93747294e-01 4.81891870e-01 7.89646506e-01 9.63122070e-01
-8.08801532e-01 5.33073246e-02 -7.71874785e-01 -4.92793590e-01
6.75182715e-02 1.12800553e-01 -6.50930524e-01 9.27345976e-02
-1.22192967e+00 1.34278631e+00 5.84477663e-01 3.43965590e-02
-2.79046863e-01 -5.48645079e-01 -6.96798503e-01 -1.06879964e-01
3.42757672e-01 -2.96480030e-01 1.22378194e+00 -1.12032723e+00
-1.02216971e+00 1.26692414e+00 -3.68206739e-01 -2.03569561e-01
3.80259126e-01 -6.30803108e-01 -5.23580611e-01 -4.16256309e-01
3.82477969e-01 4.66067851e-01 6.52515441e-02 -1.15201592e+00
-5.90020061e-01 -2.94270873e-01 -1.10033125e-01 3.58636886e-01
-6.76980019e-01 7.96076715e-01 -2.69662797e-01 -5.66907525e-01
-5.98832667e-02 -1.22411001e+00 -6.27148896e-02 -4.19508308e-01
-5.12852371e-01 -4.25292730e-01 2.74968237e-01 -8.33047926e-01
1.14671254e+00 -1.97088957e+00 -1.96918741e-01 6.90938085e-02
-2.76724756e-01 2.91603833e-01 1.48069993e-01 4.68378454e-01
-2.74280518e-01 6.17110968e-01 -1.82720304e-01 -2.53590763e-01
1.45122051e-01 2.44213864e-01 -6.00931704e-01 3.70679438e-01
3.13873231e-01 8.46840739e-01 -7.62104154e-01 -6.25346243e-01
-3.68089586e-01 4.41771418e-01 -4.60067421e-01 -3.32934618e-01
2.61536669e-02 4.13092434e-01 -4.46190715e-01 4.43049252e-01
3.14072251e-01 -1.72182158e-01 7.05772340e-01 6.98399842e-02
-3.90445352e-01 9.59042311e-01 -1.17045081e+00 1.58181059e+00
-3.50080073e-01 6.13227665e-01 -3.75803970e-02 -6.27386212e-01
8.64208698e-01 5.63790143e-01 2.27547333e-01 -3.08820337e-01
-1.97167769e-01 5.81989348e-01 2.05329910e-01 -5.38525641e-01
6.04754627e-01 -5.82262039e-01 -3.61384302e-01 6.81890547e-01
1.86970174e-01 2.60659307e-01 -2.31644530e-02 -7.85642862e-02
9.15870190e-01 4.96630490e-01 4.84648973e-01 -5.06729066e-01
3.43678236e-01 3.90184790e-01 1.14822340e+00 3.97292346e-01
-2.82417685e-01 3.09615016e-01 3.48349690e-01 -3.39823812e-01
-1.11056709e+00 -1.00839818e+00 -2.01487482e-01 1.71622169e+00
-4.17641401e-01 -3.11374307e-01 -4.73836780e-01 -1.03757548e+00
-2.69390583e-01 1.03279746e+00 -2.81287760e-01 4.33953971e-01
-9.84706163e-01 -9.73844945e-01 1.00469148e+00 6.57037556e-01
-4.92458120e-02 -1.02372003e+00 -1.91013396e-01 5.69159836e-02
-3.65871906e-01 -1.19854963e+00 -4.50608939e-01 4.61512476e-01
-5.53061068e-01 -6.39630675e-01 -6.86800241e-01 -8.98579240e-01
7.72811711e-01 3.59599777e-02 1.36503041e+00 -3.11281353e-01
4.20759141e-01 4.79202196e-02 -1.90730423e-01 -6.37003899e-01
-5.80294669e-01 3.96968633e-01 1.75219566e-01 -4.85871613e-01
1.06333911e+00 -3.62557918e-01 -8.39286968e-02 1.32706404e-01
-3.89321655e-01 -3.16626191e-01 6.03329301e-01 8.03812087e-01
6.20386779e-01 -5.79173326e-01 6.86344981e-01 -1.71125948e+00
4.36756730e-01 -5.42339504e-01 -4.71918970e-01 6.67092383e-01
-3.74727160e-01 1.12516485e-01 5.76778948e-01 -2.73545921e-01
-1.55365121e+00 2.46411562e-01 -3.14098656e-01 5.88395074e-02
-4.17532742e-01 5.66161752e-01 -5.03968954e-01 3.10935050e-01
8.93037021e-01 -1.05905339e-01 -7.16860473e-01 -6.77085519e-01
5.86040556e-01 5.30954242e-01 2.57998407e-01 -9.14394736e-01
6.09446764e-01 5.44749975e-01 -6.30187988e-01 -7.48572350e-01
-1.16672659e+00 -6.21486187e-01 -9.83686686e-01 2.28760809e-01
1.10225880e+00 -1.09912133e+00 -2.07487971e-01 -3.26353997e-01
-1.58058763e+00 -1.98170200e-01 -4.14473474e-01 7.52321482e-01
-2.31849223e-01 2.03757510e-01 -6.16272628e-01 -7.47818232e-01
-2.18528524e-01 -9.01601374e-01 7.57299602e-01 -5.20057455e-02
-5.47048509e-01 -1.16611600e+00 4.09145772e-01 3.05263549e-01
1.72047645e-01 -1.32621378e-01 8.11581850e-01 -1.41386902e+00
-1.41656265e-01 9.09969360e-02 -4.33860347e-02 2.66466767e-01
3.15327287e-01 -8.18193406e-02 -1.11152458e+00 -4.19605263e-02
1.59911662e-02 -2.35525072e-01 4.32141513e-01 3.24595235e-02
3.84957880e-01 -3.82068604e-01 -3.28922480e-01 4.43899482e-01
1.56194687e+00 -2.56380923e-02 3.58079642e-01 3.09198260e-01
7.72333682e-01 8.38578045e-01 3.61078322e-01 -1.06876105e-01
4.64952320e-01 5.01910269e-01 -4.79952991e-01 -1.54026479e-01
-1.00599691e-01 -5.09699047e-01 6.78987026e-01 1.27378786e+00
-9.50274840e-02 -8.21638703e-02 -1.53331816e+00 8.38989258e-01
-1.45457268e+00 -1.03111088e+00 -3.54345828e-01 2.17222762e+00
1.11476147e+00 2.58760452e-01 -5.68368239e-03 -5.72431147e-01
8.93158972e-01 -1.44258574e-01 -6.65056407e-02 -3.97102475e-01
-4.31471825e-01 2.57520348e-01 6.93879366e-01 4.01034594e-01
-8.62996519e-01 1.27337813e+00 7.04570961e+00 4.47793484e-01
-1.02316642e+00 6.87857330e-01 2.59843260e-01 3.15572321e-01
-6.43118799e-01 3.28155488e-01 -1.37591028e+00 3.54973733e-01
1.05542147e+00 -2.25224584e-01 -1.56768128e-01 1.01139641e+00
-1.62108094e-01 1.77102402e-01 -1.05684733e+00 4.30036038e-01
1.49244949e-01 -1.08186615e+00 -8.98151696e-02 2.45230291e-02
8.77676070e-01 4.99850810e-01 -1.36680096e-01 3.60061973e-01
9.44187284e-01 -6.68117940e-01 6.09290659e-01 8.46863762e-02
8.98221374e-01 -5.02103269e-01 7.24854946e-01 4.34916705e-01
-9.45980132e-01 3.93941700e-01 -4.37918216e-01 1.78873122e-01
5.66299200e-01 3.76636654e-01 -9.60257411e-01 3.46879452e-01
5.09602308e-01 8.78795534e-02 -5.05635977e-01 7.18336463e-01
-7.18700707e-01 1.04322016e+00 -4.68454123e-01 9.33166668e-02
9.87525061e-02 -1.03387542e-01 4.35803622e-01 1.86275530e+00
3.39619577e-01 7.21180812e-02 2.05293164e-01 5.42067230e-01
-2.68198490e-01 5.61136603e-01 -1.03038776e+00 -1.81240708e-01
5.78988552e-01 1.00441086e+00 -8.07655215e-01 -3.15895766e-01
-9.92349565e-01 7.36565292e-01 7.39133239e-01 3.85319382e-01
-5.96464634e-01 -1.38075680e-01 5.39873421e-01 -3.56242135e-02
2.23361418e-01 -5.14762521e-01 -6.67363703e-01 -1.47285223e+00
1.04507387e-01 -7.49296904e-01 6.44867599e-01 -4.99744415e-01
-1.58893883e+00 4.60114539e-01 -9.06942487e-02 -1.01176572e+00
-6.11843616e-02 -8.29512537e-01 -3.34429353e-01 1.05893219e+00
-1.40728247e+00 -1.29954827e+00 4.56274062e-01 4.42311317e-01
3.82570773e-01 -3.33790407e-02 1.20351565e+00 5.29165864e-01
-4.25175607e-01 7.03104436e-01 -6.30666018e-02 6.24052048e-01
1.14071476e+00 -1.34500325e+00 4.97536540e-01 1.26196790e+00
8.24621141e-01 1.03594398e+00 7.14327872e-01 -7.61018634e-01
-1.09561408e+00 -8.01048160e-01 1.58838081e+00 -9.22143340e-01
1.01250637e+00 -4.31670606e-01 -1.17533052e+00 1.33352077e+00
5.10353506e-01 -2.09581196e-01 1.31512415e+00 7.44574904e-01
-6.67908370e-01 3.21475863e-01 -5.94640911e-01 5.38662493e-01
9.71037745e-01 -9.18729901e-01 -1.00224733e+00 4.87785399e-01
7.58686543e-01 -1.30555227e-01 -8.34830940e-01 1.53442830e-01
2.91796982e-01 -4.74624753e-01 6.28181100e-01 -1.07327867e+00
1.62096947e-01 -2.90121555e-01 -3.86979103e-01 -1.22820354e+00
-2.04194877e-02 -5.25559068e-01 4.92476225e-01 1.78858030e+00
1.13423467e+00 -7.09834576e-01 6.95256412e-01 8.26716542e-01
-1.81266904e-01 -5.88460751e-02 -8.75152826e-01 -9.93482351e-01
6.76128685e-01 -7.37604320e-01 3.40319246e-01 1.59200907e+00
3.98079872e-01 7.10012853e-01 -1.91641599e-01 2.06184417e-01
2.88959712e-01 -6.25992566e-02 5.73686838e-01 -9.10123467e-01
-3.85555327e-01 -1.77831322e-01 1.07297875e-01 -8.41245413e-01
5.97126424e-01 -1.17912877e+00 3.83103848e-01 -1.38100731e+00
1.80613667e-01 -1.04701817e+00 -4.00434554e-01 7.95729458e-01
-3.37452352e-01 2.78592199e-01 1.54060483e-01 4.22132462e-01
-3.36779863e-01 2.61285063e-03 6.28028154e-01 2.71092981e-01
-7.22054243e-02 -1.56361386e-01 -9.11459148e-01 1.02866435e+00
9.43046629e-01 -7.14536846e-01 -8.51130695e-04 -1.00239193e+00
7.38025248e-01 -2.60576993e-01 -3.11491847e-01 -3.17824155e-01
2.40956828e-01 -4.53937083e-01 1.00120485e-01 -1.26750439e-01
1.66360527e-01 -5.22703052e-01 6.96565062e-02 1.45584047e-01
-4.75339293e-01 4.21178401e-01 1.54543012e-01 2.79622614e-01
-2.32849792e-01 -4.20552373e-01 5.16397834e-01 -3.47913146e-01
-8.07035148e-01 -1.15071952e-01 -2.69559413e-01 5.52812636e-01
7.15119779e-01 2.09517971e-01 -4.20729071e-01 7.41480738e-02
-7.11817741e-01 -1.25983253e-01 5.30762970e-01 2.98461705e-01
-1.06237777e-01 -1.29225099e+00 -9.25679982e-01 -2.31269032e-01
2.56342798e-01 -4.43177164e-01 -3.29696596e-01 7.87566721e-01
-2.45261386e-01 5.91160178e-01 -1.17703408e-01 -1.55117199e-01
-1.13861871e+00 3.98781180e-01 4.62403260e-02 -3.46685350e-01
-5.17701030e-01 1.16765523e+00 2.06581175e-01 -8.57034564e-01
-2.06933200e-01 7.96348229e-02 -7.02921674e-02 2.04982653e-01
3.55940998e-01 1.56574816e-01 1.23213723e-01 -9.48517084e-01
-5.46379507e-01 1.82188615e-01 -5.69936186e-02 -6.01251781e-01
1.27860177e+00 -2.00421512e-01 8.70020944e-04 9.20158207e-01
8.87482822e-01 8.19234133e-01 -6.51010156e-01 -4.50353414e-01
8.08713555e-01 -2.78480947e-01 -4.77128327e-01 -6.92633510e-01
-5.33500373e-01 9.53082979e-01 2.44150832e-02 1.10247424e-02
5.56082964e-01 4.51215766e-02 5.04388571e-01 6.24344945e-01
5.69710612e-01 -1.38306916e+00 -5.86228311e-01 6.48043931e-01
6.23714864e-01 -1.31479287e+00 6.47682250e-02 -4.70625103e-01
-9.69315469e-01 7.11714387e-01 8.44113052e-01 9.05822590e-02
6.39041364e-01 4.24830914e-01 3.30166429e-01 -1.15757637e-01
-6.98053956e-01 -3.24884892e-01 8.67915824e-02 4.49032396e-01
1.18399012e+00 2.66798168e-01 -7.05431938e-01 5.58184385e-01
-4.11913782e-01 -3.64555508e-01 4.36062396e-01 8.19221556e-01
-2.93455452e-01 -1.45264721e+00 -3.77468616e-01 1.69720724e-01
-1.06149793e+00 -5.51736176e-01 -5.36765814e-01 9.21080053e-01
4.09267336e-01 6.01875246e-01 -2.69640177e-01 1.48314565e-01
2.62794197e-01 7.69218326e-01 2.14721784e-01 -1.12051666e+00
-6.82634473e-01 5.70079591e-03 5.27868688e-01 -1.83040842e-01
-5.33813298e-01 -8.85937452e-01 -1.24763024e+00 -9.49189216e-02
-5.44013023e-01 2.71410078e-01 6.26464486e-01 9.81029332e-01
2.10836709e-01 -1.19428635e-01 5.84217794e-02 -1.84443831e-01
-4.97974247e-01 -8.51361513e-01 -1.61726549e-01 3.80575955e-01
-3.18824202e-01 -2.95099616e-01 -2.83236295e-01 4.02714938e-01] | [10.339458465576172, 9.852641105651855] |
0ea5394a-a891-4a37-a5af-281144469afc | hierarchical-conditional-flow-a-unified | 2108.05301 | null | https://arxiv.org/abs/2108.05301v1 | https://arxiv.org/pdf/2108.05301v1.pdf | Hierarchical Conditional Flow: A Unified Framework for Image Super-Resolution and Image Rescaling | Normalizing flows have recently demonstrated promising results for low-level vision tasks. For image super-resolution (SR), it learns to predict diverse photo-realistic high-resolution (HR) images from the low-resolution (LR) image rather than learning a deterministic mapping. For image rescaling, it achieves high accuracy by jointly modelling the downscaling and upscaling processes. While existing approaches employ specialized techniques for these two tasks, we set out to unify them in a single formulation. In this paper, we propose the hierarchical conditional flow (HCFlow) as a unified framework for image SR and image rescaling. More specifically, HCFlow learns a bijective mapping between HR and LR image pairs by modelling the distribution of the LR image and the rest high-frequency component simultaneously. In particular, the high-frequency component is conditional on the LR image in a hierarchical manner. To further enhance the performance, other losses such as perceptual loss and GAN loss are combined with the commonly used negative log-likelihood loss in training. Extensive experiments on general image SR, face image SR and image rescaling have demonstrated that the proposed HCFlow achieves state-of-the-art performance in terms of both quantitative metrics and visual quality. | ['Radu Timofte', 'Luc van Gool', 'Martin Danelljan', 'Kai Zhang', 'Andreas Lugmayr', 'Jingyun Liang'] | 2021-08-11 | null | http://openaccess.thecvf.com//content/ICCV2021/html/Liang_Hierarchical_Conditional_Flow_A_Unified_Framework_for_Image_Super-Resolution_and_ICCV_2021_paper.html | http://openaccess.thecvf.com//content/ICCV2021/papers/Liang_Hierarchical_Conditional_Flow_A_Unified_Framework_for_Image_Super-Resolution_and_ICCV_2021_paper.pdf | iccv-2021-1 | ['video-super-resolution'] | ['computer-vision'] | [ 7.11965144e-01 4.45962325e-02 -1.51562408e-01 -6.17878258e-01
-9.71337080e-01 4.93923128e-02 4.96236116e-01 -5.37505984e-01
-2.55678862e-01 7.39389002e-01 3.30872834e-01 1.61053568e-01
2.64278408e-02 -7.21915901e-01 -8.24199855e-01 -8.04194152e-01
3.75327826e-01 -4.68298681e-02 1.12368064e-02 -1.13026798e-01
-1.68312024e-02 5.76436102e-01 -1.69506955e+00 2.12972850e-01
9.23264265e-01 9.04866934e-01 3.97784144e-01 6.55768335e-01
7.63437822e-02 1.05438018e+00 -3.23094904e-01 -5.26895285e-01
4.26845402e-01 -6.37162745e-01 -6.14820659e-01 1.61965281e-01
9.68781888e-01 -5.64596951e-01 -6.11537933e-01 1.15086663e+00
6.23588681e-01 2.20441446e-01 3.55359793e-01 -1.03874898e+00
-1.17058682e+00 4.34109420e-01 -1.02532518e+00 2.32251123e-01
2.39942983e-01 1.75206065e-01 8.08324277e-01 -8.98033023e-01
5.47527850e-01 1.51207685e+00 5.19210815e-01 7.31506050e-01
-1.56032443e+00 -7.35065341e-01 8.57785270e-02 6.03009723e-02
-1.29758477e+00 -5.10583460e-01 7.82473683e-01 -2.20204636e-01
5.19407332e-01 9.53109711e-02 1.27123773e-01 1.02851582e+00
1.18388198e-01 4.36563522e-01 1.41393924e+00 -2.67008573e-01
-1.07033983e-01 9.50344354e-02 -3.21919978e-01 4.65954036e-01
-7.71134943e-02 3.71354550e-01 -6.36017084e-01 1.86023101e-01
1.23553038e+00 8.05576965e-02 -4.95869905e-01 -1.96331382e-01
-9.15646374e-01 7.06827283e-01 7.55515337e-01 2.22459331e-01
-3.47986192e-01 1.86647043e-01 3.22286156e-04 4.59295101e-02
5.21240473e-01 3.48956995e-02 -1.39544666e-01 3.43429744e-01
-9.46034193e-01 1.31286979e-01 1.45606443e-01 8.20804179e-01
8.69580448e-01 1.12843946e-01 -5.77661514e-01 1.24874902e+00
2.97815725e-02 2.93110251e-01 2.41546512e-01 -1.20025730e+00
3.82700086e-01 8.39258954e-02 1.08664744e-01 -7.94868529e-01
-2.97823418e-02 -6.81497872e-01 -1.32824385e+00 3.45138609e-01
6.27625780e-03 3.05342644e-01 -1.17786705e+00 2.07769966e+00
2.19336256e-01 4.46991652e-01 3.62031385e-02 1.09705746e+00
9.10729170e-01 7.06986845e-01 2.51966506e-01 -4.43457007e-01
1.07397962e+00 -1.03043580e+00 -9.21574712e-01 -1.30273566e-01
-2.97866791e-01 -7.36060083e-01 1.31166220e+00 1.68044671e-01
-1.49179959e+00 -1.03202426e+00 -1.01226485e+00 -6.62227392e-01
1.52676970e-01 2.32762426e-01 3.27538759e-01 3.70358318e-01
-1.33132088e+00 6.38557553e-01 -5.39945543e-01 1.91702042e-02
6.23197675e-01 -1.36108603e-02 -3.52558553e-01 -5.32719374e-01
-1.19945550e+00 7.83382416e-01 -6.72709271e-02 1.17925733e-01
-8.71514797e-01 -1.09169650e+00 -9.79406416e-01 1.07813358e-01
9.35293585e-02 -9.23479021e-01 9.51886892e-01 -8.32636535e-01
-1.64760196e+00 9.31659579e-01 -3.10778081e-01 -4.32638884e-01
6.16067827e-01 -1.93422750e-01 -5.09089649e-01 2.76235640e-01
1.99961022e-01 7.91918457e-01 1.24828541e+00 -1.59240401e+00
-5.94680250e-01 -4.02420461e-01 5.93509115e-02 4.17449027e-01
-1.80363938e-01 4.61520329e-02 -4.97223765e-01 -9.30213094e-01
-8.22773203e-02 -5.30763090e-01 -7.41694719e-02 1.37822852e-01
-2.03125119e-01 3.87119949e-02 7.30714500e-01 -9.03725803e-01
1.02904034e+00 -2.23594642e+00 3.23973894e-01 -3.43740463e-01
2.48845771e-01 2.30942428e-01 -2.38846555e-01 -1.05562493e-01
-1.23101927e-01 -1.63991768e-02 -5.00659585e-01 -5.89540839e-01
-2.78011501e-01 2.08560243e-01 -4.71466690e-01 4.47591841e-01
4.55832034e-01 8.40691328e-01 -7.89456427e-01 -5.16155720e-01
4.44192290e-01 1.20760131e+00 -3.77064586e-01 3.06785375e-01
1.79518387e-01 7.98401535e-01 1.41072394e-02 3.95587325e-01
1.03202546e+00 -3.02529871e-01 -1.46377772e-01 -6.80759370e-01
9.15340260e-02 -2.30981950e-02 -9.64342535e-01 1.57194138e+00
-8.05964887e-01 4.20033723e-01 1.29234821e-01 -5.13081312e-01
9.70797956e-01 -2.37096753e-03 4.22461361e-01 -9.52224374e-01
-2.87897289e-01 -1.78062424e-01 -3.83966863e-01 -1.84765950e-01
4.71539468e-01 -4.68966812e-01 2.23391280e-01 2.11632669e-01
1.42925575e-01 -1.17547989e-01 -1.11937568e-01 1.02553658e-01
6.17667317e-01 2.51834244e-01 1.83642015e-01 1.34125307e-01
8.71071875e-01 -6.84634924e-01 5.77259302e-01 5.82761109e-01
-8.42140391e-02 1.05533350e+00 2.52417237e-01 -1.58822834e-01
-1.19160712e+00 -1.42824590e+00 -3.62695903e-01 1.01449287e+00
3.53262335e-01 -1.54546648e-01 -7.09307611e-01 -4.83781695e-01
-2.03972787e-01 5.89320421e-01 -5.62679291e-01 -1.66626200e-01
-6.18954182e-01 -6.79799378e-01 2.77499586e-01 4.44675028e-01
9.34270740e-01 -9.74261045e-01 -2.10968077e-01 -9.26875323e-02
-3.31200749e-01 -1.50399542e+00 -7.07251012e-01 -2.38053694e-01
-6.01907313e-01 -7.58576155e-01 -8.38403404e-01 -6.93361700e-01
5.81987798e-01 3.81149590e-01 1.12358165e+00 -1.52690127e-01
-5.39869368e-01 2.04979032e-01 -1.36772990e-01 1.77711308e-01
-3.22009832e-01 -3.64993066e-01 -1.64182946e-01 4.25450116e-01
-2.59713441e-01 -5.91883957e-01 -7.75926650e-01 2.50287920e-01
-1.13955724e+00 1.29865304e-01 7.81434834e-01 1.08503449e+00
9.30930078e-01 3.80999178e-01 5.23962498e-01 -8.72011840e-01
3.01646084e-01 -1.31090343e-01 -4.46060747e-01 2.16781467e-01
-6.84211254e-01 -5.66886291e-02 6.92514837e-01 -4.42045569e-01
-1.60324144e+00 3.51037271e-02 -2.63889045e-01 -6.32061541e-01
-2.79024504e-02 -1.90345094e-01 -3.12301636e-01 -3.03134501e-01
4.21712995e-01 4.15202737e-01 -1.21084087e-01 -5.32282472e-01
6.34157240e-01 5.47901273e-01 1.02143931e+00 -3.06412309e-01
1.05147994e+00 6.52089596e-01 1.91724956e-01 -7.99795687e-01
-1.19380105e+00 -1.78291470e-01 -4.42274153e-01 -1.75386071e-01
9.79249477e-01 -1.21534169e+00 -4.63667423e-01 5.15432179e-01
-7.93424845e-01 -2.71418422e-01 -5.12410641e-01 3.55191618e-01
-7.26865590e-01 3.12624484e-01 -7.67105579e-01 -6.91487193e-01
-2.60999829e-01 -1.20527065e+00 1.31028509e+00 5.39564669e-01
4.08270001e-01 -6.94334805e-01 -2.45949626e-01 5.78541398e-01
7.27117240e-01 2.80726701e-01 7.21299052e-01 3.53938431e-01
-6.50343835e-01 1.77076086e-01 -7.74936199e-01 8.23883533e-01
2.01421708e-01 -3.41626048e-01 -1.17198384e+00 -4.69716012e-01
9.59722698e-02 -3.72544736e-01 1.05196083e+00 5.34428060e-01
1.54752827e+00 -1.84364319e-01 1.49947077e-01 1.07836056e+00
1.69153464e+00 -1.76547974e-01 1.00931215e+00 -7.65249953e-02
9.79345977e-01 6.34489000e-01 7.04606950e-01 2.37258106e-01
3.37240845e-01 9.53210771e-01 3.59473616e-01 -4.41058487e-01
-8.17122877e-01 -4.14748520e-01 3.71197253e-01 2.48364449e-01
-4.82299961e-02 7.81750903e-02 -2.55140543e-01 1.40973091e-01
-1.45104516e+00 -1.04075015e+00 1.74865469e-01 2.30479527e+00
1.03245318e+00 -1.75904810e-01 -1.52930811e-01 -5.18816411e-02
9.17508543e-01 4.88711059e-01 -7.97160685e-01 -2.63179075e-02
-5.54013014e-01 3.33556622e-01 4.64740723e-01 8.16288114e-01
-9.28995967e-01 1.01750886e+00 5.97310972e+00 9.75059271e-01
-1.14919984e+00 3.30650151e-01 1.09091878e+00 -1.23101346e-01
-3.03386748e-01 -2.30404496e-01 -9.31610584e-01 4.90709186e-01
6.67663515e-01 -3.85946892e-02 7.05688000e-01 4.33970600e-01
1.10384375e-01 3.35975140e-02 -7.89484441e-01 1.24947774e+00
2.46318474e-01 -1.17311585e+00 3.33800763e-01 -1.82449948e-02
8.98372889e-01 -1.61608368e-01 5.54244578e-01 1.54655382e-01
1.41766816e-01 -1.29686987e+00 5.59021175e-01 6.44289196e-01
1.41617072e+00 -8.06215942e-01 5.36479950e-01 -7.83058256e-02
-1.26970220e+00 -2.21128851e-01 -3.77545118e-01 3.38889152e-01
2.47929186e-01 5.52558541e-01 -7.55285770e-02 7.40499377e-01
1.02134335e+00 8.87630701e-01 -5.93841910e-01 5.47229111e-01
-4.07087266e-01 1.12284951e-01 1.24763340e-01 9.94499922e-01
-2.02685103e-01 -3.11635017e-01 4.57301795e-01 9.07289803e-01
7.36984760e-02 1.24792069e-01 -2.21471675e-02 1.10636556e+00
-3.47461671e-01 -1.86382458e-01 -3.00882488e-01 2.60118216e-01
3.46229821e-01 1.44383240e+00 -2.67387956e-01 -1.64511010e-01
-4.00839895e-01 1.18214118e+00 3.34438622e-01 6.06578827e-01
-8.64696622e-01 -2.04606578e-01 7.34446347e-01 2.55994707e-01
3.73960495e-01 5.71811348e-02 -3.55024159e-01 -1.18710327e+00
1.95932418e-01 -6.42639756e-01 3.23406786e-01 -9.86124396e-01
-1.42657304e+00 8.35179210e-01 -3.74801941e-02 -1.13300860e+00
-1.42536387e-01 -2.06757203e-01 -3.46252382e-01 1.15798676e+00
-2.13729668e+00 -1.31647551e+00 -6.41543865e-01 8.21773529e-01
5.37764251e-01 7.59917200e-02 3.35147917e-01 4.71073776e-01
-5.77341914e-01 7.58753419e-01 -9.81865376e-02 -1.29060075e-01
9.51855898e-01 -1.15254819e+00 1.88383505e-01 9.80770111e-01
-9.65477079e-02 3.73417616e-01 7.36476719e-01 -4.12253529e-01
-1.03805864e+00 -1.41920960e+00 6.27103925e-01 -1.23262390e-01
1.54750362e-01 -2.13142186e-01 -1.07854569e+00 4.78540927e-01
-1.67893674e-02 4.90356386e-01 2.36613765e-01 -5.43904722e-01
-5.86401880e-01 -4.93271410e-01 -1.47750187e+00 4.79728132e-01
1.14824653e+00 -7.14887798e-01 -2.05466315e-01 1.00573100e-01
9.08586621e-01 -3.92213970e-01 -1.08074677e+00 6.28615499e-01
3.86692107e-01 -1.22869134e+00 1.42478502e+00 -1.39689550e-01
7.26963103e-01 -5.49285293e-01 -2.56087631e-01 -1.07268226e+00
-3.82705331e-01 -5.28938591e-01 -2.17218891e-01 1.46428800e+00
-1.13742076e-01 -5.39680481e-01 4.68342185e-01 3.20017099e-01
1.15797184e-01 -7.66395628e-01 -6.48611009e-01 -6.36394322e-01
-3.88090648e-02 5.16679287e-02 5.43137789e-01 7.22945690e-01
-8.50079060e-01 4.60452855e-01 -8.45098197e-01 3.66487920e-01
1.25230086e+00 1.68168768e-01 6.44387364e-01 -8.76076639e-01
-3.80895019e-01 -3.17032695e-01 -2.10012615e-01 -9.54705536e-01
2.70923018e-01 -7.11538732e-01 6.83424622e-02 -1.48705173e+00
4.35848057e-01 -1.51952893e-01 -2.95290679e-01 1.29612446e-01
-3.15378308e-01 7.13435113e-01 2.16446742e-01 2.41630033e-01
-3.71841490e-01 7.69993544e-01 1.54883850e+00 3.29770595e-02
-1.66021556e-01 -1.71845153e-01 -8.26257527e-01 4.56343263e-01
5.91749609e-01 -1.65801853e-01 -4.90828186e-01 -3.44643176e-01
-1.81122318e-01 2.88230658e-01 6.26663685e-01 -8.85263383e-01
1.30512014e-01 -7.91190788e-02 7.43612826e-01 -4.14398134e-01
5.47292888e-01 -5.68623781e-01 1.99707285e-01 1.40831634e-01
-6.73250973e-01 -3.38385493e-01 -1.22979164e-01 8.08649600e-01
-4.89335030e-01 3.59423816e-01 1.57065129e+00 5.23727667e-03
-4.86236930e-01 7.53418565e-01 3.55000764e-01 -3.94708961e-02
8.10729504e-01 -9.97264162e-02 -4.27058816e-01 -5.51180422e-01
-6.33477867e-01 -2.15429235e-02 5.86316407e-01 5.48632741e-01
9.76812661e-01 -1.29217637e+00 -9.49711978e-01 4.30753589e-01
5.74360676e-02 4.40507755e-02 8.23859632e-01 7.28295624e-01
-1.56382158e-01 6.35659471e-02 -5.24950027e-01 -4.60913271e-01
-1.12871468e+00 5.42752922e-01 4.78460938e-01 -2.93552220e-01
-8.54708493e-01 8.06217909e-01 7.84014940e-01 -1.59701720e-01
3.35724682e-01 9.70334113e-02 -3.41508925e-01 -3.28531802e-01
8.82299483e-01 3.83703589e-01 -2.13300675e-01 -1.00373864e+00
-3.51001956e-02 8.62304389e-01 -2.50404924e-01 -3.53300981e-02
1.37382114e+00 -6.79934323e-01 -2.06171215e-01 2.78894961e-01
1.30535889e+00 -1.68676198e-01 -1.71489584e+00 -6.14340663e-01
-3.85325462e-01 -9.25321877e-01 4.35145915e-01 -7.16468394e-01
-1.36963511e+00 8.04282725e-01 7.65519261e-01 -9.15667713e-02
1.74892604e+00 8.78442004e-02 8.20532739e-01 -4.17468667e-01
2.30102852e-01 -7.55237401e-01 2.46670499e-01 1.06653750e-01
1.14576304e+00 -1.34692395e+00 1.25208929e-01 -6.16039217e-01
-7.47379780e-01 9.36181903e-01 7.03242779e-01 -2.72659540e-01
3.48043799e-01 2.62337804e-01 -1.26302421e-01 2.13864788e-01
-5.39617121e-01 -2.89615273e-01 3.54518116e-01 7.51929522e-01
3.57903391e-01 -1.24815755e-01 -1.09180145e-01 2.40426064e-01
-8.11203346e-02 1.36689261e-01 5.17391682e-01 3.64044845e-01
-1.24801397e-01 -9.62761402e-01 -1.74286693e-01 3.44846994e-01
-6.35670781e-01 -2.13245973e-01 1.26419589e-02 4.73236710e-01
-1.43689699e-02 8.40403020e-01 9.57471356e-02 -2.73504645e-01
3.12912971e-01 -4.89541054e-01 7.48993158e-01 -4.14950937e-01
-2.03741997e-01 4.54015546e-02 -4.15927708e-01 -8.99687469e-01
-5.71298540e-01 -4.41792876e-01 -8.91292453e-01 -2.30668932e-01
8.50014240e-02 -3.20579529e-01 5.09437680e-01 4.89910096e-01
3.35106611e-01 5.94507992e-01 9.30393398e-01 -1.06690800e+00
-6.44889891e-01 -9.13954139e-01 -8.12317610e-01 5.86558223e-01
7.69552052e-01 -5.96304715e-01 -5.27852476e-01 2.19151676e-01] | [11.081486701965332, -2.0184104442596436] |
6d849a17-55fd-41a4-b4df-525a8dd5df0c | s-convnet-a-shallow-convolutional-neural | 1906.03381 | null | https://arxiv.org/abs/1906.03381v1 | https://arxiv.org/pdf/1906.03381v1.pdf | S-ConvNet: A Shallow Convolutional Neural Network Architecture for Neuromuscular Activity Recognition Using Instantaneous High-Density Surface EMG Images | The concept of neuromuscular activity recognition using instantaneous high-density surface electromyography (HD-sEMG) images opens up new avenues for the development of more fluid and natural muscle-computer interfaces. However, the existing approaches employed a very large deep convolutional neural network (ConvNet) architecture and complex training schemes for HD-sEMG image recognition, which requires the network architecture to be pre-trained on a very large-scale labeled training dataset, as a result, it makes computationally very expensive. To overcome this problem, we propose S-ConvNet and All-ConvNet models, a simple yet efficient framework for learning instantaneous HD-sEMG images from scratch for neuromuscular activity recognition. Without using any pre-trained models, our proposed S-ConvNet and All-ConvNet demonstrate very competitive recognition accuracy to the more complex state of the art for neuromuscular activity recognition based on instantaneous HD-sEMG images, while using a ~ 12 x smaller dataset and reducing learning parameters to a large extent. The experimental results proved that the S-ConvNet and All-ConvNet are highly effective for learning discriminative features for instantaneous HD-sEMG image recognition especially in the data and high-end resource constrained scenarios. | ['Wei-Ping Zhu', 'Philippe Massicotte', 'Francois Nougarou', 'Daniel Massicotte', 'Md. Rabiul Islam'] | 2019-06-08 | null | null | null | null | ['gesture-recognition', 'muscle-computer-interfaces-mcis', 'low-latency-processing'] | ['computer-vision', 'robots', 'robots'] | [ 2.35561237e-01 -2.68503010e-01 -1.08912900e-01 -8.80890042e-02
-3.86713505e-01 1.57881156e-01 8.70057270e-02 -8.09693038e-01
-8.14504921e-01 8.09780896e-01 -1.26401354e-02 2.17410743e-01
-2.21129581e-01 -5.47772527e-01 -7.66738594e-01 -8.94200742e-01
-2.24762589e-01 3.25704724e-01 2.44346246e-01 -8.59662443e-02
7.87311122e-02 3.41194004e-01 -1.88655770e+00 3.65316160e-02
6.17213070e-01 1.20133471e+00 7.61719525e-01 6.09517395e-01
-5.83769567e-03 6.77024543e-01 -7.15754569e-01 1.67382255e-01
3.80604863e-01 -3.55964571e-01 -5.47355235e-01 -3.04721203e-02
4.73866194e-01 -4.91423249e-01 -7.46445894e-01 8.52311015e-01
1.00164747e+00 1.32383093e-01 4.23575073e-01 -8.15237164e-01
-3.57849687e-01 1.57777324e-01 -2.18276635e-01 5.88607252e-01
2.23801821e-01 5.68965733e-01 4.49851155e-01 -7.66061783e-01
6.97747350e-01 8.18691552e-01 9.03122783e-01 9.34272051e-01
-7.66583502e-01 -7.93434322e-01 -2.10533679e-01 3.68674755e-01
-1.18053854e+00 1.59450639e-02 7.52420306e-01 -2.38007948e-01
1.28539586e+00 2.62106955e-02 1.24630320e+00 1.46070075e+00
5.30949831e-01 9.90235031e-01 1.09460700e+00 -5.43523803e-02
2.72123218e-02 -6.11391723e-01 1.10195227e-01 5.41125536e-01
1.31226540e-01 2.06635863e-01 -7.57634282e-01 3.67627442e-01
1.28454888e+00 2.75370151e-01 -4.24851984e-01 -5.16388472e-03
-1.10400689e+00 2.74237096e-01 4.07472461e-01 5.77043176e-01
-8.08584094e-01 2.93503672e-01 7.99847007e-01 6.53326452e-01
3.43389422e-01 1.68420732e-01 -4.23077881e-01 -8.73192668e-01
-5.53256571e-01 2.14968175e-01 4.89377260e-01 5.58484435e-01
4.23604339e-01 4.00219202e-01 2.13749036e-01 1.07491136e+00
-2.21323684e-01 5.17383039e-01 8.56086493e-01 -6.59130931e-01
4.99127924e-01 7.24027693e-01 -5.09167373e-01 -7.62454927e-01
-5.44357061e-01 -4.45441037e-01 -1.06172037e+00 4.42583948e-01
3.44939053e-01 -4.15205479e-01 -1.21388531e+00 1.50065696e+00
1.26389101e-01 2.51677960e-01 -1.12266138e-01 1.42706037e+00
6.55982018e-01 4.66941416e-01 4.00088876e-02 -5.39039671e-02
8.77521753e-01 -7.53911495e-01 -7.64360487e-01 -8.73512626e-02
7.35813260e-01 -2.94052541e-01 1.19970441e+00 4.80377376e-01
-8.57219458e-01 -9.39667642e-01 -1.19078970e+00 1.10217072e-01
-3.12429428e-01 3.41560543e-01 8.27524662e-01 3.41818094e-01
-8.05771291e-01 9.95432377e-01 -1.24633920e+00 -2.34689966e-01
4.46355700e-01 9.96975362e-01 -7.59503007e-01 5.04149757e-02
-1.22648752e+00 8.13503325e-01 2.51352042e-01 4.95241076e-01
-9.57486391e-01 -4.96035278e-01 -6.64210856e-01 -1.32194325e-01
4.06849295e-01 -5.24896920e-01 8.13072920e-01 -8.78300488e-01
-1.71042240e+00 7.49233544e-01 5.01015842e-01 -3.94018173e-01
3.63822460e-01 -6.17739141e-01 -4.62540925e-01 2.10852489e-01
-2.01379254e-01 2.98557520e-01 7.32553482e-01 -3.70953679e-01
-2.89045304e-01 -5.57166934e-01 -1.06823847e-01 3.50852132e-01
-4.75007981e-01 -1.13582328e-01 -3.13646972e-01 -6.50898099e-01
-7.74667785e-02 -8.22630346e-01 -1.52132176e-02 1.30508900e-01
-1.55812323e-01 -6.17169142e-01 1.10869837e+00 -6.82105720e-01
9.46067929e-01 -2.05630255e+00 4.95743424e-01 -5.41607663e-02
3.62316698e-01 9.78326440e-01 -1.85880125e-01 3.22811872e-01
-5.08841984e-02 -3.82852882e-01 4.57357951e-02 1.01940423e-01
-3.39717299e-01 3.80070299e-01 4.62342829e-01 4.15653944e-01
7.36493096e-02 8.78242731e-01 -6.83115184e-01 -4.39871699e-01
7.38679290e-01 6.15370333e-01 -2.89410830e-01 6.25196397e-01
6.09288886e-02 3.25266898e-01 -4.18917000e-01 6.97224915e-01
3.38098943e-01 -8.57917592e-02 1.61402777e-01 -2.58566052e-01
8.43097046e-02 -4.95188013e-02 -1.09085584e+00 2.31340051e+00
-4.81872797e-01 6.43148363e-01 1.73498750e-01 -1.48992991e+00
8.83388638e-01 4.15126979e-01 9.85045552e-01 -9.07057166e-01
3.67728770e-01 4.28233743e-01 1.30071849e-01 -9.73656416e-01
-1.56462088e-01 -3.33285093e-01 7.50385374e-02 2.43747875e-01
5.32115400e-01 4.54684526e-01 6.27602413e-02 -3.64058435e-01
1.26335537e+00 4.28289145e-01 -8.23266283e-02 -1.28886670e-01
4.88909185e-01 -1.83882073e-01 7.57518411e-01 4.58090991e-01
-1.24801531e-01 6.07619941e-01 -1.18446112e-01 -8.02007496e-01
-9.56430733e-01 -1.08559620e+00 -1.77182034e-02 5.49604893e-01
2.87903845e-01 -2.85072416e-01 -8.68181944e-01 -4.36192751e-01
1.23622932e-03 -6.07495844e-01 -6.04591429e-01 -2.12111458e-01
-9.15878057e-01 -4.85255986e-01 7.01801538e-01 1.04205477e+00
9.99787629e-01 -1.48433602e+00 -6.18518174e-01 4.89946306e-01
2.21956626e-01 -8.04927886e-01 -4.07282352e-01 2.29800254e-01
-1.04199195e+00 -1.30781281e+00 -1.18375516e+00 -1.24220383e+00
3.84101123e-01 -7.03596473e-02 4.99347746e-01 2.28653923e-01
-8.11161458e-01 8.06518197e-02 -2.76896000e-01 -2.01497793e-01
1.82188362e-01 1.62836760e-01 4.85649645e-01 -2.88407385e-01
8.09406698e-01 -9.94269490e-01 -8.54781687e-01 2.94951141e-01
-5.77448845e-01 6.17328547e-02 9.99956071e-01 1.28167951e+00
6.86211050e-01 -1.83506340e-01 6.94911599e-01 -6.39692664e-01
7.22321808e-01 -1.50786877e-01 -4.40919101e-02 9.34420526e-02
-5.41327357e-01 -1.39563620e-01 9.66200292e-01 -8.27401519e-01
-6.62482500e-01 2.26713195e-02 -7.31022775e-01 -9.25535023e-01
-1.55221090e-01 6.32530034e-01 1.54859736e-04 -3.54923815e-01
6.17658973e-01 5.73631644e-01 6.24474406e-01 -8.86945486e-01
-5.61825521e-02 1.08598197e+00 9.09325600e-01 -4.47768480e-01
4.45786715e-01 6.69806972e-02 -4.55586165e-02 -1.01496947e+00
-2.44469285e-01 -5.78193843e-01 -7.69099236e-01 -5.60858428e-01
9.92628574e-01 -9.51019406e-01 -7.71830261e-01 1.16215897e+00
-7.35971808e-01 -4.84198898e-01 -3.73120099e-01 1.03665292e+00
-7.21613467e-01 5.19680023e-01 -1.05964172e+00 -5.90957463e-01
-6.74699962e-01 -9.58789766e-01 9.52246785e-01 2.99678296e-01
-1.36331003e-02 -8.31186354e-01 1.16681620e-01 4.63525206e-01
6.33658826e-01 2.61958271e-01 5.36459386e-01 -2.88287073e-01
-4.52153951e-01 -5.31739295e-01 4.55330499e-02 8.11099410e-01
3.41097236e-01 -5.47157824e-01 -5.49262285e-01 -4.65864152e-01
1.40319198e-01 -8.59972417e-01 6.43665850e-01 3.59772682e-01
1.40453577e+00 4.44305018e-02 -1.43709570e-01 7.54060507e-01
1.33322740e+00 3.11138421e-01 9.87289488e-01 1.78239837e-01
9.07517791e-01 1.40278623e-01 6.74889326e-01 1.37809068e-01
-2.03434825e-01 7.78268397e-01 9.07783359e-02 -5.36766768e-01
-4.25664693e-01 -9.60409865e-02 1.91842839e-01 1.37311995e+00
-8.73148739e-01 1.98125973e-01 -5.41749954e-01 4.28657681e-01
-1.79152715e+00 -1.07574618e+00 2.03278288e-01 1.85610902e+00
8.40191901e-01 -2.43293610e-03 3.41651499e-01 4.22843397e-01
5.32483876e-01 1.03541449e-01 -8.48949790e-01 -5.89692034e-02
5.46231791e-02 8.07870030e-01 3.03834856e-01 -3.31172079e-01
-9.08802152e-01 6.59186482e-01 6.17099524e+00 1.02242768e+00
-1.48089802e+00 4.14709337e-02 -8.84519741e-02 -4.32155728e-01
4.84421164e-01 -6.28757656e-01 -5.13196766e-01 6.73982978e-01
1.01053524e+00 -6.28977176e-03 3.53502929e-01 1.00294507e+00
-7.10921586e-02 2.02170566e-01 -7.47983038e-01 1.54498351e+00
-4.07491624e-02 -1.49411750e+00 -1.71481103e-01 2.71792740e-01
3.57979417e-01 4.21007007e-01 -4.48539495e-01 4.01769996e-01
-1.28534585e-01 -9.50409710e-01 -3.06473766e-02 5.02424121e-01
1.04075360e+00 -5.73064983e-01 1.06277609e+00 3.62862796e-01
-1.28235209e+00 -4.67393408e-03 -5.86898386e-01 -3.75237852e-01
4.56541702e-02 2.07796589e-01 -1.05376817e-01 4.94041800e-01
9.32815850e-01 1.02782214e+00 -9.62914433e-03 8.54616284e-01
2.40476072e-01 5.73208332e-01 -4.85701978e-01 -3.37667704e-01
1.74101084e-01 7.79859116e-03 3.43996912e-01 1.08674026e+00
1.77836537e-01 5.14576770e-03 2.25354061e-01 6.35208011e-01
-1.00441277e-01 -5.44529893e-02 -7.52319157e-01 -3.89711976e-01
5.74472845e-02 1.03068674e+00 -1.80961549e-01 -2.15926126e-01
-5.13881326e-01 9.44660544e-01 4.06264067e-01 1.41244292e-01
-4.48370099e-01 -6.78813279e-01 8.68832946e-01 2.32081249e-01
1.57738894e-01 -4.31329012e-01 3.45036131e-03 -1.20562172e+00
3.13802719e-01 -8.51385176e-01 1.94971755e-01 -4.22357857e-01
-1.41836107e+00 4.73697156e-01 -3.00673306e-01 -1.55980945e+00
-4.60445195e-01 -1.12539816e+00 -6.13872766e-01 7.39728868e-01
-1.04931593e+00 -9.01443899e-01 -5.75394213e-01 9.16489363e-01
8.02591980e-01 -4.65490997e-01 8.28206897e-01 5.72210193e-01
-6.06780469e-01 4.77191448e-01 1.76735017e-02 4.71229702e-01
2.44808152e-01 -1.08619416e+00 2.47986615e-01 4.63216990e-01
-2.25047581e-02 8.23862731e-01 1.45044237e-01 -6.44419372e-01
-1.95980501e+00 -6.22151077e-01 2.92672396e-01 4.91803177e-02
3.54486585e-01 -3.11188847e-01 -9.51257050e-01 3.11831713e-01
-1.35848131e-02 8.81636366e-02 7.22414970e-01 1.97978631e-01
2.68693119e-01 -2.64862984e-01 -8.82853806e-01 3.53635758e-01
1.59532464e+00 -7.37982333e-01 -7.43885458e-01 3.43983352e-01
1.86428294e-01 -5.02364218e-01 -1.48135912e+00 7.27213442e-01
8.31368864e-01 -4.15262520e-01 9.32596326e-01 -5.69950998e-01
4.50465739e-01 -2.10047111e-01 2.32356817e-01 -1.13758731e+00
-1.97419688e-01 -3.82613450e-01 -2.60158479e-01 5.51484883e-01
-9.36654955e-02 -4.69017625e-01 1.23686743e+00 3.27698559e-01
-6.63400650e-01 -1.23046350e+00 -1.32836461e+00 -9.74059284e-01
-2.06936568e-01 -1.25286564e-01 1.56194091e-01 6.67577922e-01
1.52591646e-01 1.66222796e-01 -8.17813814e-01 -4.06934559e-01
4.70872611e-01 -5.58187626e-02 9.49669063e-01 -1.08454144e+00
-2.32900172e-01 -2.56224722e-02 -1.22707689e+00 -1.35325360e+00
1.27129797e-02 -6.81849778e-01 2.19599858e-01 -1.43421102e+00
2.69860271e-02 -9.60643813e-02 -6.42371655e-01 4.01697069e-01
-3.24447080e-02 5.06909072e-01 -1.11076616e-01 3.91777068e-01
-4.04671520e-01 6.81352317e-01 1.66120982e+00 -2.41703585e-01
-2.39762198e-02 -1.25705615e-01 -3.90101448e-02 3.34537953e-01
6.12896204e-01 -1.90017894e-01 -5.26545227e-01 -3.26684475e-01
-6.08479202e-01 8.14258605e-02 2.28873014e-01 -1.67929971e+00
4.44304228e-01 3.11347619e-02 4.92219001e-01 -5.00240803e-01
5.47483087e-01 -6.64868653e-01 2.07875684e-01 5.80462039e-01
-5.20353168e-02 -2.87862778e-01 1.12629207e-02 5.70716918e-01
-5.03612399e-01 6.01354055e-02 3.36086512e-01 -4.23284709e-01
-1.21267712e+00 6.27199531e-01 -5.78971565e-01 -1.48133427e-01
9.66387451e-01 -8.98315668e-01 -1.34379327e-01 8.96133780e-02
-8.58959258e-01 1.62206829e-01 2.71177560e-01 5.95725834e-01
1.03051627e+00 -1.39402115e+00 -1.07309058e-01 6.18359983e-01
-8.67469609e-02 6.03197291e-02 7.11432815e-01 8.08935940e-01
-6.96559727e-01 3.15479845e-01 -9.91303146e-01 -7.49164939e-01
-1.25973439e+00 6.58917055e-02 3.86581898e-01 -1.44621015e-01
-1.33180833e+00 6.09493196e-01 -4.41150703e-02 -4.77530271e-01
5.11019945e-01 -1.31931946e-01 -1.06905669e-01 -6.12966359e-01
3.54548365e-01 2.19715923e-01 1.64279230e-02 -3.32777500e-01
-1.69278294e-01 7.43842602e-01 -1.29041001e-01 4.27424341e-01
1.64326310e+00 2.42895007e-01 3.64334673e-01 4.13991392e-01
1.24761283e+00 -8.15731704e-01 -1.48339701e+00 -5.21390699e-02
-2.32453302e-01 -4.66273248e-01 1.89380035e-01 -6.30524099e-01
-1.45785058e+00 1.01251435e+00 9.99122679e-01 -4.57853228e-01
1.25150859e+00 -2.67077923e-01 1.34761691e+00 6.30221665e-01
8.24053109e-01 -1.62046528e+00 3.19560230e-01 2.27146387e-01
7.98176765e-01 -1.03975224e+00 -4.64587398e-02 -2.91566104e-01
-3.18645030e-01 1.34981918e+00 1.17406952e+00 -5.24476230e-01
7.24288106e-01 3.44324499e-01 3.53390396e-01 -6.77729189e-01
-4.89823699e-01 -1.06857754e-01 2.27501348e-01 7.76914358e-01
3.99670631e-01 -5.96942715e-02 -7.38594949e-01 4.90707219e-01
1.14417315e-01 7.67128170e-01 -1.20869890e-01 1.22678459e+00
-2.69091874e-01 -1.14691341e+00 3.72348815e-01 1.01523280e+00
-5.23834765e-01 3.78792584e-01 -2.27876842e-01 1.01191235e+00
1.31792143e-01 3.11104000e-01 2.68690325e-02 -1.19887376e+00
4.96603370e-01 -2.86259186e-02 6.42934799e-01 -5.95393538e-01
-7.89698601e-01 -5.78271449e-02 1.79673374e-01 -7.52606034e-01
-6.46955192e-01 -2.46480748e-01 -1.32077813e+00 -2.79141903e-01
-4.88846838e-01 -5.47379665e-02 6.22292876e-01 1.02879572e+00
3.55148882e-01 5.65325320e-01 4.14173454e-01 -1.14352608e+00
-6.14203036e-01 -1.48830140e+00 -1.06507003e+00 7.74068475e-01
-7.74173364e-02 -1.04022563e+00 -2.43278012e-01 -2.37065122e-01] | [6.894866943359375, 0.15377454459667206] |
a6cec09b-e85a-4900-8814-afa45014121f | tribeflow-mining-predicting-user-trajectories | 1511.01032 | null | http://arxiv.org/abs/1511.01032v2 | http://arxiv.org/pdf/1511.01032v2.pdf | TribeFlow: Mining & Predicting User Trajectories | Which song will Smith listen to next? Which restaurant will Alice go to
tomorrow? Which product will John click next? These applications have in common
the prediction of user trajectories that are in a constant state of flux over a
hidden network (e.g. website links, geographic location). What users are doing
now may be unrelated to what they will be doing in an hour from now. Mindful of
these challenges we propose TribeFlow, a method designed to cope with the
complex challenges of learning personalized predictive models of
non-stationary, transient, and time-heterogeneous user trajectories. TribeFlow
is a general method that can perform next product recommendation, next song
recommendation, next location prediction, and general arbitrary-length user
trajectory prediction without domain-specific knowledge. TribeFlow is more
accurate and up to 413x faster than top competitors. | ['Christos Faloutsos', 'Jussara Almeida', 'Flavio Figueiredo', 'Bruno Ribeiro'] | 2015-11-03 | null | null | null | null | ['product-recommendation'] | ['miscellaneous'] | [-2.32074738e-01 -2.78097004e-01 -8.34618390e-01 -3.21207285e-01
-3.47501814e-01 -7.53523231e-01 2.72664875e-01 3.15470606e-01
-4.05053422e-02 7.76668131e-01 4.88465935e-01 -5.54644167e-01
-6.08721256e-01 -9.48654175e-01 -5.62101662e-01 -3.45461786e-01
-6.23838603e-01 7.64477551e-01 4.46531892e-01 -2.78906643e-01
-2.30851114e-01 3.18004817e-01 -1.59942591e+00 1.62174657e-01
1.81966022e-01 6.45912170e-01 -6.11821860e-02 7.74983525e-01
8.14200565e-02 5.82282722e-01 1.34544775e-01 -3.12432110e-01
1.45344347e-01 8.04933440e-03 -8.43313456e-01 -5.20006180e-01
2.95096487e-01 -2.44378492e-01 -6.67574823e-01 5.65815926e-01
3.55721831e-01 7.54793167e-01 3.68391275e-01 -1.27224362e+00
-4.22689497e-01 9.57296610e-01 -8.21723938e-02 6.56731963e-01
5.17251372e-01 3.81564088e-02 1.33158171e+00 -3.27758253e-01
8.26587379e-01 7.33566463e-01 9.50922370e-01 1.73911080e-01
-1.31283116e+00 -5.61625361e-01 6.82433009e-01 4.52991098e-01
-1.47790778e+00 -1.95515603e-01 3.52778822e-01 -5.38118184e-01
7.97162294e-01 8.24173331e-01 4.02770132e-01 1.36716247e+00
1.67772159e-01 7.19120324e-01 2.30670363e-01 3.60160947e-01
1.00813232e-01 4.28504534e-02 3.02009463e-01 3.14250052e-01
2.22945493e-02 7.18860328e-03 -6.42956555e-01 -4.80820566e-01
1.30896717e-01 6.75511241e-01 -4.99844886e-02 4.48731817e-02
-1.27480483e+00 6.86751723e-01 1.54465407e-01 3.32880527e-01
-6.69919670e-01 -7.79022798e-02 2.23568022e-01 3.15115511e-01
5.49118042e-01 4.23218042e-01 -1.08300626e+00 -3.81879002e-01
-9.45980370e-01 5.90082884e-01 1.34578359e+00 1.31594551e+00
8.99824739e-01 -2.75365680e-01 -1.62251949e-01 2.78614610e-01
2.82311350e-01 3.07003111e-01 5.20456612e-01 -8.94334614e-01
1.28467038e-01 1.98543862e-01 5.29095888e-01 -1.09477210e+00
-6.09017849e-01 -6.61600411e-01 -7.81995118e-01 -6.65993631e-01
5.11297941e-01 -4.70858067e-01 -4.53126580e-01 1.40820277e+00
4.77696389e-01 8.17152381e-01 -2.73662269e-01 6.10612869e-01
5.93371570e-01 1.03833485e+00 2.36786723e-01 -2.32341945e-01
1.07223570e+00 -8.32732439e-01 -4.70560968e-01 1.90518498e-01
5.70781112e-01 -8.71890724e-01 6.69648886e-01 3.51368159e-01
-8.41381371e-01 -6.94976211e-01 -5.53167045e-01 8.73338729e-02
-7.42870629e-01 -3.49072248e-01 9.13631380e-01 6.21084690e-01
-8.86241853e-01 1.18452740e+00 -9.11764741e-01 -7.28340507e-01
-1.13946244e-01 4.10287559e-01 6.77454844e-02 9.22427922e-02
-1.40111637e+00 3.94340605e-01 1.75710291e-01 -2.19512418e-01
-6.77883506e-01 -1.21714759e+00 -2.85688102e-01 2.94566810e-01
3.26174557e-01 -7.90871441e-01 1.45758116e+00 -4.52302158e-01
-1.28307295e+00 1.01929784e-01 -6.08930230e-01 -5.57224810e-01
2.84793943e-01 -2.01621085e-01 -1.28679085e+00 -5.51195443e-01
1.33335114e-01 4.57430482e-02 4.46533948e-01 -5.09265184e-01
-1.26780343e+00 -3.28355372e-01 -2.41907358e-01 8.40872526e-02
-3.29692125e-01 -3.75187129e-01 -6.41089976e-01 -2.27176473e-01
3.84210236e-02 -1.14637053e+00 -6.19823039e-01 -7.46097445e-01
-5.39980114e-01 -6.28340721e-01 9.18360889e-01 -4.73694146e-01
1.72841096e+00 -1.93425369e+00 -3.17226857e-01 5.51808178e-01
3.27057764e-02 1.05313621e-01 8.10605809e-02 9.82802629e-01
3.43698621e-01 1.87811822e-01 6.77893519e-01 2.45661195e-02
3.63002688e-01 7.82481432e-02 -3.92645687e-01 3.17029178e-01
-9.16168749e-01 7.35872805e-01 -1.23811543e+00 1.67932734e-03
-3.93100195e-02 3.45992178e-01 -4.87831116e-01 -7.62058794e-03
-4.60482150e-01 3.91692281e-01 -7.67810166e-01 4.33281124e-01
3.76454055e-01 -6.54423654e-01 3.77135277e-01 1.06672943e-01
-3.06061655e-01 5.20853877e-01 -1.06628656e+00 1.43031490e+00
-3.33521277e-01 5.66261053e-01 -3.16289276e-01 -3.51459801e-01
6.26425385e-01 4.74380493e-01 9.09051836e-01 -3.57785374e-01
-2.20591933e-01 -1.71534978e-02 -2.12205648e-01 -6.18967593e-01
7.97372162e-01 2.43866354e-01 1.65956169e-01 6.21401966e-01
-3.03514332e-01 9.88075614e-01 1.75711304e-01 3.03169906e-01
1.37151647e+00 -1.08967749e-02 1.17945716e-01 -1.69565707e-01
1.97192997e-01 3.68873440e-02 5.71895659e-01 1.08627796e+00
-1.77174300e-01 1.05653629e-02 1.87601656e-01 -7.10328519e-01
-9.56663311e-01 -1.17447710e+00 1.07316419e-01 1.79280353e+00
1.39467657e-01 -7.00062394e-01 -6.08568192e-02 -5.52969873e-01
2.12221324e-01 8.16428363e-01 -2.55990952e-01 2.58205414e-01
-4.35253024e-01 -6.48485184e-01 8.18533748e-02 9.20730233e-02
1.18645743e-01 -7.73744941e-01 2.33670220e-01 4.38019454e-01
-3.06405693e-01 -9.27105606e-01 -1.06041801e+00 -1.77457988e-01
-7.18119979e-01 -8.37687969e-01 -4.18453842e-01 -6.45858228e-01
9.96387973e-02 3.85143489e-01 1.19746578e+00 -2.97598958e-01
1.59159720e-01 6.66849613e-01 -2.22365260e-01 -1.36461571e-01
8.39303806e-02 9.01647806e-01 3.10570180e-01 5.28160155e-01
7.48449922e-01 -1.05902553e+00 -1.09846485e+00 5.44180274e-01
-3.52134168e-01 -3.47173989e-01 1.14217505e-01 1.41154692e-01
5.19039810e-01 6.15270615e-01 6.73046768e-01 -1.43287385e+00
3.81184489e-01 -1.42958283e+00 -2.72341400e-01 5.48836179e-02
-9.31213737e-01 -4.77677792e-01 7.40478933e-01 -6.05633199e-01
-7.16260910e-01 3.31221312e-01 -4.38430905e-01 6.27345033e-03
-4.56148267e-01 5.79567373e-01 7.19283149e-02 4.39999580e-01
6.39105976e-01 7.57312402e-02 -5.12666345e-01 -7.87555039e-01
4.34862643e-01 4.90161121e-01 3.02549273e-01 -2.70092264e-02
7.62427211e-01 3.91463190e-01 -9.43459943e-02 -9.20660019e-01
-7.76628852e-01 -1.11521280e+00 -6.35608733e-01 -1.14414789e-01
7.06731260e-01 -9.04941857e-01 -1.45112789e+00 -5.30741997e-02
-7.27858365e-01 -4.84639406e-01 -2.09022954e-01 6.66688859e-01
-1.49744764e-01 -5.91109358e-02 -5.41430831e-01 -6.03345931e-01
-1.62833065e-01 -3.17898095e-01 3.62149477e-01 2.48135179e-01
-5.07874191e-01 -1.46722364e+00 2.33168766e-01 8.26336220e-02
6.86635077e-01 4.04058322e-02 8.52515042e-01 -1.05225956e+00
-1.05152965e+00 -5.35380721e-01 1.24658346e-01 -3.06180149e-01
2.51550615e-01 -2.05608755e-01 -5.73206425e-01 -3.83011341e-01
-6.98649943e-01 6.92892313e-01 5.62167704e-01 5.64058244e-01
1.07733202e+00 -7.74017155e-01 -9.20898974e-01 4.91257817e-01
1.28356516e+00 2.41058663e-01 2.80233145e-01 4.02296558e-02
6.72909796e-01 3.15946907e-01 3.42506796e-01 5.57417154e-01
8.67490828e-01 9.14707303e-01 2.63657421e-01 2.79516846e-01
1.62914053e-01 -9.67020869e-01 3.87451619e-01 5.53703666e-01
-1.10370420e-01 -5.08501947e-01 -6.56801939e-01 8.61897945e-01
-2.13284159e+00 -1.54971230e+00 -5.00543535e-01 2.33991528e+00
4.56899107e-01 -2.86515392e-02 6.80569470e-01 -4.66976434e-01
6.35173440e-01 1.84551150e-01 -6.99363291e-01 4.02588807e-02
3.63797784e-01 -2.31623247e-01 9.85541523e-01 5.80955446e-01
-1.20810473e+00 8.99913907e-01 6.46859074e+00 6.91787183e-01
-1.00488651e+00 4.15639162e-01 4.76956069e-01 -2.53752977e-01
-4.11894619e-01 5.83420359e-02 -1.34532320e+00 6.64602935e-01
1.51594198e+00 -3.73438507e-01 6.15711570e-01 8.99559796e-01
4.55150306e-01 2.62585163e-01 -1.16324198e+00 8.24154496e-01
-3.10199857e-01 -1.50029206e+00 -3.39392543e-01 2.08995879e-01
9.93115842e-01 5.97247481e-01 2.94906735e-01 4.35189575e-01
8.09276164e-01 -6.73742175e-01 5.21323718e-02 1.07632709e+00
1.74274549e-01 -6.96259439e-01 2.72752076e-01 7.78036952e-01
-1.19736159e+00 -2.74818063e-01 -2.07442209e-01 -1.29482923e-02
4.22972232e-01 5.42621672e-01 -1.19680822e+00 3.78611326e-01
7.90967464e-01 1.20070803e+00 -1.63606852e-01 1.26603556e+00
2.43735164e-01 1.08678913e+00 -4.04047638e-01 -2.80324072e-01
2.34855920e-01 3.65601107e-02 6.99882269e-01 1.11838555e+00
8.22670758e-01 1.21561304e-01 4.43662226e-01 2.64642507e-01
-1.65675342e-01 2.02894405e-01 -7.61153817e-01 1.34833008e-01
5.07875323e-01 1.36522198e+00 -6.13720536e-01 -2.26401404e-01
-4.12804276e-01 9.46781218e-01 -2.60199189e-01 7.69415379e-01
-6.21119022e-01 -3.02813202e-01 1.04981911e+00 9.77209866e-01
4.85463470e-01 -1.05734080e-01 3.81853551e-01 -9.04054403e-01
-3.91935736e-01 -1.26275793e-01 6.21016741e-01 -5.74075580e-01
-1.45403206e+00 4.31319833e-01 -2.82353848e-01 -1.30801845e+00
-3.94528151e-01 -1.79572657e-01 -6.96688533e-01 5.40427387e-01
-1.47788191e+00 -1.01457930e+00 1.94963977e-01 8.08560789e-01
4.95176435e-01 -2.68612444e-01 8.77753854e-01 8.75698924e-01
-2.35022575e-01 4.28748816e-01 6.59545004e-01 -1.74611703e-01
5.91755748e-01 -1.19968450e+00 6.16967976e-01 3.70975107e-01
2.73594767e-01 7.16071844e-01 7.65690446e-01 -7.18198955e-01
-1.38732159e+00 -1.53635120e+00 1.55119586e+00 -7.89673030e-01
9.10406888e-01 -1.51241764e-01 -6.14631236e-01 1.18642271e+00
-1.34653479e-01 -2.89681405e-01 1.16873038e+00 9.11257803e-01
-1.50566980e-01 -3.95913035e-01 -5.80949306e-01 6.12680912e-01
1.09664357e+00 -4.36920255e-01 1.82518601e-01 8.25691581e-01
8.72880936e-01 -1.87055871e-01 -1.01283073e+00 -8.30850899e-02
7.80219793e-01 -7.60519445e-01 1.10385549e+00 -1.12129796e+00
-2.29661524e-01 -2.96699822e-01 -5.73502481e-02 -1.24586105e+00
-9.27468836e-01 -1.48559189e+00 -3.68286669e-01 9.85432744e-01
9.49464440e-01 -5.43002605e-01 1.16127396e+00 7.13545024e-01
4.85050268e-02 -5.35404980e-01 -7.36935318e-01 -7.08852291e-01
-3.54695886e-01 -4.66723293e-01 1.17432857e+00 9.97293949e-01
1.03607289e-01 4.22716469e-01 -1.01501429e+00 4.08544153e-01
2.58386195e-01 3.64966899e-01 7.75066078e-01 -1.62889028e+00
-3.56277108e-01 -1.79628477e-01 -7.98066258e-02 -1.53004789e+00
-1.72184139e-01 -1.08193052e+00 -2.68171757e-01 -1.29272902e+00
-1.33432254e-01 -4.74343628e-01 -6.86336875e-01 2.52502263e-01
1.39728248e-01 7.33647346e-02 -1.07321464e-01 3.64272445e-01
-1.24844623e+00 -3.36685553e-02 9.92226601e-01 2.02674732e-01
-7.62680769e-01 1.22684360e+00 -7.15376675e-01 4.78885263e-01
8.31132531e-01 -6.25900626e-01 -5.53886294e-01 1.33698851e-01
7.71868825e-01 3.90422910e-01 8.45103934e-02 -7.85054147e-01
7.72354186e-01 -2.88774937e-01 1.92780763e-01 -8.49322498e-01
3.35742742e-01 -9.76800323e-01 6.46046102e-01 1.38563901e-01
-5.66520512e-01 1.75217822e-01 -2.91140318e-01 1.15635395e+00
3.29904020e-01 1.10957704e-01 6.33429065e-02 -1.38600066e-01
-8.27512562e-01 1.11805141e+00 -7.32271850e-01 -1.48691550e-01
7.21427321e-01 1.24482781e-01 -8.48186463e-02 -8.60975444e-01
-1.55920160e+00 4.19715017e-01 -1.29645154e-01 7.75968432e-01
1.16025977e-01 -1.38446867e+00 -3.72006834e-01 -3.27742361e-02
-9.33688134e-02 -7.99397588e-01 7.29413450e-01 7.96087503e-01
1.00322880e-01 7.94844091e-01 2.26569474e-01 -1.98949754e-01
-1.21237826e+00 6.72802389e-01 1.56027317e-01 -3.95827144e-01
-6.54881418e-01 6.54847145e-01 -5.75314403e-01 -3.84036452e-01
3.91849607e-01 -7.99476821e-03 -2.72790611e-01 2.23411337e-01
6.62754834e-01 8.75989914e-01 -1.10292718e-01 -4.66627628e-01
-4.02571149e-02 7.72865787e-02 -1.90700054e-01 2.66870677e-01
1.21896446e+00 -5.83806694e-01 1.78589061e-01 8.23315144e-01
1.17094707e+00 4.04713266e-02 -1.09762359e+00 -5.36363959e-01
2.34713539e-01 -2.89028525e-01 -2.17533465e-02 -8.31264436e-01
-8.83445501e-01 3.93513739e-01 5.74958444e-01 1.01183677e+00
6.42327130e-01 -1.95790082e-02 1.45655072e+00 6.12809002e-01
4.47650135e-01 -1.03358257e+00 -5.18315792e-01 4.42486376e-01
-2.23954255e-03 -9.41334546e-01 -4.44872789e-02 -1.53315529e-01
-5.15404463e-01 7.55630136e-01 1.03586748e-01 1.87089905e-01
1.63535106e+00 -3.39187980e-01 -3.04391980e-01 -7.12626427e-02
-1.19621360e+00 -4.34290230e-01 4.53418732e-01 4.96944517e-01
4.47557032e-01 3.72802556e-01 1.34750918e-01 6.40209675e-01
-3.65968823e-01 8.63996521e-02 2.90520340e-01 1.75266489e-01
-5.24770021e-01 -1.13996959e+00 1.51069507e-01 6.88137114e-01
-5.08780003e-01 6.71461374e-02 3.31345588e-01 4.87942964e-01
2.72777706e-01 9.72544372e-01 1.17456513e-02 -7.15097606e-01
2.56318331e-01 3.20151091e-01 -2.91469038e-01 -5.93227148e-01
-6.39436781e-01 1.00182064e-01 1.60557568e-01 -6.80710554e-01
-1.36749130e-02 -1.15000856e+00 -9.42761302e-01 -9.98320639e-01
6.40896186e-02 4.46095705e-01 3.23258817e-01 7.49776542e-01
7.88632452e-01 2.86659211e-01 7.53937840e-01 -5.68227649e-01
-2.89434880e-01 -6.64557993e-01 -9.13544297e-01 2.34454408e-01
5.28374135e-01 -1.41419709e-01 -1.37727574e-01 5.58083579e-02] | [7.124205112457275, 3.089689254760742] |
2fb2c55f-f65a-4dd0-bac9-15c88575e67d | hierarchical-reinforcement-learning-for-7 | 2210.10431 | null | https://arxiv.org/abs/2210.10431v1 | https://arxiv.org/pdf/2210.10431v1.pdf | Hierarchical Reinforcement Learning for Furniture Layout in Virtual Indoor Scenes | In real life, the decoration of 3D indoor scenes through designing furniture layout provides a rich experience for people. In this paper, we explore the furniture layout task as a Markov decision process (MDP) in virtual reality, which is solved by hierarchical reinforcement learning (HRL). The goal is to produce a proper two-furniture layout in the virtual reality of the indoor scenes. In particular, we first design a simulation environment and introduce the HRL formulation for a two-furniture layout. We then apply a hierarchical actor-critic algorithm with curriculum learning to solve the MDP. We conduct our experiments on a large-scale real-world interior layout dataset that contains industrial designs from professional designers. Our numerical results demonstrate that the proposed model yields higher-quality layouts as compared with the state-of-art models. | ['Pengqian Yu', 'Xinhan Di'] | 2022-10-19 | null | null | null | null | ['hierarchical-reinforcement-learning'] | ['methodology'] | [-3.92425567e-01 -1.38064250e-01 1.92308649e-01 -1.88308969e-01
-3.52082998e-01 -3.37386906e-01 -5.73767163e-02 -1.90762773e-01
1.00838639e-01 8.55962038e-01 1.45137966e-01 -4.96053189e-01
-5.87581098e-01 -7.71240056e-01 -6.58091128e-01 -6.84092939e-01
-1.18786357e-01 3.26827675e-01 -3.57090116e-01 -2.47713953e-01
-2.61353739e-02 5.63612223e-01 -1.13202739e+00 -9.06648710e-02
9.88471270e-01 5.49893618e-01 7.51810253e-01 8.51384103e-01
1.65490925e-01 6.00806594e-01 -6.22002244e-01 -1.23808697e-01
3.75609308e-01 -1.57439873e-01 -5.52976608e-01 9.60896432e-01
1.33625403e-01 -4.37624246e-01 -4.91909474e-01 7.60036528e-01
5.16400099e-01 6.15284920e-01 6.87215388e-01 -1.44188964e+00
-8.29285383e-01 3.47063333e-01 -5.84934771e-01 -4.33926404e-01
5.83244383e-01 1.98605016e-01 9.67015028e-01 -3.32261264e-01
5.70417047e-01 1.57910907e+00 2.82245755e-01 4.94154841e-01
-1.41553652e+00 -3.42114568e-01 9.12638426e-01 -1.63311079e-01
-1.40448034e+00 3.63700271e-01 1.08415675e+00 -4.98861790e-01
5.73593676e-01 2.06283242e-01 1.14341605e+00 1.26217544e+00
6.24428689e-01 1.13860548e+00 1.30128169e+00 -6.08191937e-02
6.93664968e-01 -1.20169826e-01 -2.95865744e-01 8.83775771e-01
1.51353315e-01 1.29855379e-01 2.56432235e-01 9.02726687e-03
1.63287425e+00 3.36587787e-01 8.10492709e-02 -1.09411430e+00
-1.02331185e+00 6.77773476e-01 6.15579188e-01 -1.14133500e-01
-6.28282189e-01 4.84057993e-01 -5.81165291e-02 1.59702152e-01
-1.40248671e-01 6.95194960e-01 -4.27413255e-01 3.10225576e-01
-3.56539428e-01 7.36157775e-01 7.35987663e-01 1.29506063e+00
9.35352966e-02 1.70550331e-01 -2.78854482e-02 7.90443420e-01
8.65289688e-01 5.88077605e-01 -4.53683734e-01 -1.41966164e+00
4.78524476e-01 1.12519180e-02 8.76453459e-01 -9.05262351e-01
-3.97642344e-01 -4.35459048e-01 -1.23865044e+00 4.30994034e-01
3.18641394e-01 -4.82068747e-01 -5.79226911e-01 1.35926640e+00
4.74858552e-01 2.17379183e-01 8.37851837e-02 1.11572349e+00
5.72421610e-01 9.72698927e-01 -4.48935665e-02 -2.27793321e-01
1.08696139e+00 -1.01666653e+00 -1.01741183e+00 1.27263159e-01
5.56022348e-03 -6.98226213e-01 1.16491520e+00 8.21324766e-01
-9.65506971e-01 -1.04681325e+00 -9.84359384e-01 4.52176481e-01
1.88095391e-01 2.99073100e-01 8.34733784e-01 6.65340960e-01
-8.08985770e-01 6.00902140e-01 -6.74084604e-01 -1.92055881e-01
1.91583291e-01 2.70151645e-01 1.24295406e-01 -9.18202326e-02
-7.95583725e-01 4.55197245e-01 -5.42655364e-02 3.48209977e-01
-1.41369247e+00 -5.72067976e-01 -8.57871294e-01 -3.57131287e-02
7.26896763e-01 -1.03786540e+00 1.47461939e+00 -5.21112084e-01
-1.97836649e+00 2.53788054e-01 2.15124995e-01 1.91103980e-01
6.86429918e-01 -4.77504462e-01 -2.84434527e-01 -2.33428136e-01
4.00151499e-03 4.50994670e-01 8.58494818e-01 -2.18870902e+00
-4.12674397e-01 9.56230611e-02 5.18200815e-01 2.98163712e-01
5.38687646e-01 -5.25913298e-01 -3.33850086e-01 -6.59887850e-01
2.06610203e-01 -9.92976606e-01 -1.14739883e+00 -1.88138723e-01
-5.65164208e-01 -9.34300497e-02 4.76150393e-01 -6.66526139e-01
1.11328471e+00 -2.13325477e+00 3.47001582e-01 2.65407145e-01
-6.98090047e-02 -4.41072106e-01 -1.15854358e-02 6.40639663e-01
7.76109993e-02 -4.46457639e-02 2.45972157e-01 -4.62736964e-01
3.55295599e-01 4.77134526e-01 -2.79090315e-01 3.44063997e-01
-2.42207140e-01 7.80207694e-01 -1.38924551e+00 -4.07905310e-01
6.34960234e-01 2.55432785e-01 -9.10846353e-01 3.63602072e-01
-3.61698061e-01 7.86756516e-01 -9.52681959e-01 4.74579304e-01
5.90180755e-01 -1.71153739e-01 6.15737796e-01 1.84225038e-01
-8.55443999e-02 -4.14143682e-01 -1.86866200e+00 2.02083993e+00
-8.20166171e-01 1.37164980e-01 1.48794830e-01 -3.43353093e-01
1.20331681e+00 1.54472873e-01 4.85430628e-01 -5.49479485e-01
1.39376491e-01 -3.45568150e-01 -3.61400813e-01 -7.27408469e-01
6.08876765e-01 5.76143153e-02 -5.63379228e-01 2.95821410e-02
-4.33170050e-01 -6.52187526e-01 -1.98461637e-01 -2.14855745e-02
1.02899206e+00 6.88673794e-01 2.25951313e-03 -2.80959994e-01
3.04619610e-01 -3.75196666e-01 6.97455764e-01 9.06221807e-01
-2.18623579e-01 3.63805324e-01 4.54124242e-01 -4.50065523e-01
-1.31346059e+00 -1.73668063e+00 4.09179866e-01 5.11573255e-01
5.20638645e-01 -2.19325483e-01 -5.83189070e-01 -4.24340338e-01
7.88599178e-02 7.95915842e-01 -2.27077574e-01 1.84981868e-01
-4.65324044e-01 -3.07530016e-01 -4.82012957e-01 4.14876580e-01
8.26033831e-01 -1.12979090e+00 -7.94526339e-01 3.33503336e-01
-1.85491163e-02 -9.29467142e-01 -2.58449137e-01 -5.22242561e-02
-7.13888764e-01 -7.38982260e-01 -7.48218894e-01 -1.06505060e+00
8.53168845e-01 3.23668659e-01 1.00964344e+00 -8.57582837e-02
-3.51333261e-01 6.13741934e-01 -1.96658179e-01 -1.42834529e-01
-2.14118809e-01 -7.73439929e-02 5.84772080e-02 4.11590822e-02
-6.18769050e-01 -5.24054587e-01 -7.72461116e-01 5.48715770e-01
-3.48003298e-01 4.88735706e-01 4.74649519e-01 7.97772467e-01
7.25193441e-01 9.13376212e-01 4.40829277e-01 -5.63967228e-01
6.71139836e-01 -9.78040770e-02 -9.65357542e-01 1.95570797e-01
1.42025769e-01 -1.15808815e-01 6.82549238e-01 -3.30661207e-01
-1.53842592e+00 4.14514631e-01 -1.26163855e-01 -4.42234546e-01
-4.48436916e-01 2.52779014e-02 -5.29435873e-01 3.73401046e-01
4.12866920e-02 -2.72605598e-01 -6.20805860e-01 -5.13958633e-01
5.56062162e-01 4.31359738e-01 2.29904369e-01 -1.12508893e+00
9.30793226e-01 1.77761897e-01 1.29517376e-01 -8.72414649e-01
-7.94387639e-01 -2.39934266e-01 -7.96253622e-01 -7.03692973e-01
1.06046689e+00 -7.69832850e-01 -1.57904232e+00 1.99464560e-01
-9.33405161e-01 -1.02509129e+00 -2.29564488e-01 3.26799214e-01
-1.15585709e+00 1.50535285e-01 -6.37196898e-01 -1.38670421e+00
3.30675244e-01 -1.32712567e+00 1.20353544e+00 2.71555781e-01
-1.68882549e-01 -7.61987746e-01 1.44384921e-01 3.46930116e-01
-2.14669839e-01 7.91350543e-01 1.10650873e+00 5.94045460e-01
-1.10684252e+00 4.23326313e-01 8.86842534e-02 -8.94686505e-02
1.85799748e-01 1.14660837e-01 -5.52171826e-01 -3.62518817e-01
-3.10427815e-01 -9.32983458e-02 1.97180375e-01 7.15601385e-01
1.27630794e+00 -8.38996097e-02 -2.38985255e-01 2.90145040e-01
1.69901586e+00 6.99832559e-01 5.31350672e-01 3.96993279e-01
6.74521685e-01 5.78875959e-01 1.21308982e+00 8.91695619e-01
4.08521295e-01 4.90575194e-01 4.44866180e-01 -2.56057233e-01
2.21890450e-01 -8.01538110e-01 -1.58781360e-04 6.85808599e-01
1.36546250e-02 -2.50851363e-01 -7.14826822e-01 4.15864259e-01
-2.32870245e+00 -7.75082767e-01 -1.91707179e-01 1.85465062e+00
1.90963849e-01 2.89725542e-01 4.29762178e-04 6.83274716e-02
5.63704729e-01 1.70597527e-02 -4.38694596e-01 -5.02582133e-01
2.99025893e-01 -1.64211705e-01 2.81356722e-01 4.16862756e-01
-1.26337397e+00 8.54080379e-01 6.47850513e+00 4.41358656e-01
-2.07912102e-01 -2.22289264e-01 5.94639957e-01 4.26759049e-02
-2.42324874e-01 -1.70217037e-01 -4.64184493e-01 1.50498837e-01
4.31068212e-01 3.18538994e-01 8.21671844e-01 9.37438726e-01
8.08148324e-01 -2.65197068e-01 -1.05482781e+00 1.25949979e+00
-6.09682620e-01 -1.09291840e+00 -2.74105519e-01 1.31730452e-01
1.28325903e+00 -8.59809697e-01 1.94071680e-01 4.83251840e-01
1.04468906e+00 -7.33471632e-01 9.25773799e-01 7.08221734e-01
2.89150745e-01 -1.24444830e+00 3.20789665e-01 3.67226630e-01
-1.44986093e+00 -5.17034173e-01 -3.32090974e-01 -2.52487212e-01
5.86835146e-01 2.45770708e-01 -4.16204929e-01 7.63367712e-01
6.74623668e-01 6.23294175e-01 -8.86381883e-03 1.41545093e+00
-5.27270913e-01 -1.25275552e-03 2.94853896e-01 -4.50822532e-01
3.48571897e-01 -6.31767035e-01 2.94521630e-01 6.42205477e-01
9.56261232e-02 2.32964337e-01 8.73108685e-01 8.88369799e-01
1.86986700e-01 -9.29031074e-02 -7.95152307e-01 4.08335775e-01
2.74481684e-01 1.10822666e+00 -8.24831009e-01 -1.64971054e-01
-7.86493942e-02 1.08405828e+00 2.31859028e-01 7.24896252e-01
-1.12679982e+00 -1.34225816e-01 7.92119443e-01 -1.95032537e-01
4.13105100e-01 -8.03670883e-01 -2.40537643e-01 -8.51771176e-01
-3.55550766e-01 -7.61187136e-01 -1.47739500e-01 -1.34715807e+00
-1.26462543e+00 1.85157776e-01 6.13941699e-02 -1.04793942e+00
3.62043262e-01 -4.34540391e-01 -2.80660540e-01 4.22932625e-01
-1.19538760e+00 -8.64381611e-01 -2.73809582e-01 3.14382046e-01
9.94033873e-01 -3.23205907e-03 7.29963839e-01 2.24787518e-01
-8.02882791e-01 -7.98840914e-03 4.25403684e-01 4.10377197e-02
1.64335355e-01 -1.62915909e+00 5.82275152e-01 4.25912678e-01
-1.73970416e-01 4.97947663e-01 8.70760143e-01 -9.60131705e-01
-1.64544821e+00 -1.28772461e+00 1.57696098e-01 -4.26644236e-01
4.89947796e-01 -5.59676945e-01 -2.88245708e-01 6.55653119e-01
2.61136323e-01 -5.53335130e-01 2.64059901e-01 9.01724920e-02
3.69276553e-01 1.78342033e-02 -1.15773749e+00 1.17802703e+00
1.38398349e+00 -2.50027720e-02 -5.75661480e-01 1.81746677e-01
8.95242155e-01 -5.60766399e-01 -9.67813075e-01 5.35222515e-02
5.94639003e-01 -5.13590991e-01 1.22761774e+00 -4.95605201e-01
4.28068191e-01 -4.33313191e-01 -2.62600303e-01 -1.59329784e+00
-7.68696427e-01 -9.66324568e-01 5.59791587e-02 1.07992470e+00
-6.35742918e-02 -1.33769289e-01 8.55810761e-01 5.60864568e-01
-1.37131631e-01 -7.67772853e-01 -4.46925521e-01 -8.32789779e-01
-4.02014226e-01 -2.54969776e-01 7.49389112e-01 5.18858552e-01
-4.94887829e-01 3.44957143e-01 -4.67850506e-01 4.77003753e-01
9.64747369e-01 3.75827104e-01 9.56866324e-01 -1.07892597e+00
-5.86127520e-01 9.67259426e-03 1.61994785e-01 -1.41297257e+00
2.51970917e-01 -4.21932399e-01 4.11271602e-01 -2.12856388e+00
8.18293393e-02 -8.87353301e-01 -2.29253337e-01 -1.26056656e-01
1.89518452e-01 -4.90325153e-01 4.03070182e-01 -4.29360211e-01
-9.90023673e-01 9.03160632e-01 2.10989213e+00 -3.05459052e-01
-7.18338966e-01 2.90223241e-01 -3.48839194e-01 7.87865877e-01
7.73173153e-01 -1.69520825e-01 -7.00453818e-01 -3.27838868e-01
2.05692381e-01 7.10276008e-01 3.55448693e-01 -9.47808862e-01
-2.39268675e-01 -5.51251471e-01 7.60012627e-01 -9.21804428e-01
4.62817580e-01 -1.19054675e+00 3.16270828e-01 6.54949725e-01
-3.42171401e-01 2.27652609e-01 8.81944597e-02 8.12676370e-01
5.63052595e-01 4.45524864e-02 5.21020710e-01 -2.88308114e-01
-7.30716646e-01 1.05670452e-01 -7.31513381e-01 -3.98060590e-01
1.26418972e+00 8.17469284e-02 2.39190757e-01 -3.68359208e-01
-1.13270497e+00 5.80083132e-01 3.09982032e-01 5.77187896e-01
9.03932989e-01 -1.65463257e+00 -1.70883834e-01 2.03634202e-01
-4.11959261e-01 6.17961548e-02 6.36577249e-01 -2.50600278e-03
-6.85824037e-01 3.01636696e-01 -5.44294178e-01 -4.84946817e-01
-1.13928103e+00 1.05253744e+00 2.71113813e-01 -5.10184228e-01
-7.52468288e-01 5.21341026e-01 3.19877267e-01 -7.26626158e-01
5.69296479e-01 -6.29825294e-01 -1.25342742e-01 -4.50219661e-01
-7.89968669e-02 4.92654979e-01 -6.68706000e-01 1.26464181e-02
3.12311705e-02 5.63929796e-01 2.92742729e-01 -2.31738821e-01
1.36496508e+00 -4.46235269e-01 4.15930241e-01 5.57144225e-01
6.70843005e-01 -4.46730137e-01 -1.74222827e+00 2.86555976e-01
-3.70887309e-01 -8.35900724e-01 -1.39155230e-02 -6.05933726e-01
-8.27086687e-01 5.02654135e-01 5.49021184e-01 4.56023812e-02
8.28135729e-01 -3.85613054e-01 7.20104933e-01 6.65607750e-01
1.10103834e+00 -1.29859662e+00 7.50794470e-01 1.94984823e-01
1.04220235e+00 -7.89750338e-01 -5.62736504e-02 -5.61926544e-01
-7.84695685e-01 7.63725042e-01 8.61173034e-01 -6.50193155e-01
6.87526166e-01 7.16525316e-02 -1.56178534e-01 -7.72001892e-02
-5.73097169e-01 -1.14982598e-01 -1.65597245e-01 5.27560294e-01
-1.22107156e-01 3.99500757e-01 2.88313210e-01 5.97512066e-01
-2.38853082e-01 -3.26122418e-02 5.21867394e-01 1.23408401e+00
-3.66354078e-01 -1.11739016e+00 -6.16505682e-01 8.17173254e-03
1.71386734e-01 5.27517736e-01 1.34559259e-01 7.54892588e-01
9.72691998e-02 1.19057429e+00 -5.01413569e-02 -7.70459846e-02
9.10195768e-01 -5.38767457e-01 9.39844728e-01 -5.01420438e-01
-5.14602661e-01 4.90092486e-01 3.50989103e-02 -4.84238029e-01
1.73315462e-02 -5.99107862e-01 -1.27461326e+00 -4.19653863e-01
8.51566344e-02 6.24935180e-02 4.49306488e-01 5.98022521e-01
2.99632270e-03 1.48433685e+00 1.00521028e+00 -1.09686327e+00
-1.74417973e-01 -1.84753507e-01 -1.04695582e+00 1.59457430e-01
2.60515660e-01 -8.85111868e-01 4.38358396e-01 -1.07005812e-01] | [5.025123596191406, 0.7876526117324829] |
af949eef-3231-4a03-9a68-f0d230747e03 | unpaired-point-cloud-completion-on-real-scans | 1904.00069 | null | https://arxiv.org/abs/1904.00069v3 | https://arxiv.org/pdf/1904.00069v3.pdf | Unpaired Point Cloud Completion on Real Scans using Adversarial Training | As 3D scanning solutions become increasingly popular, several deep learning setups have been developed geared towards that task of scan completion, i.e., plausibly filling in regions there were missed in the raw scans. These methods, however, largely rely on supervision in the form of paired training data, i.e., partial scans with corresponding desired completed scans. While these methods have been successfully demonstrated on synthetic data, the approaches cannot be directly used on real scans in absence of suitable paired training data. We develop a first approach that works directly on input point clouds, does not require paired training data, and hence can directly be applied to real scans for scan completion. We evaluate the approach qualitatively on several real-world datasets (ScanNet, Matterport, KITTI), quantitatively on 3D-EPN shape completion benchmark dataset, and demonstrate realistic completions under varying levels of incompleteness. | ['Niloy J. Mitra', 'Xuelin Chen', 'Baoquan Chen'] | 2019-03-29 | null | https://openreview.net/forum?id=HkgrZ0EYwB | https://openreview.net/pdf?id=HkgrZ0EYwB | iclr-2020-1 | ['point-cloud-completion'] | ['computer-vision'] | [ 4.10731405e-01 3.98629248e-01 1.45043835e-01 -5.33581257e-01
-9.34735060e-01 -5.74649811e-01 6.25953138e-01 -1.23717897e-01
-1.20738834e-01 4.63585287e-01 -9.82977003e-02 -4.09666210e-01
-2.11644322e-01 -9.83382523e-01 -1.15973580e+00 -6.80418909e-02
-3.82565297e-02 1.22581184e+00 2.36321002e-01 -2.00964153e-01
1.11450300e-01 6.44636571e-01 -1.33892226e+00 -1.89559683e-02
6.68467224e-01 8.21508586e-01 3.69004980e-02 3.70599717e-01
-3.39070499e-01 1.11842016e-02 -1.22541018e-01 -2.56411672e-01
7.59693742e-01 1.07558124e-01 -6.94126248e-01 3.79299641e-01
6.62091553e-01 -5.12890279e-01 -2.55490452e-01 6.46863759e-01
3.20274681e-01 -2.57178187e-01 5.55676877e-01 -1.09579289e+00
-3.31948459e-01 4.03656960e-01 -8.82643819e-01 -3.19366753e-01
3.84986222e-01 2.95534223e-01 7.33322918e-01 -1.46266830e+00
7.41164386e-01 1.18997395e+00 1.05401814e+00 4.58359480e-01
-1.64457297e+00 -7.15065062e-01 -4.48908329e-01 -5.44541299e-01
-1.05913734e+00 -2.85308242e-01 7.36659169e-01 -4.17150766e-01
6.12422645e-01 1.56112318e-03 7.25903213e-01 1.02466488e+00
-3.37460071e-01 7.85263896e-01 1.20079553e+00 -2.22838104e-01
3.26281816e-01 -3.86081308e-01 -2.63741821e-01 6.41557634e-01
1.35045543e-01 4.11008567e-01 -3.43722165e-01 2.72606499e-03
1.15750158e+00 -6.85981363e-02 -1.72110513e-01 -1.13435972e+00
-1.40139854e+00 7.28002191e-01 5.29886067e-01 -1.19345367e-01
-5.48808217e-01 2.72275537e-01 1.27230808e-01 1.60338834e-01
6.35177195e-01 1.92241251e-01 -4.12268966e-01 1.60282746e-01
-1.27827466e+00 5.59994459e-01 6.96476698e-01 1.21570671e+00
9.94100749e-01 -5.15143387e-02 2.46464029e-01 5.97159207e-01
2.01593071e-01 8.04860950e-01 -1.48083836e-01 -1.00389874e+00
9.17882204e-01 5.61055660e-01 3.35347652e-01 -5.08442640e-01
-4.54073548e-01 -2.67402261e-01 -6.11479938e-01 6.68462396e-01
5.36984742e-01 -7.08832070e-02 -1.43599534e+00 1.44450271e+00
4.85408515e-01 3.72702539e-01 -3.08184437e-02 8.91028583e-01
7.59660661e-01 3.24455857e-01 -3.00800234e-01 4.93157566e-01
7.20101595e-01 -7.35578597e-01 -9.36271176e-02 -2.18224704e-01
2.63501555e-01 -7.18077004e-01 1.36831689e+00 5.27360320e-01
-1.33705652e+00 -4.75259751e-01 -1.03187311e+00 -2.50805654e-02
-9.36296117e-03 -9.79304314e-02 4.59894985e-01 4.20788616e-01
-1.14223158e+00 7.99838483e-01 -9.95308936e-01 -2.87591696e-01
7.78673112e-01 4.25567806e-01 -6.00126565e-01 -4.19828326e-01
-6.21416092e-01 6.40420556e-01 1.14888802e-01 6.36855662e-02
-1.15950167e+00 -1.03773928e+00 -7.16052353e-01 -2.08576828e-01
3.01291466e-01 -6.03975534e-01 1.36942446e+00 -6.45658076e-01
-9.88713741e-01 1.03795552e+00 2.66755193e-01 -5.29839993e-01
9.67516780e-01 -3.25969249e-01 -3.68601717e-02 2.22878046e-02
1.59802318e-01 1.16800261e+00 8.15066636e-01 -1.71731746e+00
-1.25652358e-01 -3.70438755e-01 1.08954385e-01 4.48873490e-02
1.87132522e-01 -5.04425228e-01 -4.21537757e-01 -4.49299783e-01
5.50079226e-01 -1.00577235e+00 -3.65568578e-01 6.68919921e-01
-5.40358841e-01 8.61962959e-02 1.08411396e+00 -5.49656391e-01
1.47877872e-01 -1.93526268e+00 -7.28930831e-02 5.41861355e-01
9.21516344e-02 2.55953729e-01 -4.23486441e-01 5.72094381e-01
-1.71770662e-01 1.90254629e-01 -7.57122755e-01 -7.64323235e-01
1.03069976e-01 5.08154213e-01 -5.19185543e-01 4.02555496e-01
4.13784742e-01 1.13412189e+00 -8.20814788e-01 -2.52499014e-01
4.63956773e-01 4.43494439e-01 -5.28914750e-01 1.32723242e-01
-5.33629477e-01 7.22319961e-01 -3.74836415e-01 7.05061257e-01
1.09938204e+00 -1.51435003e-01 -3.06454062e-01 -7.19543695e-02
-9.62155238e-02 4.99930009e-02 -1.09298742e+00 2.21552420e+00
-6.52980268e-01 3.90426636e-01 2.57116109e-01 -7.92383671e-01
1.00122201e+00 1.00706197e-01 6.97572947e-01 -7.85587013e-01
-2.54279584e-01 4.43567157e-01 -2.81058669e-01 -3.55909348e-01
3.90642166e-01 -1.44055754e-01 1.04007825e-01 9.00649130e-01
-2.49689117e-01 -7.20212102e-01 -9.58208516e-02 7.31046274e-02
1.24012959e+00 5.72526157e-01 -1.35537922e-01 -1.18112467e-01
5.40666021e-02 4.79567528e-01 1.40077814e-01 3.64252299e-01
2.51815677e-01 1.29197407e+00 1.52273238e-01 -6.57315612e-01
-1.68588746e+00 -1.48505139e+00 -2.70601749e-01 3.68866175e-01
3.54839295e-01 4.11039926e-02 -6.71773672e-01 -5.50006330e-01
1.65084630e-01 4.77882266e-01 -6.45482719e-01 3.73889238e-01
-8.15364122e-01 -1.45398781e-01 7.12281525e-01 6.48326039e-01
4.15048242e-01 -1.19162786e+00 -9.28727925e-01 1.13590494e-01
1.63814142e-01 -1.06354332e+00 -2.04850376e-01 4.32349928e-02
-1.26665545e+00 -1.19637942e+00 -7.53341913e-01 -4.83244419e-01
6.77097499e-01 3.56284499e-01 1.41177821e+00 2.45075420e-01
-2.64429033e-01 5.91722429e-01 -1.60628706e-01 -3.58182073e-01
-3.22053313e-01 4.17677425e-02 -2.50732124e-01 -3.20636600e-01
-1.27403259e-01 -1.03715801e+00 -7.00058222e-01 4.09658283e-01
-1.14542484e+00 2.95275003e-01 9.67110455e-01 8.78336489e-01
8.95620644e-01 -4.48503345e-01 4.68504399e-01 -9.87839758e-01
2.67616808e-01 -3.97024363e-01 -6.15114868e-01 2.41063181e-02
-4.85805959e-01 2.08850354e-01 2.48512492e-01 -7.91624114e-02
-9.34409738e-01 3.29015046e-01 -4.10111815e-01 -9.14825380e-01
-2.93053806e-01 4.41395611e-01 9.31660160e-02 -2.06053317e-01
7.51588821e-01 1.08575560e-01 1.76698431e-01 -7.10556924e-01
4.46894258e-01 1.27215728e-01 7.42197812e-01 -6.01640821e-01
1.30477607e+00 8.51530910e-01 1.93017393e-01 -6.35496855e-01
-4.91029054e-01 -3.87607843e-01 -9.05916154e-01 -3.42978351e-02
5.66434741e-01 -7.66852379e-01 -9.61205959e-02 1.96903631e-01
-1.05865157e+00 -6.33476257e-01 -5.02108216e-01 2.48059317e-01
-8.44057083e-01 4.42310035e-01 -2.53023148e-01 -3.43758225e-01
-3.23810577e-01 -1.00636184e+00 1.53369904e+00 -2.02854887e-01
-1.20223477e-01 -7.38934517e-01 1.41303778e-01 2.06042975e-01
4.42091644e-01 5.50891817e-01 1.02475297e+00 -4.03477520e-01
-9.60977495e-01 -2.50661075e-01 -3.68122041e-01 1.52973175e-01
6.58552069e-03 -2.15948280e-02 -9.16900635e-01 -2.18569607e-01
-3.04309845e-01 -7.15898335e-01 5.74032128e-01 1.47277743e-01
1.18343186e+00 9.07641426e-02 -4.58258778e-01 6.30981982e-01
1.48062122e+00 -4.27003294e-01 8.11155558e-01 -7.26702958e-02
6.05579674e-01 6.14406407e-01 6.95561111e-01 3.07935089e-01
2.29349181e-01 6.62453473e-01 1.12286115e+00 -4.44302946e-01
-3.68009210e-01 -6.47559822e-01 -2.58665860e-01 2.79931426e-01
-5.04375622e-02 -1.33984298e-01 -1.26742387e+00 5.29769123e-01
-1.56806660e+00 -5.30941665e-01 -4.97189343e-01 2.36935306e+00
4.39093262e-01 2.81344801e-01 -8.37093517e-02 6.78505749e-02
5.38863659e-01 1.60014685e-02 -9.32605684e-01 1.67454943e-01
1.50801465e-01 8.96752954e-01 4.84662324e-01 2.29462594e-01
-8.92933846e-01 5.65822542e-01 6.38596344e+00 4.53458607e-01
-9.79010403e-01 1.61154807e-01 4.40448463e-01 -8.55923537e-03
-6.85134709e-01 1.47415161e-01 -1.47111759e-01 7.25297704e-02
4.74984914e-01 2.59206414e-01 3.26421738e-01 7.96420395e-01
1.89843938e-01 -1.02605812e-01 -1.42654955e+00 1.02219868e+00
-4.59325910e-01 -1.34263146e+00 -1.52360126e-01 1.65521175e-01
8.05930138e-01 5.27654290e-01 -7.43427873e-02 2.56844997e-01
4.24341798e-01 -1.23175931e+00 8.90508413e-01 4.20751899e-01
1.12506390e+00 -4.23549384e-01 3.77443373e-01 5.62944531e-01
-9.88732994e-01 3.42154622e-01 -4.23611462e-01 3.12480420e-01
4.78444338e-01 6.94382966e-01 -1.09931576e+00 7.04789817e-01
5.45515776e-01 5.24694383e-01 -3.93484026e-01 1.29854977e+00
-2.17255592e-01 4.66654003e-01 -6.76213324e-01 5.45277357e-01
2.36762777e-01 -2.24124014e-01 4.30895567e-01 7.34176576e-01
5.72869658e-01 -1.77158922e-01 2.17562646e-01 1.34289157e+00
-1.18996978e-01 -1.38526589e-01 -7.92522490e-01 2.19976619e-01
5.09761930e-01 1.13609636e+00 -7.71959841e-01 9.11302716e-02
-4.34427202e-01 5.64913988e-01 2.34840304e-01 2.26193205e-01
-7.60215402e-01 1.43223479e-01 4.54720825e-01 6.26490533e-01
3.04048121e-01 -4.44715619e-01 -5.70760787e-01 -7.32444286e-01
3.17193538e-01 -4.74154115e-01 -2.24335697e-02 -9.35270905e-01
-1.35169375e+00 5.24520695e-01 4.82183062e-02 -1.56156743e+00
-2.81599104e-01 -3.29255760e-01 -8.00034463e-01 1.00995266e+00
-1.50513244e+00 -1.39090323e+00 -6.32175863e-01 6.15048349e-01
5.42936087e-01 3.00574601e-01 5.56755424e-01 3.36728632e-01
-1.44569129e-02 1.93009838e-01 -1.99902087e-01 4.95134294e-02
3.21327388e-01 -1.06395531e+00 1.13738286e+00 6.28072917e-01
2.58115619e-01 3.20302963e-01 5.94642997e-01 -7.01927900e-01
-1.45073712e+00 -1.15884197e+00 1.30628034e-01 -4.88136321e-01
2.96468675e-01 -5.10696590e-01 -9.97700334e-01 6.77252293e-01
-3.53856459e-02 3.33344907e-01 -1.09143645e-01 -2.24739820e-01
-2.27600768e-01 1.58589050e-01 -1.33256042e+00 2.31031269e-01
1.50394762e+00 -2.40042761e-01 -6.26119375e-01 3.90620500e-01
7.97428906e-01 -8.34135056e-01 -7.69568205e-01 7.31250226e-01
4.93595481e-01 -1.07264650e+00 1.23028827e+00 -4.49886024e-01
6.90725386e-01 -4.02371228e-01 -6.85360730e-02 -1.30578732e+00
3.32686722e-01 -9.18175876e-02 2.56290197e-01 8.34921837e-01
5.03259182e-01 -3.94558609e-01 1.25655556e+00 6.82453334e-01
-5.49580097e-01 -8.75688374e-01 -1.04917181e+00 -8.62897038e-01
6.14181273e-02 -7.16240704e-01 9.81373847e-01 7.98166573e-01
-7.21618176e-01 -1.63323641e-01 -1.96922049e-01 1.67030841e-01
7.98850119e-01 3.25599074e-01 1.23363018e+00 -1.41647756e+00
-2.21284747e-01 -1.54910982e-01 -1.98465481e-01 -1.09013772e+00
-7.50593282e-03 -7.97514200e-01 2.07669854e-01 -1.78532672e+00
-2.33511999e-01 -1.15633285e+00 4.51962233e-01 7.02404141e-01
4.25906152e-01 4.83208358e-01 -1.17331855e-01 4.47335005e-01
-3.56270336e-02 6.42772257e-01 1.41024232e+00 -1.91089109e-01
-5.78243062e-02 1.88901260e-01 -1.47944346e-01 6.84848487e-01
6.66928887e-01 -5.21172047e-01 -6.71308637e-01 -7.20980346e-01
4.66301471e-01 2.83342034e-01 7.28478670e-01 -1.21607614e+00
8.93637538e-02 -4.79587130e-02 3.64191234e-01 -1.15842962e+00
5.80238879e-01 -9.96274173e-01 5.93225300e-01 4.03828353e-01
-1.14699461e-01 1.62026718e-01 1.65684000e-01 6.79022789e-01
-1.78353544e-02 -2.38790751e-01 5.66378415e-01 -3.18282157e-01
-6.50877833e-01 6.96037769e-01 4.05420512e-01 -2.77255010e-03
9.97642457e-01 -5.43046892e-01 9.61819813e-02 -2.18595803e-01
-8.58882070e-01 3.81344646e-01 8.91602993e-01 2.84978300e-01
9.76212621e-01 -1.38636410e+00 -7.97684312e-01 4.27041471e-01
2.31323197e-01 8.29658210e-01 5.93173765e-02 6.17127538e-01
-8.20908487e-01 1.90626502e-01 -2.21763805e-01 -1.16539609e+00
-8.04996192e-01 4.32777524e-01 2.96719611e-01 -1.49763763e-01
-1.01775050e+00 5.08537352e-01 1.55006051e-01 -9.29176211e-01
9.04463679e-02 -6.38900697e-01 3.48367214e-01 -4.29916888e-01
5.89176863e-02 6.81642070e-02 5.34098566e-01 -3.75184268e-01
-6.97700083e-02 7.24033535e-01 2.45607033e-01 -4.15211588e-01
1.64593589e+00 6.66580647e-02 2.18579009e-01 -3.76535840e-02
8.89269233e-01 -1.79643720e-01 -1.75039756e+00 -3.59945208e-01
9.91166607e-02 -7.04065382e-01 -1.91045716e-01 -7.10661829e-01
-1.32133520e+00 1.05342090e+00 5.36462307e-01 9.37838107e-03
7.96874940e-01 1.37189433e-01 9.36527967e-01 3.57603848e-01
1.00213110e+00 -5.17480671e-01 2.14769170e-01 2.02471033e-01
1.13819075e+00 -1.27456522e+00 -1.09020315e-01 -5.47237277e-01
-1.53513461e-01 1.26621580e+00 5.20339489e-01 -3.24770808e-01
6.35024965e-01 3.42411637e-01 -2.00893760e-01 -7.25432694e-01
-3.80799413e-01 -8.89433846e-02 6.10523336e-02 7.72053540e-01
4.76421937e-02 3.23046111e-02 6.06630892e-02 1.17385864e-01
-2.17175692e-01 1.83288664e-01 6.33859098e-01 9.56154644e-01
-2.26864189e-01 -1.23161852e+00 -4.76213306e-01 6.65804505e-01
1.27263278e-01 1.18368417e-01 -2.89185703e-01 1.12377977e+00
-1.60229523e-02 3.99087459e-01 9.68462825e-02 -3.00018102e-01
6.26634955e-01 -1.55233398e-01 5.97457051e-01 -7.94944227e-01
-1.80931628e-01 -2.65697390e-01 2.19324082e-02 -7.59122074e-01
-4.53332871e-01 -8.33554626e-01 -1.28058386e+00 -3.30677181e-01
-1.04835585e-01 -2.15933084e-01 8.14646006e-01 8.86258483e-01
2.71263450e-01 7.21323937e-02 5.20751357e-01 -1.37305963e+00
-7.06515789e-01 -8.52806270e-01 -3.33672941e-01 5.44240415e-01
1.53587446e-01 -7.42568612e-01 -8.70030299e-02 -2.61541903e-01] | [8.417257308959961, -3.2561070919036865] |
77c20c87-59b3-4b1e-b94c-973f2c17af10 | gatector-a-unified-framework-for-gaze-object | 2112.03549 | null | https://arxiv.org/abs/2112.03549v3 | https://arxiv.org/pdf/2112.03549v3.pdf | GaTector: A Unified Framework for Gaze Object Prediction | Gaze object prediction is a newly proposed task that aims to discover the objects being stared at by humans. It is of great application significance but still lacks a unified solution framework. An intuitive solution is to incorporate an object detection branch into an existing gaze prediction method. However, previous gaze prediction methods usually use two different networks to extract features from scene image and head image, which would lead to heavy network architecture and prevent each branch from joint optimization. In this paper, we build a novel framework named GaTector to tackle the gaze object prediction problem in a unified way. Particularly, a specific-general-specific (SGS) feature extractor is firstly proposed to utilize a shared backbone to extract general features for both scene and head images. To better consider the specificity of inputs and tasks, SGS introduces two input-specific blocks before the shared backbone and three task-specific blocks after the shared backbone. Specifically, a novel Defocus layer is designed to generate object-specific features for the object detection task without losing information or requiring extra computations. Moreover, the energy aggregation loss is introduced to guide the gaze heatmap to concentrate on the stared box. In the end, we propose a novel wUoC metric that can reveal the difference between boxes even when they share no overlapping area. Extensive experiments on the GOO dataset verify the superiority of our method in all three tracks, i.e. object detection, gaze estimation, and gaze object prediction. | ['Zhijie Zhang', 'Xiaojuan Chen', 'Baoshan Li', 'Tao Hu', 'Binglu Wang'] | 2021-12-07 | null | http://openaccess.thecvf.com//content/CVPR2022/html/Wang_GaTector_A_Unified_Framework_for_Gaze_Object_Prediction_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Wang_GaTector_A_Unified_Framework_for_Gaze_Object_Prediction_CVPR_2022_paper.pdf | cvpr-2022-1 | ['gaze-estimation', 'eye-tracking'] | ['computer-vision', 'computer-vision'] | [ 2.97738731e-01 -8.09818283e-02 -5.03794029e-02 -6.36920333e-01
-3.05815730e-02 -1.70835685e-02 1.45788550e-01 -2.72062421e-01
-2.78025270e-01 4.20889348e-01 -1.49790933e-02 9.90300626e-02
-1.93474576e-01 -3.78301084e-01 -6.61253631e-01 -1.07987380e+00
5.47918260e-01 -2.78170913e-01 5.10884523e-01 -5.21251708e-02
4.19871747e-01 2.96998508e-02 -1.82552254e+00 -1.38514191e-01
1.14966154e+00 1.39581513e+00 6.73268139e-01 6.45522773e-02
4.89714481e-02 5.97102523e-01 -3.10816258e-01 -1.37391478e-01
1.35351727e-02 -6.06465161e-01 -3.17448676e-01 5.99727444e-02
5.13351619e-01 -3.07931930e-01 4.71710674e-02 1.21270275e+00
7.26141930e-01 1.14443764e-01 5.53820133e-01 -1.59043705e+00
-3.88257295e-01 3.05570155e-01 -1.00011730e+00 3.62683117e-01
1.35800928e-01 4.50973988e-01 9.43643391e-01 -7.12349892e-01
1.82876378e-01 1.05169034e+00 2.53612727e-01 7.10198343e-01
-9.79833305e-01 -1.02642608e+00 5.01816988e-01 5.03660500e-01
-1.38439643e+00 -4.57095176e-01 1.07431448e+00 -2.67994672e-01
4.00165468e-01 2.64893770e-01 5.18790781e-01 7.83926427e-01
8.06783661e-02 1.10757399e+00 1.11038196e+00 -1.81718677e-01
-4.40770648e-02 1.78414345e-01 2.84162581e-01 7.88344681e-01
2.74375528e-01 2.80559640e-02 -8.02120566e-01 4.47129726e-01
3.26785415e-01 2.54163206e-01 -8.43594074e-01 -5.97532034e-01
-1.15534627e+00 5.76834679e-01 9.68896687e-01 9.70668495e-02
-9.93329436e-02 -1.14776315e-02 7.21310079e-02 -2.00079218e-01
3.43218654e-01 7.44200349e-02 -3.69262904e-01 1.24446988e-01
-8.79467130e-01 2.39537582e-01 2.37083718e-01 8.91547263e-01
9.57723796e-01 -2.92740434e-01 -4.39271241e-01 6.60749495e-01
6.98442042e-01 3.73282045e-01 4.56783533e-01 -3.84396613e-01
5.43254972e-01 9.45202827e-01 -1.20914914e-02 -1.15244210e+00
-7.49934256e-01 -5.90227187e-01 -5.39018810e-01 1.97111547e-01
2.98810363e-01 -1.48199409e-01 -8.85911047e-01 1.92905974e+00
6.82466567e-01 3.27287674e-01 -3.12473834e-01 1.50666225e+00
1.03783524e+00 4.33847010e-01 7.92167801e-03 -3.59032661e-01
1.56452501e+00 -1.24107671e+00 -7.68627286e-01 -2.70381242e-01
5.67048073e-01 -5.52759469e-01 1.14950156e+00 1.77122653e-01
-9.88618195e-01 -5.22235930e-01 -1.08049726e+00 -3.75312299e-01
-7.93427452e-02 2.48023808e-01 5.46756804e-01 6.15761399e-01
-7.90697515e-01 8.63987878e-02 -5.50729275e-01 -2.53807485e-01
5.81227303e-01 5.75569391e-01 1.40023649e-01 8.53683352e-02
-8.44163060e-01 6.29083633e-01 4.87493783e-01 3.98636162e-01
-6.11211956e-01 -4.71388876e-01 -8.42275798e-01 4.87439245e-01
6.36645854e-01 -6.10087574e-01 1.13685465e+00 -1.06210208e+00
-1.48382664e+00 4.83227760e-01 -5.76272786e-01 -2.41198558e-02
1.25112280e-01 -8.25106055e-02 -4.20927584e-01 -8.38996693e-02
-1.01456314e-01 8.72335017e-01 1.23429716e+00 -1.11255515e+00
-1.03339815e+00 -5.59412062e-01 -6.01162994e-03 5.85255861e-01
-5.11019289e-01 1.96524993e-01 -6.02573156e-01 -4.88068581e-01
3.12742501e-01 -7.61641681e-01 2.53072530e-01 3.13145891e-02
-5.20578921e-01 -5.78195632e-01 8.31505656e-01 -3.46330583e-01
1.55250573e+00 -2.15695286e+00 2.26344258e-01 5.52360564e-02
6.86016619e-01 2.16038208e-02 8.83851573e-02 -3.31831843e-01
-1.37019843e-01 -1.84811845e-01 -1.55507162e-01 -4.42448467e-01
-9.53188241e-02 -2.14828774e-01 -9.88928005e-02 4.30659145e-01
2.44657725e-01 8.90506089e-01 -8.14737141e-01 -6.76791072e-01
1.30760306e-02 3.26001197e-01 -6.47097111e-01 1.88786387e-01
-1.45351544e-01 5.38489103e-01 -5.63115835e-01 6.41427815e-01
9.16906953e-01 -4.53360319e-01 -2.76590973e-01 -3.67154300e-01
-3.75194997e-01 2.97742456e-01 -8.79726887e-01 1.60293102e+00
-1.53516114e-01 5.25964379e-01 -3.34500745e-02 -7.17553198e-01
8.85106683e-01 -4.40190956e-02 1.44193202e-01 -7.43066609e-01
5.11140466e-01 1.09717712e-01 3.33204299e-01 -6.60273194e-01
4.04207230e-01 3.51919904e-02 1.92722321e-01 4.88268822e-01
8.19426104e-02 4.31369543e-01 -7.72029385e-02 -2.87212372e-01
5.37947595e-01 2.23828077e-01 2.52246857e-02 -2.40188330e-01
9.06191468e-01 -4.19765115e-01 8.13375175e-01 1.27999142e-01
-5.05843699e-01 7.00860918e-01 4.99404788e-01 -2.72325575e-01
-4.04885828e-01 -5.11449337e-01 -9.95015278e-02 1.26044309e+00
9.53844130e-01 -2.64399379e-01 -9.42320824e-01 -9.91426349e-01
-2.93828070e-01 5.73045492e-01 -8.17573786e-01 -3.60423058e-01
-5.26087999e-01 -8.23427498e-01 -7.55123422e-02 3.17159235e-01
5.86699724e-01 -1.17895377e+00 -9.39709008e-01 -3.39220196e-01
-2.12680459e-01 -7.77670681e-01 -8.88412237e-01 5.47497123e-02
-5.57865024e-01 -1.02417839e+00 -9.00658131e-01 -8.00245762e-01
7.95638382e-01 6.32687688e-01 6.07262254e-01 1.68052733e-01
7.21086636e-02 -1.51326150e-01 -2.07529947e-01 -6.72839224e-01
3.87724638e-01 3.95426542e-01 3.49647808e-03 5.44773638e-01
7.13397980e-01 -2.99555510e-01 -1.03077614e+00 4.85992700e-01
-5.53947270e-01 2.85331279e-01 5.91741264e-01 7.54085243e-01
3.15974683e-01 -2.18167767e-01 3.45402032e-01 -2.65237808e-01
5.07215381e-01 -5.23932099e-01 -6.82835102e-01 4.30720985e-01
-7.23670065e-01 1.25035420e-01 2.61693895e-01 -2.75505573e-01
-1.11962748e+00 5.10618165e-02 2.26685703e-01 -6.60339236e-01
-9.38619375e-02 2.27127910e-01 -5.05411625e-01 -1.21898428e-01
3.51761907e-01 3.55121464e-01 -1.46519607e-02 -5.13750196e-01
-7.30182156e-02 7.07813144e-01 1.54958457e-01 7.25738937e-03
7.68989742e-01 3.15084189e-01 4.29744981e-02 -3.29752177e-01
-9.44683611e-01 -3.89016300e-01 -4.34797913e-01 -3.22451800e-01
1.01863086e+00 -7.25764573e-01 -1.40779662e+00 6.26201808e-01
-1.10949230e+00 5.32120615e-02 2.49715641e-01 5.68001211e-01
-2.69869238e-01 6.69948012e-02 2.25198880e-01 -7.65151024e-01
-4.95165050e-01 -1.41306186e+00 1.19286108e+00 9.89786267e-01
3.18316549e-01 -4.93158549e-01 -1.78323701e-01 3.39395344e-01
2.43264273e-01 -2.03210413e-01 6.07632458e-01 -5.02963960e-01
-8.55692804e-01 3.50821346e-01 -7.34548151e-01 1.93188518e-01
1.59878731e-01 -2.64435977e-01 -1.08308148e+00 -2.15185508e-01
2.49235481e-01 -2.33152196e-01 9.95710731e-01 3.31877738e-01
1.35564351e+00 -1.19512342e-01 -6.71134353e-01 9.27961648e-01
1.17663944e+00 2.53923446e-01 3.08635920e-01 2.90181726e-01
9.47916448e-01 8.16939414e-01 6.80257380e-01 1.60381228e-01
6.14250243e-01 8.22719514e-01 6.98861122e-01 8.37876126e-02
-6.34263968e-03 -3.77455503e-02 3.39373201e-01 5.45004964e-01
5.84052131e-03 -3.70089710e-01 -8.61204863e-01 3.64979386e-01
-1.77100384e+00 -7.50614285e-01 -1.61292270e-01 2.07962513e+00
6.88205600e-01 1.42926455e-01 5.47222272e-02 -9.15103406e-02
8.96660984e-01 2.54923463e-01 -8.90336633e-01 5.37476502e-02
1.72985777e-01 -1.19692236e-01 1.93155110e-01 3.71320397e-02
-1.03628850e+00 7.01871216e-01 4.49002266e+00 9.04104829e-01
-1.48637283e+00 1.70709670e-01 3.51554990e-01 -5.13902903e-01
-1.22463800e-01 -1.14474788e-01 -1.24150538e+00 9.76179719e-01
3.33107501e-01 9.24176052e-02 3.79563838e-01 7.10660875e-01
1.86060563e-01 -1.53752208e-01 -9.67274487e-01 1.26089501e+00
4.18719113e-01 -8.47123623e-01 -2.11421520e-01 -9.13911164e-02
2.62808651e-01 -1.67237613e-02 2.37926677e-01 9.99865159e-02
-4.64589864e-01 -7.52035737e-01 8.13613534e-01 7.54601538e-01
6.56325281e-01 -6.58152759e-01 5.42651832e-01 4.44767177e-01
-1.09275723e+00 -3.08665007e-01 -2.60225147e-01 3.34882811e-02
1.09268829e-01 9.81750637e-02 -6.37505054e-01 5.30889213e-01
1.00560975e+00 8.77211869e-01 -8.75018120e-01 1.48540175e+00
-2.49839604e-01 4.49032962e-01 -2.89560944e-01 -3.43511879e-01
1.87181890e-01 -1.30325288e-01 7.14106083e-01 5.54768085e-01
3.27074289e-01 8.32628831e-02 -1.95275486e-01 1.08061218e+00
-3.16445902e-02 5.56128994e-02 -1.35430232e-01 5.29890776e-01
3.21435869e-01 1.35790265e+00 -4.80452597e-01 8.85368809e-02
-4.93607461e-01 7.58861423e-01 3.64574343e-01 4.23896283e-01
-1.01935756e+00 -7.84618378e-01 5.45481324e-01 1.40749559e-01
5.44078588e-01 2.81941354e-01 -3.59471321e-01 -1.27145123e+00
2.59656459e-01 -5.54961085e-01 2.61396915e-02 -9.78217900e-01
-9.37275350e-01 5.42263150e-01 -4.88430373e-02 -1.47637200e+00
1.77054092e-01 -4.73714918e-01 -8.31334233e-01 1.10027993e+00
-1.83931875e+00 -1.20985937e+00 -9.31420624e-01 7.69091606e-01
4.11131650e-01 -2.71974113e-02 1.24846444e-01 2.64862478e-01
-1.24055696e+00 9.62574542e-01 -1.44188374e-01 -1.55445844e-01
7.73452103e-01 -9.41077352e-01 -8.71321112e-02 8.90042424e-01
-2.06524163e-01 6.46088481e-01 4.83590871e-01 -3.77055258e-01
-9.14639950e-01 -8.93685162e-01 7.59225011e-01 -3.10523689e-01
2.93124199e-01 -5.04981577e-01 -1.02056420e+00 5.11632919e-01
2.20738336e-01 -1.34962397e-02 3.11539203e-01 8.03340375e-02
2.42747869e-02 -4.83320832e-01 -7.28919804e-01 6.87332213e-01
1.16841197e+00 -1.41665027e-01 -6.33044124e-01 9.81992707e-02
7.59841204e-01 -4.74923670e-01 -3.18931669e-01 6.32352412e-01
4.98626381e-01 -1.13944137e+00 5.47563076e-01 -2.78300852e-01
3.53985548e-01 -6.87164068e-01 3.87682050e-01 -8.70596409e-01
-1.49269179e-01 -6.54046237e-01 -4.09653246e-01 1.25310421e+00
1.79991603e-01 -6.88425362e-01 7.38443375e-01 4.38987434e-01
-2.20548645e-01 -1.15028250e+00 -6.71734631e-01 -3.86797369e-01
-5.00756443e-01 -1.43669859e-01 9.76717412e-01 6.45501733e-01
-1.27678156e-01 7.59153485e-01 -3.65108848e-01 1.01437055e-01
5.76238751e-01 3.92732143e-01 8.52849841e-01 -1.39198315e+00
1.18031114e-01 -6.34586096e-01 -2.86469579e-01 -1.48958957e+00
2.87711211e-02 -7.04111338e-01 2.49311492e-01 -1.20330071e+00
3.65496039e-01 -5.80303907e-01 -6.69187546e-01 4.50120330e-01
-7.11825967e-01 1.38061956e-01 2.81830311e-01 4.61927652e-01
-6.90727353e-01 9.00302827e-01 1.42120004e+00 2.72903647e-02
-4.62931633e-01 2.10633844e-01 -8.88084829e-01 7.75210977e-01
6.03085160e-01 -5.08819997e-01 -5.46450615e-01 -5.90793431e-01
1.88829556e-01 -3.45453590e-01 5.80678821e-01 -1.10308254e+00
5.80330551e-01 -7.83112496e-02 2.94757456e-01 -8.50346923e-01
2.50931025e-01 -7.91519225e-01 -4.37801719e-01 1.51548669e-01
-9.83613208e-02 -3.64532739e-01 5.59385680e-03 6.16102934e-01
-1.15816489e-01 -3.57173800e-01 8.18557739e-01 3.79051626e-01
-6.85645163e-01 4.86418724e-01 4.36226755e-01 -2.14343533e-01
1.21337581e+00 -4.58011389e-01 -4.92902249e-01 -3.05365375e-03
-4.22545671e-01 6.79831922e-01 5.98014951e-01 6.52006805e-01
5.41463554e-01 -1.17773259e+00 -3.78105283e-01 4.91926014e-01
3.25849801e-01 3.09068203e-01 4.42152351e-01 1.31464505e+00
-4.99801449e-02 5.00148177e-01 -3.17108184e-01 -8.23506653e-01
-1.27913582e+00 7.39888966e-01 4.79208708e-01 1.27175689e-01
-2.43396670e-01 1.12427247e+00 9.16786432e-01 -2.12400779e-02
2.45506391e-01 -4.33133513e-01 -7.23590195e-01 2.00511426e-01
5.29450715e-01 2.02112868e-01 -1.80896446e-01 -9.04128551e-01
-3.91528368e-01 8.26667905e-01 -2.34164163e-01 3.65682185e-01
1.13157547e+00 -5.74252307e-01 -1.86872751e-01 3.79970849e-01
1.06818998e+00 -1.98771134e-01 -1.51211798e+00 -3.37206721e-01
-1.87823400e-01 -4.65148360e-01 2.53798544e-01 -6.96889222e-01
-1.30991721e+00 1.25864768e+00 8.42694581e-01 3.92816588e-02
1.59222305e+00 4.80982997e-02 8.59794557e-01 1.19506404e-01
9.55978334e-02 -7.57399499e-01 -8.16995744e-04 2.33877018e-01
7.89062858e-01 -1.43616414e+00 -7.45731071e-02 -4.27463919e-01
-5.99381447e-01 8.41013253e-01 1.20850492e+00 2.58917779e-01
6.27459526e-01 -4.46176231e-01 -2.59240896e-01 -4.78969306e-01
-5.32990694e-01 -4.39919442e-01 7.68265426e-01 2.47835159e-01
4.02566567e-02 -3.02185893e-01 -4.29918170e-01 9.30078328e-01
-6.60919175e-02 4.75849696e-02 1.09942406e-01 6.66128159e-01
-5.43222725e-01 -6.45978689e-01 -2.55340844e-01 6.10494316e-01
-2.92414993e-01 -2.40267679e-01 -5.94059415e-02 4.94275749e-01
5.26674747e-01 7.33462214e-01 1.86163500e-01 -5.99394679e-01
1.69169113e-01 -7.91326631e-03 2.92907387e-01 -4.91364568e-01
-5.10050476e-01 1.36237964e-01 -5.81760585e-01 -6.57221138e-01
-6.58009410e-01 -6.83813334e-01 -1.12071037e+00 3.39134336e-02
-9.13756907e-01 -9.13637131e-02 5.05932987e-01 8.70736480e-01
5.26357353e-01 5.87912381e-01 6.90365613e-01 -9.30746078e-01
-1.97492450e-01 -1.05847383e+00 -5.74627161e-01 -5.65783028e-03
6.09179199e-01 -1.18997180e+00 -3.04238468e-01 -1.20137624e-01] | [9.832725524902344, -0.40920644998550415] |
f230dc0f-6d16-46c0-ab11-f8765a794be8 | est-ce-que-tu-me-suis-une-revue-du-suivi-de | null | null | https://aclanthology.org/2022.jeptalnrecital-recital.1 | https://aclanthology.org/2022.jeptalnrecital-recital.1.pdf | « Est-ce que tu me suis ? » : une revue du suivi de l’état du dialogue (“Do you follow me ?" : a review of dialogue state tracking ) | Tout en communiquant avec un utilisateur, un système de dialogue orienté tâche doit suivre les besoins de l’utilisateur à chaque étape selon l’historique de la conversation. Ce procédé appelé suivi de l’état du dialogue est primordial car il informe directement les actions du système. Cet article présente dans un premier temps la tâche du suivi de l’état du dialogue, les jeux de données disponibles et les approches modernes. Ensuite, compte tenu du nombre important de publications des dernières années, il vise à recenser les point saillants et les avancées des recherches. Bien que les approches neuronales aient permis des progrès notables, nous argumentons que certains aspects critiques liés aux systèmes de dialogue sont encore trop peu explorés. Pour motiver de futures études, plusieurs pistes de recherche sont proposées. | ['Léo Jacqmin'] | null | null | null | null | jep-taln-recital-2022-6 | ['dialogue-state-tracking'] | ['natural-language-processing'] | [ 7.95102268e-02 2.56087273e-01 3.11221004e-01 -4.72122014e-01
-1.07066929e-01 -1.01379919e+00 1.34378076e+00 8.59585702e-01
-4.84243482e-01 9.21589136e-01 4.15636450e-01 -1.19994819e-01
3.54545861e-01 -9.81921732e-01 -5.99635005e-01 -5.12190796e-02
3.53563279e-01 2.10935831e-01 -1.76463723e-01 -1.10477626e+00
3.85876715e-01 3.19566369e-01 -7.82765448e-01 -2.80596241e-02
5.85693002e-01 7.22395241e-01 3.52887772e-02 3.35626900e-01
-3.05684745e-01 6.55471146e-01 -4.44115460e-01 -4.90965784e-01
3.52720022e-01 -8.31855118e-01 -7.47997940e-01 -1.77096829e-01
2.08050624e-01 6.39377117e-01 -3.88509393e-01 8.01229954e-01
-1.96070686e-01 -9.25200805e-03 1.39224505e+00 -7.12858498e-01
-6.82877898e-01 6.42772615e-01 -1.98063895e-01 1.08662236e+00
7.16829002e-01 9.14865918e-03 1.10417747e+00 -7.57670045e-01
3.52309257e-01 8.55567992e-01 7.80989408e-01 4.81631428e-01
-6.64545953e-01 -1.79157943e-01 3.30048591e-01 -2.91965008e-01
-6.65399671e-01 -7.04513788e-01 -9.66619998e-02 -4.89306271e-01
1.37662083e-01 5.45977652e-01 3.59053791e-01 1.08528149e+00
7.46001959e-01 8.01300704e-01 1.30895901e+00 -1.34964705e-01
1.70609757e-01 9.45234597e-01 8.07515681e-02 6.44992828e-01
2.20602136e-02 5.76098636e-02 -2.36059338e-01 -6.77200317e-01
8.74113202e-01 2.08783850e-01 -5.50239742e-01 3.16100210e-01
-9.22138214e-01 9.97226775e-01 3.59178185e-01 8.21432054e-01
-2.08939433e-01 -1.74692884e-01 7.74216890e-01 6.19336188e-01
3.45241814e-03 8.04291308e-01 9.38091502e-02 -1.19919360e+00
-3.52891311e-02 3.03966224e-01 1.60514283e+00 6.46296740e-01
4.84975636e-01 -3.00162822e-01 7.28868604e-01 4.20817554e-01
2.61584848e-01 -1.96917847e-01 -1.52531639e-01 -4.90116239e-01
-2.23632865e-02 4.66767967e-01 4.12289985e-02 -8.10823798e-01
2.95442133e-03 -4.66686040e-01 -4.06434178e-01 3.47239934e-02
3.21006030e-01 -2.09285900e-01 3.30308437e-01 1.06732261e+00
-1.72170013e-01 -6.39259219e-01 1.02775775e-01 9.94151175e-01
-2.17345491e-01 3.32238883e-01 3.33831728e-01 -6.28397167e-01
1.41739452e+00 -8.08117688e-02 -2.92769909e-01 -2.17089150e-02
2.42511481e-01 -1.38985157e+00 6.21208966e-01 8.46038938e-01
-8.18440616e-01 -2.54479498e-01 -1.01170599e+00 7.47547075e-02
-2.61751395e-02 -1.96682140e-01 6.44538999e-02 3.26846212e-01
-7.42969573e-01 4.43583071e-01 -4.70917404e-01 -6.49322629e-01
-5.53464413e-01 4.47384417e-02 -2.68759519e-01 1.00676501e+00
-6.83440089e-01 1.35063016e+00 4.69309121e-01 6.71587348e-01
-2.69419581e-01 4.85394180e-01 -7.35304356e-01 1.09263085e-01
1.25262153e+00 -2.10450053e-01 1.78489149e+00 -1.21050513e+00
-1.37548721e+00 6.60028577e-01 -4.05047983e-01 -3.15451711e-01
9.78787363e-01 4.29569513e-01 -5.49169898e-01 1.60552442e-01
-3.57509106e-01 -8.33354741e-02 2.12480083e-01 -1.11041582e+00
-1.32465911e+00 3.69744271e-01 7.68972516e-01 2.89174974e-01
-4.40495461e-01 1.85091440e-02 -1.45896956e-01 -9.75877345e-01
-3.09825363e-03 -1.14148867e+00 3.27636510e-01 4.63495970e-01
-2.36923248e-01 -2.33998418e-01 1.23105085e+00 1.27294987e-01
8.90015125e-01 -1.92902076e+00 6.45178854e-01 -4.27663773e-01
9.65136826e-01 5.55495083e-01 -4.85551625e-01 8.22782218e-01
1.73065320e-01 1.87258556e-01 -6.30239606e-01 2.25674510e-01
7.19868839e-01 -2.90854275e-01 1.25547647e-01 1.76905990e-01
-4.11942929e-01 7.93458879e-01 -1.08827507e+00 -1.41437799e-01
-3.43310982e-01 2.29100168e-01 -2.07934439e-01 -8.48760754e-02
-4.21964616e-01 6.50741160e-01 -1.09450793e+00 -3.37967992e-01
6.53598845e-01 -5.55200517e-01 3.27726811e-01 -5.24321377e-01
-8.75146568e-01 5.93913376e-01 -9.73225534e-01 1.50652993e+00
-5.01412392e-01 2.02098310e-01 7.15639472e-01 -8.67289901e-01
1.63110936e+00 4.49904591e-01 7.76867509e-01 -4.96018492e-02
1.77185923e-01 4.70528632e-01 -2.45865192e-02 1.09715931e-01
6.03860140e-01 -4.33391601e-01 3.96295875e-01 1.59612775e+00
-6.40670657e-01 -1.31500214e-01 8.40189457e-01 -3.62143070e-01
1.15075171e+00 -1.87742133e-02 2.66793072e-01 -4.27085340e-01
1.07710505e+00 -1.92779720e-01 5.80755360e-02 9.13723279e-03
-4.03440803e-01 -2.72994608e-01 8.53051305e-01 -5.30103326e-01
-1.65975654e+00 -7.57270932e-01 -2.86617786e-01 7.34402239e-01
-6.44619584e-01 -2.67778218e-01 -3.87790471e-01 -1.36225200e+00
-1.49425223e-01 8.03866029e-01 -7.22615838e-01 1.99961945e-01
-2.87583590e-01 -4.40264553e-01 3.51561248e-01 3.98718476e-01
7.59413183e-01 -3.39729339e-01 -9.83105958e-01 3.41242164e-01
-2.02537343e-01 -7.99801409e-01 -2.59010017e-01 -4.45125192e-01
-2.89287180e-01 -9.72525239e-01 -7.85622954e-01 8.11403170e-02
3.04913193e-01 3.73461872e-01 9.93396401e-01 -4.03936297e-01
-8.39234218e-02 5.52872360e-01 -7.78015137e-01 -3.87582213e-01
-1.57421637e+00 3.75787288e-01 -8.66904035e-02 -2.32114017e-01
6.57421291e-01 -3.68703187e-01 -9.13436234e-01 6.36802673e-01
-8.61508727e-01 -1.94370240e-01 4.92114753e-01 4.22513694e-01
-3.72002393e-01 -5.65922558e-01 4.27583754e-01 -1.00384653e+00
1.12488258e+00 -5.54771245e-01 -1.37117002e-02 -1.14633106e-01
-3.28524619e-01 -8.34499300e-02 1.09716880e+00 5.75461745e-01
-1.23029041e+00 -6.05801046e-01 -3.60934764e-01 4.88710731e-01
-5.70173144e-01 8.46033871e-01 9.90350172e-02 6.93383574e-01
1.13469517e+00 6.07572794e-01 -1.19753227e-01 -6.63245499e-01
2.10906699e-01 8.45373452e-01 1.23726107e-01 -7.64352739e-01
4.24252391e-01 4.03850377e-01 -8.58661890e-01 -4.39074486e-01
1.07760280e-01 6.79666027e-02 -7.27068782e-01 -2.06467047e-01
9.19382572e-01 -5.95892310e-01 -9.25722361e-01 -2.87498217e-02
-1.21153831e+00 -3.59401047e-01 5.09930365e-02 6.68991327e-01
-2.24257246e-01 -2.61516094e-01 -9.16522205e-01 -6.18574023e-01
5.23925424e-01 -7.23245859e-01 5.82183182e-01 -4.87544499e-02
-4.56480622e-01 -1.14928865e+00 5.78499019e-01 1.29978508e-01
2.97210813e-01 -1.61228180e-01 7.62845099e-01 -8.07099976e-03
-1.10730219e+00 -1.42902270e-01 2.35869795e-01 1.21773630e-01
1.05514181e+00 -3.68681103e-02 -3.80139440e-01 -4.38716322e-01
-3.47973675e-01 4.11971658e-01 1.20543912e-01 -5.02923846e-01
4.09100592e-01 -6.50905132e-01 9.87849664e-03 -3.51161420e-01
1.47721410e+00 6.34277344e-01 1.69523343e-01 8.83153677e-02
-4.86903936e-01 9.59254622e-01 1.02602196e+00 -1.00542828e-01
4.77913707e-01 8.30820024e-01 -2.00156379e-03 6.67917848e-01
1.08178616e+00 6.54003918e-02 1.01723246e-01 1.25531399e+00
-5.66163838e-01 -1.44100368e-01 -1.05992973e+00 3.95228565e-01
-1.71973205e+00 -1.58810299e-02 -4.85472620e-01 1.92108774e+00
8.30259502e-01 2.60420859e-01 7.43574858e-01 -1.78344831e-01
5.54005921e-01 8.83997142e-01 -2.81024635e-01 -1.01390719e+00
-4.27798592e-02 -3.50217193e-01 -1.61375821e-01 7.88293123e-01
-3.25189680e-01 4.01022196e-01 5.12519693e+00 7.17094019e-02
-1.46523857e+00 3.44229676e-02 7.40815282e-01 2.75762856e-01
-7.49311864e-01 -4.90614921e-02 -4.33323443e-01 5.56908548e-01
1.23334515e+00 -6.75744712e-01 6.03434086e-01 2.97653228e-01
4.43962187e-01 -2.61338532e-01 -1.24849272e+00 5.08151948e-01
3.57887268e-01 -1.46891940e+00 -1.26810217e+00 -2.67331034e-01
-1.67587951e-01 2.43451759e-01 3.21897119e-02 6.44999564e-01
4.51539338e-01 -9.30249751e-01 5.49567759e-01 4.11580980e-01
4.34440613e-01 -6.83163822e-01 -5.60272872e-01 1.90516010e-01
-6.45207644e-01 6.11996166e-02 -1.53875545e-01 -1.56587631e-01
6.62745714e-01 -1.39614746e-01 -5.88742733e-01 9.87698317e-01
-3.62449467e-01 6.91854119e-01 -2.15825573e-01 9.75951076e-01
-1.62301853e-01 5.78383863e-01 -3.64011526e-01 -7.56242394e-01
6.80574238e-01 -6.35620117e-01 7.63895810e-01 1.08166659e+00
3.37124437e-01 8.54431331e-01 -5.00847876e-01 1.00235629e+00
3.49976182e-01 7.25295305e-01 1.90841649e-02 -6.28855884e-01
-1.18578905e-02 4.78217602e-01 -1.07194853e+00 2.83888370e-01
-5.00591099e-01 7.08208501e-01 8.68364945e-02 -1.22951828e-01
-2.57628292e-01 -5.90040028e-01 4.01256591e-01 -1.78180531e-01
-3.97529870e-01 -8.92565668e-01 -1.15927093e-01 -1.16859460e+00
-6.48754478e-01 -1.43498123e+00 -3.26043457e-01 -5.10506332e-01
-8.13146412e-01 7.74035513e-01 2.40489200e-01 -5.63974321e-01
-1.02975976e+00 -6.29483610e-02 -5.38578928e-01 5.37512064e-01
-6.80038989e-01 -3.59229296e-01 5.60984731e-01 -4.20470722e-02
5.16340196e-01 -1.39902964e-01 1.51978374e+00 2.85307646e-01
-7.97613621e-01 -3.77802998e-01 1.99479282e-01 -6.52331561e-02
7.16546774e-01 -1.05771434e+00 9.56530392e-01 2.43378386e-01
4.46266353e-01 9.49649692e-01 9.26656783e-01 -5.73063076e-01
-1.08525062e+00 -4.45923775e-01 9.18512404e-01 -3.84475917e-01
1.02714622e+00 -3.36290240e-01 -8.99166644e-01 1.88229060e+00
7.25836217e-01 -9.67689157e-01 -1.23984367e-02 4.09397483e-01
2.45301381e-01 -1.67723775e-01 -2.70702124e-01 8.25923979e-01
-2.64046997e-01 -4.49416131e-01 -4.80382204e-01 -1.93237528e-01
-1.15419358e-01 -8.03547859e-01 -1.03889716e+00 2.85005480e-01
-9.63672847e-02 -1.59496343e+00 -1.72588885e-01 -2.18863469e-02
4.82158959e-01 -3.13009262e-01 4.42536324e-01 -9.09165561e-01
6.94068491e-01 -9.84883606e-01 1.26265928e-01 1.04682279e+00
2.40943238e-01 -7.40476727e-01 2.97804445e-01 1.80994257e-01
2.78451502e-01 -9.45057273e-02 -1.04321146e+00 2.42753610e-01
4.59174395e-01 -8.07266653e-01 5.91864884e-01 9.64268625e-01
2.22119555e-01 6.88670993e-01 5.58140501e-02 -5.29388189e-01
-1.15429074e-01 -1.04599208e-01 9.05665636e-01 -8.87135327e-01
-5.61231434e-01 -3.56477082e-01 -1.46338344e-01 -1.41071606e+00
-3.10560942e-01 -5.24614036e-01 -9.90106285e-01 -1.37761140e+00
6.43876567e-02 -1.25241414e-01 -1.92930356e-01 2.41678223e-01
3.35416079e-01 1.30658522e-01 1.28280067e+00 2.31829479e-01
-5.95633805e-01 3.52136701e-01 1.21809733e+00 1.75752446e-01
1.41817793e-01 8.85586217e-02 -8.09450507e-01 8.49476457e-01
4.45208758e-01 -2.51144648e-01 -3.80383998e-01 4.04773559e-03
8.00722778e-01 -1.23360254e-01 7.55585078e-03 -6.28103018e-01
4.00674731e-01 -5.71326852e-01 2.33824566e-01 -5.46260357e-01
1.07990801e+00 -4.97758776e-01 2.72525668e-01 3.40828039e-02
-6.30481243e-01 7.66424881e-03 3.55931491e-01 1.94496095e-01
-3.73621762e-01 -2.72491664e-01 7.57688820e-01 -3.89716983e-01
-7.05412567e-01 6.09687030e-01 -9.74956691e-01 3.82598311e-01
9.07558382e-01 -3.80459368e-01 7.39621697e-03 -2.53850102e-01
-7.69989491e-01 2.54210174e-01 7.97222018e-01 3.86581481e-01
1.88267618e-01 -5.68492711e-01 -1.50200665e+00 -1.84033006e-01
-4.01287287e-01 -4.08303231e-01 1.08838844e+00 5.26312172e-01
-1.00437355e+00 5.87527752e-01 -4.50304598e-01 -9.05679643e-01
-1.68874431e+00 1.73726499e-01 4.35233265e-01 7.94833660e-01
-1.08817410e+00 7.20356345e-01 -2.43835285e-01 7.31948810e-03
-6.02266729e-01 1.00805417e-01 2.80311406e-01 -2.87076712e-01
9.52311382e-02 6.10603809e-01 -6.67732656e-01 -4.48362559e-01
2.45386306e-02 3.15413177e-01 -5.32365203e-01 -8.54861557e-01
1.11803830e+00 -6.02468848e-01 -8.01633179e-01 9.51942444e-01
1.20100474e+00 7.31042802e-01 -1.78813100e+00 8.45678627e-01
-2.78839827e-01 -1.88930079e-01 -4.54506278e-01 -1.18037927e+00
-5.69945514e-01 7.78217554e-01 4.37426984e-01 6.72841251e-01
5.28179109e-01 -3.66578728e-01 4.95316625e-01 3.67093146e-01
5.43580234e-01 -8.87191355e-01 -5.34992039e-01 8.25327575e-01
1.33100641e+00 -1.13957870e+00 -3.20535719e-01 -1.57553758e-02
-6.02352321e-01 1.55599415e+00 -2.06518278e-01 3.10264647e-01
4.82932568e-01 2.18763798e-02 7.14145124e-01 -2.80434519e-01
-7.87140548e-01 1.86930373e-01 1.62132174e-01 9.83207077e-02
1.10256433e+00 -3.65846157e-01 -1.49832797e+00 1.56743601e-01
-8.25224519e-01 2.89171308e-01 9.21538115e-01 1.02268469e+00
-9.89293866e-03 -1.35076153e+00 -6.17140055e-01 -1.60625353e-01
-3.09124589e-01 1.33521706e-01 -8.35295141e-01 8.05945754e-01
-1.27263637e-02 1.12508047e+00 9.12373811e-02 -3.73472333e-01
7.78859377e-01 -3.20311636e-01 1.72581673e-01 -7.40747809e-01
-1.32226861e+00 3.70040894e-01 2.40884289e-01 6.79479912e-02
-8.53461146e-01 -5.92501462e-01 -3.47360283e-01 -1.28921378e+00
-3.33340555e-01 1.12199104e+00 6.44261420e-01 1.40392566e+00
8.12648758e-02 5.38280129e-01 2.36256495e-01 -3.25748801e-01
-5.47373533e-01 -1.04653108e+00 -1.12634706e+00 -3.32549989e-01
4.88396794e-01 -2.89409995e-01 1.96317002e-01 -4.72641796e-01] | [14.102492332458496, 13.31672191619873] |
ebb3a513-9916-4e5f-82f4-e94b0cd21a01 | online-robust-mpc-based-emergency-maneuvering | 2109.11959 | null | https://arxiv.org/abs/2109.11959v2 | https://arxiv.org/pdf/2109.11959v2.pdf | Online Robust MPC based Emergency Maneuvering System for Autonomous Vehicles | Nonlinear Robust Model Predictive Control (RMPC) provides a very promising solution to the problem of automatic emergency maneuvering, which is capable of handling multiple possibly conflicting objectives of robustness and performance. Even though RMPC gives a suboptimal solution, the key challenge in real-time implementation is that it is computationally very demanding. In this paper a real-time capable robust tube MPC based framework for steering control during emergency obstacle avoidance maneuver is presented. The novelty of this framework lies in the robust integration of path planning and path following tasks of autonomous vehicles. A simulation study showcases the robust performance improvements due to the proposed strategy over a non-robust MPC in different extreme maneuvering scenarios. | ['Vivek Bithar', 'Shawn Midlam-Mohler', 'Punit Tulpule'] | 2021-09-24 | null | null | null | null | ['steering-control'] | ['computer-vision'] | [ 3.58357638e-01 2.67630160e-01 7.57699609e-02 6.78289682e-02
-7.26550937e-01 -5.87528288e-01 8.32313657e-01 -1.00786239e-01
-4.73259985e-01 1.14092028e+00 -3.07702690e-01 -8.14432859e-01
-7.53131211e-01 -6.79420829e-01 -4.18128014e-01 -1.04526246e+00
-2.89431512e-01 7.39993334e-01 5.11572361e-01 -9.96316433e-01
2.61893272e-01 1.10920846e+00 -1.42690849e+00 -3.90308619e-01
1.05767512e+00 7.76043475e-01 3.33211243e-01 9.97459650e-01
5.91192901e-01 1.74542278e-01 2.97616661e-01 2.69829065e-01
5.19840717e-01 2.28546649e-01 -3.97424370e-01 3.64541262e-01
-1.99378818e-01 5.04588112e-02 -1.50512248e-01 8.61491024e-01
3.91185760e-01 8.16294372e-01 7.90688097e-01 -1.50417840e+00
7.84565091e-01 -2.93366164e-01 1.08250296e-02 -8.84521008e-03
2.64820866e-02 4.73383397e-01 4.34306771e-01 -6.68322921e-01
4.46931124e-01 1.13759482e+00 3.38729650e-01 3.19369137e-01
-1.15334594e+00 -8.92914832e-03 -1.28396317e-01 3.67302060e-01
-1.41985416e+00 -4.87979650e-01 1.64018974e-01 -4.12630349e-01
8.91362250e-01 5.82724929e-01 2.86357343e-01 3.49586993e-01
7.47761786e-01 1.76629737e-01 9.60107446e-01 1.08339064e-01
3.07696044e-01 2.32080862e-01 -1.48816600e-01 6.73631787e-01
4.98225898e-01 7.15346396e-01 4.46197450e-01 -6.58519268e-02
2.87741601e-01 -2.99070597e-01 -3.34525168e-01 -5.89997113e-01
-9.92889166e-01 8.32273602e-01 3.71394992e-01 -1.91022813e-01
-6.12621009e-01 -2.34227423e-02 3.59854072e-01 3.05852562e-01
-1.04934022e-01 3.38388562e-01 -6.78797781e-01 -2.04964474e-01
-4.88865584e-01 9.16724384e-01 8.43592465e-01 1.13192499e+00
1.79053679e-01 8.16584706e-01 1.22006811e-01 1.94988310e-01
3.29508483e-01 7.07591712e-01 -1.65660098e-01 -1.24842739e+00
6.82072163e-01 3.87525231e-01 9.23467398e-01 -1.17367089e+00
-8.95123005e-01 -1.70658246e-01 -6.41204834e-01 9.81638730e-01
2.72222757e-01 -5.53803027e-01 -5.87616444e-01 1.31058443e+00
4.59842086e-01 -3.15082699e-01 5.89597464e-01 8.05673182e-01
1.39938638e-01 9.88915741e-01 -1.53342605e-01 -5.39603949e-01
7.97958374e-01 -7.72402704e-01 -8.34005654e-01 -2.16587350e-01
2.50573248e-01 -4.97147948e-01 2.80705184e-01 6.18495107e-01
-8.81347954e-01 -2.88441330e-01 -1.05715454e+00 4.90881473e-01
-4.01769072e-01 -6.79419637e-02 -1.37955248e-01 4.02894139e-01
-9.63796496e-01 6.55574441e-01 -5.74532866e-01 -5.50564289e-01
-3.03289413e-01 6.97169304e-01 -5.43744445e-01 1.34641334e-01
-9.37548220e-01 1.63270521e+00 7.82701135e-01 2.37557158e-01
-8.12984228e-01 -3.43310475e-01 -8.41329396e-01 -1.21873446e-01
8.34979117e-01 -2.27460608e-01 9.51318860e-01 -2.93417484e-01
-1.87044811e+00 6.99042976e-02 -2.64046460e-01 -7.52358913e-01
9.97079968e-01 -4.57171053e-02 -2.79731035e-01 1.50646986e-02
-2.97366202e-01 3.52308989e-01 7.38412976e-01 -1.41024041e+00
-9.24260736e-01 7.88158178e-02 -1.62205547e-01 2.93211699e-01
4.84148353e-01 -2.94931591e-01 7.24779516e-02 -1.03090949e-01
3.10104489e-02 -1.43819857e+00 -9.58339214e-01 -2.64398515e-01
-2.29004592e-01 1.96561530e-01 1.12105918e+00 -4.76701289e-01
9.14863944e-01 -1.55585563e+00 6.18236780e-01 3.78413528e-01
-4.76359606e-01 5.16948700e-01 8.23785737e-03 9.89646018e-01
1.07124411e-01 -2.67953992e-01 -4.61355001e-01 2.63832361e-01
-4.11070809e-02 7.08348453e-01 -7.64735281e-01 7.14885592e-01
2.60487884e-01 4.96248811e-01 -7.40160882e-01 -1.72935501e-01
6.20253563e-01 2.90659636e-01 -2.03704506e-01 1.18788160e-01
-1.87337875e-01 7.52119422e-01 -4.24831659e-01 4.95707691e-01
6.35189056e-01 7.89886534e-01 -7.63368011e-02 1.69438243e-01
-1.00378227e+00 -5.52690029e-01 -1.56131828e+00 7.12358594e-01
-2.76198447e-01 6.89717293e-01 8.74080837e-01 -1.06925535e+00
8.55160356e-01 3.48582834e-01 3.98856610e-01 -2.77269423e-01
4.67004865e-01 5.28742194e-01 -1.25610933e-01 -7.46621311e-01
1.02091503e+00 -3.05400997e-01 -4.94136438e-02 -3.25141191e-01
-5.61087430e-01 -7.92308569e-01 1.67413220e-01 -2.09313080e-01
8.69154513e-01 9.19027403e-02 6.33874834e-01 -7.39919782e-01
1.29167485e+00 6.14645362e-01 7.41397321e-01 1.93398133e-01
-5.75694501e-01 2.51693815e-01 3.67730349e-01 -1.81778878e-01
-1.10973787e+00 -8.29432309e-01 -1.15430102e-01 6.16002381e-01
2.41071790e-01 1.98950246e-01 -1.66771501e-01 -1.17085353e-01
1.42295331e-01 1.10123765e+00 -2.74842560e-01 -4.87292418e-03
-1.07484579e+00 -6.68689728e-01 1.35704160e-01 5.77583723e-03
2.23161895e-02 -4.22151566e-01 -8.87679577e-01 5.05551636e-01
8.37887898e-02 -1.15736878e+00 -2.75195036e-02 1.25813141e-01
-8.82235050e-01 -1.25434661e+00 -1.39491916e-01 -6.63705587e-01
6.09916449e-01 5.58810651e-01 2.02626541e-01 -9.29649696e-02
-2.08408281e-01 4.25852895e-01 -2.07402587e-01 -4.97392148e-01
-7.30638623e-01 -4.97429132e-01 3.60173285e-01 1.05276264e-01
-6.31001413e-01 -1.59397319e-01 -6.13089725e-02 7.15106487e-01
-6.82561576e-01 -2.09664688e-01 3.68157506e-01 6.66128039e-01
6.44368351e-01 5.23825765e-01 6.31876886e-01 -3.45968693e-01
7.27475107e-01 -6.10101223e-01 -1.31519032e+00 -2.73169458e-01
-6.71013117e-01 -9.05218720e-02 9.54281211e-01 -1.41791096e-02
-1.01326346e+00 4.60907489e-01 -1.82664439e-01 -1.63475014e-02
-1.96353748e-01 3.42232883e-01 -5.87592088e-02 -6.27173305e-01
4.04474080e-01 2.83964932e-01 4.60605443e-01 -1.03239410e-01
2.87810773e-01 2.14286074e-01 7.03365684e-01 -1.34218156e-01
1.44346452e+00 5.45602024e-01 1.08885300e+00 -1.26963031e+00
8.38324055e-02 -6.25053763e-01 -6.86259329e-01 -5.67927539e-01
7.60898829e-01 -5.47352612e-01 -7.22491086e-01 2.06772983e-03
-9.22188580e-01 -3.45762819e-01 -2.83035457e-01 2.19306663e-01
-1.14368510e+00 4.41900045e-01 8.15884471e-02 -1.32483435e+00
-5.03859296e-02 -1.23411667e+00 3.09268266e-01 3.11210930e-01
1.31592661e-01 -1.11112726e+00 -2.95276158e-02 -2.52998304e-02
8.74820650e-01 9.93286669e-01 5.69246590e-01 -4.14868683e-01
-4.79954898e-01 -8.42601299e-01 1.72066569e-01 8.77392367e-02
-3.87282073e-01 2.25070357e-01 -3.26987028e-01 -7.07792044e-01
2.98270434e-02 1.22776423e-02 3.26588571e-01 2.45850340e-01
1.29693002e-01 -6.11134052e-01 -5.65364540e-01 2.50885159e-01
1.84094679e+00 4.85702962e-01 2.19324425e-01 5.87869883e-01
2.72603959e-01 1.09371591e+00 1.48978388e+00 3.23886871e-01
3.12377214e-01 9.10551369e-01 1.02625203e+00 2.55271286e-01
6.06539428e-01 1.93196088e-01 5.66313326e-01 2.87555546e-01
-1.77098066e-01 -2.62492746e-01 -9.38550711e-01 6.65126681e-01
-2.14914608e+00 -1.03247535e+00 -8.74312878e-01 2.36761689e+00
-1.38669973e-02 7.48732314e-02 -8.09703693e-02 3.19981575e-01
5.66127419e-01 -1.43335953e-01 -8.14338624e-02 -9.66078043e-01
4.62763524e-03 -5.35812736e-01 1.16303849e+00 1.13145041e+00
-1.19094992e+00 5.02146900e-01 6.06551695e+00 7.59575248e-01
-8.91637802e-01 -1.44350722e-01 -2.47271821e-01 1.83199793e-01
2.44728908e-01 1.44209802e-01 -7.42691159e-01 2.18146309e-01
1.37189424e+00 -6.17264271e-01 2.50766486e-01 9.15965140e-01
1.14852715e+00 -5.14353514e-01 -3.94693881e-01 1.85291231e-01
-2.72463918e-01 -1.06945574e+00 -2.69582987e-01 1.75735250e-01
9.15285051e-01 3.45181935e-02 -2.86412388e-02 1.50541976e-01
1.16504893e-01 -8.32477629e-01 6.52665913e-01 6.11085534e-01
2.33283639e-01 -1.26128209e+00 9.01067317e-01 8.34842920e-01
-1.18708944e+00 -5.69091618e-01 -3.68222713e-01 -1.47362173e-01
9.05946672e-01 -1.36620015e-01 -1.06220472e+00 8.52846205e-01
-3.88845541e-02 1.96014956e-01 -4.34351675e-02 1.52784503e+00
-2.21043855e-01 1.33864045e-01 -4.04637575e-01 -1.14974953e-01
8.37404728e-01 -5.10026753e-01 1.52049541e+00 1.24536490e+00
2.85444677e-01 3.28078389e-01 4.39919293e-01 3.69424194e-01
1.23761642e+00 3.38493362e-02 -1.14137185e+00 4.93647635e-01
1.88537203e-02 1.34564400e+00 -3.64685506e-01 -1.52149588e-01
4.90440838e-02 3.41331095e-01 -2.28983566e-01 2.97318280e-01
-8.47090304e-01 -5.29603779e-01 7.01621652e-01 3.39494079e-01
2.36970901e-01 -8.15447628e-01 -2.19980046e-01 -3.89742136e-01
-3.59825730e-01 -2.80072451e-01 9.54790972e-03 -3.95464003e-01
-4.93919253e-01 5.24729550e-01 3.79711211e-01 -1.80335498e+00
-4.92355019e-01 -7.49120116e-01 -8.39205921e-01 8.98653567e-01
-1.73924017e+00 -9.49813724e-01 -8.74489918e-02 3.10161740e-01
8.97579849e-01 -2.60423988e-01 6.14848614e-01 -7.80250877e-02
-4.71076906e-01 -2.04151958e-01 6.26237512e-01 -1.00082016e+00
1.80739775e-01 -1.16496968e+00 -1.71162829e-01 1.23936415e+00
-1.18731785e+00 2.13574827e-01 1.66571939e+00 -6.25666142e-01
-1.95910740e+00 -1.44001663e+00 8.63122463e-01 -1.16995469e-01
8.30166638e-01 4.87520784e-01 -7.57635772e-01 3.39575887e-01
1.05460659e-01 -4.25849229e-01 -1.83654815e-01 -9.66177762e-01
7.65152454e-01 1.04879081e-01 -1.26310158e+00 8.68280768e-01
2.03908637e-01 3.55546951e-01 -4.90346402e-01 3.40873599e-01
6.27953708e-01 -4.54555750e-01 -5.22058010e-01 8.18947852e-01
1.09090172e-01 -5.96009672e-01 7.84686983e-01 -4.20737267e-01
-4.65971500e-01 -8.24874997e-01 -2.63371199e-01 -1.22678220e+00
-1.25338808e-01 -1.12872183e+00 6.89027980e-02 7.75115967e-01
4.96998221e-01 -8.69235814e-01 5.07432580e-01 6.82783604e-01
-4.85439569e-01 -6.86038673e-01 -1.58240509e+00 -1.09177136e+00
5.94355911e-02 -5.84933579e-01 -3.32361728e-01 1.16271056e-01
3.18743028e-02 -1.36919722e-01 -5.92438996e-01 7.87889123e-01
6.01092041e-01 -2.32989311e-01 8.80697250e-01 -1.14045393e+00
2.28378266e-01 -3.24481010e-01 -3.81586790e-01 -5.48780560e-01
1.22928888e-01 -4.75688487e-01 6.25543714e-01 -1.65312588e+00
-6.93980277e-01 -1.79763466e-01 3.10476482e-01 -1.58582702e-01
1.55470232e-02 -1.86791584e-01 2.24067137e-01 -1.22555479e-01
-2.15871364e-01 6.68378294e-01 9.81240869e-01 2.98392102e-02
-4.18509543e-01 7.48216093e-01 1.35315970e-01 7.96829998e-01
1.15576375e+00 -3.14443529e-01 -6.01450562e-01 1.04506128e-01
-1.15179278e-01 7.96221197e-01 1.95012823e-01 -1.16989565e+00
4.74254221e-01 -7.85830021e-01 -6.72681272e-01 -9.49751556e-01
4.76932973e-01 -1.32855892e+00 2.58667558e-01 1.18918502e+00
1.65301234e-01 4.05603856e-01 5.33621430e-01 1.09905756e+00
-2.28558689e-01 -3.70421648e-01 1.31795228e+00 1.43545106e-01
-1.00494480e+00 1.37266845e-01 -1.18312490e+00 -3.80728453e-01
1.64667356e+00 -3.86362761e-01 2.83154510e-02 -3.67861211e-01
-7.22620606e-01 7.26107001e-01 4.75114495e-01 3.38384658e-01
7.02880621e-01 -7.53022492e-01 -6.02656186e-01 -8.30769837e-02
-2.65853882e-01 -2.78767973e-01 2.62882560e-01 1.20802701e+00
-8.42294812e-01 1.04105091e+00 -4.27586496e-01 -3.94723415e-01
-1.39230704e+00 6.61995888e-01 4.23388600e-01 4.50029084e-03
-4.54506814e-01 2.49827608e-01 -3.74140233e-01 -6.48820579e-01
-2.43215561e-02 3.25537361e-02 -5.65248311e-01 -6.66385293e-02
3.18720549e-01 1.16765010e+00 -9.41770300e-02 -1.06810069e+00
-2.32152328e-01 6.93203509e-01 6.34539127e-01 -5.51666379e-01
1.18280864e+00 -5.77308655e-01 -4.93035875e-02 1.78035989e-01
7.89309859e-01 -1.38109207e-01 -1.55328476e+00 4.32085484e-01
2.06044331e-01 -1.98311165e-01 -2.36984883e-02 -5.07387400e-01
-4.63249981e-01 4.72376108e-01 3.51305068e-01 1.24689467e-01
7.29780853e-01 -8.65119874e-01 4.94845659e-01 6.74691379e-01
5.75216413e-01 -1.21143818e+00 -5.03921032e-01 1.04750979e+00
1.37306607e+00 -1.11324465e+00 2.01422989e-01 -6.96380556e-01
-8.98665428e-01 1.44699633e+00 4.70542878e-01 -5.27444184e-01
4.37697202e-01 5.67779243e-02 -2.01574206e-01 2.41739184e-01
-1.05993712e+00 -4.42962199e-01 2.24701121e-01 6.79696143e-01
-3.54406595e-01 1.69081047e-01 -1.03425217e+00 -5.39200902e-02
2.22781062e-01 -4.63098139e-01 9.67910230e-01 1.20110953e+00
-1.13780117e+00 -7.73202598e-01 -7.28457510e-01 3.45948078e-02
2.47082859e-02 5.65464795e-01 1.20306775e-01 1.26946414e+00
-1.10201508e-01 1.43411672e+00 -4.36769992e-01 -4.16155577e-01
7.32914031e-01 -1.87053904e-01 -1.60700276e-01 -9.08749551e-02
-3.84596497e-01 -9.10867229e-02 5.89173794e-01 -7.42649317e-01
1.10683277e-01 -7.51149952e-01 -1.57618380e+00 -1.66774839e-01
-2.23315120e-01 3.69172722e-01 9.36919689e-01 8.78293395e-01
1.43830895e-01 3.06621045e-01 8.39053929e-01 -1.31587255e+00
-8.67067873e-01 -4.58063394e-01 -2.69212782e-01 -2.87917376e-01
4.97000962e-01 -8.31874192e-01 -3.83764893e-01 -3.90198052e-01] | [5.257544040679932, 2.088841676712036] |
27d7be43-ff22-4fe4-b35c-a0f90f405da9 | did-you-mean-confidence-based-trade-offs-in | 2303.16857 | null | https://arxiv.org/abs/2303.16857v2 | https://arxiv.org/pdf/2303.16857v2.pdf | Did You Mean...? Confidence-based Trade-offs in Semantic Parsing | We illustrate how a calibrated model can help balance common trade-offs in task-oriented parsing. In a simulated annotator-in-the-loop experiment, we show that well-calibrated confidence scores allow us to balance cost with annotator load, improving accuracy with a small number of interactions. We then examine how confidence scores can help optimize the trade-off between usability and safety. We show that confidence-based thresholding can substantially reduce the number of incorrect low-confidence programs executed; however, this comes at a cost to usability. We propose the DidYouMean system which better balances usability and safety. | ['Benjamin Van Durme', 'Elias Stengel-Eskin'] | 2023-03-29 | null | null | null | null | ['semantic-parsing'] | ['natural-language-processing'] | [-1.92988571e-02 5.14010489e-01 -4.09922078e-02 -7.76958704e-01
-1.26998353e+00 -9.81728733e-01 8.72623525e-04 6.15903080e-01
-6.89798295e-01 5.43428779e-01 4.90016770e-03 -6.68078125e-01
2.15300456e-01 -3.93829286e-01 -6.98996902e-01 -3.93785127e-02
2.22467273e-01 3.67390037e-01 5.78800440e-01 -1.72049664e-02
2.83466637e-01 -8.03730711e-02 -1.37674034e+00 3.44693452e-01
1.25121379e+00 3.51144761e-01 2.11245552e-01 9.99998868e-01
1.94541865e-03 7.90295124e-01 -9.04783249e-01 -5.83187103e-01
3.67289446e-02 -4.07543182e-01 -8.63249660e-01 -4.55772102e-01
4.17044699e-01 -1.60091683e-01 3.98169309e-01 1.10139227e+00
4.30861801e-01 -2.05135047e-01 4.61515009e-01 -1.31654286e+00
-4.13570970e-01 1.03278422e+00 -4.52111632e-01 1.12033084e-01
4.31475252e-01 3.45543832e-01 9.78237629e-01 -2.25229442e-01
4.23834532e-01 1.19287980e+00 8.92325997e-01 3.70002329e-01
-1.38401067e+00 -3.49868864e-01 2.14346141e-01 -4.01128620e-01
-1.20740187e+00 -5.65326333e-01 2.17321247e-01 -5.00017047e-01
1.31419420e+00 2.98201382e-01 3.13495874e-01 6.52856112e-01
1.63899407e-01 3.31059813e-01 9.20297563e-01 -8.06379855e-01
2.29724050e-01 2.08092242e-01 7.02904820e-01 9.96159852e-01
7.10594594e-01 -9.25805047e-02 -5.54898918e-01 -3.86153370e-01
3.57546210e-01 -6.72707260e-01 -3.09683025e-01 -3.98935676e-01
-7.67000377e-01 7.11153567e-01 -8.45914185e-02 2.06040040e-01
1.14782333e-01 4.38636005e-01 5.30441463e-01 4.61832792e-01
2.56170511e-01 1.01791608e+00 -8.26926708e-01 -6.08498752e-01
-5.30503035e-01 3.05624574e-01 1.17162073e+00 1.10282850e+00
4.25117761e-01 -3.69998783e-01 -3.91330212e-01 8.09062183e-01
3.20979446e-01 4.11886454e-01 1.85556665e-01 -1.35185492e+00
7.03718841e-01 4.62795287e-01 6.36756897e-01 -5.92972934e-01
-4.75987256e-01 6.70449212e-02 2.71870196e-01 4.24779594e-01
9.99692261e-01 -3.85081708e-01 -7.23654985e-01 1.93754840e+00
3.92166935e-02 -7.64223635e-01 -2.66349554e-01 7.09882319e-01
-4.31668982e-02 6.64512217e-02 6.17113650e-01 -4.62592721e-01
1.68685997e+00 -7.65264869e-01 -9.36907709e-01 -6.68413937e-01
1.28896582e+00 -8.63090813e-01 1.75452352e+00 2.54944056e-01
-1.26827228e+00 -2.03438684e-01 -1.11708474e+00 -2.67944753e-01
1.04973502e-01 3.16353478e-02 5.02070546e-01 1.17202353e+00
-9.33614552e-01 7.90463626e-01 -1.12836409e+00 -4.29485887e-01
-7.07131103e-02 2.33311817e-01 -8.67653042e-02 3.75629008e-01
-7.32538879e-01 1.26940501e+00 4.11075473e-01 -2.95051545e-01
-2.44318530e-01 -4.01501566e-01 -9.18990254e-01 2.68194824e-01
7.83240378e-01 -5.27432501e-01 2.05564189e+00 -6.38029337e-01
-1.22355926e+00 7.58453906e-01 -9.11285952e-02 -2.01543316e-01
6.20659173e-01 -6.40742779e-01 6.44080266e-02 -4.08709913e-01
3.33481401e-01 5.41633368e-01 2.38441646e-01 -1.01847804e+00
-6.69243574e-01 -3.11146796e-01 3.01572233e-01 2.20727667e-01
-3.03971618e-01 2.78553873e-01 -5.81111312e-01 -3.79823446e-01
-5.59051782e-02 -1.15554416e+00 -1.51842371e-01 8.57337844e-03
-1.03112392e-01 -1.53409436e-01 2.81333059e-01 -7.52192318e-01
1.48109210e+00 -2.09715366e+00 -1.44951791e-01 -7.49101862e-02
9.77925360e-02 1.26791015e-01 -7.71004111e-02 -6.01771437e-02
1.47007689e-01 5.86409390e-01 -1.29039720e-01 -1.42295077e-01
-1.21990658e-01 1.12108402e-01 3.60493362e-01 1.26677260e-01
2.39196956e-01 6.84293509e-01 -8.64638090e-01 -7.38570571e-01
-2.29763508e-01 8.43967795e-02 -9.40247297e-01 4.83473778e-01
-4.34548616e-01 8.33890885e-02 -3.23386103e-01 4.37657684e-01
3.25628042e-01 -1.34072736e-01 7.28571773e-01 -9.02569368e-02
-1.56132817e-01 5.13261080e-01 -9.28151309e-01 1.47583890e+00
-5.67574143e-01 4.82806414e-01 2.62185544e-01 -4.35601085e-01
5.53716421e-01 1.04030296e-01 -3.14339638e-01 -6.01218581e-01
1.61607459e-01 2.90598065e-01 3.48735064e-01 -6.36535347e-01
5.14276445e-01 6.56419024e-02 -3.92913938e-01 6.70034111e-01
-3.20139110e-01 -3.38085085e-01 2.17899293e-01 1.16888188e-01
1.21587300e+00 2.60210782e-01 5.32116413e-01 -5.87505102e-01
-2.13168263e-01 3.04381311e-01 5.15941501e-01 1.00630164e+00
-4.19818252e-01 2.32969716e-01 1.03918040e+00 -3.51981163e-01
-1.21694529e+00 -5.75695217e-01 6.12317100e-02 1.56344235e+00
-6.95521906e-02 -6.82752788e-01 -1.31833339e+00 -9.61972296e-01
-1.31058246e-01 1.08081269e+00 -3.08279067e-01 -1.43562853e-01
-7.40928650e-01 -5.07961750e-01 7.27606654e-01 7.51079082e-01
6.70788065e-02 -7.84862995e-01 -1.34621835e+00 1.24281757e-01
-3.37792784e-01 -9.98498857e-01 -7.61062264e-01 7.79190540e-01
-7.70094454e-01 -1.24302351e+00 -1.85481906e-01 -6.03833854e-01
7.39675999e-01 7.57606560e-03 1.63315797e+00 4.19971764e-01
-2.01646592e-02 1.39830932e-01 -4.54485238e-01 -4.28218722e-01
-8.20648134e-01 2.03933969e-01 -1.10126607e-01 -1.20177436e+00
4.55912650e-01 -2.27032915e-01 -1.06838226e-01 4.42702711e-01
-4.38068509e-01 -8.26741084e-02 2.80038267e-01 8.83216619e-01
1.72548309e-01 -3.89200717e-01 2.66125172e-01 -1.22805536e+00
1.09389496e+00 7.84641355e-02 -9.61946428e-01 6.99611187e-01
-9.85383272e-01 7.51063645e-01 3.22948575e-01 -5.61489701e-01
-7.73929834e-01 3.44957918e-01 -2.46104095e-02 1.25985980e-01
1.21758722e-01 3.83535951e-01 -1.79341003e-01 -8.52959454e-02
1.00772643e+00 -5.92810750e-01 -2.05570832e-01 -3.48678946e-01
4.82386738e-01 6.45197809e-01 4.09212559e-01 -9.92441177e-01
1.50948033e-01 -5.38488448e-01 -6.14160538e-01 -1.81548461e-01
-7.87389576e-01 2.10592747e-02 -3.88359517e-01 1.26088887e-01
7.13426769e-01 -7.34616697e-01 -9.96223211e-01 5.56770079e-02
-1.30560660e+00 -6.61352813e-01 -2.31650770e-01 5.44577353e-02
-5.67109764e-01 4.15114373e-01 -6.69578731e-01 -1.06162190e+00
-3.60693276e-01 -1.37862861e+00 1.03627300e+00 2.01354697e-01
-8.68218064e-01 -4.36030775e-01 -2.84184143e-02 1.88508958e-01
3.13095927e-01 -1.64595053e-01 1.11258996e+00 -8.09786439e-01
-1.85373276e-01 -1.73014600e-03 -2.26189330e-01 1.93929672e-01
-2.06466392e-01 -3.88403907e-02 -8.79260182e-01 -1.15411624e-01
-1.04055934e-01 -3.74847978e-01 2.43378103e-01 3.09137464e-01
9.28014219e-01 -4.68274891e-01 -3.62395287e-01 2.51394719e-01
1.05266750e+00 3.53524625e-01 3.48050445e-01 4.33212340e-01
6.16787672e-01 4.98621285e-01 9.46867645e-01 1.42452553e-01
2.96723425e-01 1.04266357e+00 -1.10918917e-01 2.61234909e-01
1.32664829e-01 -1.79062054e-01 1.24270901e-01 5.66082001e-01
-2.74709822e-03 -4.07622576e-01 -1.30512762e+00 4.90296096e-01
-2.25405383e+00 -4.86802965e-01 -1.31769180e-01 2.31200671e+00
1.21674347e+00 6.77646399e-01 3.80589575e-01 6.07248805e-02
6.59165323e-01 -3.24046552e-01 -3.02541733e-01 -7.99959242e-01
5.52411258e-01 -3.48783098e-02 6.16462231e-01 9.94013488e-01
-8.61933887e-01 9.60095584e-01 7.80214739e+00 4.33730334e-01
-5.43162584e-01 3.67009401e-01 6.24841273e-01 -5.82097210e-02
-2.63542295e-01 1.34334251e-01 -7.65305340e-01 5.68986297e-01
1.40467227e+00 -2.47971013e-01 2.76970148e-01 1.37809432e+00
2.53486242e-02 -3.52277637e-01 -1.56420946e+00 5.88281929e-01
-1.60721421e-01 -7.58096397e-01 -6.86671853e-01 -1.60394788e-01
1.43739000e-01 -2.32723385e-01 -4.72882271e-01 4.13019925e-01
7.50536084e-01 -7.12847054e-01 9.83800471e-01 9.27218702e-03
7.66006470e-01 -6.59399927e-01 8.59683573e-01 4.57854718e-01
-9.46058095e-01 -8.95463750e-02 -3.84647660e-02 -1.20717973e-01
3.30322117e-01 5.46019256e-01 -1.05460453e+00 -1.64210901e-01
8.33608091e-01 -4.05034095e-01 -6.70628428e-01 8.50157857e-01
-3.70297581e-01 3.76496673e-01 -3.47203046e-01 -2.47709930e-01
-3.16022307e-01 1.39274165e-01 1.38753340e-01 1.45611215e+00
1.55082583e-01 3.25575680e-01 2.00473785e-01 7.56466269e-01
3.81539725e-02 -9.23922583e-02 -4.49032485e-01 -6.91701546e-02
9.09332633e-01 8.93026054e-01 -8.92035246e-01 -4.90851730e-01
-8.03412795e-02 8.02951932e-01 7.62915313e-01 1.01776682e-02
-9.73079324e-01 -5.29608309e-01 4.56867397e-01 2.41299853e-01
1.11046895e-01 -3.33796769e-01 -6.06206596e-01 -9.61181164e-01
2.53471762e-01 -1.19452024e+00 2.71188617e-01 -7.02196896e-01
-7.99860179e-01 5.51709771e-01 1.55484945e-01 -6.61021709e-01
-2.59080261e-01 -5.76077223e-01 -1.69898838e-01 8.01024675e-01
-7.86055863e-01 -5.93420386e-01 -2.29833603e-01 -2.42084190e-01
4.96243924e-01 4.80729491e-01 9.71173346e-01 2.00465634e-01
-4.88591343e-01 9.51461256e-01 -4.89696085e-01 -1.42870760e-02
8.90895128e-01 -1.48988175e+00 9.26626265e-01 9.41325903e-01
-2.50131816e-01 1.05408835e+00 1.09066308e+00 -8.13304424e-01
-1.07040262e+00 -6.56625748e-01 9.16324735e-01 -8.94154906e-01
3.97811919e-01 -3.99673343e-01 -9.92272913e-01 8.14519405e-01
-2.51625348e-02 -1.61617219e-01 3.57174426e-01 6.80726826e-01
-8.19626153e-01 2.28656337e-01 -1.32261777e+00 6.52877629e-01
1.03972232e+00 -6.66842818e-01 -5.42443454e-01 3.83001536e-01
1.00738108e+00 -6.73144162e-01 -8.14634919e-01 2.54556477e-01
7.78973401e-01 -8.61448705e-01 5.00412822e-01 -3.77660513e-01
1.05062917e-01 -2.27593571e-01 -1.03599261e-02 -1.34017873e+00
-2.30782256e-01 -5.69398820e-01 3.56368572e-02 1.17517948e+00
8.08433115e-01 -4.28877741e-01 5.66169679e-01 1.50052655e+00
-1.36576831e-01 -2.86636382e-01 -5.16122341e-01 -8.04463923e-01
-7.53437877e-02 -4.48391825e-01 3.50390285e-01 7.03628659e-01
6.41539097e-01 5.97234190e-01 -1.63767502e-01 5.18835112e-02
2.91824847e-01 -5.59174061e-01 5.83729148e-01 -1.05093169e+00
-6.26682401e-01 -2.20437586e-01 4.71881405e-02 -8.10429990e-01
-2.31768116e-01 -1.39056832e-01 7.53242671e-01 -9.32959378e-01
4.62791175e-01 -6.78911984e-01 1.03206582e-01 9.34945226e-01
-7.60243595e-01 -2.05916122e-01 3.77690703e-01 7.27033243e-02
-9.03447092e-01 -3.40277046e-01 5.35669744e-01 1.18828803e-01
-4.34726447e-01 -2.84899384e-01 -1.05145466e+00 7.69106686e-01
5.60049355e-01 -9.03972387e-01 -3.74858171e-01 -9.11017537e-01
5.08499861e-01 2.58248299e-01 -3.23364824e-01 -1.01918995e+00
2.04188839e-01 -1.19864091e-01 -1.41950116e-01 -7.76158227e-03
-4.19252031e-02 -6.78804338e-01 2.46527642e-02 4.68865931e-01
-6.81540072e-01 6.30576789e-01 3.49099755e-01 4.61850017e-01
2.23200887e-01 -6.29237592e-01 7.08557904e-01 -1.41913831e-01
-2.85887331e-01 -6.32681787e-01 -4.55916584e-01 2.80987084e-01
1.10456753e+00 -1.45475753e-02 -6.86073661e-01 -1.45536557e-01
-5.33052146e-01 2.54520595e-01 8.46325219e-01 2.21381843e-01
-1.98353738e-01 -7.78790534e-01 -2.18056768e-01 -6.95615560e-02
3.71627808e-01 -3.01869541e-01 -4.57113385e-01 4.90021527e-01
-8.01679850e-01 2.73554891e-01 -4.19649333e-02 -6.56093597e-01
-1.68649638e+00 4.54679996e-01 3.54314238e-01 -3.12680840e-01
-2.18835026e-01 9.50042129e-01 -1.62408903e-01 -4.86682594e-01
5.88226080e-01 -5.63315749e-01 1.37676910e-01 -2.49185979e-01
4.81429756e-01 1.39350712e-01 2.80189186e-01 1.15146391e-01
-6.43768609e-01 4.67257380e-01 -2.70947665e-01 -5.07386088e-01
9.33616817e-01 7.28705227e-02 1.31259575e-01 3.66459489e-01
5.94745874e-01 1.27904832e-01 -1.08979440e+00 3.83504122e-01
6.40666604e-01 -5.18784821e-01 -6.57382682e-02 -1.16644359e+00
-5.10672450e-01 6.52002275e-01 6.68346643e-01 5.11526406e-01
8.06517124e-01 -1.04101352e-01 4.48929399e-01 7.29010224e-01
6.69255316e-01 -1.21233165e+00 -6.17970973e-02 3.47318947e-01
3.05552483e-01 -1.30726969e+00 1.25733957e-01 -7.66745627e-01
-1.01667297e+00 7.97642469e-01 1.02252281e+00 3.03533733e-01
1.36947885e-01 9.66728806e-01 2.76527792e-01 -1.46453947e-01
-8.43952179e-01 1.56983182e-01 -3.23352993e-01 7.72141516e-01
1.14853740e+00 3.74203354e-01 -7.05741107e-01 8.02781582e-01
-1.72653809e-01 1.40386708e-02 6.34653687e-01 1.41116476e+00
-6.92474604e-01 -1.38115478e+00 -5.23106992e-01 3.29829901e-01
-5.34390092e-01 -1.39862001e-01 -3.98983508e-01 7.82676935e-01
-5.92552163e-02 1.12949133e+00 3.28216329e-02 -3.01536351e-01
7.17115104e-01 3.20652902e-01 6.24181688e-01 -9.07487810e-01
-9.31724131e-01 1.38901785e-01 9.11620855e-01 -7.45026767e-01
-4.61647585e-02 -3.38859528e-01 -1.08264649e+00 -2.42151260e-01
-7.91329563e-01 4.07611281e-01 6.48014724e-01 8.45728219e-01
4.61543500e-01 5.26495397e-01 -3.57677303e-02 -4.41098988e-01
-9.33572829e-01 -9.50356126e-01 -1.31261021e-01 3.47614408e-01
1.64054483e-01 -4.58491594e-01 -2.35200360e-01 1.41859666e-01] | [10.675456047058105, 8.727056503295898] |
930d6df9-82f1-44fa-acb4-a1d5fa257ab7 | is-more-data-better-using-transformers-based | null | null | https://openreview.net/forum?id=f0KsTiVPZWZ | https://openreview.net/pdf?id=f0KsTiVPZWZ | Is More Data Better? Using Transformers-Based Active Learning for Efficient and Effective Detection of Abusive Language | Annotating abusive language content can cause psychological harm; yet, most machine learning research has prioritized efficacy (i.e., F1 or accuracy scores) while little research has analyzed data efficiency (i.e., how to minimize annotation requirements).In this paper, we use a series of simulated experiments over two datasets at varying percentages of abuse to demonstrate that transformers-based active learning is a promising approach that maintains high efficacy but substantially raises efficiency, requiring a fraction of labeled data to reach equivalent performance to passive training over the full dataset. | ['Anonymous'] | 2022-01-16 | null | null | null | acl-arr-january-2022-1 | ['abusive-language'] | ['natural-language-processing'] | [ 6.28267378e-02 4.15693372e-01 -9.88586366e-01 -4.11066294e-01
-8.44546676e-01 -6.28637493e-01 2.73501247e-01 8.03409457e-01
-9.78281796e-01 7.92245865e-01 2.35970795e-01 -4.41087276e-01
-2.20493138e-01 -3.35873187e-01 -2.88653314e-01 -2.71489948e-01
-1.60996258e-01 2.57420003e-01 3.61062624e-02 1.08910382e-01
5.87191105e-01 5.38583219e-01 -7.27421105e-01 -8.83454382e-02
9.21189249e-01 2.87883043e-01 -6.53320849e-01 8.38398412e-02
-3.86523791e-02 1.39384437e+00 -9.88867283e-01 -1.07387817e+00
1.33303404e-01 -1.75152659e-01 -9.70178366e-01 -1.35346785e-01
3.55947286e-01 -1.00991201e+00 -1.57461748e-01 9.35113549e-01
6.12032890e-01 1.62813231e-01 3.33969444e-01 -1.34823847e+00
-7.76647389e-01 6.40946329e-01 -4.96854037e-01 7.77206957e-01
3.86362731e-01 4.25086677e-01 9.26944852e-01 -4.52161580e-01
6.35959208e-01 1.01432109e+00 6.26264870e-01 7.43993521e-01
-1.51206577e+00 -8.64022493e-01 3.00255958e-02 -4.43629362e-02
-9.44235742e-01 -8.59148145e-01 8.11857104e-01 -4.47951496e-01
9.36751902e-01 1.60459340e-01 5.72394609e-01 1.54279697e+00
1.23171108e-02 6.46215439e-01 1.11115730e+00 -2.65017986e-01
5.77010989e-01 4.04164940e-01 9.20877516e-01 5.00561833e-01
1.01170647e+00 2.22761612e-02 -6.70636714e-01 -8.63149464e-01
2.30171740e-01 -2.36124933e-01 1.46246180e-01 -2.60158122e-01
1.36329576e-01 1.38215089e+00 8.58530519e-04 2.71764278e-01
-2.64074713e-01 -1.67433709e-01 6.31723166e-01 1.42869696e-01
8.50260794e-01 8.66241276e-01 -1.97907746e-01 -6.64740860e-01
-8.31614554e-01 2.38298059e-01 7.96799898e-01 3.35518062e-01
3.01574022e-01 2.85977870e-01 -5.33268005e-02 1.07227182e+00
6.29089400e-02 1.63623005e-01 1.38232470e-01 -1.06738412e+00
5.78744650e-01 6.91568315e-01 1.86035007e-01 -1.06330049e+00
-4.06218171e-01 -1.12846889e-01 8.95617157e-02 1.21156864e-01
3.78833622e-01 -4.96935010e-01 -5.54881454e-01 1.96675968e+00
2.58230362e-02 -2.14258790e-01 -5.51898181e-01 4.25754040e-01
3.38260889e-01 1.94374248e-01 1.04921639e+00 -8.69346201e-01
4.94054258e-01 -3.39892328e-01 -1.05062103e+00 -7.64269054e-01
1.03258955e+00 -3.00854176e-01 1.18781674e+00 3.96771222e-01
-1.39868712e+00 2.12166309e-01 -7.35858083e-01 -3.17008980e-02
-1.77599132e-01 -8.26571465e-01 9.71619785e-01 1.46659923e+00
-6.39699638e-01 2.76946098e-01 -8.31553161e-01 -2.44783804e-01
9.76603150e-01 3.65389407e-01 -3.26803952e-01 1.13509014e-01
-1.18072486e+00 1.24316466e+00 2.33141571e-01 -4.15410012e-01
-8.78593624e-01 -8.81361902e-01 -7.85755634e-01 6.02810495e-02
4.78751183e-01 -7.29272664e-02 1.16094589e+00 -9.93405879e-01
-1.02290940e+00 8.23635817e-01 2.46675849e-01 -5.61085224e-01
2.10028350e-01 -6.64441526e-01 -3.49256694e-01 1.87485456e-01
7.40583427e-03 3.36957574e-01 3.94873083e-01 -1.15965128e+00
-8.58504847e-02 -6.39619946e-01 3.64561111e-01 1.86412007e-01
-1.33915699e+00 7.94441700e-01 3.04374993e-01 -5.35245895e-01
-6.24012709e-01 -7.36471534e-01 -1.21792309e-01 -8.20461810e-02
-1.28371924e-01 -3.60026866e-01 7.65057385e-01 -7.98742771e-01
1.63229549e+00 -2.17291117e+00 -3.29338670e-01 1.65786371e-01
3.78569901e-01 6.08858287e-01 1.54064208e-01 2.60995418e-01
-2.24362299e-01 8.17036092e-01 -3.44972730e-01 -4.06518728e-01
-3.86267960e-01 2.41360143e-01 -2.44735047e-01 7.12989211e-01
2.24401891e-01 8.59554231e-01 -9.02823806e-01 -6.34295166e-01
1.85256097e-02 1.90330613e-02 -7.34721363e-01 3.54072422e-01
3.04893494e-01 4.40054908e-02 -1.49306521e-01 4.82497454e-01
4.21444207e-01 1.23031944e-01 3.12451690e-01 4.12754506e-01
-3.48297246e-02 4.62879717e-01 -3.48978162e-01 1.10558021e+00
-2.63733208e-01 4.77280259e-01 1.98173285e-01 -1.05887556e+00
7.54972935e-01 3.15874428e-01 8.26570332e-01 -1.15173864e+00
3.33443671e-01 -7.95256346e-02 1.89219043e-01 -8.12783301e-01
4.43643659e-01 -2.43013456e-01 -2.40132064e-01 6.28929317e-01
3.91755812e-02 3.35310817e-01 1.16398353e-02 3.39776516e-01
1.49725425e+00 -4.57817018e-01 3.73700023e-01 -1.55470043e-01
-3.81874681e-01 1.79684490e-01 5.97432256e-01 8.92645359e-01
-7.29774594e-01 -2.10096285e-01 6.22209013e-01 -4.50886637e-02
-1.18781304e+00 -1.09024513e+00 -3.07053596e-01 1.24679089e+00
9.63114202e-03 -3.99308532e-01 -9.99275208e-01 -8.66124451e-01
-1.39852762e-01 1.30741680e+00 -7.01723099e-01 -6.92216814e-01
-5.70950925e-01 -1.10997605e+00 9.08332229e-01 3.60017985e-01
3.10668170e-01 -9.57750142e-01 -6.92381442e-01 4.87287194e-02
-1.00533580e-02 -7.91435421e-01 4.82602715e-02 1.54548243e-01
-1.00482607e+00 -1.10288668e+00 8.46982151e-02 -8.89685154e-02
4.75371152e-01 7.29751363e-02 9.76695120e-01 3.52922529e-01
-8.14463198e-02 3.72641057e-01 -4.63095456e-01 -5.27794302e-01
-1.21396102e-01 1.91715546e-02 -1.48896256e-03 -4.96634215e-01
7.86969304e-01 -4.73377138e-01 -2.75408387e-01 -2.22450286e-01
-9.35728490e-01 -6.51124537e-01 3.33975017e-01 6.53611422e-01
-1.34068519e-01 -2.90715471e-02 8.09500158e-01 -1.55302286e+00
1.38089359e+00 -1.05782700e+00 -1.15596786e-01 1.28548175e-01
-1.02298164e+00 -6.78736866e-01 4.17194426e-01 -7.69574881e-01
-1.18164647e+00 -3.12501997e-01 -1.65273950e-01 1.71103314e-01
-3.97352017e-02 4.65048462e-01 -1.63463503e-02 7.47145787e-02
1.15826225e+00 -5.22563994e-01 -7.60790259e-02 -4.31568950e-01
1.09437965e-01 6.85944140e-01 8.99340361e-02 -5.08965671e-01
6.26192868e-01 8.36564898e-02 -5.44511795e-01 -5.66646755e-01
-1.09886909e+00 -2.72426665e-01 -4.57638830e-01 -2.63945669e-01
5.15743554e-01 -4.40342873e-01 -5.55399656e-01 1.84527680e-01
-6.78682029e-01 -4.99680042e-01 -2.48042032e-01 3.34098130e-01
-3.52137685e-01 2.23551214e-01 -7.23014653e-01 -1.30284119e+00
-4.89001960e-01 -5.37518859e-01 3.34536254e-01 9.41478610e-02
-9.61350083e-01 -1.03898394e+00 5.33178031e-01 6.30592823e-01
2.92404234e-01 2.12916434e-01 1.09258592e+00 -1.24316728e+00
2.61087000e-01 -4.14573342e-01 1.67095453e-01 1.94288433e-01
1.80421919e-01 -1.08846519e-02 -9.27791476e-01 -5.09320378e-01
2.11158320e-01 -9.23624218e-01 2.78637081e-01 1.31271139e-01
1.28036261e+00 -8.86154056e-01 -1.10136792e-01 -1.23412974e-01
1.14345491e+00 6.47709489e-01 7.31918454e-01 1.66416556e-01
3.20374280e-01 6.14340186e-01 4.51337844e-01 3.83518726e-01
-1.67807981e-01 5.32109737e-01 -3.02103255e-02 -4.15303893e-02
8.75181928e-02 -2.87873000e-01 4.43585753e-01 1.83939889e-01
3.22004944e-01 -2.22053260e-01 -1.10432923e+00 5.32262564e-01
-1.49864876e+00 -1.35212910e+00 1.95893452e-01 2.21542311e+00
1.05560195e+00 7.77169168e-01 4.98640537e-01 3.36287439e-01
6.32679641e-01 1.31860003e-01 -6.67640090e-01 -8.61863375e-01
2.73878664e-01 4.75965351e-01 2.68464744e-01 5.63351989e-01
-7.71621406e-01 6.30628049e-01 7.91031075e+00 7.19912767e-01
-9.18817282e-01 7.30445385e-01 8.96481276e-01 -5.09691954e-01
-3.06950718e-01 6.89744204e-02 -3.58978748e-01 7.18529046e-01
1.28374314e+00 -3.95645946e-01 1.42012194e-01 1.07235146e+00
3.51197660e-01 -2.36043036e-01 -7.29344130e-01 5.12067676e-01
2.56516814e-01 -7.66559303e-01 -1.32847711e-01 2.25185692e-01
4.96637821e-01 -4.31225747e-01 5.14491042e-03 4.49622542e-01
4.80820656e-01 -9.84139204e-01 6.75101519e-01 1.81358993e-01
6.82308793e-01 -8.63136292e-01 6.06973946e-01 7.24789500e-01
6.58656210e-02 -6.46025419e-01 -1.34558737e-01 -4.14561421e-01
6.93690404e-02 4.40798879e-01 -8.19732249e-01 -1.57075092e-01
6.02386415e-01 1.86233863e-01 -8.62203479e-01 1.06192613e+00
-2.37563983e-01 1.40588701e+00 -1.12808146e-01 -1.03735618e-01
-5.59522025e-02 2.94015586e-01 2.92778075e-01 1.07766616e+00
-4.34273273e-01 3.91785651e-01 7.59386048e-02 5.61767161e-01
-3.05070668e-01 1.67752042e-01 -1.00086105e+00 -4.73784357e-01
9.47634161e-01 1.12524474e+00 -4.21804219e-01 -6.04539216e-02
-3.75108480e-01 3.36853772e-01 8.95380318e-01 7.70795271e-02
-6.37945712e-01 -1.35491136e-02 3.50322247e-01 5.32243311e-01
-7.46836424e-01 -2.11534590e-01 -7.96246946e-01 -7.16624022e-01
-4.69657481e-01 -8.09798539e-01 6.39079988e-01 -5.40438890e-01
-1.47720933e+00 8.71574581e-02 4.26226795e-01 -5.12718678e-01
-2.60728747e-01 -3.21923107e-01 -5.74819922e-01 5.88642836e-01
-4.93225753e-01 -8.68644655e-01 -4.84167300e-02 2.14853972e-01
3.72164130e-01 7.89576471e-02 9.42877710e-01 4.38935250e-01
-1.01274574e+00 1.02298832e+00 -4.23481286e-01 4.72359240e-01
6.48054838e-01 -8.08050096e-01 1.25799878e-02 8.72955978e-01
-2.10233197e-01 1.15089703e+00 5.86940289e-01 -1.13432205e+00
-8.50739002e-01 -6.63230300e-01 5.57896554e-01 -8.55997920e-01
6.81607127e-01 -4.78828192e-01 -1.01671517e+00 9.76481557e-01
9.73552465e-02 -3.95678431e-01 1.19261193e+00 5.39543033e-01
-4.39796776e-01 8.45660046e-02 -1.64374399e+00 6.19323611e-01
1.19762838e+00 -4.92323726e-01 -5.59789658e-01 1.25038058e-01
4.51892167e-01 5.79908118e-02 -7.14046299e-01 4.77331638e-01
3.44436496e-01 -6.72695041e-01 7.50468552e-01 -1.12397349e+00
3.91424090e-01 7.96359718e-01 1.78823963e-01 -1.05565608e+00
-5.09936333e-01 -3.49623591e-01 -4.02542263e-01 1.53380239e+00
5.81111252e-01 -4.22887862e-01 7.87936687e-01 1.39131176e+00
7.21392483e-02 -8.24965239e-01 -8.55474830e-01 -6.11465991e-01
4.46819603e-01 -2.90649474e-01 2.97568254e-02 1.58980632e+00
3.62711877e-01 4.00243819e-01 -4.67501730e-01 -1.96108595e-01
6.93236351e-01 -6.62252665e-01 4.75025684e-01 -1.12204754e+00
1.47327140e-01 -2.61644900e-01 -1.16856262e-01 -1.66325018e-01
4.25989062e-01 -4.99153614e-01 -3.56452525e-01 -1.21938443e+00
8.24514568e-01 -4.93237615e-01 -7.36901239e-02 9.86273348e-01
-5.08843541e-01 3.27368617e-01 -6.74112514e-02 1.89001232e-01
-4.75953579e-01 1.99095771e-01 4.77861196e-01 -1.60450079e-02
-2.80516833e-01 -4.88794446e-01 -1.21812963e+00 7.25975096e-01
1.05521822e+00 -8.85583401e-01 -7.52618849e-01 -3.28847826e-01
3.54053944e-01 -8.11402313e-03 2.92938381e-01 -5.35701692e-01
-7.48081552e-03 -7.57259250e-01 4.80614036e-01 1.84466973e-01
3.97237062e-01 -6.82695448e-01 -5.00616282e-02 4.62318122e-01
-8.95544052e-01 1.38572842e-01 3.09175909e-01 2.11517304e-01
2.89054275e-01 -8.67204905e-01 8.66163850e-01 -1.37590514e-02
-3.17716002e-01 2.38807946e-01 -5.16973436e-01 4.08256799e-01
1.12622809e+00 -1.45092249e-01 -8.85536730e-01 -4.01420265e-01
-5.13469875e-01 -2.83985555e-01 4.91529375e-01 3.41999292e-01
3.75051469e-01 -1.04520845e+00 -4.32471365e-01 -3.62914503e-01
1.45594195e-01 -7.42397726e-01 1.50535405e-02 6.93546534e-01
-1.77170709e-01 1.26748696e-01 -3.05770159e-01 6.90748766e-02
-1.33161330e+00 7.61739671e-01 2.19236121e-01 -1.21237896e-01
-1.60008803e-01 5.49741745e-01 -3.57732236e-01 -1.09192975e-01
2.31390014e-01 9.32500362e-01 -1.61859736e-01 9.86147374e-02
4.62183446e-01 6.42857671e-01 9.86621454e-02 -3.86600852e-01
-2.76976943e-01 -4.02346641e-01 -4.26910132e-01 -1.96242347e-01
1.38579190e+00 1.98622361e-01 -3.32725304e-03 3.96142125e-01
1.14139020e+00 4.90914881e-01 -8.60156357e-01 3.35145295e-02
4.91210401e-01 -9.58526552e-01 2.14910507e-01 -1.01041770e+00
-6.66864097e-01 5.13083935e-01 6.19627059e-01 4.92623121e-01
9.70687211e-01 -1.26368701e-01 3.24596137e-01 2.85928696e-01
4.67502147e-01 -1.28976822e+00 6.75697327e-01 -2.04323288e-02
5.11002302e-01 -8.79366040e-01 3.42790127e-01 -5.03497601e-01
-7.14057565e-01 4.31603193e-01 1.05657625e+00 -2.22286601e-02
3.81202519e-01 5.23041844e-01 -1.10938519e-01 -4.09138918e-01
-5.60917497e-01 2.71813482e-01 -3.99302661e-01 5.19932866e-01
6.09155297e-01 -7.51830414e-02 -1.00095510e+00 6.58557177e-01
-2.63223320e-01 -1.79792225e-01 5.28719485e-01 1.18181467e+00
-4.69212830e-01 -1.12204802e+00 -1.26308456e-01 8.95010054e-01
-9.69604969e-01 1.21415677e-02 -1.16261792e+00 9.84627247e-01
7.06535578e-02 1.38168609e+00 2.24750146e-01 -3.07290643e-01
2.36080974e-01 2.97890902e-01 5.36904275e-01 -9.33093190e-01
-9.52264547e-01 -4.27121162e-01 6.33892655e-01 -5.46789706e-01
-4.40477192e-01 -6.43649995e-01 -9.34242845e-01 -8.33333850e-01
-6.14043117e-01 2.39632785e-01 3.17374110e-01 9.21183109e-01
1.11629151e-01 -2.16881528e-01 7.43638456e-01 3.65174800e-01
-6.41857326e-01 -1.21313465e+00 -4.34417456e-01 9.33575630e-01
-1.16170309e-01 -8.72074306e-01 -4.91317332e-01 -3.72442067e-01] | [8.709134101867676, 10.452536582946777] |
6397ac78-cdcd-42cb-894e-0eb2f200d830 | graph-attention-networks | 1710.10903 | null | http://arxiv.org/abs/1710.10903v3 | http://arxiv.org/pdf/1710.10903v3.pdf | Graph Attention Networks | We present graph attention networks (GATs), novel neural network
architectures that operate on graph-structured data, leveraging masked
self-attentional layers to address the shortcomings of prior methods based on
graph convolutions or their approximations. By stacking layers in which nodes
are able to attend over their neighborhoods' features, we enable (implicitly)
specifying different weights to different nodes in a neighborhood, without
requiring any kind of costly matrix operation (such as inversion) or depending
on knowing the graph structure upfront. In this way, we address several key
challenges of spectral-based graph neural networks simultaneously, and make our
model readily applicable to inductive as well as transductive problems. Our GAT
models have achieved or matched state-of-the-art results across four
established transductive and inductive graph benchmarks: the Cora, Citeseer and
Pubmed citation network datasets, as well as a protein-protein interaction
dataset (wherein test graphs remain unseen during training). | ['Pietro Liò', 'Arantxa Casanova', 'Petar Veličković', 'Yoshua Bengio', 'Guillem Cucurull', 'Adriana Romero'] | 2017-10-30 | graph-attention-networks-1 | https://openreview.net/forum?id=rJXMpikCZ | https://openreview.net/pdf?id=rJXMpikCZ | iclr-2018-1 | ['node-classification-on-non-homophilic', 'graph-regression'] | ['graphs', 'graphs'] | [ 5.70019305e-01 5.06777585e-01 -2.70349473e-01 -3.56839269e-01
-2.07632303e-01 -4.96313602e-01 5.90490699e-01 6.60726845e-01
-3.46184283e-01 7.67944038e-01 3.50372553e-01 -7.69435644e-01
-9.89673361e-02 -1.16936064e+00 -1.19487286e+00 -5.43128669e-01
-4.59176749e-01 6.64400935e-01 -9.50215757e-02 -3.05534214e-01
-3.99943180e-02 4.71736521e-01 -1.03054535e+00 4.22625154e-01
7.99089134e-01 6.55082703e-01 -3.04139882e-01 6.32897258e-01
-2.83560276e-01 9.65561926e-01 -1.53629228e-01 -4.33754444e-01
-1.45213649e-01 -2.58660257e-01 -9.23550546e-01 -3.30953479e-01
5.52958786e-01 9.53103080e-02 -7.20628440e-01 1.05819142e+00
3.33124191e-01 2.40055621e-01 7.99645483e-01 -9.82640982e-01
-1.45822763e+00 9.03799355e-01 -4.31410909e-01 2.93755144e-01
3.45841050e-01 2.27491498e-01 1.45463264e+00 -8.31162632e-01
7.58413255e-01 1.12414896e+00 9.91810739e-01 4.38086390e-01
-1.69868815e+00 -4.71294135e-01 3.65462393e-01 3.12232431e-02
-1.25357866e+00 -3.85842234e-01 7.71193027e-01 -3.61871094e-01
1.29045820e+00 1.48984373e-01 7.07894325e-01 1.18950546e+00
9.19344649e-02 6.09106481e-01 6.06498182e-01 -2.95721620e-01
1.37831524e-01 -4.92985219e-01 3.59887838e-01 1.06400299e+00
3.30258757e-01 -1.71772972e-01 -2.99245298e-01 -3.23780209e-01
6.33084476e-01 7.26910830e-02 -5.20991027e-01 -4.23994899e-01
-1.32377851e+00 9.97187436e-01 1.20179486e+00 2.36593962e-01
-4.43009228e-01 5.47480166e-01 4.14934009e-01 3.00387233e-01
7.38194942e-01 6.80662155e-01 -2.92878181e-01 4.47031051e-01
-5.21439552e-01 -4.61661406e-02 7.62827694e-01 7.89311945e-01
9.45026159e-01 -2.86653284e-02 -3.33719492e-01 4.82611626e-01
2.93906093e-01 2.86283772e-02 9.17793289e-02 -2.47280315e-01
4.34072882e-01 9.38458622e-01 -3.45394760e-01 -9.65435565e-01
-6.47509038e-01 -4.06700104e-01 -1.18743527e+00 -2.00195059e-01
3.13538671e-01 1.98446307e-02 -1.39531934e+00 1.88480699e+00
1.91324055e-01 4.69959170e-01 -2.15919361e-01 6.14847004e-01
1.40192449e+00 3.98462147e-01 3.36221457e-01 1.46532431e-02
1.13868129e+00 -8.45383346e-01 -5.85988820e-01 -2.41078064e-01
9.31890249e-01 -1.88899875e-01 1.12915277e+00 1.56208659e-02
-9.34580028e-01 -2.14319825e-01 -1.00573277e+00 -4.65123206e-01
-8.97003472e-01 -3.75647634e-01 1.12323892e+00 1.73050478e-01
-1.53744757e+00 7.71481812e-01 -7.21468747e-01 -4.68024582e-01
7.43413448e-01 7.16913998e-01 -6.17179215e-01 -2.48505190e-01
-1.24856055e+00 6.84287071e-01 3.18730146e-01 2.84275889e-01
-7.46985257e-01 -8.96493196e-01 -9.99407291e-01 3.69894862e-01
3.65898281e-01 -1.03479612e+00 7.53082335e-01 -8.33633363e-01
-1.18874609e+00 9.62232411e-01 4.07479554e-02 -7.20839977e-01
-4.76716496e-02 2.11182777e-02 -2.75510997e-01 -1.08432798e-02
-1.24771267e-01 6.68277085e-01 4.71016735e-01 -8.37103367e-01
2.84864932e-01 -2.51407236e-01 3.72143656e-01 1.83391813e-02
-3.48625392e-01 -2.25083038e-01 -3.27431947e-01 -6.06467485e-01
-4.18936275e-02 -8.60930026e-01 -4.26704079e-01 -4.54316325e-02
-8.40478480e-01 -4.46880102e-01 5.94195187e-01 -2.71238267e-01
9.54984367e-01 -1.99588907e+00 4.89490032e-01 3.56683403e-01
1.00251567e+00 3.39721918e-01 -4.52215880e-01 6.21454000e-01
-4.93754297e-01 4.73903418e-01 -3.79614353e-01 -7.20667765e-02
-6.79816231e-02 2.01296523e-01 -5.54720461e-02 4.51893330e-01
5.76700389e-01 1.57779062e+00 -1.14783561e+00 -2.73091823e-01
8.33491087e-02 4.93693382e-01 -7.62153387e-01 7.15445653e-02
-6.03586972e-01 1.80067793e-01 -3.95095587e-01 6.42979920e-01
3.55424792e-01 -1.01273620e+00 4.41595554e-01 -4.98306125e-01
4.55766648e-01 5.22407472e-01 -4.59943891e-01 1.74801922e+00
-1.93621978e-01 4.58947331e-01 2.48856023e-02 -1.45562351e+00
5.92083752e-01 1.23398364e-01 3.89117479e-01 -3.93049479e-01
6.95105568e-02 -7.74821192e-02 1.67741880e-01 -9.49683785e-02
1.45222008e-01 4.63522561e-02 1.46022037e-01 4.91106272e-01
3.51460308e-01 2.54178762e-01 1.57312706e-01 5.97805560e-01
1.54086936e+00 -4.06362191e-02 2.07573399e-01 -4.71960813e-01
2.51367420e-01 -1.11697316e-01 5.90943247e-02 9.04380143e-01
1.69371828e-01 3.30299705e-01 8.28262508e-01 -6.84245288e-01
-6.51192129e-01 -1.13353157e+00 1.86175585e-01 1.27901804e+00
-2.43527040e-01 -5.21627367e-01 -4.45949763e-01 -7.97118366e-01
1.12488948e-01 3.35252434e-01 -1.22694838e+00 -5.30259609e-01
-4.71099496e-01 -8.92217815e-01 6.06769204e-01 5.54868937e-01
5.25519960e-02 -1.11810267e+00 3.36539984e-01 2.83733487e-01
1.81951687e-01 -8.99299383e-01 -5.65079987e-01 6.30642116e-01
-7.76248157e-01 -1.23615313e+00 -3.53424639e-01 -8.67502153e-01
9.48489010e-01 1.36413276e-01 1.60561264e+00 3.25124770e-01
-3.26691121e-01 1.32296637e-01 -8.02454352e-02 -2.10734293e-01
-1.51969418e-01 4.65089172e-01 -1.71067759e-01 -6.74977228e-02
3.58304948e-01 -8.93849552e-01 -4.58375305e-01 -2.72693813e-01
-8.71984959e-01 2.13695969e-02 5.66880286e-01 1.08984709e+00
6.07536435e-01 -5.23988545e-01 6.99770033e-01 -1.41538012e+00
9.06234264e-01 -7.63871253e-01 -4.99049813e-01 3.99158239e-01
-6.25037134e-01 4.46445227e-01 6.84723794e-01 -4.01080042e-01
-3.57579082e-01 -3.41991824e-03 1.33815333e-01 -4.94567245e-01
1.87528670e-01 9.83196735e-01 1.80210695e-01 -3.41732472e-01
8.18800747e-01 7.05233589e-03 1.61348417e-01 -2.69732445e-01
7.30599582e-01 1.16676651e-01 5.25746584e-01 -5.46441138e-01
7.01234341e-01 3.40124488e-01 3.60301256e-01 -4.74552780e-01
-8.85354996e-01 -2.85722822e-01 -6.17702484e-01 1.28247842e-01
9.66205418e-01 -7.54660130e-01 -1.06709814e+00 1.98344156e-01
-1.00303519e+00 -3.69570643e-01 -1.86782673e-01 2.18713984e-01
-9.04165804e-02 2.44753063e-01 -9.72631752e-01 -2.77643323e-01
-5.70092678e-01 -9.89692211e-01 8.52100611e-01 -1.38660207e-01
-5.49627095e-02 -1.42723560e+00 -4.62982804e-02 2.00152379e-02
5.61253548e-01 5.42679727e-01 1.45737541e+00 -8.63306224e-01
-6.57977402e-01 -1.55539632e-01 -5.58500290e-01 -7.25356638e-02
3.13329101e-01 -6.35889173e-03 -8.82864773e-01 -3.39369684e-01
-9.47329104e-01 -6.43820584e-01 1.37991524e+00 3.63110185e-01
1.33497858e+00 -4.00103599e-01 -6.01958811e-01 8.03911984e-01
1.15970087e+00 -3.80281687e-01 5.14915347e-01 -1.82129323e-01
1.31860805e+00 1.78089693e-01 -4.05540377e-01 -1.32435247e-01
3.52311999e-01 3.07863504e-01 6.90512359e-01 -5.49741983e-01
-3.61633375e-02 -4.29813892e-01 6.04894757e-02 7.64177620e-01
-1.02702551e-01 -6.01220548e-01 -9.00717080e-01 5.33439219e-01
-1.93623865e+00 -7.24337816e-01 -3.04428518e-01 1.89313281e+00
9.15903568e-01 9.53206345e-02 -6.76378459e-02 -3.95371825e-01
7.02453434e-01 3.85770112e-01 -7.75337279e-01 -4.61481184e-01
-1.55632421e-01 7.03543901e-01 5.20278811e-01 4.78298992e-01
-1.13120544e+00 1.07192624e+00 7.06487703e+00 4.36873198e-01
-1.12227237e+00 -1.48891523e-01 6.98129177e-01 -1.33062080e-01
-7.48706639e-01 7.98733309e-02 -7.83113018e-02 8.53818431e-02
1.01380718e+00 -6.37292787e-02 9.45634663e-01 4.08217877e-01
-2.94040203e-01 4.21156168e-01 -1.41474485e+00 6.90155745e-01
-2.20946595e-01 -2.01975894e+00 3.33375305e-01 1.83662489e-01
5.78519404e-01 6.97019398e-01 -2.37756521e-02 4.38793272e-01
9.34736252e-01 -1.57671106e+00 1.71548948e-02 5.64076245e-01
9.01756942e-01 -4.25732106e-01 5.20554543e-01 -8.51890668e-02
-1.13335001e+00 3.94969255e-01 -2.26549745e-01 -1.36716604e-01
-2.72529930e-01 7.66344726e-01 -1.05443704e+00 6.38412774e-01
5.24050772e-01 1.03951168e+00 -5.48007786e-01 4.87907559e-01
-2.20585167e-01 7.77959824e-01 -3.08543831e-01 -2.15889037e-01
5.28833091e-01 -1.94254830e-01 2.65860170e-01 1.18031085e+00
-1.22562528e-01 5.13601117e-03 2.95995861e-01 1.28040171e+00
-8.64319801e-01 4.88314182e-02 -8.59511197e-01 -5.57744265e-01
2.82437235e-01 1.26107919e+00 -5.90045035e-01 -2.87551433e-01
-5.48220992e-01 8.07353199e-01 1.03167772e+00 7.40885556e-01
-6.49442375e-01 -4.38035011e-01 6.15546286e-01 2.13444963e-01
2.53023684e-01 -7.18148192e-03 9.57693066e-03 -1.17373323e+00
-1.70886531e-01 -6.39213204e-01 6.15510046e-01 -7.63610184e-01
-1.55645692e+00 4.69088316e-01 -3.40279549e-01 -5.32075822e-01
9.40855667e-02 -7.83616364e-01 -6.70054436e-01 1.00059021e+00
-1.44373417e+00 -1.37610972e+00 -8.78550410e-02 6.87368095e-01
-3.62663150e-01 4.60104495e-02 9.98678267e-01 3.42926830e-01
-4.96773452e-01 5.76559722e-01 -8.98853466e-02 4.00424093e-01
5.16640484e-01 -1.32626688e+00 8.87433052e-01 4.53576922e-01
3.26839596e-01 9.67335045e-01 2.80280799e-01 -5.36576688e-01
-1.63337743e+00 -1.36488950e+00 8.16846073e-01 -4.12139028e-01
1.00446022e+00 -6.63319767e-01 -1.22971678e+00 1.08493578e+00
1.79235891e-01 7.36594975e-01 5.96135378e-01 7.54020214e-01
-6.50467992e-01 1.59475133e-01 -8.52434576e-01 5.01693249e-01
1.60345852e+00 -8.69771898e-01 -1.98997855e-01 7.65893757e-01
1.03699827e+00 -3.14589262e-01 -1.02544606e+00 4.89789128e-01
1.72559202e-01 -5.67728698e-01 1.08008981e+00 -1.39070857e+00
4.50723648e-01 -1.06523238e-01 1.82642758e-01 -1.39143336e+00
-8.48306477e-01 -7.83424199e-01 -2.50326514e-01 8.35604727e-01
7.63685465e-01 -9.00014341e-01 7.71982491e-01 3.19180608e-01
-4.25351620e-01 -9.33814526e-01 -8.32485616e-01 -2.04741642e-01
1.98209528e-02 -4.13852967e-02 6.41534984e-01 1.35452700e+00
1.10851951e-01 6.53482676e-01 -8.21592435e-02 8.85734186e-02
3.92576367e-01 3.43584716e-02 5.12833774e-01 -1.38763869e+00
-3.62729669e-01 -5.53140640e-01 -4.80465978e-01 -7.68098056e-01
4.82315779e-01 -1.52173913e+00 -1.81476280e-01 -1.80007696e+00
2.35530630e-01 -3.84167284e-01 -7.62337208e-01 1.00552607e+00
-4.71473694e-01 3.98397297e-01 -3.65209460e-01 -3.64075191e-02
-7.04737484e-01 5.06428897e-01 1.29990768e+00 -5.34877300e-01
4.28141356e-02 -4.87052441e-01 -1.10066366e+00 4.34910178e-01
5.57016551e-01 -4.00819451e-01 -5.40162623e-01 -6.25666738e-01
5.53269327e-01 -7.82460868e-02 5.22730589e-01 -5.13967514e-01
2.47032046e-01 1.99661870e-02 3.40935975e-01 -2.39525571e-01
1.44668356e-01 -2.74503708e-01 2.19929218e-02 4.12334234e-01
-5.78639328e-01 4.99004833e-02 3.44758987e-01 9.14593399e-01
4.26533520e-02 3.67814541e-01 4.59486723e-01 -1.74658582e-01
-3.63360018e-01 8.00279140e-01 -1.24793909e-01 2.25484535e-01
6.46149874e-01 7.33646676e-02 -6.85682714e-01 -3.50748807e-01
-8.57592106e-01 2.41038039e-01 3.49643111e-01 2.94052213e-01
5.22703230e-01 -1.34658587e+00 -7.32642353e-01 1.86099246e-01
3.79658550e-01 1.79391921e-01 -6.81592971e-02 8.11126709e-01
-4.27918434e-01 3.50117683e-01 5.16390987e-02 -5.47572911e-01
-7.06990898e-01 8.97982657e-01 3.76230717e-01 -4.54003721e-01
-5.67805707e-01 1.11611104e+00 4.09623832e-01 -8.45110416e-01
-6.96455827e-03 -7.18313694e-01 -3.93466204e-02 -2.83828765e-01
-1.38666900e-02 -2.78955400e-01 2.01604143e-01 -2.60295898e-01
-5.56682765e-01 1.95174754e-01 -2.63826668e-01 6.16719544e-01
1.37938225e+00 4.38064218e-01 -5.70551157e-01 2.55532295e-01
1.27906048e+00 -2.32576787e-01 -7.73449600e-01 -5.46659172e-01
-7.22564161e-02 2.43506253e-01 2.19839633e-01 -7.24553764e-01
-1.07437360e+00 8.08041811e-01 1.30196540e-02 3.09214562e-01
7.85625756e-01 2.11402729e-01 6.13518476e-01 7.51065671e-01
5.02295084e-02 -5.55767059e-01 -7.03076944e-02 5.15020370e-01
7.63345778e-01 -1.23396158e+00 1.30740985e-01 -4.95282888e-01
-4.43203337e-02 9.29259956e-01 4.49977309e-01 -2.09822342e-01
7.59829938e-01 1.24354087e-01 -4.08373445e-01 -7.61341631e-01
-1.02696848e+00 -2.83524305e-01 5.61922967e-01 5.83457887e-01
7.88988411e-01 1.64542630e-01 1.17663093e-01 9.66511369e-02
2.99571484e-01 -4.58211415e-02 2.19560817e-01 6.41420007e-01
-4.25456017e-02 -9.38628614e-01 2.25977778e-01 8.88990402e-01
-3.41033578e-01 -7.86718845e-01 -8.23313534e-01 7.41303623e-01
-3.41499418e-01 5.99165201e-01 7.14209974e-02 -4.18715954e-01
4.67668287e-02 -5.17449081e-02 4.66233015e-01 -9.04088438e-01
-6.46661341e-01 -3.83495927e-01 3.98625523e-01 -5.67148268e-01
-4.67994124e-01 -1.48455501e-01 -1.35383737e+00 -4.77948546e-01
-4.00897861e-01 1.72901556e-01 1.90778032e-01 8.09165716e-01
7.88920403e-01 9.36633527e-01 6.88234866e-02 -8.03900242e-01
-2.73192674e-01 -1.09034359e+00 -2.89719731e-01 6.06016934e-01
3.82632375e-01 -4.93717641e-01 -1.86304018e-01 -2.12934896e-01] | [6.869557857513428, 6.286044597625732] |
4d0ce8f0-71b6-4db7-9597-3c669cad1dc5 | attacking-pre-trained-recommendation | 2305.03995 | null | https://arxiv.org/abs/2305.03995v1 | https://arxiv.org/pdf/2305.03995v1.pdf | Attacking Pre-trained Recommendation | Recently, a series of pioneer studies have shown the potency of pre-trained models in sequential recommendation, illuminating the path of building an omniscient unified pre-trained recommendation model for different downstream recommendation tasks. Despite these advancements, the vulnerabilities of classical recommender systems also exist in pre-trained recommendation in a new form, while the security of pre-trained recommendation model is still unexplored, which may threaten its widely practical applications. In this study, we propose a novel framework for backdoor attacking in pre-trained recommendation. We demonstrate the provider of the pre-trained model can easily insert a backdoor in pre-training, thereby increasing the exposure rates of target items to target user groups. Specifically, we design two novel and effective backdoor attacks: basic replacement and prompt-enhanced, under various recommendation pre-training usage scenarios. Experimental results on real-world datasets show that our proposed attack strategies significantly improve the exposure rates of target items to target users by hundreds of times in comparison to the clean model. | ['Qing He', 'Yongjun Xu', 'Jie zhou', 'Fuzhen Zhuang', 'Yongchun Zhu', 'Zhao Zhang', 'Ruobing Xie', 'Yiqing Wu'] | 2023-05-06 | null | null | null | null | ['sequential-recommendation'] | ['miscellaneous'] | [ 1.48915350e-01 -3.59615028e-01 -4.51051861e-01 -5.26500382e-02
-2.45483294e-01 -1.24110532e+00 4.53209519e-01 -3.03608268e-01
5.66182323e-02 2.98790783e-01 2.35320166e-01 -9.82713044e-01
-4.16234940e-01 -8.89830887e-01 -7.63707697e-01 -5.94083965e-01
-3.29023093e-01 -2.01948792e-01 2.85452276e-01 -5.96525848e-01
4.12243694e-01 3.08698207e-01 -1.27068007e+00 4.44594979e-01
7.95534790e-01 6.52672470e-01 -1.06245382e-02 4.56590056e-01
3.24379921e-01 2.23082289e-01 -5.54892361e-01 -7.34235466e-01
7.79314756e-01 -1.34521469e-01 -3.36396158e-01 -4.76723671e-01
5.39862216e-01 -6.51115298e-01 -6.39103591e-01 1.13095713e+00
2.87796021e-01 2.47669369e-01 2.61216015e-01 -9.75038528e-01
-9.05076265e-01 1.22912157e+00 -4.03699368e-01 3.47581416e-01
2.83212632e-01 -2.68158615e-01 1.13120735e+00 -6.62082374e-01
8.96315463e-03 8.53621781e-01 6.46299064e-01 4.80440199e-01
-8.20837438e-01 -1.11399114e+00 5.61347008e-01 -1.81285799e-01
-1.10936332e+00 -2.36422509e-01 2.55794019e-01 -1.04881160e-01
5.02550364e-01 7.28154540e-01 2.34833449e-01 1.43452597e+00
6.83173016e-02 4.65646148e-01 9.32249665e-01 -1.60053328e-01
-6.90676644e-02 5.61769724e-01 7.51364291e-01 3.27895522e-01
9.17465150e-01 7.73613393e-01 -1.96115121e-01 -7.58339942e-01
7.02732921e-01 7.39463449e-01 -5.03322184e-01 -7.80852791e-03
-8.15183043e-01 8.67647827e-01 2.61244982e-01 1.67359412e-01
1.07837707e-01 -3.10872555e-01 3.83848071e-01 5.62506318e-01
1.65978044e-01 5.34575462e-01 -5.23051083e-01 4.40648139e-01
-6.04044020e-01 2.44393602e-01 8.04879725e-01 1.06445658e+00
1.60611555e-01 9.33086500e-03 -1.48984212e-02 3.12744886e-01
3.43643129e-01 6.02434218e-01 3.53923023e-01 -1.90767929e-01
5.45057654e-01 1.30132928e-01 2.38559693e-01 -1.22179973e+00
7.15329172e-03 -1.00770426e+00 -7.33972013e-01 -3.65115315e-01
4.07190472e-01 -4.48992133e-01 -4.39026535e-01 1.50609267e+00
4.11411881e-01 8.08125436e-01 1.02840953e-01 8.73970032e-01
3.60235423e-01 7.96020806e-01 -2.36091137e-01 -1.84001699e-01
1.26662827e+00 -1.23380256e+00 -4.40969497e-01 1.84684753e-01
7.27635503e-01 -8.33073676e-01 1.07789004e+00 9.57724452e-01
-4.92237151e-01 -7.22341120e-01 -1.33059502e+00 5.89776337e-01
-3.42806339e-01 3.19160074e-02 8.94864738e-01 1.53389049e+00
-3.91354293e-01 7.03086197e-01 -2.74621695e-01 -2.89356977e-01
1.08534589e-01 5.84888518e-01 -1.85710654e-01 -1.12839371e-01
-1.40346217e+00 3.61584455e-01 8.27636719e-02 2.19628081e-01
-1.02105641e+00 -8.99427950e-01 -9.90578085e-02 2.54116267e-01
7.49115288e-01 -5.04912317e-01 1.04205298e+00 -4.20965940e-01
-1.67646790e+00 -2.44065791e-01 5.10573447e-01 -4.28868234e-01
6.37761056e-02 -9.00473356e-01 -1.09868062e+00 -3.46160144e-01
-4.93648082e-01 -5.67921221e-01 1.13987541e+00 -1.18610764e+00
-9.85765755e-01 -3.09774518e-01 6.27057612e-01 -6.65517598e-02
-9.75965202e-01 1.67502075e-01 -1.19555548e-01 -8.11536372e-01
-2.57787883e-01 -1.04743731e+00 -5.39890528e-01 -7.36494482e-01
-5.07373452e-01 1.84756935e-01 9.47276831e-01 -2.99595505e-01
1.73123765e+00 -2.24318743e+00 -2.86923856e-01 7.42144644e-01
2.01838359e-01 8.35489869e-01 -2.73123443e-01 6.91173315e-01
1.08735813e-02 3.68785053e-01 5.03989100e-01 2.01751478e-02
-7.68788084e-02 9.50361937e-02 -1.44157922e+00 5.50279856e-01
-6.75900519e-01 3.26622337e-01 -9.56636071e-01 4.66290146e-01
5.38234971e-02 3.83033097e-01 -9.74273264e-01 4.68508303e-01
6.03161715e-02 2.76414126e-01 -7.93998301e-01 3.04663509e-01
1.03210485e+00 4.77540828e-02 4.71535623e-01 -1.78866148e-01
-1.33059353e-01 6.78034246e-01 -1.20585191e+00 1.19318497e+00
-2.51250684e-01 -2.70869970e-01 -1.01454273e-01 -6.91497266e-01
7.34589994e-01 1.90012485e-01 3.01517313e-03 -3.31386417e-01
1.50239870e-01 1.21458285e-02 9.05364081e-02 -1.98285788e-01
7.72418737e-01 2.37960413e-01 -5.87521382e-02 8.30504060e-01
-1.95962891e-01 8.11942518e-01 -2.74915993e-01 5.22341609e-01
9.25050855e-01 -1.34599775e-01 -3.56265046e-02 -3.34774315e-01
6.14807069e-01 -5.34649134e-01 2.77862877e-01 1.34167457e+00
3.14651430e-01 1.65624946e-01 2.60743536e-02 -4.14414287e-01
-4.44069058e-01 -9.45695758e-01 -8.92093033e-02 1.46658278e+00
4.61214483e-01 -1.20189047e+00 -4.83235717e-01 -1.41636932e+00
1.64412826e-01 6.06159270e-01 -4.74996358e-01 -3.87525111e-01
-5.07676244e-01 -6.29520357e-01 7.36661792e-01 2.25570768e-01
1.18284389e-01 -3.56633186e-01 4.59288545e-02 1.38452277e-01
2.50779003e-01 -8.46037924e-01 -8.86129677e-01 -5.77070937e-02
-9.69734073e-01 -1.07830012e+00 -2.34483451e-01 -3.61336529e-01
8.61760855e-01 1.30170619e+00 5.07174492e-01 5.33935308e-01
3.65430772e-01 1.86883852e-01 -7.93595850e-01 -4.97735515e-02
-2.76447147e-01 1.14249058e-01 6.20244026e-01 1.99707448e-01
1.50610939e-01 -6.46071494e-01 -7.65677214e-01 8.32431912e-01
-8.26529086e-01 -5.12366414e-01 2.76798576e-01 6.76033318e-01
4.66071814e-02 2.25833714e-01 6.15074754e-01 -1.50465786e+00
6.62074924e-01 -8.42543483e-01 -6.08946860e-01 3.58406365e-01
-1.00258994e+00 -2.70882517e-01 1.25289166e+00 -8.38948905e-01
-9.86374497e-01 -5.93088090e-01 -2.94983357e-01 -1.94722444e-01
2.94434149e-02 4.29017007e-01 -1.78900644e-01 -2.48438507e-01
7.14037597e-01 9.70367715e-02 -3.51409376e-01 -9.14529860e-01
7.77390122e-01 7.55565763e-01 4.77383286e-01 -6.84170663e-01
1.19173646e+00 3.82978737e-01 -3.68684828e-01 -4.40076321e-01
-8.90180528e-01 -5.40152431e-01 -9.01566632e-03 1.96214929e-01
-1.09495491e-01 -7.79721200e-01 -6.72828913e-01 1.20264359e-01
-6.69786394e-01 8.96743536e-02 1.18500821e-01 6.13096952e-01
2.70251155e-01 8.88858140e-01 -5.93377948e-01 -4.48985636e-01
-7.06213295e-01 -9.80434299e-01 2.58997798e-01 3.02822053e-01
2.34480381e-01 -8.09016168e-01 2.35552251e-01 3.57760459e-01
4.90156353e-01 -4.93340641e-01 6.71097994e-01 -1.15007186e+00
-4.75064546e-01 -3.75865579e-01 4.10524793e-02 4.18468505e-01
1.26165450e-01 -1.59658983e-01 -7.30608642e-01 -8.35969210e-01
1.86444581e-01 3.27166915e-01 4.96327698e-01 -1.53374374e-01
1.14367890e+00 -7.34466612e-01 -6.38646066e-01 8.99273157e-01
1.18692839e+00 2.01158702e-01 6.40554965e-01 8.15565586e-02
7.20110655e-01 1.75812349e-01 8.15552294e-01 4.81503159e-01
-2.13460580e-01 6.35911942e-01 3.64620149e-01 1.99543491e-01
4.58090901e-01 -7.09233046e-01 7.10920930e-01 8.19819093e-01
-3.04593772e-01 -6.24483645e-01 9.54340026e-03 -3.72408554e-02
-1.73179352e+00 -1.03975248e+00 -3.20709884e-01 2.68440652e+00
3.32203686e-01 3.41227740e-01 2.77310044e-01 -6.57538846e-02
6.87617421e-01 -1.14498228e-01 -2.08844364e-01 -3.24360698e-01
3.67259473e-01 3.03212643e-01 7.73099363e-01 1.62305936e-01
-1.06934428e+00 9.02910948e-01 6.55210876e+00 8.26325715e-01
-1.22346413e+00 2.67460048e-01 6.85401857e-02 -1.56309158e-01
-4.19009060e-01 3.36253971e-01 -1.31979632e+00 5.57497859e-01
1.09475982e+00 -2.59057403e-01 4.59579319e-01 9.94379699e-01
-8.21612179e-02 6.64295197e-01 -9.49249268e-01 5.64772248e-01
8.36244747e-02 -1.26551557e+00 3.07952613e-01 4.70650792e-01
7.04893708e-01 -1.97943464e-01 5.25824487e-01 7.09043741e-01
4.47673470e-01 -5.55277109e-01 4.13941234e-01 -3.10074422e-03
5.41293383e-01 -8.85133445e-01 4.83334273e-01 4.37676370e-01
-9.62932289e-01 -6.07467175e-01 -8.06422174e-01 -2.00543374e-01
8.42061713e-02 4.82700944e-01 -5.90788424e-01 1.02319980e+00
5.77733576e-01 5.93490183e-01 -4.10484344e-01 1.06601250e+00
-2.44070977e-01 1.29085684e+00 -3.98042083e-01 1.24011837e-01
1.50297225e-01 -4.31845158e-01 6.17580652e-01 9.92989182e-01
5.19853294e-01 3.67610872e-01 2.01459408e-01 2.35781029e-01
-1.07340023e-01 3.39427263e-01 -6.06389701e-01 -8.16493183e-02
5.92254698e-01 1.41729939e+00 -4.02393132e-01 -1.13254853e-01
-5.29746950e-01 6.91912889e-01 1.62252486e-01 3.17171901e-01
-1.20786059e+00 -1.71934098e-01 7.79394209e-01 2.12165505e-01
4.40146238e-01 -1.25789791e-01 8.47812742e-02 -1.38745511e+00
-3.86971474e-01 -1.40777910e+00 5.75787008e-01 7.25382939e-02
-1.40858817e+00 5.83178520e-01 -2.48186737e-01 -1.37014794e+00
2.85985351e-01 -4.34483379e-01 -7.92474151e-01 6.45444632e-01
-1.13064265e+00 -9.82086182e-01 2.25274667e-01 8.10419083e-01
1.58971712e-01 -5.22600889e-01 8.52920651e-01 7.38577604e-01
-9.05668199e-01 1.47595012e+00 6.50832772e-01 -1.63757622e-01
9.91166472e-01 -6.49722695e-01 4.99159485e-01 1.25209439e+00
6.01356447e-01 1.74079776e+00 6.14084244e-01 -6.47177041e-01
-1.79083490e+00 -1.10916853e+00 1.77401066e-01 -6.42981887e-01
5.59073150e-01 -5.54462194e-01 -6.66233957e-01 9.55765188e-01
-5.79612851e-02 -2.88999140e-01 1.07076335e+00 6.32426858e-01
-9.80096579e-01 -1.95859998e-01 -9.28167045e-01 9.10230339e-01
1.25982535e+00 -3.45915645e-01 -4.69203502e-01 3.40203732e-01
9.17000413e-01 -1.68907642e-01 -7.37994850e-01 3.97709697e-01
9.38936651e-01 -8.97416174e-01 1.21857905e+00 -1.13628149e+00
-1.20791279e-01 -3.39036196e-01 -1.17929406e-01 -9.54195142e-01
-7.05932975e-01 -1.36999953e+00 -5.41192174e-01 1.18100810e+00
2.96086699e-01 -8.34175527e-01 6.01563454e-01 2.08739221e-01
-2.55399287e-01 -5.37885249e-01 -2.32485324e-01 -7.15008318e-01
-6.90938085e-02 -2.36468896e-01 8.98595273e-01 9.92071986e-01
2.71309555e-01 2.99743176e-01 -1.12562931e+00 8.56500924e-01
4.88704324e-01 2.15115249e-01 1.16426849e+00 -9.73561227e-01
-9.16199565e-01 -4.64423522e-02 5.88524081e-02 -1.72586620e+00
-3.71151745e-01 -8.63151670e-01 -5.78682840e-01 -6.63059950e-01
-1.25778904e-02 -8.05515409e-01 -8.12104881e-01 3.10940146e-01
-2.05785245e-01 2.01369673e-01 1.58968419e-01 2.53159225e-01
-3.82110566e-01 -1.67686231e-02 9.87389207e-01 8.70243162e-02
-2.77703285e-01 6.44406617e-01 -1.43874943e+00 4.38301444e-01
5.78200579e-01 -8.19388449e-01 -9.48573768e-01 -8.50424394e-02
4.25085008e-01 -2.72796899e-01 -1.12608306e-01 -7.74745584e-01
1.45853892e-01 -1.18007250e-01 -1.61470786e-01 -4.46195900e-01
-2.05166996e-01 -1.01679945e+00 3.27285409e-01 4.12101388e-01
-3.34844589e-01 -1.78731918e-01 1.06404265e-02 9.41004455e-01
5.11722922e-01 -2.87848741e-01 2.27224514e-01 3.08390141e-01
-2.23001644e-01 5.42985618e-01 -8.64537284e-02 -5.62936842e-01
8.83656859e-01 -2.39154929e-03 -6.80371702e-01 -1.74498856e-01
-5.11126637e-01 -3.05936113e-02 9.16810185e-02 6.92248225e-01
4.92149204e-01 -9.12064433e-01 -3.86041343e-01 5.82479835e-01
-1.52380437e-01 -8.48141015e-01 5.18676519e-01 6.46596909e-01
1.83892879e-03 4.23269570e-01 -7.56585132e-03 1.51362633e-02
-1.37093973e+00 1.25433922e+00 4.20828834e-02 -4.16018993e-01
-7.74005949e-01 7.82791555e-01 4.46264744e-01 -3.36276799e-01
3.71912539e-01 -2.53866613e-01 -3.54698241e-01 -5.24967074e-01
1.08113110e+00 3.52862239e-01 -9.68753770e-02 -2.07679063e-01
-9.63575691e-02 2.63484120e-01 -9.02339876e-01 4.26096976e-01
9.35460687e-01 -1.21354669e-01 1.02198727e-01 -2.85200477e-01
8.71876478e-01 9.80867445e-01 -6.05823755e-01 -3.30503106e-01
-4.73167837e-01 -8.48721385e-01 3.88882682e-02 -7.18060255e-01
-1.08628201e+00 4.97041643e-01 4.41658229e-01 5.52437425e-01
7.82067418e-01 -5.82443535e-01 1.26731980e+00 5.82899988e-01
7.44896531e-01 -4.01469976e-01 -1.93301648e-01 3.88625771e-01
3.75225961e-01 -6.17413282e-01 1.72994927e-01 -7.54597127e-01
-3.52151811e-01 8.97573054e-01 5.24012029e-01 -4.80786771e-01
1.11945820e+00 1.83044195e-01 -1.24530882e-01 9.27077085e-02
-6.46166384e-01 3.21547985e-01 2.27624744e-01 4.39195484e-01
1.22225888e-01 3.68599072e-02 -4.60229129e-01 1.56825757e+00
-1.02476105e-01 -1.25120163e-01 5.66507578e-01 6.99332118e-01
-4.41810608e-01 -1.67733502e+00 -4.86242980e-01 5.74705005e-01
-8.24337065e-01 -1.89313084e-01 -6.44493802e-03 4.68862206e-01
3.61833610e-02 1.25113797e+00 -4.87386763e-01 -1.21718860e+00
4.07143146e-01 -3.21411729e-01 2.86163598e-01 -7.10141599e-01
-1.17986655e+00 7.83891380e-02 1.82398230e-01 -4.14745599e-01
1.86219841e-01 -1.72976345e-01 -6.61910295e-01 -6.46970332e-01
-1.12802100e+00 7.94036806e-01 2.16070160e-01 8.18797708e-01
6.54174507e-01 1.05637759e-01 1.19922745e+00 -4.63794649e-01
-1.30180025e+00 -7.04737604e-01 -7.50763297e-01 2.67833680e-01
7.41826668e-02 -6.94369555e-01 -5.86770356e-01 -3.25516284e-01] | [5.862672328948975, 7.1338725090026855] |
617a2d63-9b1f-4dd7-a9e2-0e2308a19d95 | dior-cvae-diffusion-priors-in-variational | 2305.15025 | null | https://arxiv.org/abs/2305.15025v1 | https://arxiv.org/pdf/2305.15025v1.pdf | Dior-CVAE: Diffusion Priors in Variational Dialog Generation | Conditional variational autoencoders (CVAEs) have been used recently for diverse response generation, by introducing latent variables to represent the relationship between a dialog context and its potential responses. However, the diversity of the generated responses brought by a CVAE model is limited due to the oversimplified assumption of the isotropic Gaussian prior. We propose, Dior-CVAE, a hierarchical CVAE model with an informative prior produced by a diffusion model. Dior-CVAE derives a series of layer-wise latent variables using attention mechanism and infusing them into decoder layers accordingly. We propose memory dropout in the latent infusion to alleviate posterior collapse. The prior distribution of the latent variables is parameterized by a diffusion model to introduce a multimodal distribution. Overall, experiments on two popular open-domain dialog datasets indicate the advantages of our approach over previous Transformer-based variational dialog models in dialog response generation. We publicly release the code for reproducing Dior-CVAE and all baselines at https://github.com/SkyFishMoon/Latent-Diffusion-Response-Generation. | ['Iryna Gurevych', 'Thy Thy Tran', 'Tianyu Yang'] | 2023-05-24 | null | null | null | null | ['response-generation', 'open-domain-dialog'] | ['natural-language-processing', 'natural-language-processing'] | [-2.33814865e-01 4.50693697e-01 1.28473714e-01 -5.89257360e-01
-7.89773166e-01 -6.52925551e-01 1.01581478e+00 -4.63476002e-01
-1.74638376e-01 8.49384308e-01 8.46690893e-01 2.44258679e-02
3.82946044e-01 -7.30573356e-01 -4.35337424e-01 -7.01015651e-01
6.55227900e-01 9.37729597e-01 4.43801172e-02 -4.47493941e-01
5.94576038e-02 -1.99829027e-01 -9.28514063e-01 4.70903784e-01
8.37431729e-01 5.54989100e-01 5.04687130e-01 7.03608274e-01
-3.64692330e-01 9.47578490e-01 -8.36875200e-01 -7.72904754e-01
-1.78798199e-01 -7.57290244e-01 -9.08598304e-01 6.81963786e-02
-1.80538759e-01 -5.87007344e-01 -5.70818424e-01 9.01947558e-01
4.90362704e-01 3.83455813e-01 1.00616980e+00 -1.38679039e+00
-1.46470845e+00 1.14273751e+00 -1.53320864e-01 -2.41809025e-01
2.56881922e-01 3.06307465e-01 1.13123333e+00 -9.66852486e-01
7.54427552e-01 1.82765901e+00 1.81564674e-01 1.18805099e+00
-1.55881548e+00 -5.64361513e-01 2.51810968e-01 -1.90644041e-02
-1.10038412e+00 -4.64217365e-01 9.50882316e-01 -5.51518440e-01
8.93449485e-01 1.34611964e-01 2.39577249e-01 2.07084799e+00
5.71484081e-02 1.08429170e+00 8.37867796e-01 -5.17925657e-02
2.47093678e-01 6.34052932e-01 1.97286040e-01 2.84587651e-01
-7.17306972e-01 -1.09504387e-01 -5.38402736e-01 -3.77603620e-01
7.80512154e-01 -3.80080305e-02 -3.30984801e-01 -1.80022120e-01
-9.07201469e-01 1.39115977e+00 4.23576146e-01 3.56337465e-02
-4.81833309e-01 1.47670982e-02 2.05496788e-01 3.24720383e-01
5.47593355e-01 2.06998467e-01 -2.73941636e-01 -4.93965335e-02
-5.78485310e-01 5.97788572e-01 9.42790985e-01 1.22618902e+00
6.29546821e-01 4.21612144e-01 -8.75410616e-01 1.21149433e+00
7.28114486e-01 3.50531518e-01 5.80066442e-01 -1.37507153e+00
3.83701771e-01 4.90864038e-01 2.16975346e-01 -5.52251518e-01
2.15006739e-01 7.22251907e-02 -8.33247542e-01 -8.54488276e-03
2.07985789e-01 -4.75316852e-01 -8.06198597e-01 2.10691905e+00
1.70112744e-01 -3.07758003e-01 2.17792347e-01 1.03193629e+00
1.16313183e+00 9.65841353e-01 2.23589107e-01 -5.96886538e-02
1.15032411e+00 -1.15571129e+00 -1.06387901e+00 -1.43220678e-01
3.38708088e-02 -6.63653970e-01 1.36389828e+00 3.56145054e-02
-1.24750483e+00 -5.69679081e-01 -6.71073794e-01 -4.14889634e-01
-1.44853339e-01 1.41458586e-02 4.45039719e-01 2.95961916e-01
-1.21086752e+00 1.04978181e-01 -6.63988531e-01 -7.35596642e-02
-3.09356246e-02 3.00231576e-01 2.58449744e-02 5.10463059e-01
-1.51020610e+00 7.73786306e-01 7.10482821e-02 2.82426942e-02
-1.32324564e+00 -5.29572308e-01 -8.78432095e-01 2.54426360e-01
5.92532381e-02 -9.93636012e-01 1.65127373e+00 -8.25688601e-01
-2.15899920e+00 5.20888507e-01 -3.60508144e-01 -2.93195516e-01
6.55588806e-01 -2.44173229e-01 6.87949285e-02 -2.16165513e-01
-1.54851541e-01 1.18172538e+00 8.84944558e-01 -1.45667875e+00
-1.25102736e-02 3.13741378e-02 1.76626116e-01 2.55008042e-01
-2.01153740e-01 -8.52461159e-02 -4.38277721e-01 -6.68766797e-01
-3.41602921e-01 -1.07535195e+00 -2.06318006e-01 -3.38860810e-01
-4.85385388e-01 -6.23943150e-01 8.07244837e-01 -6.23634636e-01
1.19154263e+00 -2.06377959e+00 8.17356586e-01 -4.54855651e-01
4.42888200e-01 -7.75341541e-02 -1.35739580e-01 7.69559741e-01
3.12501490e-01 8.04266036e-02 -3.90117258e-01 -9.70044255e-01
4.78911787e-01 3.26981664e-01 -7.76040852e-01 -9.41663161e-02
2.15986833e-01 9.99896884e-01 -6.28178120e-01 -3.16780120e-01
1.31661758e-01 9.14800346e-01 -7.46506631e-01 7.55290329e-01
-8.18132937e-01 6.79219186e-01 -4.96825188e-01 2.49790922e-01
6.69058204e-01 -2.98566788e-01 2.34033242e-01 3.15400809e-02
7.02327937e-02 3.93136114e-01 -6.94607198e-01 1.72607696e+00
-4.06427115e-01 6.02107644e-01 4.78800803e-01 -3.27119321e-01
1.08229911e+00 5.20359278e-01 -3.09798084e-02 -2.91770726e-01
3.43981206e-01 -9.86968428e-02 -1.36299714e-01 -3.30410928e-01
9.52544332e-01 -1.19967073e-01 -3.74563962e-01 5.09054124e-01
4.12395030e-01 -2.56527483e-01 -6.35831878e-02 7.71598279e-01
4.50787693e-01 2.79965818e-01 -3.12809855e-01 -1.14735015e-01
5.04701853e-01 -2.30580062e-01 5.63378632e-01 6.50417507e-01
-2.10283756e-01 7.60580838e-01 6.93339229e-01 1.49562746e-01
-1.08326733e+00 -1.24211144e+00 6.15384849e-03 1.26042652e+00
-1.10511869e-01 -3.26690853e-01 -8.51195872e-01 -4.36286867e-01
-8.03206787e-02 1.20820379e+00 -6.72663093e-01 -7.21207783e-02
-4.35613424e-01 -5.55800736e-01 5.63426614e-01 5.76671183e-01
3.58411998e-01 -1.21954370e+00 -1.49843305e-01 3.20449561e-01
-6.00553870e-01 -8.66053820e-01 -6.38937593e-01 1.45311337e-02
-5.85727751e-01 -2.86641330e-01 -1.04954839e+00 -6.57514155e-01
3.58348489e-01 -1.47400245e-01 1.14507699e+00 -3.71487051e-01
2.19264612e-01 3.78305465e-01 -3.07707965e-01 -4.13591601e-02
-9.24556136e-01 -7.52002522e-02 -4.67657372e-02 -5.51110506e-02
5.46566367e-01 -5.26096702e-01 -6.02752686e-01 1.16092943e-01
-7.91204751e-01 8.24454501e-02 1.19880013e-01 1.13511944e+00
2.62311965e-01 -7.36214757e-01 5.15315831e-01 -9.15152490e-01
1.35573649e+00 -8.59113395e-01 -4.98596370e-01 2.02826753e-01
-3.50350022e-01 3.73751670e-01 3.58034521e-01 -7.61058629e-01
-1.76330674e+00 -1.91519588e-01 -2.40499020e-01 -6.46750629e-01
-1.64254278e-01 2.82012343e-01 -1.78504318e-01 7.63420463e-01
4.15763319e-01 2.80458689e-01 -5.65349907e-02 -4.87119079e-01
8.97786736e-01 9.64450002e-01 3.99490207e-01 -7.04065979e-01
2.73865491e-01 8.84819031e-02 -7.90589333e-01 -5.74608386e-01
-3.53540003e-01 1.67837497e-02 -3.16634893e-01 -1.27847180e-01
1.32503033e+00 -9.91548240e-01 -7.08028495e-01 6.09275460e-01
-1.57154071e+00 -7.25692391e-01 -1.05087608e-01 2.53885657e-01
-5.30056536e-01 2.54452854e-01 -1.11613607e+00 -9.95785773e-01
-4.13555324e-01 -1.44839668e+00 9.56779718e-01 3.60077530e-01
-5.90803981e-01 -1.17015398e+00 3.46192420e-01 5.01380146e-01
7.25473881e-01 -1.62316978e-01 9.32884097e-01 -6.63988948e-01
-6.76304758e-01 2.99800962e-01 9.52970758e-02 2.18246743e-01
-1.18437164e-01 1.75592266e-02 -1.21914434e+00 -5.15085459e-02
2.14782998e-01 -7.15564728e-01 8.74011040e-01 5.05462110e-01
7.26342678e-01 -4.74454761e-01 -2.26099581e-01 3.72080088e-01
1.00689816e+00 2.14699417e-01 4.05296981e-01 -2.01398075e-01
5.79973757e-01 7.15518832e-01 8.09380785e-02 6.80169046e-01
8.13949108e-01 6.34279072e-01 2.90985852e-01 1.56637490e-01
-2.49410477e-02 -5.00957429e-01 5.94658017e-01 1.18021703e+00
3.88200790e-01 -8.37010324e-01 -6.00854695e-01 6.32906139e-01
-1.96326888e+00 -8.65955532e-01 -3.29436928e-01 1.67560756e+00
1.13940370e+00 -1.91128582e-01 1.15029059e-01 -7.52665341e-01
6.97035909e-01 4.01584595e-01 -6.35103106e-01 -5.96233547e-01
-2.46893272e-01 -7.24834353e-02 -1.39884129e-01 1.12542915e+00
-6.40506387e-01 1.34377623e+00 5.80655813e+00 5.09718955e-01
-8.89745653e-01 4.40441251e-01 4.98185724e-01 -9.42929834e-02
-9.06339347e-01 1.31637707e-01 -9.10230875e-01 5.08065403e-01
9.70728755e-01 -1.24579966e-01 3.50183427e-01 8.21685910e-01
1.92751497e-01 2.50969470e-01 -1.00599802e+00 5.68846345e-01
-4.75544576e-03 -1.16612256e+00 2.35802591e-01 -1.40831433e-02
7.15029359e-01 -1.14409789e-01 4.76479739e-01 6.90573215e-01
1.16304946e+00 -9.34588373e-01 6.82873368e-01 5.36708593e-01
5.28194547e-01 -3.78687918e-01 4.03652161e-01 2.51382798e-01
-8.15746069e-01 -1.11742802e-02 -5.70102572e-01 1.37745291e-01
4.92613792e-01 -3.13906781e-02 -8.39980721e-01 1.32628113e-01
4.19794410e-01 3.01611990e-01 -6.86891526e-02 1.91485465e-01
-5.47077417e-01 7.41371632e-01 6.94882125e-02 -1.49238691e-01
3.17900598e-01 -4.51682597e-01 7.77987838e-01 1.16775489e+00
7.82471746e-02 1.10762902e-01 -5.51647022e-02 1.84581661e+00
-1.44193575e-01 -2.08156183e-01 -2.65650839e-01 -1.67510688e-01
8.69336009e-01 1.06724417e+00 -1.03144841e-02 -3.55518907e-01
-2.47680530e-01 1.35417151e+00 3.86816949e-01 7.78168023e-01
-1.01279581e+00 -8.65694210e-02 6.12637460e-01 -3.42458546e-01
3.69912416e-01 -1.29776120e-01 -1.19524591e-01 -1.31857347e+00
-3.07921976e-01 -8.31775248e-01 1.94729224e-01 -9.38045859e-01
-1.50459242e+00 1.06384540e+00 2.27002740e-01 -6.90612018e-01
-8.40411127e-01 -2.83559978e-01 -6.95018172e-01 1.49588335e+00
-9.46127117e-01 -1.29225969e+00 -1.78207561e-01 7.50283420e-01
1.19562829e+00 -2.93643624e-01 1.02285755e+00 -5.21148629e-02
-5.34216344e-01 6.79321587e-01 6.14160486e-02 1.88908204e-02
9.21535015e-01 -1.42934167e+00 4.51896846e-01 4.46109861e-01
-2.82951683e-01 9.58466709e-01 1.01897669e+00 -7.41976678e-01
-1.08592606e+00 -7.39355683e-01 9.81394529e-01 -7.41983473e-01
5.98109841e-01 -7.89396226e-01 -1.14762521e+00 9.62876022e-01
1.17115545e+00 -8.35056365e-01 7.25875378e-01 3.13240699e-02
-2.21292764e-01 5.52541733e-01 -8.16562235e-01 7.60610580e-01
5.48527837e-01 -7.01288223e-01 -7.79817879e-01 1.22061074e-01
1.20311069e+00 -3.57473791e-01 -7.38098502e-01 -7.54571520e-03
3.24774206e-01 -8.47535491e-01 7.40810573e-01 -4.54219699e-01
7.18573570e-01 7.49620572e-02 -2.55286098e-01 -1.40133846e+00
-4.82228398e-01 -8.51778626e-01 -3.79499823e-01 1.71708632e+00
5.09144425e-01 -3.07327032e-01 6.36896133e-01 9.42947447e-01
-1.73344478e-01 -4.25855100e-01 -5.61331570e-01 -1.88334599e-01
5.87641001e-01 -6.07037842e-02 5.08487642e-01 9.62409258e-01
-4.12977673e-02 7.94519365e-01 -7.63364494e-01 -1.11095645e-01
4.14277583e-01 -1.04228206e-01 8.39688599e-01 -9.00323093e-01
-6.60663784e-01 -4.53118116e-01 5.48331201e-01 -1.69186783e+00
5.20514667e-01 -7.33062029e-01 1.12023674e-01 -1.52404189e+00
2.98093081e-01 -1.00729018e-01 5.67487516e-02 1.34266138e-01
-3.67238015e-01 -3.52053076e-01 2.53270656e-01 2.99747139e-01
-1.90298915e-01 1.34062636e+00 1.07751131e+00 -2.35177010e-01
-5.80701292e-01 -8.88488144e-02 -6.58766866e-01 4.09130275e-01
7.38896251e-01 -5.39280117e-01 -7.78869390e-01 -7.18951225e-01
-3.25200967e-02 6.18686616e-01 2.65658140e-01 -2.14165598e-01
3.78668427e-01 -2.02309415e-01 7.80538619e-02 -6.31257176e-01
9.40130234e-01 -2.12203041e-01 2.86266714e-01 4.89677005e-02
-9.56213176e-01 6.84103146e-02 -1.89747605e-02 6.01335406e-01
-2.99983621e-01 -1.60061225e-01 5.65002918e-01 -2.85871357e-01
-4.41482604e-01 1.01041228e-01 -7.70161808e-01 1.73902884e-01
5.00050128e-01 1.26316652e-01 -4.19538707e-01 -9.34436381e-01
-8.60474527e-01 4.55235928e-01 2.97430903e-01 7.41041362e-01
7.62326479e-01 -1.32389843e+00 -9.36918378e-01 7.76861683e-02
-1.32188678e-01 -1.57982200e-01 4.81843472e-01 4.81958419e-01
-4.67680395e-02 3.56340110e-01 -1.62223697e-01 -4.50448811e-01
-1.13551795e+00 4.39545423e-01 2.96014130e-01 -2.23667994e-01
-4.45816100e-01 1.36251283e+00 6.28695965e-01 -7.69783497e-01
4.23266172e-01 -3.31335589e-02 -2.75113165e-01 7.58734718e-02
2.28776887e-01 9.87011641e-02 -7.29756176e-01 -6.59318924e-01
-6.91819638e-02 -2.05384448e-01 -2.85768390e-01 -7.97765851e-01
1.09077942e+00 -4.74942952e-01 -6.22329302e-02 6.36516690e-01
9.88120019e-01 2.78381500e-02 -1.65611327e+00 -2.76200444e-01
-4.07191515e-01 -1.06997766e-01 -8.00663307e-02 -8.01964283e-01
-8.35740328e-01 1.19911826e+00 2.46974304e-01 3.62196825e-02
6.66104257e-01 7.42065385e-02 7.68698692e-01 3.18779200e-02
-6.82098418e-02 -9.59570050e-01 3.60643893e-01 8.63542199e-01
1.22047293e+00 -1.17887950e+00 -5.94017386e-01 -2.36522779e-02
-1.52496314e+00 9.09396529e-01 9.07092214e-01 -6.61308914e-02
4.91801113e-01 1.89071313e-01 4.45860922e-01 -6.96577579e-02
-1.33637524e+00 1.97021365e-01 3.47856283e-02 5.87513208e-01
7.87772894e-01 6.98256567e-02 -1.83824241e-01 9.18630540e-01
-3.64138395e-01 -2.89465845e-01 6.02435470e-01 4.46701735e-01
-1.08392417e-01 -1.43467677e+00 -1.39014229e-01 -2.27041647e-01
-3.38072896e-01 -2.74868518e-01 -8.90999496e-01 4.37981933e-01
-3.80085289e-01 1.26697421e+00 4.17084098e-02 -5.08090198e-01
-5.79411276e-02 3.90955895e-01 2.85565406e-02 -4.93934155e-01
-7.78994441e-01 4.56199288e-01 4.65084203e-02 -2.87456334e-01
-9.53834429e-02 -5.15569508e-01 -1.13066506e+00 -2.59665430e-01
-2.58686244e-01 2.98221529e-01 3.61321121e-01 5.57556033e-01
4.69861239e-01 6.90615714e-01 4.74166095e-01 -5.81372499e-01
-1.06084192e+00 -1.47417164e+00 -2.34910727e-01 5.31717002e-01
2.42206171e-01 -5.61117053e-01 -3.78059238e-01 1.88661236e-02] | [12.560476303100586, 8.385916709899902] |
7b51c1c2-71ac-48eb-9062-b76356b6314d | food-ingredients-recognition-through-multi-1 | 2210.14147 | null | https://arxiv.org/abs/2210.14147v1 | https://arxiv.org/pdf/2210.14147v1.pdf | Food Ingredients Recognition through Multi-label Learning | The ability to recognize various food-items in a generic food plate is a key determinant for an automated diet assessment system. This study motivates the need for automated diet assessment and proposes a framework to achieve this. Within this framework, we focus on one of the core functionalities to visually recognize various ingredients. To this end, we employed a deep multi-label learning approach and evaluated several state-of-the-art neural networks for their ability to detect an arbitrary number of ingredients in a dish image. The models evaluated in this work follow a definite meta-structure, consisting of an encoder and a decoder component. Two distinct decoding schemes, one based on global average pooling and the other on attention mechanism, are evaluated and benchmarked. Whereas for encoding, several well-known architectures, including DenseNet, EfficientNet, MobileNet, Inception and Xception, were employed. We present promising preliminary results for deep learning-based ingredients detection, using a challenging dataset, Nutrition5K, and establish a strong baseline for future explorations. | ['Zhaorui Yuan', 'Rameez Ismail'] | 2022-10-24 | null | null | null | null | ['multi-label-learning'] | ['methodology'] | [ 2.76666135e-01 -1.79978400e-01 -1.34209663e-01 -3.57994437e-01
-5.33280134e-01 -7.30377555e-01 3.46700013e-01 8.48742306e-01
-4.20810044e-01 1.84950948e-01 2.69222409e-01 4.98140976e-02
1.07389733e-01 -8.29491854e-01 -8.57305348e-01 -5.89156985e-01
-2.87981331e-01 1.47524923e-01 -1.44800395e-01 -9.95603651e-02
-1.98585689e-02 9.13616195e-02 -1.78615308e+00 8.97916794e-01
5.52394032e-01 1.11384261e+00 2.85246134e-01 7.72953153e-01
-4.36764136e-02 1.02816904e+00 -3.45494241e-01 -4.06967849e-01
1.50049523e-01 -4.36607480e-01 -4.66590136e-01 4.08667177e-02
6.79293811e-01 -4.52709585e-01 6.72378317e-02 1.32903349e+00
8.85125160e-01 -2.56516516e-01 7.43856192e-01 -8.67677271e-01
-1.09838867e+00 9.91788089e-01 -1.80097714e-01 5.81263639e-02
5.27398825e-01 1.98266193e-01 8.63540232e-01 -5.47244549e-01
1.99575037e-01 1.07123876e+00 1.04753971e+00 4.19864565e-01
-1.07616460e+00 -4.09798563e-01 1.51892409e-01 2.22340375e-01
-1.07932878e+00 -3.73942971e-01 5.96796811e-01 -6.34543598e-01
9.84989703e-01 2.12991044e-01 9.02484894e-01 1.16665530e+00
3.32860261e-01 1.01037014e+00 1.24734497e+00 -6.86111152e-01
2.24782407e-01 1.51871800e-01 4.14050579e-01 1.01826489e+00
6.35159612e-01 1.38447300e-01 -3.57044309e-01 1.20623298e-01
5.98347962e-01 -4.79546525e-02 -1.18167505e-01 -4.98241276e-01
-1.05480373e+00 1.13240850e+00 7.15253770e-01 3.93940330e-01
-6.02399647e-01 2.15064645e-01 5.54499507e-01 1.75036967e-01
5.50671935e-01 4.33987588e-01 -4.57402170e-01 6.96520209e-01
-9.40359235e-01 1.34202838e-01 1.07126176e+00 6.53813362e-01
3.24359000e-01 -2.43735701e-01 -4.34436947e-01 6.86773479e-01
7.35730648e-01 4.35284734e-01 5.72839797e-01 -4.62489635e-01
3.61228622e-02 7.37995863e-01 -4.09416631e-02 -8.92273664e-01
-9.37963068e-01 -4.28459287e-01 -8.95558417e-01 5.25570571e-01
3.03123951e-01 -2.02588677e-01 -1.20436478e+00 1.45341253e+00
2.33610272e-01 -1.28977895e-01 1.50120154e-01 9.99319494e-01
1.39188921e+00 3.27264637e-01 5.92688441e-01 3.28119576e-01
1.92510974e+00 -1.46984112e+00 -7.13156164e-01 -4.06094417e-02
4.87270832e-01 -8.64404261e-01 7.74905860e-01 6.43498719e-01
-1.25790417e+00 -8.14846218e-01 -1.63763225e+00 -2.22862139e-01
-9.84333932e-01 5.89594483e-01 5.55982530e-01 8.98460805e-01
-1.15097392e+00 6.84377432e-01 -7.69911230e-01 -5.88011444e-01
1.92356482e-01 3.65178108e-01 -1.65441394e-01 -1.69564068e-01
-1.14112616e+00 9.21339869e-01 6.62935257e-01 2.72953302e-01
-1.22051787e+00 -4.42932814e-01 -1.17118084e+00 1.81112573e-01
1.91175882e-02 -1.02191091e+00 1.44293952e+00 -1.12222004e+00
-1.63402045e+00 1.27854395e+00 6.11333609e-01 -7.16552615e-01
3.36651325e-01 -3.57126534e-01 -6.44439578e-01 3.39049041e-01
-8.14466774e-02 9.43843663e-01 5.17331302e-01 -9.69269216e-01
-5.72690248e-01 -2.49035880e-01 4.25795823e-01 1.50269464e-01
-1.92017332e-01 2.03567937e-01 3.07313669e-02 -5.72820008e-01
-3.62158865e-01 -6.76913559e-01 -1.27288297e-01 2.51785934e-01
-4.75857675e-01 -1.56874612e-01 -3.29989605e-02 -7.47185171e-01
1.08449972e+00 -1.89086998e+00 3.91681865e-02 -9.06916410e-02
2.69313663e-01 3.10761541e-01 -3.16019565e-01 2.18479156e-01
-9.74184722e-02 -1.22338627e-02 2.35423502e-02 -1.29934624e-01
4.13296372e-01 -1.86543033e-01 6.10417664e-01 7.33107269e-01
2.22388193e-01 1.02344728e+00 -9.37422395e-01 -2.60410666e-01
7.07300603e-01 6.42573953e-01 -5.84583580e-01 1.18356422e-01
-6.99938297e-01 -1.26968592e-01 -1.50049150e-01 8.33667874e-01
6.55676842e-01 -4.97041106e-01 3.92651707e-01 -7.19756782e-01
-5.04606068e-01 3.67950611e-02 -1.29895723e+00 2.09392500e+00
-1.10988528e-01 2.41835341e-02 3.80468875e-01 -8.03384721e-01
5.93421042e-01 4.79079366e-01 6.14408374e-01 -8.06642532e-01
5.01103401e-01 2.37784520e-01 8.24165344e-02 -7.73276508e-01
5.12518644e-01 2.19456241e-01 8.88371766e-02 1.76812455e-01
6.10387027e-01 7.42696941e-01 6.38051033e-01 -4.13661182e-01
8.56763363e-01 3.94123673e-01 7.12613225e-01 -8.01042080e-01
5.46460330e-01 2.62645930e-01 -2.60887798e-02 6.05877221e-01
-4.59440678e-01 6.12133443e-01 3.06384154e-02 -8.19950044e-01
-9.69840467e-01 -7.26023912e-01 9.99685898e-02 1.37274432e+00
6.09464459e-02 -3.37201476e-01 -8.45941484e-01 -6.52064621e-01
4.29128818e-02 3.77871245e-01 -8.87268186e-01 -3.33862901e-02
-4.69538897e-01 -1.05211437e+00 4.81690615e-01 5.23115754e-01
8.64066362e-01 -1.24770308e+00 -1.02526534e+00 6.53883636e-01
-3.45917679e-02 -6.00574493e-01 -4.27038282e-01 7.71965563e-01
-3.00697654e-01 -1.53050172e+00 -1.03823602e+00 -1.20008266e+00
1.23870775e-01 2.86375076e-01 1.62074912e+00 -1.28254965e-01
-4.93923306e-01 3.02108586e-01 -3.55635256e-01 -5.56959867e-01
-5.49054623e-01 2.08528817e-01 -5.37847996e-01 -2.83754587e-01
8.61559272e-01 6.24000393e-02 -1.21718454e+00 -3.60467425e-03
-9.07566309e-01 -2.20606513e-02 7.52269924e-01 4.19731855e-01
6.85109556e-01 -1.16872668e-01 2.44875863e-01 -6.74765348e-01
4.58162487e-01 -5.90269566e-01 -3.98686737e-01 4.88370717e-01
-4.41282213e-01 2.46081784e-01 3.32151651e-01 -4.78583723e-01
-5.97838104e-01 5.38627326e-01 -7.15184808e-01 5.22680841e-02
-5.89079201e-01 6.20810688e-01 -2.27510631e-01 -6.41830117e-02
7.89183855e-01 -5.13378531e-02 -1.35824576e-01 -8.38918388e-01
6.46871150e-01 5.03458977e-01 3.93801033e-01 -2.95161545e-01
-1.38860404e-01 1.43213376e-01 -3.27734262e-01 -5.58005571e-01
-8.94429445e-01 -7.92264402e-01 -5.30047476e-01 -3.07497263e-01
1.42975342e+00 -1.34829390e+00 -6.86819375e-01 7.94012725e-01
-9.90176499e-01 -5.32705247e-01 -1.03915773e-01 5.33317327e-01
-3.78588647e-01 8.91737342e-02 -9.90745246e-01 -3.95776600e-01
-8.82837713e-01 -1.10479963e+00 1.04439962e+00 1.28462985e-01
-5.32702766e-02 -1.04114151e+00 4.07062471e-01 3.87634262e-02
4.98773158e-01 5.87707043e-01 8.54361713e-01 -6.25469029e-01
-2.09204797e-02 2.57386416e-01 -4.19772387e-01 3.17317933e-01
8.09885785e-02 -3.06506306e-01 -1.07647765e+00 -3.34017932e-01
-5.44560775e-02 -4.89314944e-01 1.23819602e+00 7.32182086e-01
7.26953089e-01 -6.59172013e-02 -1.71473116e-01 5.10429263e-01
1.87502658e+00 6.67924583e-02 5.02830207e-01 6.27973139e-01
5.09509861e-01 2.56475866e-01 -7.04008639e-02 1.32126629e-01
6.66919410e-01 5.13021410e-01 7.93379128e-01 -7.04094648e-01
-8.00071478e-01 -5.05230315e-02 5.16364157e-01 6.61713541e-01
1.64391249e-01 -5.87444603e-01 -3.51446211e-01 3.75570238e-01
-1.60296810e+00 -6.95882142e-01 -3.73671472e-01 1.84054577e+00
7.05457926e-01 -2.41374090e-01 5.37531257e-01 3.23241726e-02
7.76812732e-01 -2.77274698e-02 -5.93533099e-01 -3.74064744e-01
-2.59414285e-01 3.90417099e-01 7.08234549e-01 5.03518395e-02
-1.79217386e+00 3.97230029e-01 6.99668741e+00 2.76874542e-01
-1.08301699e+00 2.92851031e-01 5.24383545e-01 1.19151689e-01
3.51801217e-01 -1.10052967e+00 -7.97449172e-01 3.65252644e-01
1.17383742e+00 6.47365928e-01 4.41490144e-01 8.97319496e-01
-2.06389338e-01 1.27737790e-01 -1.21565342e+00 7.31645465e-01
5.23695827e-01 -9.95141208e-01 -1.22747026e-01 -1.55072242e-01
5.46553314e-01 4.26621497e-01 -2.57134326e-02 3.34317207e-01
6.13730490e-01 -8.52568150e-01 1.08164716e+00 2.94722974e-01
6.38779104e-01 -3.65636081e-01 7.07594872e-01 -2.95120090e-01
-1.64104593e+00 -1.12819754e-01 -4.51654524e-01 6.86097071e-02
-1.21186204e-01 4.60603833e-01 -2.17113346e-01 6.32328987e-01
6.47434890e-01 7.55242646e-01 -6.76163852e-01 1.52555430e+00
-2.66270220e-01 2.73369998e-01 -4.19944763e-01 -1.24385849e-01
5.05697727e-01 6.71961978e-02 -1.31212041e-01 1.74191570e+00
4.47186738e-01 -2.73310155e-01 5.32420039e-01 6.32005870e-01
-1.26258329e-01 5.68239987e-01 -2.93033391e-01 -1.70329865e-02
-3.79797488e-01 1.56063282e+00 -9.79193866e-01 -3.73913378e-01
-8.34838569e-01 1.18975973e+00 2.31434137e-01 3.12787434e-03
-9.13611233e-01 1.72235332e-02 5.11088192e-01 -2.17951804e-01
5.59604824e-01 6.80241138e-02 7.20635951e-02 -1.05872536e+00
-3.83765996e-01 -1.08808792e+00 8.22065473e-01 -6.62242413e-01
-1.37955999e+00 5.49551606e-01 -3.17820609e-01 -8.64309371e-01
2.25094542e-01 -1.19332469e+00 -1.47719085e-01 6.33194625e-01
-2.01740718e+00 -1.44287860e+00 -5.13325870e-01 4.52768683e-01
5.74557841e-01 1.56559914e-01 1.31556857e+00 8.85438085e-01
-5.66715956e-01 4.17661399e-01 7.89184868e-02 1.96573228e-01
5.73996067e-01 -1.63536668e+00 3.14341962e-01 5.34684122e-01
2.90351063e-01 1.23451091e-01 4.37990934e-01 -5.68602622e-01
-1.38080764e+00 -1.47855377e+00 6.91698968e-01 -3.54894996e-02
3.19837332e-01 -3.65316451e-01 -5.19793272e-01 3.90565813e-01
7.97988832e-01 -4.03720230e-01 9.87567425e-01 5.91168590e-02
-3.75149697e-01 2.39284620e-01 -1.36217773e+00 2.02787355e-01
9.27673340e-01 -1.65328667e-01 -4.45330352e-01 5.12675405e-01
4.50229108e-01 -4.13844258e-01 -9.60450292e-01 3.21759820e-01
9.29998219e-01 -9.51136351e-01 1.18816543e+00 -4.06684816e-01
5.09359419e-01 -2.42549524e-01 -2.78005153e-01 -1.37977684e+00
-8.90664399e-01 2.87412722e-02 -4.22270238e-01 8.61761093e-01
3.32265258e-01 -2.10486457e-01 6.09378338e-01 -1.77587066e-02
-5.01541972e-01 -3.74950826e-01 -2.40982294e-01 -3.54538977e-01
9.73661095e-02 1.76247075e-01 6.08672500e-01 8.34593952e-01
-1.57640949e-01 4.49881345e-01 -4.29096758e-01 3.72178018e-01
5.41766644e-01 1.74714088e-01 1.02373131e-01 -1.16051161e+00
-3.31142038e-01 -6.41233325e-01 -2.63367057e-01 -9.59018648e-01
-3.68808180e-01 -1.14460289e+00 2.22208098e-01 -1.92275596e+00
4.05126840e-01 1.79380149e-01 -9.39257264e-01 5.09405911e-01
-9.74901244e-02 6.26998484e-01 1.72108442e-01 -2.14977384e-01
-8.77792835e-01 -5.88734075e-02 9.02520180e-01 -6.29511178e-01
-1.07132472e-01 -4.64388072e-01 -8.28784823e-01 6.72867119e-01
1.04542398e+00 -3.99187863e-01 -1.88737940e-02 -6.67002916e-01
3.18884671e-01 -6.90410137e-01 5.10608137e-01 -1.46927786e+00
4.35785018e-02 4.99873370e-01 6.72131419e-01 -6.52948916e-01
-6.00031111e-03 -7.51396894e-01 3.30924362e-01 9.67907310e-01
-5.74758053e-01 1.23055084e-02 2.76185215e-01 3.24309945e-01
8.81842673e-02 -4.27945912e-01 6.59830093e-01 -5.99773586e-01
-9.71711278e-01 9.68624577e-02 -3.74664158e-01 -4.57700104e-01
7.59724557e-01 1.78393841e-01 -3.48531783e-01 4.34881955e-01
-9.81392443e-01 1.57233268e-01 1.97151437e-01 2.32258797e-01
1.81081906e-01 -1.43515086e+00 -9.65770483e-01 2.42776304e-01
3.21806371e-01 -8.24798882e-01 2.87125707e-01 5.01231790e-01
-8.46089125e-01 5.66837788e-01 -5.35854518e-01 -4.17222708e-01
-1.12593794e+00 1.06829703e+00 6.16684854e-01 -6.54660344e-01
-6.27138555e-01 6.02295220e-01 1.77463681e-01 -2.28911653e-01
5.00955105e-01 -1.08222389e+00 -6.56002522e-01 2.32983544e-01
8.88718605e-01 3.12053144e-01 3.86948109e-01 -3.38579953e-01
-6.26389906e-02 4.54546452e-01 1.30781382e-01 6.11239195e-01
1.24766243e+00 -1.51102722e-01 2.51506835e-01 2.44157076e-01
9.90969718e-01 -6.70574605e-01 -1.10592687e+00 5.40978946e-02
-2.82057542e-02 2.86649257e-01 4.00143087e-01 -1.51214623e+00
-1.20897853e+00 8.05520296e-01 1.52927399e+00 6.16387665e-01
1.12483811e+00 -2.60749429e-01 7.63568163e-01 1.18673623e-01
1.79194897e-01 -8.61541867e-01 -7.10310740e-03 8.60235170e-02
7.18466759e-01 -1.50185287e+00 -2.12231755e-01 -4.37054485e-01
-3.72641198e-02 1.33568108e+00 2.58518368e-01 -2.53305316e-01
7.09374011e-01 3.43839139e-01 2.12378323e-01 -6.51446462e-01
-1.79355472e-01 -7.34770298e-01 4.32941079e-01 5.89493573e-01
8.74340534e-01 2.60171086e-01 -3.33038777e-01 5.84549427e-01
2.77686000e-01 4.68327284e-01 -2.76582409e-02 7.42075980e-01
-6.62595332e-01 -9.31073308e-01 -2.34646425e-01 2.85249859e-01
-7.37791836e-01 -3.88740540e-01 -1.99076146e-01 7.61418521e-01
7.51015007e-01 1.06795740e+00 -1.83355004e-01 -4.80837226e-02
4.85615373e-01 2.10074842e-01 8.19473565e-01 -6.44461811e-01
-1.31831622e+00 3.05174321e-01 4.25736308e-01 -4.65796798e-01
-1.02863252e+00 -5.71084976e-01 -5.49577773e-01 -9.48263109e-02
-1.76211193e-01 -2.32826516e-01 8.31559598e-01 5.66286147e-01
6.70416132e-02 9.38322067e-01 -6.64353464e-03 -9.09303665e-01
-5.52182555e-01 -1.16315699e+00 -6.80643320e-01 4.10905600e-01
1.97301432e-01 -5.52302241e-01 6.59420416e-02 3.25220585e-01] | [11.5648832321167, 4.397433757781982] |
61381627-c841-40ca-b2af-a784f61fab3d | privacy-preserving-deep-learning-via-weight | 1809.03272 | null | http://arxiv.org/abs/1809.03272v3 | http://arxiv.org/pdf/1809.03272v3.pdf | Privacy-Preserving Deep Learning via Weight Transmission | This paper considers the scenario that multiple data owners wish to apply a
machine learning method over the combined dataset of all owners to obtain the
best possible learning output but do not want to share the local datasets owing
to privacy concerns. We design systems for the scenario that the stochastic
gradient descent (SGD) algorithm is used as the machine learning method because
SGD (or its variants) is at the heart of recent deep learning techniques over
neural networks. Our systems differ from existing systems in the following
features: {\bf (1)} any activation function can be used, meaning that no
privacy-preserving-friendly approximation is required; {\bf (2)} gradients
computed by SGD are not shared but the weight parameters are shared instead;
and {\bf (3)} robustness against colluding parties even in the extreme case
that only one honest party exists. We prove that our systems, while
privacy-preserving, achieve the same learning accuracy as SGD and hence retain
the merit of deep learning with respect to accuracy. Finally, we conduct
several experiments using benchmark datasets, and show that our systems
outperform previous system in terms of learning accuracies. | ['Tran Thi Phuong', 'Le Trieu Phong'] | 2018-09-10 | null | null | null | null | ['privacy-preserving-deep-learning', 'privacy-preserving-deep-learning'] | ['methodology', 'natural-language-processing'] | [-4.63221669e-02 2.58495867e-01 -1.46594375e-01 -6.78100109e-01
-7.18510747e-01 -8.38423193e-01 3.54336232e-01 1.82658896e-01
-1.02185929e+00 9.83835399e-01 -2.36289218e-01 -4.80631769e-01
-9.19094309e-02 -8.83602858e-01 -1.08318770e+00 -1.20906365e+00
-2.47195587e-01 9.02684480e-02 -2.57739544e-01 1.21731617e-01
-1.90630957e-01 7.48804152e-01 -9.53663349e-01 1.95305236e-02
5.47018051e-01 1.36850047e+00 -4.45629746e-01 3.17432821e-01
2.46050626e-01 6.60441637e-01 -6.22176290e-01 -1.06858933e+00
1.11134481e+00 -7.59660825e-02 -7.77035594e-01 -4.90269065e-01
5.12578249e-01 -8.91342700e-01 -4.53859359e-01 1.53496420e+00
4.08405095e-01 -1.45146057e-01 1.84237450e-01 -1.68167233e+00
-5.96044779e-01 7.28535354e-01 -6.02449358e-01 -4.67521369e-01
-3.56004208e-01 -1.24697126e-01 1.05977929e+00 -2.68964499e-01
4.75495577e-01 7.61297107e-01 5.93938112e-01 8.87245238e-01
-1.29892826e+00 -9.29581702e-01 6.17109686e-02 -1.03860542e-01
-1.35263336e+00 -5.20959437e-01 7.26768911e-01 3.75377424e-02
5.81512213e-01 5.94665170e-01 1.75512940e-01 8.31821918e-01
1.76872924e-01 1.06292808e+00 1.10158300e+00 -1.63913205e-01
5.71421087e-01 6.27952158e-01 9.22345072e-02 5.77786803e-01
6.04319632e-01 -8.97541270e-03 -5.04752159e-01 -9.12343860e-01
3.58930558e-01 3.38703007e-01 -4.69964772e-01 -8.04532230e-01
-6.01455331e-01 1.11725354e+00 3.93940836e-01 5.80378659e-02
-3.14521223e-01 3.50798517e-01 4.64986622e-01 9.04309154e-01
4.95691001e-01 1.54011786e-01 -5.85324168e-01 4.53734159e-01
-8.78435433e-01 5.91935337e-01 1.11951947e+00 1.07653582e+00
7.85120726e-01 -1.85968444e-01 6.59701973e-02 3.09069097e-01
-4.44276966e-02 2.36298367e-01 3.79221559e-01 -1.28724253e+00
7.13419437e-01 2.92510390e-01 3.05434048e-01 -9.97320771e-01
9.76957381e-02 -3.95316511e-01 -1.19655633e+00 5.66933215e-01
4.54013318e-01 -6.43809676e-01 -2.79722482e-01 2.13630414e+00
1.80488303e-01 -3.36827993e-01 4.41931367e-01 7.23111749e-01
3.53728592e-01 3.88058543e-01 -2.49351099e-01 -1.93243086e-01
9.59208012e-01 -6.29129648e-01 -4.77782398e-01 3.59710716e-02
8.36406350e-01 6.78216442e-02 5.53641200e-01 2.82494336e-01
-1.14063275e+00 2.24258438e-01 -1.03911233e+00 -2.56163567e-01
-4.81098592e-01 -2.67029721e-02 6.59704387e-01 1.04914057e+00
-1.25371122e+00 8.10942948e-01 -6.02919459e-01 1.20998196e-01
1.19369233e+00 9.02300835e-01 -9.03861046e-01 8.89897496e-02
-1.11496699e+00 2.96768486e-01 2.12238178e-01 3.96650396e-02
-7.62982666e-01 -6.29248440e-01 -5.78892171e-01 1.42869547e-01
9.10005718e-02 -4.91619438e-01 1.10703635e+00 -9.86752987e-01
-1.19296670e+00 1.21879780e+00 -6.68565417e-03 -1.03039134e+00
1.30083454e+00 -5.48023693e-02 1.93648040e-01 -8.23569819e-02
-1.98085278e-01 2.90405780e-01 5.73984802e-01 -1.08280754e+00
-9.47087944e-01 -9.39427853e-01 2.84314781e-01 -9.14055761e-03
-8.87833595e-01 -1.92369744e-01 2.65266486e-02 -3.02371204e-01
-1.32536978e-01 -9.31821704e-01 -2.80447900e-01 7.26132631e-01
-5.28482258e-01 -7.17963353e-02 1.12774813e+00 -5.06750226e-01
8.02979469e-01 -2.22605848e+00 -2.09812671e-01 4.31349516e-01
5.54647744e-01 4.19163853e-01 1.78838655e-01 1.82513878e-01
3.13602269e-01 3.14583004e-01 -4.96022403e-01 -9.24532831e-01
2.71010071e-01 2.07329914e-01 -4.19113278e-01 9.75814342e-01
-4.52749550e-01 8.91205788e-01 -5.03484070e-01 -3.01648647e-01
-2.64740795e-01 1.32680729e-01 -3.94191712e-01 1.23782076e-01
7.61089250e-02 1.14461102e-01 -6.73588455e-01 1.74273804e-01
1.22436666e+00 -1.19792111e-01 2.11123705e-01 1.38931274e-01
2.33608708e-01 -5.97640872e-02 -1.18552518e+00 1.35755050e+00
-1.88221574e-01 5.56446016e-01 6.31670952e-01 -9.78466332e-01
7.84642458e-01 3.63558412e-01 4.78010863e-01 -3.45291495e-01
2.50273108e-01 2.21811771e-01 -3.78877133e-01 -9.29108337e-02
3.09086181e-02 2.37226292e-01 -8.82219449e-02 8.85783792e-01
-1.81464478e-01 3.25287163e-01 -6.50494039e-01 1.90362051e-01
1.05379832e+00 -5.02907515e-01 4.47326675e-02 -3.70323360e-01
3.55678529e-01 -3.70124191e-01 8.05515707e-01 1.14287305e+00
-3.47612053e-01 4.43785697e-01 7.67589271e-01 -6.96189821e-01
-1.02116430e+00 -6.65036559e-01 -5.24454610e-03 1.02890456e+00
1.03072427e-01 2.81569418e-02 -9.82421398e-01 -1.02767658e+00
4.35971290e-01 6.20682955e-01 -8.55410635e-01 -7.79717341e-02
-8.66784453e-02 -7.12592304e-01 8.24252963e-01 2.40700558e-01
9.28021848e-01 -6.34755015e-01 -7.22034693e-01 -1.97861597e-01
2.28770629e-01 -6.84535384e-01 -5.61260283e-01 3.58988523e-01
-7.34984994e-01 -7.35861957e-01 -7.90129185e-01 -5.54656982e-01
8.69063377e-01 2.44547635e-01 4.54753190e-01 -3.57199796e-02
1.47629455e-01 6.76357746e-02 2.27437764e-01 -7.62158096e-01
-1.59756497e-01 1.36347353e-01 1.11200465e-02 5.21969736e-01
4.26211864e-01 -7.21467316e-01 -6.40042007e-01 -1.29622489e-01
-1.04143023e+00 -2.20869333e-01 2.49389246e-01 6.54370189e-01
5.52922070e-01 1.27934471e-01 3.84007782e-01 -1.60137284e+00
8.40995073e-01 -3.87321174e-01 -9.66359437e-01 3.54749739e-01
-7.34428048e-01 6.19219430e-02 1.04216504e+00 -3.67907703e-01
-8.45171332e-01 2.83860981e-01 1.95569158e-01 -6.94439113e-01
-1.44619839e-02 -1.07714601e-01 -4.20557022e-01 -4.63471115e-01
3.75823408e-01 2.95099884e-01 1.69998139e-01 -5.91139555e-01
2.40787625e-01 7.94404268e-01 3.99333209e-01 -3.30866963e-01
6.82950318e-01 8.52386653e-01 3.37926894e-01 -2.75765359e-01
-5.06868541e-01 1.94618881e-01 -4.19400930e-01 4.68129039e-01
4.70562547e-01 -9.30859506e-01 -1.39370573e+00 8.59628141e-01
-1.11789310e+00 3.29208821e-02 -4.55357283e-01 3.28700662e-01
-4.31014299e-01 2.94383019e-01 -5.86664498e-01 -1.22258806e+00
-7.22368121e-01 -1.07479358e+00 5.42253017e-01 2.90628076e-01
1.05766065e-01 -9.85820055e-01 -4.54550296e-01 3.62130761e-01
4.86373782e-01 5.20677984e-01 7.21713126e-01 -1.22586286e+00
-5.04692852e-01 -7.05434442e-01 -7.77598023e-02 7.07097590e-01
-4.77713458e-02 -6.36049032e-01 -1.27818215e+00 -6.77575588e-01
4.35498983e-01 -3.85009438e-01 7.89708138e-01 8.67351517e-02
1.69305909e+00 -1.17975807e+00 -1.66277409e-01 9.34129238e-01
1.36873400e+00 2.67565213e-02 1.43668249e-01 1.17931433e-01
6.15847528e-01 6.72975719e-01 -5.84902093e-02 6.40144765e-01
2.74527788e-01 3.05748969e-01 7.27916718e-01 -1.25311449e-01
8.30057621e-01 -2.94324338e-01 3.10220420e-01 -1.39974728e-01
2.58513302e-01 -1.21227115e-01 -3.58203143e-01 5.82584679e-01
-2.07272458e+00 -7.90171981e-01 1.11132979e-01 2.62442231e+00
8.76242220e-01 -1.94352746e-01 9.56258476e-02 -9.28579718e-02
4.82181221e-01 2.33155668e-01 -1.07600558e+00 -6.09354019e-01
-3.59571576e-01 8.27320963e-02 1.26270700e+00 3.49334210e-01
-1.01667297e+00 4.93149459e-01 5.50203133e+00 7.01048195e-01
-1.04784167e+00 2.91917920e-01 1.14408350e+00 -6.68405771e-01
-3.40427607e-01 -9.11657289e-02 -7.40612328e-01 5.03137708e-01
5.86577415e-01 -3.86358291e-01 5.21704614e-01 1.09986627e+00
-3.15160632e-01 2.36787081e-01 -1.31148815e+00 9.53503966e-01
-2.34953552e-01 -1.43226731e+00 -2.78333515e-01 5.36525607e-01
7.25045860e-01 6.97617531e-02 4.24903244e-01 -2.03675687e-01
8.58224571e-01 -8.90279770e-01 7.33944714e-01 3.41882020e-01
5.73884428e-01 -1.19350755e+00 8.34755480e-01 7.77900994e-01
-4.92391229e-01 -3.17251384e-01 -6.21420085e-01 2.70629287e-01
-4.28471029e-01 5.53846300e-01 -5.50228000e-01 5.33867896e-01
9.93593872e-01 2.90484447e-02 -1.52737245e-01 6.41631424e-01
-1.34486184e-02 4.85330433e-01 -7.74378598e-01 -2.50500411e-01
3.04679930e-01 -3.22640032e-01 3.85577440e-01 8.34263861e-01
2.39380911e-01 6.54506534e-02 -1.69902340e-01 1.09898674e+00
-9.07514215e-01 8.75968039e-02 -8.38271856e-01 3.21993887e-01
6.64996386e-01 9.31584895e-01 -1.92789920e-02 1.17018130e-02
-3.53224456e-01 1.30754185e+00 5.20944655e-01 3.95141214e-01
-2.42549986e-01 -5.26603818e-01 1.06555986e+00 -2.01052368e-01
2.99410731e-01 1.19020097e-01 -5.06436944e-01 -1.01698804e+00
4.71242636e-01 -6.63768828e-01 5.58093607e-01 -1.08112618e-01
-1.48789108e+00 4.02889639e-01 -3.90081972e-01 -7.90398717e-01
-3.41509022e-02 -3.26266348e-01 -5.08802116e-01 1.04851699e+00
-1.51559782e+00 -1.22947252e+00 3.81142765e-01 1.00151932e+00
-3.67754787e-01 -4.41644162e-01 9.47229505e-01 5.09018265e-02
-5.15551686e-01 1.34986007e+00 8.43282044e-01 4.79977310e-01
2.49974653e-01 -1.11187243e+00 2.26745084e-01 7.50587881e-01
2.13835940e-01 7.29942560e-01 3.73903751e-01 -3.27049673e-01
-1.64848316e+00 -1.26249349e+00 1.02968156e+00 -2.14729145e-01
1.80314094e-01 -8.45148087e-01 -7.37001181e-01 9.43613529e-01
-6.64908532e-03 5.68759382e-01 8.26239109e-01 -1.93085596e-01
-4.45226431e-01 -6.92567885e-01 -2.03773952e+00 4.40812737e-01
7.70995021e-01 -6.40496075e-01 1.45905063e-01 3.98822248e-01
7.49411285e-01 -2.82703370e-01 -6.88746154e-01 -1.69873491e-01
6.50510848e-01 -9.69161510e-01 5.80664039e-01 -1.03326809e+00
-1.62465841e-01 4.09146305e-03 -4.44367737e-01 -7.51246810e-01
1.23872980e-01 -1.16648948e+00 -3.53937387e-01 1.28208876e+00
5.56851387e-01 -1.16254997e+00 1.30847502e+00 1.51835620e+00
8.38395774e-01 -7.64934778e-01 -1.53738678e+00 -7.48325944e-01
3.90442818e-01 -1.45769134e-01 1.13027310e+00 9.40474987e-01
-3.70482802e-01 -3.87181282e-01 -8.11187208e-01 3.07076454e-01
1.15489352e+00 3.59187752e-01 8.80206227e-01 -1.11971688e+00
-2.81289339e-01 -1.52492836e-01 -2.33783230e-01 -8.53492260e-01
4.98189092e-01 -8.89671564e-01 -3.37368697e-01 -9.15540576e-01
2.00080752e-01 -5.62660635e-01 -6.85153306e-01 9.95628476e-01
2.89034724e-01 7.93551654e-02 3.31514031e-01 1.48554549e-01
-3.37133050e-01 4.79652017e-01 7.56618500e-01 -8.40354338e-02
-8.19813758e-02 7.50955760e-01 -1.30043781e+00 4.30781245e-01
6.68527424e-01 -7.21079648e-01 -3.86599094e-01 -5.44369340e-01
8.61522630e-02 -3.71219702e-02 5.19781232e-01 -7.08897471e-01
7.48107374e-01 -6.29897639e-02 3.49937469e-01 -1.04049355e-01
3.30298215e-01 -1.39897418e+00 2.36195728e-01 4.54410434e-01
-7.53255069e-01 -2.37483025e-01 -2.30789050e-01 6.15134239e-01
2.01790899e-01 -2.44884476e-01 9.60828662e-01 6.21193927e-03
4.99649458e-02 8.83252680e-01 1.21487845e-02 -2.15199649e-01
1.28390801e+00 -9.05102789e-02 -1.23926565e-01 -8.04161310e-01
-3.39110851e-01 4.02880758e-01 6.50658786e-01 -2.40397751e-02
3.44498783e-01 -1.15396941e+00 -8.15141976e-01 3.15806478e-01
-1.84584588e-01 2.04814106e-01 3.13425548e-02 4.45221245e-01
-1.91974699e-01 2.44735390e-01 -1.74553078e-02 4.52416856e-03
-1.36490774e+00 4.76142973e-01 4.76772815e-01 -2.75151163e-01
-5.80649376e-01 1.12127602e+00 2.51651198e-01 -5.21052480e-01
6.99175835e-01 1.66031599e-01 4.35880184e-01 -1.49785817e-01
5.73531926e-01 2.97961742e-01 1.43827423e-01 -3.65570426e-01
-1.25089347e-01 -6.09835573e-02 -4.22581851e-01 -1.35424390e-01
1.74300265e+00 -2.32241243e-01 -1.93319246e-01 3.82460654e-02
1.76081491e+00 -4.91038561e-02 -1.30105579e+00 -6.40452862e-01
-3.50106835e-01 -6.58611059e-01 1.56327024e-01 -5.63780129e-01
-1.65590549e+00 9.92858171e-01 7.81361580e-01 2.20011905e-01
1.26318634e+00 -2.09428921e-01 9.19043541e-01 7.16456294e-01
5.91248393e-01 -9.85948563e-01 -1.00830960e+00 1.12972902e-02
4.02862072e-01 -1.03592551e+00 4.12641168e-02 1.70807943e-01
-6.87865853e-01 7.84236610e-01 2.91454166e-01 -9.40656140e-02
7.80997217e-01 3.45021665e-01 -1.01953179e-01 2.52687812e-01
-6.50972366e-01 5.59034228e-01 -3.42093259e-01 5.27990997e-01
-2.94546276e-01 -2.85795517e-02 -1.93641528e-01 8.90201449e-01
-1.98426947e-01 -1.83619428e-02 2.05579281e-01 1.00964856e+00
8.04249272e-02 -1.27561080e+00 1.01129107e-01 6.19867623e-01
-9.43083227e-01 1.86813712e-01 -6.19889796e-01 4.76501793e-01
1.29490212e-01 6.42371118e-01 -2.28208438e-01 -1.58311605e-01
1.00818604e-01 -8.20825994e-02 4.18174602e-02 -5.29123135e-02
-8.76558423e-01 -6.02993369e-01 -3.10672075e-01 -6.12920046e-01
-4.68949974e-02 -6.62549853e-01 -9.60963666e-01 -8.71527612e-01
-3.00437450e-01 2.71717310e-01 7.44838357e-01 6.01569593e-01
7.09834218e-01 -4.51375812e-01 1.27975535e+00 -9.39077139e-02
-1.06926274e+00 -2.94179201e-01 -1.16623032e+00 5.17002821e-01
8.05592716e-01 2.88781166e-01 -5.32328606e-01 -2.59851545e-01] | [5.87021017074585, 6.714230537414551] |
04fb5b6e-5ab0-425a-a827-1bfb4aba60da | 4dsr-gcn-4d-video-point-cloud-upsampling | 2306.01081 | null | https://arxiv.org/abs/2306.01081v1 | https://arxiv.org/pdf/2306.01081v1.pdf | 4DSR-GCN: 4D Video Point Cloud Upsampling using Graph Convolutional Networks | Time varying sequences of 3D point clouds, or 4D point clouds, are now being acquired at an increasing pace in several applications (e.g., LiDAR in autonomous or assisted driving). In many cases, such volume of data is transmitted, thus requiring that proper compression tools are applied to either reduce the resolution or the bandwidth. In this paper, we propose a new solution for upscaling and restoration of time-varying 3D video point clouds after they have been heavily compressed. In consideration of recent growing relevance of 3D applications, %We focused on a model allowing user-side upscaling and artifact removal for 3D video point clouds, a real-time stream of which would require . Our model consists of a specifically designed Graph Convolutional Network (GCN) that combines Dynamic Edge Convolution and Graph Attention Networks for feature aggregation in a Generative Adversarial setting. By taking inspiration PointNet++, We present a different way to sample dense point clouds with the intent to make these modules work in synergy to provide each node enough features about its neighbourhood in order to later on generate new vertices. Compared to other solutions in the literature that address the same task, our proposed model is capable of obtaining comparable results in terms of quality of the reconstruction, while using a substantially lower number of parameters (about 300KB), making our solution deployable in edge computing devices such as LiDAR. | ['Alberto del Bimbo', 'Marco Bertini', 'Stefano Berretti', 'Lorenzo Berlincioni'] | 2023-06-01 | null | null | null | null | ['graph-attention', 'edge-computing'] | ['graphs', 'time-series'] | [ 1.82669953e-01 1.58365861e-01 3.43180120e-01 1.10992370e-02
-4.24958348e-01 -5.09710550e-01 7.07701683e-01 2.14808449e-01
-4.16071862e-01 4.38108563e-01 -2.66914397e-01 -5.97611964e-01
3.95280942e-02 -1.25503349e+00 -1.12080991e+00 -3.83867294e-01
-2.41906047e-01 6.73397958e-01 3.34953040e-01 -2.40975887e-01
2.77529627e-01 1.38080812e+00 -2.01435971e+00 -1.23349935e-01
7.78136015e-01 1.14271104e+00 3.28244358e-01 7.95013189e-01
-3.04668546e-01 2.98638076e-01 -4.49134737e-01 -4.52292562e-01
6.25841379e-01 1.46593928e-01 -2.94135958e-01 1.56592190e-01
7.08550513e-01 -4.42884237e-01 -6.61426425e-01 8.45943034e-01
6.48263216e-01 1.22351259e-01 5.04076004e-01 -1.46749282e+00
-3.92805099e-01 2.45869592e-01 -3.81403863e-01 5.43277785e-02
1.97864607e-01 5.34355164e-01 4.67285007e-01 -1.00702083e+00
7.49518037e-01 1.20248675e+00 8.16898108e-01 5.17929673e-01
-1.12038004e+00 -7.07033396e-01 -4.46195230e-02 2.59829670e-01
-1.38321269e+00 -4.08449560e-01 9.88141894e-01 -2.54841775e-01
1.20727932e+00 2.20319748e-01 1.00891066e+00 9.55307305e-01
2.69517064e-01 3.66028845e-01 5.91752768e-01 7.59833213e-03
3.60411286e-01 -1.89248756e-01 -5.38547635e-01 4.23086137e-01
3.31679910e-01 2.18105137e-01 -4.67475444e-01 -1.89589173e-01
9.28267419e-01 2.19487578e-01 -1.41888604e-01 -6.77408278e-01
-1.12735689e+00 7.97606647e-01 8.42299938e-01 9.57748815e-02
-6.42483532e-01 7.36615300e-01 4.01193917e-01 3.18582982e-01
6.54777348e-01 -2.65971627e-02 -2.32853126e-02 -1.52852848e-01
-1.06045067e+00 4.47257191e-01 5.15250087e-01 1.28493798e+00
8.75856578e-01 2.04498887e-01 2.00277660e-02 1.16955578e-01
2.02567041e-01 8.44798028e-01 9.01397318e-02 -8.32042336e-01
6.45269454e-01 5.49340189e-01 6.57929257e-02 -1.18820572e+00
-3.10547143e-01 -1.72623679e-01 -1.03317487e+00 7.20577121e-01
2.12544538e-02 9.41003487e-02 -1.15805805e+00 1.51899827e+00
5.38917422e-01 8.90976667e-01 -3.27959545e-02 8.97740245e-01
7.73477018e-01 6.94057703e-01 -1.99330151e-01 2.09667176e-01
1.17629302e+00 -2.76206136e-01 -3.45734566e-01 -1.59914419e-01
2.67631859e-01 -5.76356411e-01 7.94818997e-01 1.41804963e-01
-1.19807172e+00 -6.12893641e-01 -1.16386116e+00 -2.93868244e-01
-5.37994027e-01 -5.76972544e-01 5.95021725e-01 6.01342201e-01
-1.42636943e+00 9.47370231e-01 -9.73304272e-01 -2.95346528e-01
6.40147984e-01 6.02602780e-01 -5.23678899e-01 -2.00364441e-01
-8.88355851e-01 6.92367077e-01 -7.74320401e-03 8.41440186e-02
-6.65747166e-01 -9.01462853e-01 -8.18698883e-01 2.19946876e-01
6.75871596e-02 -1.18647563e+00 7.00854957e-01 -3.78228962e-01
-1.14569128e+00 7.43085742e-01 6.65665492e-02 -7.88480043e-01
6.15512490e-01 2.20462866e-03 -1.21631458e-01 2.06931591e-01
-3.39846909e-02 9.29672301e-01 1.20409286e+00 -1.22006989e+00
-4.71812606e-01 -5.86099088e-01 1.06352881e-01 1.56090766e-01
-2.10992936e-02 -5.06363392e-01 -5.49586475e-01 -3.81000727e-01
1.73180535e-01 -1.05826271e+00 -4.25651371e-01 3.67471784e-01
-1.04486920e-01 -1.84836499e-02 1.30993903e+00 -2.00736046e-01
3.81873339e-01 -2.22516131e+00 1.35252178e-01 1.75265014e-01
5.09418964e-01 3.30510527e-01 -9.63973626e-02 4.72733080e-01
1.08221993e-02 1.05333984e-01 -4.36954498e-01 -7.61296391e-01
-5.44900857e-02 3.80261898e-01 -4.41373914e-01 4.95112091e-01
5.70104778e-01 1.01881528e+00 -9.07232046e-01 -2.87770867e-01
6.80707157e-01 1.07909048e+00 -7.27385521e-01 8.08649957e-02
-3.39470357e-01 4.75632757e-01 -3.92551720e-01 5.62071860e-01
1.14185631e+00 -7.24951476e-02 -3.54368061e-01 -1.44095287e-01
-5.48815988e-02 5.52683659e-02 -1.18995118e+00 2.03019309e+00
-6.26547396e-01 5.84428668e-01 3.22719276e-01 -7.65523374e-01
1.00836492e+00 2.32105404e-01 7.55664706e-01 -4.73439932e-01
2.61558861e-01 2.20742121e-01 -3.39134604e-01 -1.49354279e-01
9.08534884e-01 -1.19865953e-03 8.97067320e-03 2.23083258e-01
-1.90875128e-01 -8.79804492e-01 -1.22443430e-01 3.34929287e-01
1.36616194e+00 3.81254591e-02 -1.63477853e-01 1.34218112e-01
3.31414640e-01 9.59780961e-02 6.40678778e-02 4.95805621e-01
6.21079840e-02 8.12857747e-01 1.58958167e-01 -4.32442456e-01
-1.34742880e+00 -9.77648020e-01 -4.70813848e-02 1.90975934e-01
3.00585508e-01 -2.30120212e-01 -5.30024707e-01 -4.92789805e-01
5.00467300e-01 6.41157806e-01 -3.37515503e-01 -1.37163892e-01
-6.63224816e-01 -1.43613353e-01 4.05766726e-01 3.53756994e-01
4.24014151e-01 -1.02835774e+00 -9.65599060e-01 2.76362032e-01
3.06640148e-01 -1.16718709e+00 -1.95661053e-01 1.00448266e-01
-1.20643532e+00 -7.64670193e-01 -5.03001273e-01 -2.75844574e-01
5.98087311e-01 6.67294085e-01 1.30796874e+00 1.57596976e-01
-2.04345465e-01 6.55218840e-01 -3.54863226e-01 -5.06754458e-01
-3.44397008e-01 -7.04745110e-03 3.96046899e-02 -4.57499549e-02
3.14381242e-01 -1.28037035e+00 -6.16785347e-01 -2.31517509e-01
-1.34551573e+00 -1.71766821e-02 4.15989339e-01 5.70046008e-01
6.83888674e-01 -1.17101401e-01 2.24919438e-01 -6.71498775e-01
3.53614122e-01 -7.07508326e-01 -8.68826747e-01 -5.29158413e-01
-4.91488814e-01 -1.12594284e-01 6.91441774e-01 -2.51492769e-01
-3.08180243e-01 2.38201916e-01 -3.90359432e-01 -1.23746812e+00
-9.78541896e-02 2.19646841e-01 -1.61425009e-01 -4.19791222e-01
4.65485930e-01 9.28871259e-02 2.23642647e-01 -2.92077899e-01
6.20052755e-01 4.84752387e-01 6.36328459e-01 -2.76362509e-01
1.25913894e+00 7.65096068e-01 6.00425661e-01 -6.26028419e-01
2.37013102e-02 -2.21362293e-01 -6.63998187e-01 -3.27787071e-01
7.25597203e-01 -9.39045310e-01 -8.50914896e-01 2.92000592e-01
-1.42670596e+00 -1.90002695e-01 -7.04994977e-01 2.79843748e-01
-8.17893565e-01 3.41110021e-01 -1.83083206e-01 -6.61593497e-01
-4.28186059e-01 -1.14073372e+00 1.46673214e+00 -1.37412455e-02
2.88384527e-01 -7.10116565e-01 -1.36799246e-01 -1.06506683e-02
6.33906722e-01 6.66760325e-01 6.06232107e-01 -2.15764552e-01
-1.32041681e+00 -3.54526252e-01 -2.65112042e-01 2.34976530e-01
-1.51896983e-01 -1.05254024e-01 -1.00665832e+00 -3.73603970e-01
-1.10016456e-02 1.50555193e-01 5.65385103e-01 1.89177915e-01
1.27152169e+00 3.98108996e-02 -3.31124783e-01 1.00647509e+00
1.48737228e+00 -1.95001826e-01 8.49589169e-01 3.40369567e-02
9.02909517e-01 2.87040025e-01 5.29138625e-01 5.51416039e-01
3.33928078e-01 7.19946384e-01 1.28859246e+00 -4.73929271e-02
-3.92811507e-01 -2.31863469e-01 1.04782209e-01 6.19561076e-01
-2.02584103e-01 -3.21814537e-01 -8.08901012e-01 6.73481345e-01
-1.67430174e+00 -9.47523236e-01 -1.55520648e-01 2.31879640e+00
1.73886031e-01 1.00773647e-01 -1.76489696e-01 1.71892077e-01
6.06671453e-01 3.02320749e-01 -7.76519716e-01 -4.23829913e-01
9.96240601e-02 6.04710698e-01 8.36587191e-01 2.93974757e-01
-6.10767305e-01 7.20558763e-01 5.06881762e+00 5.92717409e-01
-1.46153855e+00 -5.57117388e-02 2.50216275e-01 -4.03499603e-01
-5.54378211e-01 -4.97544035e-02 -5.59486389e-01 5.75731814e-01
1.18127596e+00 -3.69261801e-01 4.41769123e-01 8.76190722e-01
1.76393047e-01 2.27374032e-01 -1.01599789e+00 1.27292573e+00
-6.28750324e-02 -1.66838443e+00 2.07677767e-01 4.07352805e-01
3.67292076e-01 5.22506416e-01 8.32326040e-02 3.78325470e-02
2.00041085e-01 -9.77266014e-01 7.99739182e-01 5.28728426e-01
1.07692540e+00 -9.79055583e-01 4.82134730e-01 4.49822575e-01
-1.16145062e+00 2.44119391e-01 -5.60733795e-01 3.54969054e-02
4.77448970e-01 1.00172353e+00 -1.02007699e+00 7.64945805e-01
8.04922819e-01 6.71582937e-01 -3.92111719e-01 1.00675440e+00
1.81147456e-01 1.54399693e-01 -7.56285429e-01 1.71105489e-01
1.79058522e-01 -2.55904555e-01 8.56788993e-01 7.51632631e-01
9.79287565e-01 1.09796273e-02 -8.48639682e-02 9.56262350e-01
-3.13104928e-01 -1.23708554e-01 -1.28037441e+00 2.41376281e-01
6.63907528e-01 1.23391354e+00 -6.63442135e-01 -3.91920537e-01
-4.91935104e-01 1.06577492e+00 2.62087554e-01 2.55929702e-03
-8.70224655e-01 -3.36608976e-01 1.04561782e+00 5.84402800e-01
6.50988817e-01 -6.50533676e-01 -2.58785754e-01 -9.13215160e-01
1.79534256e-01 -4.26420540e-01 -1.44974202e-01 -9.59655344e-01
-1.09550083e+00 7.81217158e-01 -1.81572258e-01 -1.64753234e+00
-3.71674120e-01 -4.62278098e-01 -4.66573924e-01 1.01549399e+00
-1.66529191e+00 -1.14219368e+00 -6.76941812e-01 7.67293513e-01
2.66103387e-01 1.11546472e-01 6.40512824e-01 6.04259014e-01
1.23931818e-01 1.75341979e-01 -1.93474010e-01 -3.05446953e-01
3.11834663e-01 -1.00677836e+00 1.21922052e+00 8.62589836e-01
2.15722367e-01 3.68877977e-01 6.52461827e-01 -8.18419516e-01
-2.08696413e+00 -1.33566439e+00 6.51116431e-01 -4.60244119e-01
4.43015099e-01 -5.11973023e-01 -7.85017550e-01 5.60870528e-01
-7.84421936e-02 4.78775084e-01 1.83077604e-02 -5.50030231e-01
-1.22109033e-01 -1.15878500e-01 -1.49028671e+00 5.19033074e-01
1.49568868e+00 -3.50614786e-01 -8.81195441e-02 3.99924725e-01
1.04026818e+00 -8.20251584e-01 -8.78734171e-01 5.42698681e-01
-3.05620786e-02 -1.20197320e+00 1.16012371e+00 -1.99802771e-01
3.82960290e-01 -6.12608790e-01 -7.39439800e-02 -1.31388056e+00
-2.53023565e-01 -6.71345770e-01 -4.16306317e-01 8.87061834e-01
-9.04310048e-02 -6.98705018e-01 1.05655456e+00 4.83198106e-01
-5.64857006e-01 -5.58359802e-01 -1.46782279e+00 -6.59641266e-01
-2.81018287e-01 -7.87080705e-01 1.18649864e+00 6.69882476e-01
-6.70254290e-01 1.23237416e-01 -1.74394608e-01 4.64435875e-01
6.54142201e-01 2.23816767e-01 1.24205887e+00 -1.29211056e+00
4.63542379e-02 -2.26650789e-01 -1.13042951e+00 -1.15277064e+00
2.11916827e-02 -1.03663146e+00 -2.02123582e-01 -1.53673482e+00
-5.73763430e-01 -8.15514147e-01 1.64422438e-01 1.20742366e-01
1.02895916e-01 5.73923588e-01 3.52763563e-01 3.59693356e-02
7.64047652e-02 6.81600869e-01 1.09424806e+00 -2.12518498e-01
-8.35823938e-02 7.58420527e-02 -3.68518353e-01 2.04324007e-01
4.67396796e-01 -4.13879335e-01 -4.56494600e-01 -7.07108617e-01
1.97657973e-01 1.18106343e-01 7.63286293e-01 -1.40579891e+00
3.04605484e-01 1.35410100e-01 2.69990623e-01 -1.04054511e+00
8.03914964e-01 -1.45268500e+00 6.42105162e-01 3.15764725e-01
4.87523347e-01 4.27196860e-01 4.23723310e-01 6.68139338e-01
-1.03131980e-01 -6.66491985e-02 4.92653072e-01 -1.32889494e-01
-7.14704216e-01 9.12762165e-01 -1.62853561e-02 -4.95670617e-01
1.16535211e+00 -4.31120455e-01 -3.07043623e-02 -4.75275546e-01
-5.13442814e-01 -6.78669810e-02 9.11102831e-01 4.98856187e-01
8.12566698e-01 -1.39660573e+00 -7.41979301e-01 4.74816471e-01
4.53042015e-02 6.08742714e-01 4.25722748e-01 4.92740661e-01
-8.70844424e-01 2.14137048e-01 -1.76076144e-01 -8.95454824e-01
-9.56019938e-01 8.20975542e-01 -3.36498842e-02 2.44334377e-02
-1.18048894e+00 5.98171055e-01 -2.63205647e-01 -3.10017318e-01
-1.54214073e-02 -4.48374808e-01 2.58193374e-01 -1.13793120e-01
2.98239559e-01 2.51597494e-01 5.56841195e-01 -5.75756669e-01
-2.34508395e-01 6.01060569e-01 3.79954875e-01 6.98362738e-02
1.54148626e+00 3.26338522e-02 -4.37119696e-03 1.04236409e-01
9.70927477e-01 2.23925799e-01 -1.40863466e+00 -4.95225750e-03
-3.87936562e-01 -7.34593213e-01 1.05274379e-01 -3.27065140e-02
-1.38791752e+00 8.24917078e-01 6.03961229e-01 4.21047658e-01
1.05667758e+00 -3.02238334e-02 1.08612037e+00 7.00620264e-02
8.03521395e-01 -4.48392034e-01 -4.38757867e-01 3.87863487e-01
7.74424791e-01 -9.59088206e-01 1.16684750e-01 -5.68428576e-01
-1.78334042e-01 8.94799829e-01 1.39906362e-01 -5.62798798e-01
5.72904527e-01 5.48916221e-01 -2.13013053e-01 -3.49438876e-01
-6.49956286e-01 -2.01861665e-01 -4.58132364e-02 1.03979409e+00
-1.81934446e-01 -4.53560539e-02 1.35214671e-01 -7.96927288e-02
-4.30212229e-01 2.85982620e-02 5.32430291e-01 8.19424748e-01
-2.29439437e-01 -1.11221218e+00 -5.09344041e-01 4.52222794e-01
6.50493987e-03 -1.31673902e-01 -8.59340206e-02 7.73946583e-01
6.37142286e-02 6.98360920e-01 5.00729442e-01 -4.45110619e-01
6.31620705e-01 -9.31977481e-02 4.22096074e-01 -4.97923166e-01
-5.77824175e-01 -3.48251015e-01 -1.43968388e-01 -8.81383359e-01
-3.61241192e-01 -5.77310801e-01 -1.17834127e+00 -7.46859848e-01
1.44698285e-02 -1.47100538e-01 1.18011260e+00 3.38258833e-01
1.03616798e+00 5.17413020e-01 6.04949951e-01 -1.64209747e+00
-2.08206251e-01 -7.24260092e-01 -4.95298237e-01 5.37530422e-01
3.03462535e-01 -6.69624388e-01 -4.19545740e-01 -1.48439616e-01] | [8.487367630004883, -3.6283373832702637] |
5c205e60-caae-4e4d-bd13-a1e4d4218a52 | towards-open-set-3d-learning-a-benchmark-on | 2207.11554 | null | https://arxiv.org/abs/2207.11554v3 | https://arxiv.org/pdf/2207.11554v3.pdf | 3DOS: Towards 3D Open Set Learning -- Benchmarking and Understanding Semantic Novelty Detection on Point Clouds | In recent years there has been significant progress in the field of 3D learning on classification, detection and segmentation problems. The vast majority of the existing studies focus on canonical closed-set conditions, neglecting the intrinsic open nature of the real-world. This limits the abilities of robots and autonomous systems involved in safety-critical applications that require managing novel and unknown signals. In this context exploiting 3D data can be a valuable asset since it provides rich information about the geometry of perceived objects and scenes. With this paper we provide the first broad study on 3D Open Set learning. We introduce 3DOS: a novel testbed for semantic novelty detection that considers several settings with increasing difficulties in terms of semantic (category) shift, and covers both in-domain (synthetic-to-synthetic, real-to-real) and cross-domain (synthetic-to-real) scenarios. Moreover, we investigate the related 2D Open Set literature to understand if and how its recent improvements are effective on 3D data. Our extensive benchmark positions several algorithms in the same coherent picture, revealing their strengths and limitations. The results of our analysis may serve as a reliable foothold for future tailored 3D Open Set methods. | ['Tatiana Tommasi', 'Francesco Cappio Borlino', 'Antonio Alliegro'] | 2022-07-23 | null | null | null | null | ['open-set-learning'] | ['miscellaneous'] | [ 2.24603698e-01 1.24438696e-01 2.67220922e-02 -2.20746726e-01
-4.77428973e-01 -8.73433650e-01 6.69377267e-01 1.72158971e-01
-1.98128864e-01 4.76140440e-01 -3.18664938e-01 -8.15334246e-02
-3.65100265e-01 -5.46403289e-01 -6.75575495e-01 -7.27391779e-01
-5.25463343e-01 6.27823114e-01 5.21828532e-01 -4.43573296e-01
2.34153748e-01 8.33758831e-01 -2.13623166e+00 9.77374837e-02
5.62714100e-01 1.11019611e+00 3.11770409e-01 2.70297080e-01
-5.80552705e-02 -9.54810306e-02 -4.81948942e-01 -1.02553137e-01
1.00238121e+00 1.60516705e-02 -6.41935587e-01 2.22790644e-01
6.08401537e-01 2.68411160e-01 -2.49011874e-01 1.30637240e+00
7.05821335e-01 1.45395592e-01 6.80224299e-01 -1.37225401e+00
-3.02712739e-01 1.25466079e-01 -5.14589965e-01 3.34034950e-01
5.71528196e-01 1.97689340e-01 8.69848728e-01 -9.07521129e-01
9.39955294e-01 1.22900426e+00 8.07169795e-01 3.46197635e-01
-1.07762742e+00 -2.99467474e-01 2.17778310e-01 5.05280912e-01
-1.17510784e+00 -2.59410203e-01 9.94341016e-01 -5.71686149e-01
7.49477983e-01 1.71430364e-01 6.58878267e-01 1.36784077e+00
1.01303030e-02 7.88788855e-01 1.38352513e+00 -6.36390924e-01
4.43503290e-01 3.51998478e-01 3.85444537e-02 3.08458954e-01
3.04722071e-01 4.53007489e-01 -3.28743577e-01 1.48443058e-01
5.66961944e-01 -2.81963617e-01 -2.96716541e-01 -1.27413034e+00
-1.22866654e+00 6.68454766e-01 4.50368613e-01 3.61121655e-01
5.79813821e-03 -3.86291087e-01 4.96013820e-01 6.08107090e-01
5.73089242e-01 6.50066972e-01 -5.33680439e-01 -3.78901958e-02
-6.53399587e-01 4.69128937e-01 8.89250338e-01 1.21569848e+00
7.07244515e-01 -3.39073576e-02 2.88548082e-01 7.96103716e-01
-1.18302992e-02 3.43252927e-01 4.38774019e-01 -9.30319309e-01
1.84502065e-01 3.20907772e-01 -4.66609672e-02 -1.05068803e+00
-7.19490409e-01 -7.26884782e-01 -5.95903099e-01 5.35289943e-01
5.60454309e-01 3.28327537e-01 -5.50370753e-01 1.50900018e+00
6.56890392e-01 1.90236390e-01 -1.18791319e-01 9.43288624e-01
5.91375470e-01 1.49147406e-01 -6.01383090e-01 -1.65441036e-01
1.05133593e+00 -5.80286562e-01 -2.58280724e-01 -6.78231493e-02
6.47438467e-01 -7.23276138e-01 1.00645602e+00 4.22074854e-01
-7.17046201e-01 -6.16142929e-01 -1.05357981e+00 2.33713001e-01
-7.76029110e-01 -3.87597561e-01 4.17482376e-01 7.37220466e-01
-7.58146167e-01 6.08226717e-01 -3.52307975e-01 -7.64147878e-01
7.13158786e-01 7.08517209e-02 -4.32044268e-01 -2.02395082e-01
-9.67743158e-01 1.03523386e+00 4.69317347e-01 -2.31053144e-01
-6.37460887e-01 -6.90426886e-01 -9.63896096e-01 -4.73499537e-01
9.06537592e-01 -6.31348312e-01 1.11225092e+00 -5.97025096e-01
-1.12788117e+00 1.28977942e+00 4.92608845e-01 -5.42214096e-01
9.99711692e-01 -7.61461630e-02 -4.13482040e-01 2.67960653e-02
2.06832811e-01 7.22433865e-01 8.26001167e-01 -1.51036775e+00
-5.52798152e-01 -7.66160786e-01 1.84502810e-01 1.14980370e-01
1.37786195e-01 -4.64496762e-01 2.04644814e-01 -7.82094717e-01
5.16653538e-01 -9.02091503e-01 -1.66071624e-01 1.70853645e-01
-1.85887113e-01 -1.56335816e-01 8.76865327e-01 -7.01854052e-03
6.21888697e-01 -2.12450218e+00 1.55405164e-01 1.86185956e-01
1.23064667e-01 2.45302588e-01 8.34817365e-02 3.40576917e-01
-2.86760151e-01 -1.52868688e-01 -5.08684397e-01 -7.15790763e-02
1.56310111e-01 3.12152833e-01 -3.50595623e-01 7.22435951e-01
6.22847229e-02 6.07495010e-01 -9.00974452e-01 -2.67568082e-01
5.79639435e-01 -8.10284093e-02 -4.36217755e-01 -2.43296608e-01
-3.98950040e-01 6.83862209e-01 -3.05158794e-01 6.67167187e-01
7.89711595e-01 1.75700113e-01 -4.61588055e-01 -5.29256091e-02
-2.35757962e-01 -4.85046282e-02 -1.51676226e+00 2.08363867e+00
-2.68101186e-01 5.86648524e-01 1.37603488e-02 -1.50219476e+00
1.00777054e+00 -3.98152471e-02 6.22138798e-01 -6.94809020e-01
3.23470622e-01 3.55562150e-01 -3.58729288e-02 -8.37351263e-01
4.26033735e-01 -2.80034214e-01 -2.51626551e-01 1.93722084e-01
2.32192367e-01 -5.48518538e-01 8.26217532e-02 -2.17904821e-01
9.79473054e-01 1.98234111e-01 4.60898131e-01 -4.05715793e-01
4.88410473e-01 7.05739558e-02 4.28098053e-01 9.27553535e-01
-5.60490787e-01 7.06931829e-01 3.26021403e-01 -4.55528677e-01
-1.05907845e+00 -1.29150951e+00 -6.13032758e-01 4.14829493e-01
5.40240645e-01 1.14009432e-01 -5.44864297e-01 -7.79112637e-01
2.00372547e-01 6.39779091e-01 -5.28222084e-01 -2.44934872e-01
-3.01912427e-01 -4.28563595e-01 4.90582496e-01 1.49676695e-01
4.46827114e-01 -7.92555451e-01 -9.53763723e-01 -1.03346258e-01
-7.18757808e-02 -1.41950452e+00 1.07264407e-01 2.86504656e-01
-8.79824460e-01 -1.22192657e+00 -5.75850487e-01 -7.51220047e-01
1.99173927e-01 5.89871407e-01 1.05209935e+00 -6.95353508e-01
-7.09708691e-01 6.75332963e-01 -6.16203547e-01 -6.49714351e-01
-4.45410550e-01 -2.10910395e-01 3.54320943e-01 7.78482258e-02
3.53831470e-01 -7.57097065e-01 -8.12056959e-02 5.72201729e-01
-7.28008866e-01 -2.99833357e-01 4.25644159e-01 6.88998699e-01
6.47339761e-01 2.63702542e-01 6.79521441e-01 -5.47609150e-01
2.55006641e-01 -5.63732862e-01 -7.63789415e-01 -4.18364033e-02
-3.96090925e-01 -9.69059467e-02 2.79258728e-01 -3.30888599e-01
-8.19226503e-01 4.34733443e-02 -3.51883564e-03 -6.39143944e-01
-7.18803108e-01 1.61244953e-03 -4.12900865e-01 -2.30607197e-01
9.01728749e-01 -1.97270196e-02 -4.13436070e-02 -6.02802217e-01
4.88531917e-01 6.00228727e-01 5.34225583e-01 -4.09014255e-01
9.68416572e-01 1.01359272e+00 3.60108018e-01 -1.23647797e+00
-8.95493507e-01 -8.04946601e-01 -1.13765502e+00 -2.73091435e-01
5.16706884e-01 -6.31125450e-01 -1.95810854e-01 4.72762823e-01
-9.83476043e-01 -9.58176795e-03 -9.82611418e-01 4.41792279e-01
-9.70225453e-01 5.77304065e-01 1.13185719e-01 -7.81666040e-01
3.57549071e-01 -9.86521959e-01 1.07551122e+00 -1.53139040e-01
-2.43032470e-01 -1.01724482e+00 1.60032921e-02 2.70071775e-01
-1.21791467e-01 5.11790752e-01 8.12262714e-01 -7.72107542e-01
-5.19186854e-01 -1.29193023e-01 -4.53468598e-02 4.94160146e-01
3.96019081e-03 -5.38681448e-01 -1.20780468e+00 -1.23107314e-01
3.97984445e-01 -1.89958006e-01 8.29134464e-01 3.44341487e-01
9.25826728e-01 2.21533224e-01 -2.72246808e-01 4.49025333e-01
1.27926481e+00 -2.98755765e-02 2.58181661e-01 3.24945867e-01
4.13328826e-01 1.02925313e+00 9.54538703e-01 6.07194185e-01
5.75392544e-02 8.44351232e-01 7.89194047e-01 8.37127864e-02
-2.95209944e-01 -3.55761684e-02 1.26890123e-01 3.19429755e-01
3.68939221e-01 -6.73042759e-02 -7.76567280e-01 6.44097805e-01
-1.71058750e+00 -1.05274320e+00 -1.22733928e-01 2.37518549e+00
2.97271311e-01 4.84956920e-01 2.34008849e-01 5.96210778e-01
8.74861836e-01 1.69503726e-02 -6.33193970e-01 -5.30644767e-02
-5.30847251e-01 6.43320605e-02 3.28844696e-01 9.01805535e-02
-1.21561110e+00 5.72993994e-01 5.68051672e+00 8.73240769e-01
-8.44497859e-01 6.28448129e-02 2.24917501e-01 -1.44771427e-01
-1.81839719e-01 -8.40763077e-02 -5.59540570e-01 2.20074117e-01
2.36776546e-01 1.19985960e-01 5.31945825e-02 8.27283025e-01
1.10843748e-01 -3.81030232e-01 -1.54691935e+00 1.46250689e+00
3.72388214e-01 -1.22126997e+00 -1.11890107e-01 1.20086722e-01
7.64699340e-01 2.13639215e-01 1.53505638e-01 2.40193717e-02
-2.06784919e-01 -5.45941770e-01 9.43293512e-01 3.71393681e-01
5.36378920e-01 -4.97147858e-01 6.49629414e-01 7.09642768e-01
-8.59560013e-01 -1.53126255e-01 -2.27167815e-01 -2.89319307e-01
1.83720291e-01 8.70025635e-01 -8.90218496e-01 7.57674158e-01
9.62390721e-01 1.04498553e+00 -4.66221094e-01 1.28783941e+00
1.06304944e-01 6.01397157e-02 -5.60044646e-01 1.10663705e-01
1.48995906e-01 -2.12335706e-01 1.33630073e+00 9.35477257e-01
2.84044713e-01 -2.27119103e-01 3.66311997e-01 9.06269133e-01
1.90110415e-01 -6.90114647e-02 -1.21918213e+00 3.75583887e-01
1.75656721e-01 8.48950624e-01 -1.01832545e+00 3.58292721e-02
-4.82568204e-01 7.38814831e-01 -8.75993595e-02 -7.53226632e-04
-6.14519835e-01 -3.24286699e-01 6.41586244e-01 2.51548320e-01
4.61347401e-01 -3.30179483e-01 -4.02435422e-01 -1.10985041e+00
3.39599490e-01 -5.55810213e-01 4.47564721e-01 -6.14119887e-01
-1.52329111e+00 1.69040695e-01 3.61581564e-01 -1.55206776e+00
-7.03295832e-03 -8.44387472e-01 -1.69530526e-01 2.27610752e-01
-1.52253771e+00 -8.45921457e-01 -2.87562072e-01 5.28604269e-01
8.08348715e-01 -1.73855945e-01 4.85587984e-01 3.09653133e-01
-7.10952803e-02 2.49767542e-01 3.27620715e-01 -2.57919431e-01
6.40415132e-01 -1.17784631e+00 2.71932513e-01 7.52779722e-01
3.86913270e-01 -4.86148596e-02 8.80500853e-01 -4.37054276e-01
-1.10144329e+00 -9.64076638e-01 4.37552989e-01 -7.99940705e-01
5.35510480e-01 -6.46390140e-01 -8.09825063e-01 2.83333391e-01
-4.91715670e-01 2.16554776e-01 4.36966628e-01 -9.26983282e-02
-3.77905577e-01 -2.56326050e-02 -1.34904909e+00 5.06481171e-01
1.61359715e+00 -3.15111279e-01 -8.60943556e-01 3.20100635e-01
7.95991242e-01 -3.89524817e-01 -7.05178142e-01 6.35355651e-01
4.21264052e-01 -1.32515728e+00 1.10572624e+00 -5.38032234e-01
-1.40636982e-02 -2.30995551e-01 -6.16920352e-01 -1.32300949e+00
-1.43693490e-02 -6.10893667e-01 2.07064617e-02 9.29923534e-01
-7.06179291e-02 -6.74075961e-01 6.88142180e-01 -1.84136301e-01
-5.38010418e-01 -6.28022909e-01 -1.33592689e+00 -1.40038681e+00
6.58989176e-02 -1.02425253e+00 3.32458436e-01 9.30012047e-01
-1.59968093e-01 2.38040879e-01 4.01300518e-03 1.70335442e-01
7.57278442e-01 3.17699701e-01 7.91097879e-01 -1.63683832e+00
-2.65749305e-01 -5.98392785e-01 -1.11516273e+00 -1.08916700e+00
2.00165555e-01 -1.09475613e+00 -2.04918161e-01 -9.34819818e-01
-2.80206949e-01 -7.40999103e-01 -1.84557047e-02 -1.15773007e-01
5.09459972e-01 1.76628768e-01 1.78796127e-01 1.20625854e-01
-6.88764751e-01 7.09228516e-01 1.20145476e+00 -1.97668776e-01
-1.69570342e-01 2.72999227e-01 -2.89096177e-01 1.03241336e+00
7.08855510e-01 -3.88000250e-01 -4.71150368e-01 -1.38581172e-01
2.10545421e-01 -4.13523257e-01 7.34142125e-01 -1.29775965e+00
-3.88233699e-02 5.05429879e-02 7.44378269e-02 -7.38968790e-01
5.19543648e-01 -1.15422916e+00 -2.47954562e-01 3.29023182e-01
-1.81457952e-01 -2.92894572e-01 2.51444370e-01 8.96222532e-01
-1.43194109e-01 -4.02525276e-01 9.20241892e-01 -4.62131739e-01
-1.16003489e+00 3.50480556e-01 -7.50987902e-02 5.15411615e-01
1.50057828e+00 -9.20181096e-01 1.50011897e-01 -1.17305897e-01
-8.50889981e-01 1.34682715e-01 6.46245062e-01 7.98308074e-01
5.09080291e-01 -1.14976537e+00 -5.46207547e-01 4.50425386e-01
7.10916221e-01 2.47703716e-01 4.14345503e-01 8.10900569e-01
-1.91356048e-01 3.13471973e-01 -3.80837500e-01 -1.14060640e+00
-1.12325513e+00 7.78324723e-01 4.45789903e-01 2.60234177e-01
-7.75775552e-01 7.96833515e-01 3.72046947e-01 -7.94836998e-01
5.41060627e-01 -3.48306358e-01 -1.14240702e-02 2.47889817e-01
7.71450996e-02 7.30654895e-01 3.74885499e-01 -7.04194367e-01
-2.62758225e-01 7.59187043e-01 2.46733665e-01 6.94695786e-02
1.26457739e+00 -3.40651751e-01 1.85501829e-01 1.02398849e+00
1.25165820e+00 -2.73831874e-01 -1.15304482e+00 -4.78751421e-01
1.26028076e-01 -5.61671674e-01 -1.36043787e-01 -4.78478402e-01
-5.03244936e-01 1.02659225e+00 9.38459218e-01 3.54402989e-01
8.61460686e-01 5.23120761e-01 4.90694791e-01 4.26492095e-01
7.88001657e-01 -1.42551148e+00 1.81592852e-01 5.81933439e-01
9.34097111e-01 -1.46918738e+00 -1.56896636e-01 -6.37836516e-01
-2.41886809e-01 9.56756711e-01 3.49245131e-01 -9.47885662e-02
8.94960344e-01 -1.00071914e-01 -2.13104576e-01 -2.68513888e-01
-2.83481121e-01 -4.13857549e-01 1.26127779e-01 1.25480163e+00
-3.28652859e-01 -1.00205354e-02 2.39748120e-01 2.58849531e-01
-3.83137822e-01 -3.58018279e-01 6.29047215e-01 9.35205221e-01
-5.80767572e-01 -9.11733449e-01 -6.45954072e-01 3.55565041e-01
3.55564505e-02 3.50257993e-01 -4.26166117e-01 1.03683507e+00
6.35401964e-01 7.84120440e-01 -7.21751973e-02 -2.18245968e-01
6.51998222e-01 2.14884952e-02 9.95952904e-01 -6.42466426e-01
1.45631414e-02 -1.71758205e-01 -8.91028792e-02 -5.28832436e-01
-4.76484805e-01 -1.14438415e+00 -9.77890670e-01 1.77965790e-01
-5.10065615e-01 -2.60300994e-01 7.65756428e-01 8.98737490e-01
3.74262959e-01 4.32992190e-01 6.24769449e-01 -1.04133916e+00
-8.28353703e-01 -6.80098653e-01 -8.16516280e-01 5.70492685e-01
3.98603380e-01 -1.24619913e+00 -5.11759102e-01 -1.15866400e-01] | [8.117250442504883, -2.980351686477661] |
33830dff-3266-43bd-9fb3-2c5ac885ad5e | robust-tensor-recovery-using-low-rank-tensor | 1904.00435 | null | https://arxiv.org/abs/1904.00435v3 | https://arxiv.org/pdf/1904.00435v3.pdf | Robust Low-Rank Tensor Ring Completion | Low-rank tensor completion recovers missing entries based on different tensor decompositions. Due to its outstanding performance in exploiting some higher-order data structure, low rank tensor ring has been applied in tensor completion. To further deal with its sensitivity to sparse component as it does in tensor principle component analysis, we propose robust tensor ring completion (RTRC), which separates latent low-rank tensor component from sparse component with limited number of measurements. The low rank tensor component is constrained by the weighted sum of nuclear norms of its balanced unfoldings, while the sparse component is regularized by its l1 norm. We analyze the RTRC model and gives the exact recovery guarantee. The alternating direction method of multipliers is used to divide the problem into several sub-problems with fast solutions. In numerical experiments, we verify the recovery condition of the proposed method on synthetic data, and show the proposed method outperforms the state-of-the-art ones in terms of both accuracy and computational complexity in a number of real-world data based tasks, i.e., light-field image recovery, shadow removal in face images, and background extraction in color video. | ['Yipeng Liu', 'Ce Zhu', 'Huyan Huang'] | 2019-03-31 | null | null | null | null | ['shadow-removal'] | ['computer-vision'] | [ 1.17737658e-01 -4.69834507e-01 -4.56897207e-02 6.47542328e-02
-4.31356490e-01 -3.49091113e-01 1.64040431e-01 -4.72510964e-01
-1.40545100e-01 4.52109486e-01 3.37085515e-01 -3.71560194e-02
-4.72202510e-01 -1.45217374e-01 -3.94865960e-01 -1.15489697e+00
-2.67017186e-01 4.53261435e-02 -2.15673074e-01 -2.42930427e-01
1.20475277e-01 2.93139309e-01 -1.26370072e+00 2.18633965e-01
8.72440755e-01 1.11411691e+00 -5.90841174e-02 1.97015718e-01
1.38127476e-01 1.02475357e+00 -1.24107867e-01 -3.57818067e-01
3.76001269e-01 5.01251146e-02 -6.27105176e-01 6.97464049e-01
3.12431335e-01 -3.90243351e-01 -5.24094641e-01 1.28873932e+00
3.13453436e-01 1.05325334e-01 3.69166225e-01 -1.18606150e+00
-5.64692259e-01 1.79090366e-01 -1.20428765e+00 1.58465534e-01
3.13517362e-01 -3.64314765e-01 9.78852928e-01 -1.48793006e+00
5.11929154e-01 1.40068471e+00 3.17968786e-01 1.53121591e-01
-1.41872716e+00 -5.00382721e-01 1.11355104e-01 1.49529964e-01
-1.48369706e+00 -6.06123030e-01 1.11813283e+00 -7.19760895e-01
1.90316752e-01 4.82338220e-01 2.16168761e-01 7.21447408e-01
1.23936422e-02 8.59986484e-01 1.35741353e+00 -1.04142092e-01
-5.64692589e-03 -2.95060009e-01 2.59116918e-01 8.47358346e-01
5.53666174e-01 -2.15132758e-01 -6.07667625e-01 -4.58844066e-01
7.48216391e-01 3.97360533e-01 -5.92161655e-01 -3.46631229e-01
-1.67710054e+00 6.64693117e-01 1.93514645e-01 2.45439276e-01
-5.84237576e-01 -1.19861744e-01 2.60364115e-01 1.40909776e-01
5.65412819e-01 -9.75540280e-02 -1.33753240e-01 2.99365222e-01
-8.49125147e-01 1.35210350e-01 4.75899965e-01 7.85674393e-01
8.25503290e-01 5.01232862e-01 -1.33724123e-01 1.12436485e+00
4.59381759e-01 8.09526145e-01 4.21008229e-01 -9.10202265e-01
6.27108216e-01 3.50945026e-01 1.10684864e-01 -1.45164394e+00
-1.90083578e-01 -5.47533095e-01 -1.45645595e+00 -2.57414907e-01
4.17670786e-01 3.09939701e-02 -5.84849954e-01 1.45503414e+00
5.74268818e-01 3.57172459e-01 -3.55477519e-02 1.31384230e+00
5.77868044e-01 6.34689271e-01 -6.56245232e-01 -7.47372091e-01
1.59640837e+00 -5.89250028e-01 -1.01028943e+00 6.08783886e-02
2.40431890e-01 -1.16901863e+00 4.25304264e-01 7.33137786e-01
-7.93983340e-01 -3.51765007e-01 -8.33184063e-01 -7.03545958e-02
4.76423204e-01 6.09430015e-01 9.70302343e-01 4.28490043e-01
-5.86051285e-01 4.04610306e-01 -8.29947054e-01 -3.03528998e-02
1.42608836e-01 1.70057520e-01 -7.22243428e-01 -6.29601181e-01
-7.01921880e-01 1.46546230e-01 -6.28691912e-02 7.45464206e-01
-8.59760225e-01 -5.29302835e-01 -6.12478435e-01 -3.19606543e-01
6.04779780e-01 -4.85582560e-01 6.02935314e-01 -5.42415679e-01
-1.31197715e+00 4.67350334e-01 -3.14806521e-01 2.80710340e-01
1.84705228e-01 -2.70577133e-01 -3.04812193e-01 4.28126127e-01
2.09452376e-01 -1.96589082e-01 1.48004854e+00 -1.08497727e+00
-1.47252768e-01 -7.57836759e-01 -6.85813278e-02 6.46318793e-02
-4.24556851e-01 1.61828492e-02 -4.62287724e-01 -9.89884377e-01
9.28051114e-01 -1.07176507e+00 -2.82490402e-01 -1.23985715e-01
-4.82782960e-01 -5.19743450e-02 7.82724440e-01 -9.57141101e-01
1.09779453e+00 -2.33619595e+00 7.55322754e-01 1.31709620e-01
5.10788798e-01 -7.55089298e-02 -5.81781790e-02 4.52868164e-01
-4.29984301e-01 -3.20341915e-01 -2.51201242e-01 -4.45705891e-01
-3.20605785e-01 3.00992817e-01 -4.85861033e-01 9.79632080e-01
1.70703799e-01 1.53297991e-01 -8.04785132e-01 -5.24006248e-01
-9.66038257e-02 6.65436208e-01 -4.14822906e-01 2.48555079e-01
3.48151326e-01 6.35767579e-01 -6.26123309e-01 9.43964124e-01
1.19793832e+00 -2.92949200e-01 1.52501225e-01 -8.92733753e-01
-9.78929400e-02 -1.40125692e-01 -1.71763289e+00 1.69801521e+00
-1.61853284e-01 2.31726170e-01 7.14844584e-01 -1.16185176e+00
7.45855451e-01 4.07032281e-01 9.16420281e-01 -3.30757529e-01
-4.19964828e-02 2.94570386e-01 1.94767129e-03 -7.01766908e-01
4.70281363e-01 -1.84623197e-01 4.18494076e-01 3.99837822e-01
-1.52572945e-01 3.17763150e-01 4.55175430e-01 3.98141235e-01
1.08714354e+00 9.35739651e-03 -1.21872030e-01 -3.47689927e-01
8.88999104e-01 -5.20194232e-01 9.01364028e-01 -5.59616014e-02
7.03827217e-02 5.50025761e-01 5.15151262e-01 -3.75139564e-01
-7.91540384e-01 -5.85051298e-01 -2.85923809e-01 7.60941803e-01
-5.47321439e-02 -4.49464500e-01 -5.12531698e-01 -3.75563771e-01
-1.11249648e-01 -4.99617606e-02 -4.55893964e-01 2.41243437e-01
-5.92493892e-01 -1.29346085e+00 3.03752311e-02 -8.64441618e-02
4.41996723e-01 -2.58976549e-01 -5.93784265e-02 1.29434261e-02
-6.42119348e-01 -1.52522314e+00 -4.77936327e-01 -3.95256072e-01
-1.07744277e+00 -1.33746147e+00 -1.04878104e+00 -3.48172009e-01
1.07933068e+00 1.03713465e+00 5.63444555e-01 7.80553147e-02
-1.82214662e-01 4.38914299e-01 -4.64481711e-01 4.34927195e-01
2.61900902e-01 -4.14342642e-01 3.65671724e-01 1.15962374e+00
-1.98813722e-01 -7.12157547e-01 -5.37639558e-01 5.36620557e-01
-1.27241015e+00 -9.71248373e-03 7.77067661e-01 9.71197009e-01
5.37834287e-01 3.57723653e-01 -1.02069788e-01 -5.23111641e-01
6.38275623e-01 -4.11521286e-01 -7.53424823e-01 1.91373490e-02
-3.88387442e-01 3.39490026e-01 4.37489122e-01 -4.71082807e-01
-8.75997424e-01 1.17533498e-01 4.17226583e-01 -7.88543046e-01
6.15418971e-01 7.80205905e-01 -4.85545248e-02 -2.48829305e-01
2.44162276e-01 4.66162533e-01 6.55360296e-02 -9.36200321e-01
2.30901808e-01 2.96182603e-01 3.74391258e-01 -9.65277493e-01
1.35860431e+00 8.65221798e-01 4.92031068e-01 -1.23048937e+00
-7.66980231e-01 -6.82969391e-01 -5.35225093e-01 -1.64255619e-01
5.33941031e-01 -1.04949450e+00 -9.73735750e-01 5.25349081e-01
-1.23291540e+00 4.63022530e-01 1.01866685e-01 8.33947122e-01
-5.27186282e-02 9.19886887e-01 -6.46502733e-01 -9.93850946e-01
-1.45478219e-01 -1.19115627e+00 1.05052698e+00 -3.30156386e-01
6.37771428e-01 -5.46284020e-01 2.85900421e-02 6.10402584e-01
-1.81194544e-02 3.40028882e-01 6.34122133e-01 1.26712844e-01
-7.74938941e-01 -1.68169647e-01 -3.82014066e-01 6.29740000e-01
-2.80272737e-02 -1.13990091e-01 -7.50765502e-01 -6.49035037e-01
4.60847557e-01 -6.72198758e-02 8.65483582e-01 1.46501228e-01
8.99483442e-01 -7.14708149e-01 -1.72984116e-02 7.89120495e-01
1.42835844e+00 -4.56240624e-01 6.28690124e-01 -1.07150316e-01
1.27515650e+00 7.93488741e-01 5.80856800e-01 6.00115240e-01
2.11067811e-01 7.98422515e-01 4.94374335e-01 4.93878797e-02
6.84660897e-02 1.65129423e-01 5.65166771e-01 1.52564502e+00
-5.65261602e-01 4.33355749e-01 -5.29061913e-01 4.14092630e-01
-2.09759855e+00 -8.36324215e-01 -6.00172698e-01 2.54237318e+00
5.12174249e-01 -3.10425669e-01 1.80803433e-01 5.67275047e-01
6.49903357e-01 4.07104075e-01 -2.54469275e-01 2.00580478e-01
-2.35789582e-01 -1.81047320e-01 5.94911337e-01 4.58088130e-01
-8.22297275e-01 4.56074566e-01 5.30303621e+00 8.49738061e-01
-9.36512589e-01 2.93172926e-01 2.56291240e-01 1.66024789e-01
-2.67309278e-01 2.30035961e-01 -4.08999413e-01 3.14971268e-01
2.62691170e-01 2.01062262e-01 8.49658251e-01 6.25756085e-01
3.06945741e-01 1.73405837e-02 -7.80706048e-01 1.33535123e+00
2.62712508e-01 -9.63035345e-01 7.87608847e-02 4.02214408e-01
6.74354851e-01 -2.27311835e-01 1.66044921e-01 -1.11027472e-01
-1.72091767e-01 -6.89413905e-01 5.45248389e-01 4.88873631e-01
7.20368266e-01 -5.24541199e-01 5.45338392e-01 1.81980714e-01
-1.29585314e+00 -7.49181062e-02 -6.46175683e-01 -1.41712546e-01
-3.04506235e-02 1.37339580e+00 -3.47140491e-01 9.04490829e-01
4.75722462e-01 9.64071453e-01 -3.00051868e-01 8.25218320e-01
-1.44599184e-01 5.87215960e-01 -3.60898316e-01 5.17040908e-01
-1.48445172e-02 -9.51830447e-01 9.00191426e-01 8.10724735e-01
2.36782998e-01 4.44829911e-01 3.87091845e-01 4.72983688e-01
9.37759727e-02 2.53399521e-01 -3.14035118e-01 -1.07317120e-01
1.70274042e-02 1.65274620e+00 -5.41244626e-01 -4.40982021e-02
-3.40360105e-01 9.52549696e-01 -8.67541283e-02 7.74666250e-01
-3.39394778e-01 -2.00878263e-01 7.28808105e-01 1.88023955e-01
3.37430626e-01 -6.29729450e-01 7.22280219e-02 -1.89221299e+00
4.75847423e-01 -1.04822421e+00 1.92340031e-01 -6.13237500e-01
-1.24716353e+00 6.02526367e-01 -1.83365699e-02 -1.50350332e+00
2.64196340e-02 -7.39539266e-01 -1.80390075e-01 7.52140641e-01
-1.55274677e+00 -1.29643846e+00 -3.18138361e-01 1.07007408e+00
1.16761856e-01 -1.03730716e-01 4.39618140e-01 7.86833763e-01
-1.09996080e+00 3.43615055e-01 3.29518110e-01 2.89732099e-01
4.40875292e-01 -8.38491321e-01 -2.96309143e-01 1.23963833e+00
-1.20387353e-01 8.77613962e-01 6.41629159e-01 -5.22137403e-01
-2.21658349e+00 -8.05488348e-01 5.05020916e-01 5.51544353e-02
8.53645205e-01 -2.86748707e-01 -7.45586574e-01 5.55177033e-01
-3.30720901e-01 5.11256635e-01 5.23747623e-01 1.32905781e-01
-7.72495270e-01 -5.65687716e-01 -9.46657002e-01 2.78410077e-01
7.05683470e-01 -3.99718404e-01 -1.51551247e-01 6.70731127e-01
5.69861472e-01 -2.42622405e-01 -1.05982888e+00 3.43865335e-01
4.47886348e-01 -8.61018896e-01 1.08028281e+00 -4.60909098e-01
3.92057776e-01 -8.13997328e-01 -6.26532972e-01 -9.48793590e-01
-5.26929796e-01 -1.10396755e+00 -2.85428137e-01 1.17895293e+00
-1.42159998e-01 -4.85159576e-01 5.73813975e-01 2.45299265e-01
3.87945473e-02 -5.83213985e-01 -1.10846138e+00 -7.83928156e-01
-6.86683536e-01 -3.32655996e-01 2.57034719e-01 1.14928114e+00
-1.97445333e-01 5.97932637e-01 -1.03414750e+00 4.28545922e-01
1.30809617e+00 1.27964467e-01 8.07905853e-01 -1.17031217e+00
-4.43104208e-01 1.40379772e-01 -4.71898824e-01 -1.08929801e+00
2.48246267e-02 -7.66263902e-01 -3.05637538e-01 -1.10102177e+00
3.80665272e-01 -3.56492072e-01 -1.88795179e-01 2.72324264e-01
-6.75888211e-02 3.58668000e-01 2.96020985e-01 6.80723071e-01
-4.37326103e-01 7.60093331e-01 1.39687037e+00 -2.74902135e-01
8.71558040e-02 -9.59977880e-02 -5.98229110e-01 6.11762047e-01
5.88083304e-02 -4.21280891e-01 -3.33578587e-01 -5.61713398e-01
5.42622447e-01 3.67422402e-01 1.35303438e-01 -8.42938006e-01
3.34274210e-02 -3.16182762e-01 7.58289993e-02 -4.83090937e-01
6.40115440e-01 -8.85087848e-01 1.68618008e-01 2.86699235e-01
1.25061437e-01 7.97450170e-02 -3.08978379e-01 6.27141654e-01
-2.72364318e-01 -1.15568347e-01 7.19547391e-01 1.24798402e-01
-1.56041101e-01 5.35573006e-01 -7.92349800e-02 -1.44592762e-01
4.27686483e-01 6.77190302e-03 -3.04695725e-01 -3.11549038e-01
-5.05001724e-01 -1.62626848e-01 1.11320242e-01 6.54872805e-02
9.19201851e-01 -1.64299440e+00 -1.12488770e+00 3.36240560e-01
1.67714074e-01 -5.46656037e-03 3.72027457e-01 1.47806406e+00
-2.89474159e-01 2.41273940e-01 -1.99309993e-03 -7.54041374e-01
-1.17536139e+00 7.73907959e-01 -4.55308631e-02 -3.67101640e-01
-4.78917301e-01 4.92174864e-01 3.51652265e-01 -1.17047697e-01
4.81171943e-02 -2.34605059e-01 -2.34418139e-01 7.06145391e-02
7.83962190e-01 7.50025213e-01 1.63113818e-01 -1.11734509e+00
-3.63080382e-01 9.42974746e-01 -9.52120647e-02 -1.17261432e-01
1.39354634e+00 -1.26499996e-01 -9.13471043e-01 2.44614437e-01
1.47097361e+00 2.67250925e-01 -9.17084515e-01 -5.35461366e-01
-2.47942850e-01 -8.77661884e-01 4.53338832e-01 -1.09290674e-01
-1.41168654e+00 7.75530577e-01 4.90338504e-01 6.43631443e-02
1.27550101e+00 -5.23415983e-01 7.64587820e-01 4.77459341e-01
4.00211841e-01 -5.87872505e-01 3.12444985e-01 3.51488620e-01
1.19629312e+00 -1.07428670e+00 6.55886292e-01 -7.27657437e-01
-3.89663130e-01 1.05619740e+00 5.82705401e-02 -3.17670614e-01
6.81169629e-01 -3.69907260e-01 -3.19243461e-01 -3.03940326e-01
-4.63435441e-01 -6.50756955e-02 5.11611283e-01 4.01318371e-01
3.41064930e-01 1.42377809e-01 -5.21029592e-01 9.05286223e-02
1.07687749e-01 -1.85442567e-01 6.40706182e-01 5.79483628e-01
-9.53842029e-02 -1.01174366e+00 -1.14053607e+00 2.51309931e-01
-6.48909688e-01 -6.54258803e-02 3.22225466e-02 2.73241520e-01
-7.05897808e-02 1.25946414e+00 -4.56891567e-01 -4.01336193e-01
8.24621618e-02 -4.50050980e-01 4.89627093e-01 -2.73042828e-01
4.63965349e-02 7.69422114e-01 -2.73522101e-02 -7.05805421e-01
-6.22952878e-01 -7.73219228e-01 -7.29458034e-01 -4.77995694e-01
-2.83242166e-01 3.82079601e-01 8.25492322e-01 8.51357996e-01
9.39152837e-02 2.26296559e-01 1.11183202e+00 -8.77487898e-01
-7.03267932e-01 -9.44951713e-01 -9.81540978e-01 5.74066401e-01
6.34177208e-01 -7.32456267e-01 -4.76658434e-01 8.71677846e-02] | [7.434358596801758, 4.4486775398254395] |
7a73e60f-0c03-4512-925a-356ad43a60b1 | spleeter-a-fast-and-state-of-the-art-music | null | null | https://archives.ismir.net/ismir2019/latebreaking/ | https://archives.ismir.net/ismir2019/latebreaking/000036.pdf | Spleeter: A Fast And State-of-the Art Music Source Separation Tool With Pre-trained Models | We present and release a new tool for music source separation with pre-trained models called Spleeter.Spleeter was designed with ease of use, separation performance and speed in mind. Spleeter is based onTensorflow [1] and makes it possible to:•separate audio files into2,4or5stems with a single command line using pre-trained models.•train source separation models or fine-tune pre-trained ones with Tensorflow (provided you have a dataset of isolated sources).The performance of the pre-trained models are very close to the published state of the art and is, to the authors knowledge, the best performing4stems separation model on the common musdb18 benchmark [6]to be publicly released. Spleeter is also very fast as it can separate a mix audio file into4stems100timesfaster than real-time1on a single Graphics Processing Unit (GPU) using the pre-trained4-stems model. Spleeter is packaged within Docker which makes it usable as is on various platforms. | ['Manuel Moussallam', 'Romain Hennequin', 'Felix Voituret', 'Anis Khlif'] | 2019-11-04 | null | null | null | ismir-2019-late-breaking-demo-2019-11 | ['music-source-separation'] | ['music'] | [-8.81820768e-02 -2.96953201e-01 -2.20163584e-01 9.47827846e-02
-1.04847193e+00 -7.81030834e-01 2.54863381e-01 2.27685392e-01
-2.66725898e-01 2.21733898e-01 4.58073080e-01 7.18201175e-02
-3.19579303e-01 -2.64802396e-01 -5.56319714e-01 -3.31238568e-01
-1.61955550e-01 4.42237675e-01 3.47437024e-01 -8.83096531e-02
1.71607450e-01 2.56881684e-01 -1.86480963e+00 6.36477053e-01
4.03716326e-01 9.21348035e-01 6.06308654e-02 1.31254947e+00
1.73037216e-01 7.97690690e-01 -5.44529676e-01 -6.28007725e-02
4.09281462e-01 -4.48787987e-01 -9.39317763e-01 -1.54585019e-01
4.51778889e-01 -5.73108792e-02 1.15754545e-01 7.17538595e-01
6.56215966e-01 -2.08967492e-01 5.66433668e-02 -1.05199969e+00
-6.28787875e-02 1.39213884e+00 -3.70616972e-01 4.94378895e-01
3.58229041e-01 -1.71714723e-01 1.11842883e+00 -8.20032239e-01
3.46334994e-01 1.04937220e+00 1.12816072e+00 -8.12430382e-02
-1.70342314e+00 -8.72355521e-01 -3.59624654e-01 1.86198205e-01
-1.14905465e+00 -1.01260686e+00 5.97469926e-01 -5.76437056e-01
1.17190146e+00 6.13439739e-01 4.82051879e-01 1.18021679e+00
-2.60937065e-01 7.54886448e-01 9.78228450e-01 -5.42535901e-01
1.43431455e-01 8.88327230e-03 4.62671667e-01 2.22614020e-01
-2.60523647e-01 -7.95295984e-02 -9.78653312e-01 -4.37246531e-01
7.66295075e-01 -5.15820146e-01 -3.50189984e-01 1.84764117e-01
-1.89615631e+00 5.36180973e-01 -1.15289479e-01 5.28899848e-01
-3.46356779e-01 3.14845294e-01 8.89276803e-01 5.81970215e-01
2.09631607e-01 4.82633233e-01 -6.91532135e-01 -7.47520685e-01
-1.51750529e+00 4.09775555e-01 9.98755097e-01 9.61914480e-01
3.21482062e-01 4.25298572e-01 8.24754089e-02 1.01983404e+00
1.02739625e-01 3.42517346e-01 8.95526588e-01 -1.26245093e+00
3.21823627e-01 -5.03696762e-02 -2.57875264e-01 -6.68081164e-01
-4.30467188e-01 -6.81819201e-01 -5.03183067e-01 4.74261403e-01
6.34959400e-01 -1.97143272e-01 -5.86317778e-01 1.22657275e+00
1.96538776e-01 4.78454322e-01 -1.56726956e-01 8.39635730e-01
1.09307241e+00 5.87817967e-01 -5.07514000e-01 -2.25462373e-02
1.26032341e+00 -1.28700233e+00 -6.28800094e-01 -2.63339113e-02
3.05219233e-01 -1.47098434e+00 9.90046859e-01 1.12752700e+00
-1.35746753e+00 -9.19559896e-01 -1.24387467e+00 9.98719111e-02
-8.67237225e-02 5.06160915e-01 6.41494393e-01 8.22658777e-01
-9.87714231e-01 1.13730872e+00 -1.08199346e+00 -1.76134810e-01
-3.37684266e-02 2.95819759e-01 -3.66517901e-01 6.45285487e-01
-8.13221097e-01 3.74641687e-01 6.28792346e-01 -3.47477973e-01
-7.65067399e-01 -1.03050137e+00 -4.52127904e-01 5.56993261e-02
3.53367388e-01 -4.37814951e-01 1.54942834e+00 -1.29112530e+00
-1.95137715e+00 6.29790545e-01 2.26013258e-01 -5.21138370e-01
2.47436002e-01 -3.42819571e-01 -6.80790663e-01 2.43011355e-01
5.71000800e-02 2.88316250e-01 1.14964986e+00 -6.64081514e-01
-4.55695331e-01 -3.54660898e-02 -2.82078713e-01 -1.83516294e-02
-1.76910266e-01 5.09484351e-01 -3.31889659e-01 -1.22965479e+00
-1.69344544e-01 -9.61750090e-01 2.54969686e-01 -4.31868851e-01
-6.97233140e-01 1.94757789e-01 3.90029103e-01 -8.49274576e-01
1.08378899e+00 -2.36154318e+00 2.56936073e-01 1.00538701e-01
9.58303809e-02 3.19019407e-01 -2.37085462e-01 5.43934047e-01
-6.95772290e-01 -1.83696374e-01 8.56171325e-02 -4.89862055e-01
2.48559430e-01 -2.78690666e-01 -5.14243186e-01 3.62931937e-01
-2.82079041e-01 2.65120089e-01 -7.03862071e-01 -1.82042927e-01
2.43751958e-01 4.93174881e-01 -7.09485888e-01 1.40278473e-01
-5.40481471e-02 2.87375450e-01 2.76648641e-01 4.09773082e-01
6.79838359e-01 1.05496198e-01 -2.04629656e-02 -8.89534652e-02
-4.62944299e-01 7.03732669e-01 -1.87383962e+00 2.14503384e+00
-4.04588580e-01 9.40761328e-01 4.41483170e-01 -4.97633368e-01
6.50889218e-01 6.11109793e-01 6.09107077e-01 1.88309699e-01
1.55658454e-01 4.92624164e-01 1.26307443e-01 -1.92541942e-01
3.23078036e-01 6.86259121e-02 6.43421859e-02 7.64114380e-01
7.85888672e-01 -2.10889474e-01 4.77492630e-01 -7.26115308e-04
1.05931842e+00 3.08366686e-01 6.50590435e-02 -3.85110885e-01
3.13405544e-01 -8.96509811e-02 2.03228787e-01 6.40452743e-01
1.68862551e-01 9.76751387e-01 3.63151014e-01 4.57472466e-02
-8.24222982e-01 -1.12337410e+00 -2.12378308e-01 1.55770898e+00
-6.13462090e-01 -1.18476772e+00 -8.32188487e-01 -7.63807073e-02
-1.96105927e-01 6.03133440e-01 -3.15369606e-01 3.41743231e-01
-1.60032496e-01 -6.83486462e-01 1.01672804e+00 4.62663770e-01
2.43777439e-01 -1.17752016e+00 -5.43716371e-01 4.24847156e-01
-1.11930512e-01 -9.67247605e-01 -5.66106379e-01 4.78495926e-01
-7.42101967e-01 -9.68970358e-01 -4.78155702e-01 -6.46275997e-01
-4.42102313e-01 1.24558628e-01 1.27826786e+00 -3.97796988e-01
-2.25063950e-01 4.33073431e-01 -5.07950544e-01 -6.30615950e-01
-2.59258687e-01 2.86853731e-01 -2.02391036e-02 -9.19855759e-03
1.68436527e-01 -9.73695278e-01 -2.48010352e-01 3.55576091e-02
-6.40821159e-01 -5.49919065e-03 4.40767817e-02 4.52062905e-01
5.84257185e-01 4.13503230e-01 4.44119245e-01 -5.64952791e-01
7.36739695e-01 -4.69448358e-01 -5.96459746e-01 -3.73471737e-01
-4.26804781e-01 -1.21939614e-01 7.32602298e-01 -3.00907046e-01
-6.53907895e-01 2.23016828e-01 -3.80153745e-01 -6.52608275e-01
-3.58620077e-01 4.24448699e-01 1.99287772e-01 1.22121170e-01
7.21960068e-01 -1.07872307e-01 -1.33205727e-01 -1.11676753e+00
3.04108322e-01 7.30082214e-01 8.49805295e-01 -4.04364437e-01
4.62878823e-01 2.82233417e-01 -4.43104684e-01 -9.44549203e-01
-7.30213761e-01 -7.61883199e-01 -6.28987551e-01 1.55861422e-01
6.77303255e-01 -1.07826889e+00 -7.20184207e-01 6.86733902e-01
-9.68604386e-01 -2.53314078e-01 -4.04740512e-01 7.87591338e-01
-4.88805771e-01 7.67948925e-02 -5.66298187e-01 -5.06361783e-01
-4.55809623e-01 -8.68121445e-01 1.01697826e+00 4.67820056e-02
-6.30306005e-01 -9.30496693e-01 5.57347953e-01 2.92730451e-01
1.96674183e-01 1.20338373e-01 3.18339169e-01 -8.12723875e-01
-1.20188855e-01 -1.14287697e-01 1.35339573e-01 6.98096752e-01
-7.75706917e-02 4.28166717e-01 -1.53960621e+00 6.77045062e-03
1.32067174e-01 -1.88245341e-01 9.73324418e-01 4.24132317e-01
1.08752751e+00 3.60163711e-02 2.77306270e-02 7.89444029e-01
9.92399395e-01 -2.34366134e-01 4.14439529e-01 4.03487444e-01
5.70544243e-01 3.02938104e-01 1.61493644e-01 6.39697075e-01
1.80374399e-01 8.15207481e-01 2.11392924e-01 -1.10271119e-03
-3.97268325e-01 8.17125589e-02 6.68313324e-01 1.19832623e+00
-3.95038158e-01 2.06920683e-01 -6.86876774e-01 4.23028976e-01
-1.69238424e+00 -1.18595016e+00 -8.20134044e-01 2.33550358e+00
1.00485563e+00 2.74967283e-01 7.29635060e-01 8.42378914e-01
3.72485042e-01 1.81101009e-01 -2.48478223e-02 -4.07502204e-01
3.70594747e-02 8.54815900e-01 3.30363810e-01 5.10051787e-01
-1.37258208e+00 4.42368567e-01 6.90327215e+00 8.79008651e-01
-1.28796303e+00 3.93634558e-01 -1.64478004e-01 -6.76191628e-01
2.18879491e-01 1.14640370e-01 -6.47641361e-01 6.41828537e-01
1.46959281e+00 -1.56872511e-01 8.80661488e-01 8.10590804e-01
2.07459882e-01 -7.98541214e-03 -1.22858596e+00 1.45759010e+00
1.10102154e-01 -1.29613328e+00 -6.43185258e-01 -2.28086650e-01
1.57058984e-01 5.48824072e-01 -1.01316767e-03 3.04363370e-01
1.96934164e-01 -8.61650407e-01 1.16613078e+00 3.48422348e-01
5.07955790e-01 -9.95524049e-01 5.52339673e-01 2.44169667e-01
-1.14757490e+00 5.03477007e-02 -1.26109019e-01 -7.28463903e-02
5.28463274e-02 7.04384565e-01 -8.85151505e-01 6.44732237e-01
1.10342181e+00 8.03272188e-01 -6.97399795e-01 1.19233584e+00
-2.88685947e-03 1.11906683e+00 -5.25119483e-01 7.52792776e-01
-4.46850806e-03 -2.45532002e-02 9.46633756e-01 1.50018728e+00
4.04410064e-01 -3.77830952e-01 1.91492945e-01 5.69901645e-01
2.24813193e-01 3.10811192e-01 1.31444573e-01 -2.64040768e-01
1.59502193e-01 1.56640768e+00 -8.14035177e-01 -2.90804923e-01
-1.39143094e-01 8.21583688e-01 -2.38094166e-01 -1.78885281e-01
-8.22726488e-01 -9.11501229e-01 6.48714602e-01 3.62964392e-01
5.76920331e-01 -8.36164728e-02 -4.14926767e-01 -1.40787673e+00
-2.35041335e-01 -1.27721155e+00 5.16548872e-01 -9.98588026e-01
-1.07476091e+00 1.04702199e+00 -5.85078634e-02 -1.27399182e+00
-2.89609462e-01 -6.01608992e-01 -6.73104703e-01 7.44317114e-01
-1.14258015e+00 -7.24261165e-01 -3.35123718e-01 7.18949676e-01
5.95398486e-01 -3.24841797e-01 1.33964634e+00 4.93515790e-01
-5.41478753e-01 4.12775278e-01 1.89411104e-01 -1.71260256e-02
1.05538762e+00 -1.47775030e+00 3.23874325e-01 7.59027719e-01
6.81946158e-01 4.87608165e-01 7.64442801e-01 -2.97237754e-01
-1.16308892e+00 -7.68041074e-01 7.41888881e-01 -4.20102060e-01
1.17335188e+00 -6.23760939e-01 -8.38834465e-01 7.25701034e-01
7.92781889e-01 -1.54950082e-01 1.08453512e+00 4.66060489e-01
-5.97457647e-01 -3.04668993e-01 -5.24205208e-01 4.36111204e-02
6.82863295e-01 -5.32508731e-01 -7.28303254e-01 4.49576557e-01
3.48801225e-01 -7.35764861e-01 -1.07155228e+00 5.21376380e-04
5.62911689e-01 -1.52018607e+00 1.02168012e+00 -2.37340815e-02
6.45912811e-02 -3.15914303e-01 -9.04381424e-02 -1.53807473e+00
-3.40806425e-01 -1.35634184e+00 -2.10210323e-01 1.39818919e+00
2.97212303e-01 -4.12401766e-01 2.54528284e-01 -2.06065446e-01
-2.62023807e-01 -3.03606130e-02 -7.10911989e-01 -9.74281073e-01
-2.64695436e-01 -9.11842644e-01 6.57557607e-01 8.68223369e-01
2.11483017e-01 6.40412807e-01 -3.81581545e-01 1.41064832e-02
5.54383099e-01 3.16665620e-01 9.99354362e-01 -1.41418684e+00
-1.07596934e+00 -4.98417020e-01 -4.75306153e-01 -5.46210706e-01
8.34261905e-03 -1.14707732e+00 -2.81865746e-01 -8.45352530e-01
-1.47241250e-01 -3.41622978e-01 -5.11788487e-01 8.91870201e-01
2.42086485e-01 6.30786359e-01 3.94728571e-01 2.31730446e-01
-3.30476642e-01 2.15690747e-01 6.41790271e-01 1.05520943e-02
-5.45876086e-01 3.53918672e-01 -7.90926516e-01 8.09251368e-01
6.43609703e-01 -6.93582892e-01 -5.30748844e-01 -3.10553551e-01
1.14028670e-01 -2.23603591e-01 4.57347810e-01 -1.56369507e+00
-1.07421249e-01 4.30947512e-01 4.05758470e-01 -6.74186349e-01
4.91151035e-01 -5.80143809e-01 5.00418901e-01 8.44939798e-02
-2.72793174e-01 -1.59276485e-01 4.92943823e-01 1.99148133e-01
-2.93948710e-01 -3.92254680e-01 6.64712310e-01 6.14658333e-02
-3.85773957e-01 -1.67730942e-01 -4.63552386e-01 1.15355939e-01
7.41822839e-01 4.01462615e-02 3.23137268e-02 -4.43051219e-01
-1.18856716e+00 -3.41390103e-01 1.27383024e-01 5.29689848e-01
8.39584693e-02 -1.42468679e+00 -9.44192052e-01 3.46671015e-01
-9.43806767e-02 -3.47150326e-01 2.50217795e-01 1.13245010e+00
-7.16798544e-01 1.43009365e-01 -2.34694168e-01 -7.70227790e-01
-1.53593588e+00 5.36819577e-01 2.11034566e-01 6.01352304e-02
-7.58323193e-01 9.10745382e-01 -4.25291628e-01 -5.31645834e-01
3.31182212e-01 -6.53394580e-01 -1.60469353e-01 4.73571897e-01
9.03368294e-01 7.29030252e-01 5.77242672e-01 -5.13870060e-01
-4.49359328e-01 2.65674055e-01 2.38418862e-01 -4.18511033e-01
1.65386093e+00 2.61194676e-01 -2.10664913e-01 9.24847782e-01
9.67318773e-01 5.76558173e-01 -1.04853201e+00 7.92082921e-02
-1.01185545e-01 -4.65415776e-01 5.18045425e-01 -6.31756783e-01
-1.04637492e+00 8.07331443e-01 5.27897596e-01 5.16720772e-01
1.19937932e+00 -2.80201823e-01 6.99557543e-01 6.32984415e-02
3.38603333e-02 -1.22927952e+00 -6.32337704e-02 5.54214597e-01
9.82747018e-01 -7.42183328e-01 1.46217160e-02 -3.49081069e-01
-6.31104827e-01 1.47529876e+00 2.27606893e-02 -5.94004571e-01
1.01271224e+00 8.39295447e-01 2.47228578e-01 4.55651842e-02
-8.07013869e-01 -1.57387853e-01 6.18735254e-01 5.57063997e-01
6.95075631e-01 4.44785878e-02 2.83588111e-01 8.66011083e-01
-8.67991626e-01 1.22728445e-01 6.12589121e-01 7.40785956e-01
2.68864259e-02 -1.29901409e+00 -8.14770103e-01 1.69363737e-01
-8.45593154e-01 -3.82831901e-01 -3.39145511e-01 5.00123084e-01
3.03990692e-01 1.17284906e+00 -1.05792612e-01 -4.14085209e-01
2.75951833e-01 -1.33136278e-02 6.01411343e-01 -7.66897857e-01
-1.33217692e+00 8.51761281e-01 2.93641478e-01 -6.27213538e-01
-4.68683213e-01 -8.25838983e-01 -1.05163896e+00 -3.36328954e-01
-3.52444977e-01 1.67150617e-01 7.81795859e-01 7.74177074e-01
6.01769745e-01 7.02533484e-01 3.31794888e-01 -1.25276983e+00
-3.09498668e-01 -1.20095754e+00 -7.67838538e-01 -4.80383299e-02
2.91642606e-01 -2.98907131e-01 -3.18342119e-01 4.97597009e-01] | [15.644262313842773, 5.4791951179504395] |
b341c6c5-f012-4434-8876-fdf45f5c6a1f | transfer-learning-with-jukebox-for-music | 2111.14200 | null | https://arxiv.org/abs/2111.14200v3 | https://arxiv.org/pdf/2111.14200v3.pdf | Transfer Learning with Jukebox for Music Source Separation | In this work, we demonstrate how a publicly available, pre-trained Jukebox model can be adapted for the problem of audio source separation from a single mixed audio channel. Our neural network architecture, which is using transfer learning, is quick to train and the results demonstrate performance comparable to other state-of-the-art approaches that require a lot more compute resources, training data, and time. We provide an open-source code implementation of our architecture (https://github.com/wzaielamri/unmix) | ['A. Melnik', 'H. Ritter', 'O. Tautz', 'W. Zai El Amri'] | 2021-11-28 | null | null | null | null | ['audio-source-separation', 'music-source-separation'] | ['audio', 'music'] | [-1.83049470e-01 -4.30090427e-01 6.46709725e-02 -2.55090028e-01
-1.52783334e+00 -5.88835299e-01 1.30286708e-01 -5.71009934e-01
-9.31752399e-02 4.54287052e-01 3.50246787e-01 -2.99670935e-01
1.20652407e-01 -2.95293897e-01 -7.55378008e-01 -4.77849066e-01
-3.33440870e-01 1.63682446e-01 1.60164252e-01 -2.98384368e-01
-2.94374317e-01 -1.70529053e-01 -1.51102901e+00 6.33702993e-01
1.98255941e-01 1.17728150e+00 -6.93312883e-02 1.29662693e+00
2.71802843e-01 1.11511838e+00 -6.24797463e-01 -1.08730465e-01
1.72967449e-01 -6.52627885e-01 -8.99365842e-01 -3.58156174e-01
7.25630164e-01 -5.12008786e-01 -6.22250497e-01 9.79572535e-01
1.10508108e+00 2.27336898e-01 2.83191770e-01 -1.38174069e+00
-3.66423815e-01 9.61960673e-01 -6.64218664e-02 4.34445202e-01
3.77710521e-01 5.56416214e-02 9.85145092e-01 -1.11142457e+00
5.86418845e-02 1.21390045e+00 9.01896834e-01 4.90660876e-01
-1.08927310e+00 -1.17835855e+00 -1.43896893e-01 3.52663606e-01
-1.47962451e+00 -1.13432598e+00 7.72728503e-01 -4.43961442e-01
9.14317608e-01 2.68474221e-01 3.49815041e-01 1.32732558e+00
-4.19487745e-01 8.75763237e-01 8.64392579e-01 -4.76219177e-01
1.63309649e-01 -4.96506281e-02 4.35346010e-04 4.80384320e-01
-5.07612705e-01 3.74386050e-02 -9.17931736e-01 -4.72207665e-01
5.73662281e-01 -2.92255759e-01 -5.79204142e-01 2.45983973e-01
-1.03906178e+00 6.56768084e-01 3.46413255e-01 2.79153585e-01
5.28916977e-02 5.38249910e-01 5.99781811e-01 5.54371238e-01
8.24715793e-01 4.88437302e-02 -6.54692054e-01 -6.37064993e-01
-1.22926939e+00 1.36604339e-01 9.47551489e-01 1.06321132e+00
4.53239858e-01 5.83545446e-01 2.53604680e-01 1.05070555e+00
2.49996424e-01 3.01482379e-01 7.13834524e-01 -1.38930631e+00
3.57947439e-01 -2.52763093e-01 -2.82199204e-01 -4.14688677e-01
-1.01566061e-01 -4.04182941e-01 -7.29966819e-01 1.22780539e-01
3.91589135e-01 -7.05170512e-01 -8.29072535e-01 1.61310506e+00
9.53796431e-02 1.08698809e+00 8.76399875e-03 1.01155782e+00
1.34040415e+00 9.17653799e-01 -4.07999188e-01 3.60390484e-01
1.09130430e+00 -1.65657592e+00 -6.22730196e-01 -2.63233006e-01
1.01424515e-01 -1.23493862e+00 9.15795743e-01 6.55044258e-01
-1.39944017e+00 -6.68408334e-01 -1.04411542e+00 -1.53474942e-01
-2.23537102e-01 3.67505848e-01 5.63692868e-01 4.79338020e-01
-1.30512190e+00 7.87684500e-01 -7.43945837e-01 -4.51466255e-02
1.81103498e-01 1.73599005e-01 -1.36771217e-01 1.90577477e-01
-1.24062324e+00 2.99331993e-01 1.89950868e-01 2.53268760e-02
-1.55437613e+00 -7.98322976e-01 -6.79675460e-01 1.92498729e-01
4.34337109e-01 -3.14703375e-01 2.00138879e+00 -1.01900959e+00
-2.12033629e+00 3.32918346e-01 -1.23016775e-01 -3.66320014e-01
1.36304051e-01 -6.10563755e-01 -7.47233152e-01 4.12114263e-01
-2.32289612e-01 4.65328574e-01 1.21540093e+00 -1.00366175e+00
-5.63430488e-01 2.31233731e-01 -1.40212432e-01 -9.03644711e-02
-2.96861172e-01 6.61921620e-01 -4.00476575e-01 -9.15354192e-01
-4.02708530e-01 -9.79899943e-01 8.27991664e-02 -2.03976944e-01
-3.28455150e-01 9.70711634e-02 5.88241398e-01 -9.14997280e-01
1.23685420e+00 -2.57082176e+00 4.66228500e-02 -8.46007690e-02
5.55498116e-02 3.70283604e-01 -4.29072291e-01 5.50425947e-01
-3.19791168e-01 6.54317951e-03 -2.37427473e-01 -4.81150985e-01
1.27988741e-01 -3.35298955e-01 -7.80439436e-01 2.47485995e-01
-1.12011358e-01 3.63221020e-01 -8.19023490e-01 -1.22826159e-01
-7.20918551e-02 6.15591347e-01 -6.68382287e-01 5.58695436e-01
2.34596860e-02 3.42799217e-01 3.42017025e-01 7.35048831e-01
5.72587967e-01 -1.60977662e-01 -9.15667415e-02 4.57228115e-03
-1.60083491e-02 8.37990224e-01 -1.43161988e+00 2.02582383e+00
-4.66321290e-01 1.03264451e+00 7.61400998e-01 -6.77382767e-01
4.60913122e-01 1.00290644e+00 3.63559216e-01 3.60722952e-02
3.16667467e-01 5.04787862e-01 -8.46298877e-03 -2.60421097e-01
2.97006994e-01 2.28795037e-01 1.12863041e-01 6.99310303e-01
8.43797922e-01 -3.47635061e-01 1.80981621e-01 2.18163952e-01
1.00431705e+00 6.47847056e-02 -1.45206124e-01 -8.87952968e-02
1.89435944e-01 -3.67082506e-01 3.55076075e-01 5.31642079e-01
-1.41160414e-01 8.24558198e-01 6.74117655e-02 -1.42633587e-01
-7.90630698e-01 -1.10014081e+00 7.89787434e-03 1.56153297e+00
-4.36968803e-01 -8.09855521e-01 -1.02714849e+00 -4.91146073e-02
-1.35877058e-01 1.64410323e-01 -9.02369693e-02 1.07290305e-01
-5.41721404e-01 -4.30145532e-01 1.33605111e+00 6.29104614e-01
3.19636524e-01 -9.50828135e-01 -1.44463211e-01 3.06596398e-01
-3.41724843e-01 -1.05578804e+00 -5.74913085e-01 4.27009642e-01
-6.77047431e-01 -8.80899906e-01 -9.28861201e-01 -8.30907524e-01
-7.22481161e-02 2.14451090e-01 1.34490168e+00 -8.16192999e-02
-2.20515430e-01 4.06542748e-01 -3.94683599e-01 -7.91261137e-01
-2.82632619e-01 2.79488504e-01 1.78337440e-01 -1.39190763e-01
2.46262535e-01 -8.65174353e-01 -5.29952884e-01 1.94782168e-01
-8.66731226e-01 -2.49049351e-01 5.27973212e-02 5.68528414e-01
2.86796421e-01 -7.23900273e-02 7.31166482e-01 -4.40100670e-01
8.53650630e-01 -7.05510914e-01 -4.17169541e-01 -2.30916336e-01
-1.71694711e-01 -3.77375752e-01 4.97570813e-01 -7.02242851e-01
-7.98214555e-01 -1.23203229e-02 -6.99370027e-01 -5.46393633e-01
-3.77906233e-01 3.99407297e-01 1.79999232e-01 8.94181896e-03
5.61059773e-01 -8.27129632e-02 -1.43787246e-02 -9.02552247e-01
5.83621800e-01 1.19601786e+00 6.40976608e-01 -4.98844057e-01
5.38927495e-01 4.13953245e-01 -6.61492944e-01 -7.38353610e-01
-8.23813736e-01 -6.24105275e-01 -3.31488341e-01 -4.13834006e-02
5.16303480e-01 -1.62598038e+00 -4.55835372e-01 7.31853306e-01
-1.04447353e+00 -8.20069015e-01 -5.36706075e-02 7.69415855e-01
-6.34624779e-01 -4.54908013e-02 -1.14147711e+00 -7.02275515e-01
-5.04071295e-01 -8.93159568e-01 1.03542209e+00 3.99079025e-02
-2.47962862e-01 -7.71604717e-01 5.03452659e-01 1.43009320e-01
6.00905299e-01 -1.24449849e-01 2.50864863e-01 -6.18823707e-01
-4.48489308e-01 -2.73048468e-02 2.93983929e-02 8.66836369e-01
2.90225819e-02 1.97704718e-01 -1.67271948e+00 -4.03507471e-01
4.92791720e-02 -6.59555554e-01 9.73844767e-01 3.48950446e-01
1.25444889e+00 -4.08043921e-01 9.14036408e-02 8.81974101e-01
9.58980620e-01 8.98362510e-03 2.97009081e-01 1.49033204e-01
4.52855200e-01 9.61789712e-02 6.15599081e-02 4.36149269e-01
3.10596198e-01 4.52086091e-01 2.93883413e-01 -4.25033718e-01
-3.29605043e-01 -1.18121393e-01 6.83347464e-01 1.46925962e+00
-7.06841797e-02 -4.04372588e-02 -7.02988148e-01 7.02309608e-01
-1.81046510e+00 -9.75804150e-01 3.68023515e-02 1.67034101e+00
1.28484321e+00 -2.80951560e-01 4.50482458e-01 4.22953814e-01
5.89897811e-01 2.04521477e-01 -5.35700284e-02 -2.68077284e-01
1.34974241e-01 6.71624124e-01 1.77240800e-02 7.08668232e-01
-1.29708171e+00 7.92450368e-01 7.93920374e+00 9.81526792e-01
-1.31357455e+00 4.19044971e-01 2.40467057e-01 -6.84381545e-01
2.37581015e-01 -1.89205438e-01 -5.45031548e-01 6.02070630e-01
1.70403028e+00 -2.12785625e-03 9.45226252e-01 8.78219187e-01
-8.21840391e-02 2.32043952e-01 -1.13632035e+00 1.06911755e+00
6.83978051e-02 -1.12325835e+00 -5.14722049e-01 -3.31178904e-01
4.39252138e-01 6.01040483e-01 1.44147292e-01 4.41110313e-01
4.89375591e-01 -8.36770535e-01 9.78481472e-01 3.93305495e-02
8.16889942e-01 -6.98324800e-01 4.81551379e-01 1.35039493e-01
-1.40720332e+00 -1.02283165e-01 -3.98836792e-01 -2.81392187e-01
2.77099669e-01 6.59950078e-01 -5.97045481e-01 4.76660073e-01
1.26316094e+00 5.64317703e-01 -3.84758353e-01 1.18820882e+00
-3.73135805e-01 1.31421220e+00 -3.85700583e-01 4.09501791e-01
1.70131952e-01 2.06748933e-01 5.72843492e-01 1.59658813e+00
5.91875255e-01 -1.66129455e-01 1.47095352e-01 6.55992508e-01
-3.18526059e-01 -3.89206223e-02 -3.59175950e-01 1.05229076e-02
5.17795742e-01 1.35918784e+00 -1.67729557e-01 -5.06972432e-01
-4.28426296e-01 6.97458982e-01 1.76373437e-01 5.39739847e-01
-9.42102194e-01 -7.09900916e-01 9.36859131e-01 1.25431111e-02
5.17217934e-01 -1.76793188e-01 1.75883323e-01 -1.30969155e+00
-1.11906327e-01 -1.26245785e+00 4.46615398e-01 -1.09057927e+00
-1.28350365e+00 9.92186666e-01 -3.41007821e-02 -1.39843810e+00
-6.30639315e-01 -6.92836642e-01 -7.85482585e-01 7.94410348e-01
-1.55588412e+00 -8.41497898e-01 -2.04051793e-01 9.01933730e-01
5.07246196e-01 -4.50066835e-01 1.08412230e+00 8.67560923e-01
-4.10744488e-01 5.74693263e-01 1.42063633e-01 4.37913656e-01
1.17140484e+00 -1.23262954e+00 6.06785297e-01 7.23264873e-01
6.07262015e-01 4.57005590e-01 5.27982891e-01 4.80801240e-03
-1.05778193e+00 -9.58235562e-01 3.74178588e-01 -3.82750660e-01
1.07337546e+00 -5.54714382e-01 -7.81642973e-01 9.14763451e-01
9.08662915e-01 8.81626531e-02 1.31878150e+00 3.12105477e-01
-6.15018070e-01 -2.01494262e-01 -5.67134619e-01 2.28250608e-01
7.03075707e-01 -9.73920703e-01 -4.26439673e-01 9.19246152e-02
8.38136435e-01 -5.91172099e-01 -7.61765718e-01 2.56462753e-01
5.42388022e-01 -8.27113986e-01 1.03960717e+00 -4.84953612e-01
4.01367456e-01 -2.65810013e-01 -1.17316134e-01 -1.64875293e+00
-2.16498792e-01 -1.08472276e+00 -5.58917999e-01 1.30723071e+00
5.38959861e-01 -4.92743671e-01 2.38394275e-01 1.61596343e-01
-2.20642954e-01 -4.63040113e-01 -9.73844111e-01 -9.39341843e-01
1.57340139e-01 -8.50373507e-01 7.16313064e-01 7.91854024e-01
2.33315036e-01 3.83951008e-01 -5.96679688e-01 1.13454312e-01
3.13261449e-01 2.43379306e-02 8.90809596e-01 -9.94819582e-01
-8.17580819e-01 -3.14700156e-01 -9.26905591e-03 -1.02062726e+00
2.36964762e-01 -8.71374965e-01 1.85606122e-01 -1.16187990e+00
-2.78232098e-01 -8.34598690e-02 -5.81888616e-01 5.99775195e-01
4.92011793e-02 8.49704325e-01 4.79706615e-01 2.20043451e-01
-5.42528629e-01 2.41993561e-01 5.85119486e-01 -2.72051603e-01
-1.63082197e-01 4.54989523e-02 -7.74205148e-01 9.54792380e-01
1.04611540e+00 -7.07077146e-01 -1.81372151e-01 -8.48848581e-01
-8.89878441e-03 -1.45798743e-01 4.15076196e-01 -1.31338692e+00
2.27214321e-01 3.62259626e-01 7.90718645e-02 -2.85038084e-01
7.78781116e-01 -6.25116169e-01 9.07117650e-02 -1.23954080e-02
-3.36885005e-01 -1.76048651e-01 5.31922400e-01 9.27119777e-02
-6.94153309e-01 -4.14936364e-01 5.72947681e-01 -1.62235901e-01
-3.99445742e-01 1.86263099e-01 -5.39504349e-01 1.60008460e-01
4.69870031e-01 5.58301270e-01 -4.50145155e-01 -9.06137109e-01
-7.79968441e-01 -1.62285790e-01 -1.44016296e-01 5.15253127e-01
4.56169814e-01 -1.55724275e+00 -9.46069717e-01 1.13984890e-01
-1.96235672e-01 -2.06035957e-01 7.41597265e-02 6.95338488e-01
-4.73153859e-01 2.13410363e-01 1.24064162e-02 -4.59696203e-01
-1.34651971e+00 3.81967664e-01 3.96968007e-01 8.39192048e-02
-4.44659501e-01 1.19900084e+00 -2.94198513e-01 -5.44384956e-01
4.48953092e-01 -3.63727003e-01 2.17795387e-01 -1.62330177e-02
9.61618721e-01 7.19327390e-01 1.83840826e-01 -4.40816283e-01
-3.22376698e-01 2.71665245e-01 3.52742672e-01 -5.17382801e-01
1.34822595e+00 1.66869432e-01 -2.48273388e-01 8.16557944e-01
1.25511861e+00 3.05573702e-01 -1.12925899e+00 -3.38933140e-01
-4.16188598e-01 -2.94711083e-01 3.60473901e-01 -6.21766329e-01
-1.15897274e+00 1.23882163e+00 6.64634526e-01 4.31177229e-01
1.32568824e+00 1.60354331e-01 9.39245284e-01 3.80029261e-01
1.01832092e-01 -1.13260376e+00 7.82480463e-02 6.51274085e-01
8.68080199e-01 -9.55888748e-01 -2.27659762e-01 -2.82892317e-01
-4.80342358e-01 1.08564043e+00 3.83706927e-01 -4.48213220e-01
1.15633583e+00 9.86506104e-01 6.08025074e-01 1.65255770e-01
-9.28664625e-01 -2.14391455e-01 3.70360643e-01 5.24479508e-01
8.04636955e-01 3.72209512e-02 5.36163628e-01 8.89661551e-01
-5.40034056e-01 1.01507530e-01 5.10275304e-01 8.79300117e-01
-2.73724914e-01 -1.12917650e+00 -4.73557740e-01 6.19154200e-02
-9.48553443e-01 -5.22208750e-01 -3.35358679e-01 2.33939275e-01
-3.11329570e-02 1.19691348e+00 4.98474538e-02 -4.97831732e-01
1.13426253e-01 3.49325299e-01 2.80441403e-01 -6.17153049e-01
-8.11776519e-01 6.54905498e-01 5.57098128e-02 -5.99409521e-01
-6.55972958e-01 -1.69086754e-01 -9.14108813e-01 -3.76302868e-01
-2.81622261e-01 2.80387729e-01 6.34259522e-01 5.45304298e-01
7.23366678e-01 8.18140984e-01 5.30384541e-01 -1.28037751e+00
-4.04617578e-01 -1.39198720e+00 -5.25258780e-01 -4.14846987e-02
7.21116781e-01 -3.54151845e-01 -5.58265746e-01 4.18031096e-01] | [15.370015144348145, 5.428928375244141] |
349c93c6-6939-4592-85c1-c533f11bafb2 | unsupervised-one-class-learning-for-automatic | null | null | http://openaccess.thecvf.com/content_cvpr_2014/html/Liu_Unsupervised_One-Class_Learning_2014_CVPR_paper.html | http://openaccess.thecvf.com/content_cvpr_2014/papers/Liu_Unsupervised_One-Class_Learning_2014_CVPR_paper.pdf | Unsupervised One-Class Learning for Automatic Outlier Removal | Outliers are pervasive in many computer vision and pattern recognition problems. Automatically eliminating outliers scattering among practical data collections becomes increasingly important, especially for Internet inspired vision applications. In this paper, we propose a novel one-class learning approach which is robust to contamination of input training data and able to discover the outliers that corrupt one class of data source. Our approach works under a fully unsupervised manner, differing from traditional one-class learning supervised by known positive labels. By design, our approach optimizes a kernel-based max-margin objective which jointly learns a large margin one-class classifier and a soft label assignment for inliers and outliers. An alternating optimization algorithm is then designed to iteratively refine the classifier and the labeling, achieving a provably convergent solution in only a few iterations. Extensive experiments conducted on four image datasets in the presence of artificial and real-world outliers demonstrate that the proposed approach is considerably superior to the state-of-the-arts in obliterating outliers from contaminated one class of images, exhibiting strong robustness at a high outlier proportion up to 60%. | ['Wei Liu', 'John R. Smith', 'Gang Hua'] | 2014-06-01 | null | null | null | cvpr-2014-6 | ['one-class-classifier'] | ['methodology'] | [ 3.01253170e-01 3.49193290e-02 8.13956261e-02 -4.09452230e-01
-9.86485064e-01 -2.68830836e-01 5.08358777e-01 5.36659777e-01
-6.19600952e-01 5.34080982e-01 -3.85135919e-01 -5.98983392e-02
-1.74386173e-01 -2.75337011e-01 -8.97904098e-01 -8.38239670e-01
-5.62786385e-02 4.11606640e-01 3.04811001e-01 3.97459000e-01
4.12120044e-01 5.88364661e-01 -1.57198942e+00 -8.81495401e-02
1.18317556e+00 9.69441533e-01 -2.48711228e-01 4.69501913e-01
2.44096979e-01 6.64824069e-01 -6.26175642e-01 -1.31096244e-01
5.61195254e-01 -2.14544564e-01 -2.71551251e-01 7.99201429e-01
7.25896239e-01 4.45852466e-02 1.12747110e-01 1.23083138e+00
2.86572576e-01 4.63775620e-02 7.86569834e-01 -1.26280534e+00
-3.87330174e-01 6.03452250e-02 -7.65359700e-01 1.31202102e-01
1.17621675e-01 1.80209458e-01 6.12535000e-01 -1.21524072e+00
3.58696729e-01 7.75796592e-01 6.73072875e-01 7.58950114e-02
-1.36295247e+00 -3.90895158e-01 2.23351300e-01 1.18175996e-02
-1.44800603e+00 -3.28263402e-01 8.47182870e-01 -6.58771276e-01
4.47737575e-01 2.49209166e-01 3.30216080e-01 7.35177636e-01
9.44692716e-02 7.12467968e-01 1.05636978e+00 -4.65923518e-01
6.27229929e-01 2.18787998e-01 1.59886241e-01 5.09079874e-01
7.14220107e-01 8.92635062e-02 -6.96483552e-01 -5.36508262e-01
3.73292387e-01 2.43406892e-01 -3.61065596e-01 -7.03950107e-01
-1.20992064e+00 6.34885728e-01 3.51279169e-01 1.09996703e-02
-4.02072281e-01 -6.34334758e-02 2.52132446e-01 4.35582429e-01
6.77913845e-01 2.80272245e-01 -4.24690694e-01 2.50660151e-01
-1.02042437e+00 5.88320717e-02 7.77361512e-01 8.33709717e-01
9.51244414e-01 2.31346384e-01 2.22511232e-01 8.50745976e-01
4.60681140e-01 6.57488167e-01 3.44726145e-01 -5.43579996e-01
2.20262095e-01 6.36604786e-01 2.18007401e-01 -9.60014284e-01
-3.47017616e-01 -6.18540764e-01 -8.33546340e-01 5.48447490e-01
6.20520890e-01 8.14670473e-02 -1.00269198e+00 1.16918921e+00
6.66468799e-01 5.31079054e-01 -6.06797598e-02 8.50506186e-01
3.52439106e-01 3.73987943e-01 -3.66163641e-01 -5.53196430e-01
6.76586390e-01 -5.90681791e-01 -5.33623636e-01 -2.27072567e-01
4.61256742e-01 -1.03133500e+00 8.49477530e-01 9.66708004e-01
-5.11125982e-01 -3.61247361e-01 -1.08832145e+00 4.64212507e-01
-1.73728108e-01 2.22661257e-01 2.10129157e-01 6.54018402e-01
-4.62643087e-01 5.61801672e-01 -7.71929383e-01 -2.87779808e-01
5.14627934e-01 5.26816726e-01 -5.05955040e-01 -1.65599272e-01
-1.77424803e-01 5.72266042e-01 4.65898305e-01 2.63493419e-01
-7.52534926e-01 -6.30866408e-01 -8.33955288e-01 -4.24093068e-01
6.74362123e-01 -2.42085829e-01 8.99853230e-01 -1.15534186e+00
-1.33509016e+00 9.93935227e-01 -1.50747493e-01 -5.24458170e-01
8.21838796e-01 -5.20193040e-01 -3.71786743e-01 -2.10055299e-02
1.90259919e-01 7.56664202e-02 1.65176380e+00 -1.60595775e+00
-6.04618371e-01 -5.28693378e-01 -8.26847792e-01 -1.76852375e-01
-2.75236517e-01 -4.67261337e-02 -2.27957413e-01 -7.27735519e-01
6.18291974e-01 -9.63368833e-01 -3.97427320e-01 8.56782794e-02
-5.55320382e-01 2.23625675e-02 1.02098548e+00 -2.33546808e-01
7.62143016e-01 -2.23432565e+00 -1.66531205e-01 7.04997063e-01
2.55994797e-01 3.18702668e-01 6.02060631e-02 2.63933331e-01
-1.27849892e-01 -3.41580957e-01 -6.47857606e-01 -4.98948067e-01
-1.59079656e-01 2.12908670e-01 -5.23028672e-01 1.35951042e+00
3.31264973e-01 3.06397229e-01 -9.73027825e-01 -1.20143592e-01
4.91230637e-01 2.66836822e-01 -3.73389274e-01 3.91466111e-01
3.80273871e-02 6.44351602e-01 -1.32876456e-01 9.50867116e-01
7.93505907e-01 -2.58602723e-02 -3.09200883e-01 2.43202925e-01
-4.66013439e-02 -5.77548444e-01 -1.80826688e+00 1.21245861e+00
-1.04160821e-02 5.55466771e-01 4.35508713e-02 -1.06601536e+00
1.19169652e+00 1.19468644e-01 6.19759381e-01 -3.24597567e-01
2.10166335e-01 7.10508764e-01 -2.26248667e-01 -4.57278699e-01
3.01088363e-01 1.01939723e-01 6.21557683e-02 1.98483750e-01
9.45873633e-02 -3.78739201e-02 1.17477775e-01 -3.96181233e-02
1.12370801e+00 -1.08785816e-01 4.26548451e-01 -2.92853326e-01
7.27477372e-01 -2.39715680e-01 9.13098693e-01 9.67995644e-01
-3.25491548e-01 9.02010381e-01 2.55121797e-01 -6.00266397e-01
-9.04363751e-01 -1.16229331e+00 -3.08050811e-01 5.32606483e-01
2.21886039e-01 -1.01146780e-01 -5.60907245e-01 -7.38511622e-01
2.65365452e-01 3.71857584e-01 -2.79910147e-01 -1.47792965e-01
-5.16414762e-01 -8.98432553e-01 1.03290379e-01 6.91078529e-02
1.16248675e-01 -8.94099295e-01 -5.29543042e-01 1.42489091e-01
2.50275105e-01 -1.17623067e+00 -3.06232184e-01 2.81629741e-01
-1.04730475e+00 -1.47183013e+00 -3.72841328e-01 -6.87636673e-01
1.25430882e+00 2.01383650e-01 8.64385784e-01 1.94061220e-01
-5.79630554e-01 3.61577511e-01 -3.51503193e-01 -6.06247842e-01
-2.60266393e-01 -2.56582588e-01 4.01986539e-01 7.83292294e-01
4.70642239e-01 -4.39488083e-01 -2.72335500e-01 2.85861880e-01
-1.13381648e+00 -6.39706910e-01 6.04899824e-01 8.76712263e-01
7.13352382e-01 1.60407573e-01 5.19828677e-01 -1.01618636e+00
3.17713886e-01 -6.66651905e-01 -1.09514606e+00 -1.47552013e-01
-6.56770945e-01 -3.03177349e-02 8.98775399e-01 -5.08065879e-01
-6.22325540e-01 5.59547544e-01 3.64825934e-01 -5.17571509e-01
-5.51343501e-01 1.32574201e-01 -1.23367228e-01 -4.51727360e-01
8.41630280e-01 1.69533610e-01 -2.16920897e-01 -5.15806437e-01
2.13877082e-01 6.23470128e-01 8.78133655e-01 -5.30679941e-01
1.34439158e+00 8.27744126e-01 1.05712108e-01 -1.27920294e+00
-7.61971712e-01 -1.09919643e+00 -8.63291204e-01 -3.60745370e-01
3.20860535e-01 -8.29039276e-01 -4.67001766e-01 8.58504772e-01
-8.85511935e-01 1.11096226e-01 -4.47517425e-01 6.12405419e-01
-6.36346161e-01 5.34022748e-01 -1.80251494e-01 -9.48580921e-01
-3.75856124e-02 -1.00217938e+00 1.04997611e+00 2.09836558e-01
-2.33940035e-02 -7.23240256e-01 1.39967754e-01 2.43167505e-01
-7.27340113e-03 6.25975728e-01 3.86922002e-01 -1.05259788e+00
-7.73722351e-01 -6.78443491e-01 -9.27882269e-02 5.69051921e-01
1.24789961e-01 2.03331962e-01 -9.93289709e-01 -6.42531931e-01
1.58514798e-01 -2.43188053e-01 8.44664156e-01 2.55646706e-01
9.42447424e-01 -2.45911002e-01 -5.47864772e-02 7.09954619e-01
1.66394329e+00 -1.78862140e-01 3.66634697e-01 4.75144923e-01
6.47092104e-01 4.15456772e-01 7.72889197e-01 7.04855978e-01
-1.19535774e-01 2.42425606e-01 8.68918598e-01 -2.69755363e-01
3.26584220e-01 1.18048519e-01 2.52156913e-01 5.44411242e-01
3.17214102e-01 1.68494936e-02 -9.52268004e-01 8.71778905e-01
-2.09346747e+00 -6.57043934e-01 -5.17992795e-01 2.81435728e+00
6.48055255e-01 3.31845760e-01 -4.42660041e-02 4.90478098e-01
6.78934276e-01 -1.28229475e-02 -5.86295009e-01 2.60563083e-02
-1.40971988e-01 2.46904865e-02 6.67477608e-01 5.28020382e-01
-1.43235576e+00 6.54892802e-01 5.71991825e+00 4.18342769e-01
-1.03065920e+00 -2.07808316e-01 3.06928903e-01 -1.80614442e-02
1.81879580e-01 1.53944924e-01 -4.76803899e-01 5.34877419e-01
5.91875672e-01 1.91095233e-01 9.15739760e-02 9.65175211e-01
2.41240963e-01 -3.91733050e-01 -1.02652192e+00 1.12113929e+00
4.39427406e-01 -9.66792047e-01 -1.56754106e-01 1.46505877e-01
1.02955067e+00 1.40187502e-01 -2.77969707e-02 -3.56820077e-01
1.32101089e-01 -7.36969113e-01 7.97567606e-01 6.75211370e-01
2.20057786e-01 -8.36073458e-01 7.51753390e-01 5.15910804e-01
-6.63101256e-01 -2.57736087e-01 -3.15731734e-01 -5.30700246e-03
-7.19611645e-02 1.26717699e+00 -6.89320266e-01 4.47696060e-01
8.35287213e-01 7.84559906e-01 -6.27677441e-01 1.66755390e+00
-2.95618862e-01 6.93814039e-01 -5.74200451e-01 3.72298270e-01
8.88294950e-02 -4.86673057e-01 8.65877211e-01 1.01499164e+00
1.67854458e-01 -1.14113778e-01 6.02339208e-01 4.44127113e-01
-1.66750640e-01 2.74331838e-01 -7.56043077e-01 4.62146342e-01
3.35085690e-01 1.08942318e+00 -9.89240348e-01 -1.67578682e-01
-3.62908483e-01 1.08185410e+00 3.39079648e-01 5.09153128e-01
-5.37516296e-01 -4.59347606e-01 4.74688798e-01 1.89583022e-02
4.95346874e-01 -3.66050601e-01 -4.60299671e-01 -1.28411126e+00
6.18054509e-01 -7.98384130e-01 2.98768193e-01 -2.22484812e-01
-1.50191844e+00 3.90229613e-01 -4.04996902e-01 -1.69008255e+00
6.97603524e-02 -6.82857931e-01 -6.41012192e-01 3.40335518e-01
-1.72856617e+00 -9.20972824e-01 -4.60077941e-01 5.44296861e-01
3.85194540e-01 -2.78557718e-01 4.26448286e-01 2.21872061e-01
-6.17308557e-01 4.07206833e-01 6.38370156e-01 9.32003632e-02
1.17834675e+00 -1.30725050e+00 -1.58152714e-01 1.47441137e+00
4.80234623e-01 3.81139934e-01 9.55700815e-01 -6.35823727e-01
-1.15173650e+00 -1.25280142e+00 7.57139266e-01 -3.40408653e-01
5.86008847e-01 -2.19790369e-01 -1.13386595e+00 6.66747451e-01
-1.43777356e-01 6.24838352e-01 7.01932669e-01 -2.09180862e-01
-4.30286705e-01 -4.27180707e-01 -1.20586216e+00 2.82793999e-01
5.56037724e-01 -1.82049155e-01 -5.41771650e-01 5.86433649e-01
2.23971680e-01 -4.09181476e-01 -7.05208421e-01 4.81630981e-01
-3.25292423e-02 -1.07070208e+00 7.94989288e-01 -3.38197112e-01
-7.76460171e-02 -9.42693591e-01 -1.18752465e-01 -1.21159673e+00
1.75808012e-01 -9.57240403e-01 -2.96248257e-01 1.06421518e+00
1.64935619e-01 -8.91625285e-01 7.00839102e-01 2.99317360e-01
-1.71810985e-01 -4.69476849e-01 -1.00897264e+00 -1.11165369e+00
-3.78811568e-01 -5.35808206e-01 5.93488254e-02 8.63116324e-01
-4.60195929e-01 -1.66738942e-01 -5.33390462e-01 6.79259837e-01
1.43826866e+00 -5.90472780e-02 1.20776796e+00 -1.51918423e+00
-1.02064712e-02 -6.48410618e-02 -7.79951274e-01 -6.42139673e-01
1.26402363e-01 -6.07528567e-01 3.47810626e-01 -8.92993927e-01
-1.80904299e-01 -4.94576305e-01 -2.69690901e-01 3.34876806e-01
-3.29590261e-01 5.07370353e-01 1.92707200e-02 4.29394096e-01
-6.57514274e-01 6.11280739e-01 4.28677082e-01 -6.01754934e-02
-2.46636942e-01 3.29392850e-01 -3.06209892e-01 1.12697542e+00
6.26669228e-01 -9.87360835e-01 -1.47992328e-01 -2.90735632e-01
1.57562196e-02 -6.25291526e-01 5.40887177e-01 -1.26684940e+00
4.95461255e-01 -2.19771713e-01 4.16652411e-01 -6.11644089e-01
3.17060649e-02 -1.21176469e+00 -1.81037381e-01 3.59219790e-01
-3.26844230e-02 -1.95290282e-01 -1.02318287e-01 9.92858052e-01
-3.32829326e-01 -2.62520015e-01 1.04556632e+00 1.29682735e-01
-5.36937296e-01 3.03199798e-01 -2.95892656e-01 1.69555679e-01
1.34930730e+00 -3.91393661e-01 -6.22028001e-02 -8.58799443e-02
-5.54861188e-01 1.46157995e-01 7.28479624e-01 3.25859517e-01
7.41284013e-01 -1.12370455e+00 -8.62620234e-01 5.82686782e-01
5.40282607e-01 2.13696748e-01 -2.41904959e-01 8.34279478e-01
-4.73679274e-01 -8.93015880e-03 2.56185979e-01 -1.08919334e+00
-1.21612489e+00 4.62729305e-01 1.32152215e-01 1.65348332e-02
-6.21316969e-01 5.72965980e-01 -4.33885157e-01 -5.49954593e-01
4.77970779e-01 -3.04694504e-01 3.31902653e-01 -9.77910534e-02
5.30977249e-01 5.28985500e-01 3.94603014e-01 -8.24578702e-01
-3.38785321e-01 6.73362255e-01 -6.30616546e-02 2.34940425e-01
1.30318975e+00 6.57128319e-02 -2.73443460e-01 6.22063339e-01
1.17897356e+00 2.49704301e-01 -1.46089578e+00 -6.92100942e-01
6.02201104e-01 -9.15257633e-01 -2.35028222e-01 -3.17959726e-01
-9.11101580e-01 3.04261744e-01 4.81027991e-01 1.34552628e-01
1.19391716e+00 -8.72317627e-02 6.02874458e-01 5.31693041e-01
3.44126403e-01 -1.23082471e+00 2.76477665e-01 3.04749399e-01
5.85357010e-01 -1.76772940e+00 3.01398426e-01 -4.51167107e-01
-3.28305513e-01 1.09294450e+00 5.26479840e-01 -7.31271148e-01
6.71127975e-01 1.51988909e-01 3.94267768e-01 -4.48320061e-02
-3.72178406e-01 -2.05104575e-01 4.02990907e-01 5.77260196e-01
-5.10218292e-02 -1.23726860e-01 -7.43378848e-02 -9.36879367e-02
1.04539633e-01 -2.43138924e-01 7.24088430e-01 1.09726381e+00
-7.30942488e-01 -9.07021105e-01 -9.18790102e-01 5.22646964e-01
-4.35574114e-01 2.67949313e-01 -3.25039029e-01 6.14871025e-01
1.41373530e-01 1.08279324e+00 -1.25750631e-01 7.52811953e-02
4.30673629e-01 -9.17495340e-02 6.80397497e-03 -5.98930240e-01
-4.35778856e-01 1.63680688e-01 -4.55193520e-01 -8.04657102e-01
-5.24263620e-01 -9.02607024e-01 -1.15049422e+00 1.48316652e-01
-6.82173073e-01 7.61720017e-02 5.76900721e-01 1.08596039e+00
8.77992660e-02 1.83750883e-01 1.01890063e+00 -8.12469482e-01
-7.16196120e-01 -5.39452732e-01 -6.21298730e-01 5.64777434e-01
8.46110761e-01 -4.25575495e-01 -7.79664278e-01 3.25022936e-01] | [7.687137603759766, 2.3307273387908936] |
e59700cf-336a-4fd5-930a-55d7b0f45952 | impact-of-business-analytics-and-decision | 2212.00016 | null | https://arxiv.org/abs/2212.00016v1 | https://arxiv.org/pdf/2212.00016v1.pdf | Impact of Business Analytics and Decision Support Systems on e-commerce in SMEs | With the advancement in the marketing channel, the use of e-commerce has increased tremendously therefore the basic objective of this study is to analyze the impact of business analytics and decision support systems on e-commerce in small and medium enterprises. Small and medium enterprises are becoming a priority for economies as by implementing some policies and regulations these businesses could encourage gain development on an international level. The objective of this study is to analyze the impact of business analytics and decision support systems on e-commerce in small and medium enterprises that investigate the relationship between business analytics and decision support systems in e-commerce businesses. To evaluate the impact of both on e-commerce the, descriptive analysis approach is adopted that reviews the research of different scholars who adopted different plans and strategies to predict the relationship between e-commerce and business analytics. The study contributes to the literature by examining the impact of business analytics in SMEs and provides a comprehensive understanding of its relationship with the decision support system. After analyzing the impact of business analytics and decision support system in SMEs, the research also highlights some limitations and provide future recommendations that are helpful to overcome these limitations. | ['Shah J Miah'] | 2022-11-30 | null | null | null | null | ['marketing'] | ['miscellaneous'] | [-9.23234046e-01 5.41267283e-02 -5.04357815e-01 -1.72544405e-01
4.26928014e-01 -5.51758647e-01 5.78141689e-01 5.19559622e-01
-3.18064272e-01 -1.71198100e-01 1.17483236e-01 -1.00441122e+00
-4.60520923e-01 -1.06322348e+00 -3.96153241e-01 -3.79919708e-01
1.43916160e-01 1.43540218e-01 -4.68254872e-02 -5.64182937e-01
1.14587057e+00 4.09108877e-01 -1.12505198e+00 3.36638093e-01
4.28332239e-01 1.14216578e+00 3.42990279e-01 2.30804950e-01
-3.14470083e-01 1.04880524e+00 -3.58541578e-01 -6.61778569e-01
6.49717569e-01 -1.90188348e-01 -2.45924950e-01 1.21027991e-01
-3.90965551e-01 -7.13124752e-01 2.63167948e-01 8.31901252e-01
1.04358397e-01 1.25655219e-01 5.73132575e-01 -1.23695350e+00
-8.76566410e-01 4.67392653e-01 -2.83288240e-01 7.34923899e-01
-9.08880159e-02 -4.81084101e-02 5.27548552e-01 -1.00628769e+00
6.79682016e-01 6.98958158e-01 6.38419569e-01 -3.85409981e-01
-7.22963452e-01 -9.86040711e-01 -1.37395948e-01 9.96514559e-02
-6.33240938e-01 -1.78777099e-01 3.50661218e-01 -5.22897899e-01
1.69618678e+00 -2.61850238e-01 1.06169844e+00 1.07029341e-02
6.21723473e-01 -1.18822262e-01 1.31069589e+00 -9.25412118e-01
2.53999740e-01 1.19856548e+00 7.38929152e-01 -2.11517781e-01
9.81749952e-01 -4.39408310e-02 -3.26244533e-02 2.03999981e-01
9.17013645e-01 7.41364717e-01 8.03600609e-01 2.07283601e-01
-9.02804136e-01 1.18789709e+00 3.42236638e-01 7.41422772e-01
-9.94969487e-01 -5.38209140e-01 3.98154140e-01 8.79835308e-01
1.78396419e-01 3.14112186e-01 -4.69002694e-01 -7.07676232e-01
-2.28299722e-01 -2.03881994e-01 1.32909071e+00 1.01753438e+00
8.68832096e-02 1.33272424e-01 6.24702334e-01 2.31876180e-01
7.50203788e-01 2.24897712e-01 3.91343266e-01 -6.61152780e-01
4.82360989e-01 9.20279741e-01 1.58222660e-01 -1.26205540e+00
1.08458428e-02 -3.11262399e-01 -1.90475434e-01 4.89069670e-02
3.96742940e-01 -2.42761254e-01 -3.57388288e-01 2.75477201e-01
-4.79662679e-02 -7.44899333e-01 8.14682841e-02 5.70480287e-01
7.72707939e-01 6.79778636e-01 3.96910608e-01 -2.24975497e-01
9.29367483e-01 -1.19033611e+00 -1.07426250e+00 3.87589894e-02
4.30322051e-01 -7.87989736e-01 8.92378867e-01 5.93751669e-01
-9.42519784e-01 -4.83586818e-01 -8.11659455e-01 3.57822776e-01
-8.50840628e-01 -3.31599772e-01 9.30797637e-01 8.59218419e-01
-6.82767391e-01 2.28235811e-01 -8.87951374e-01 -6.03892028e-01
-2.90709585e-02 4.81557697e-01 -1.33326888e-01 2.91654974e-01
-9.34918046e-01 1.31834102e+00 4.21043843e-01 -2.28113428e-01
1.44166037e-01 -2.10962549e-01 -4.13602263e-01 2.19256043e-01
-1.36541743e-02 -1.14100128e-01 9.56075907e-01 -1.26948392e+00
-8.95923436e-01 4.79670405e-01 3.91005307e-01 -6.32599175e-01
1.88482240e-01 1.85516223e-01 -6.30321681e-01 4.07593101e-02
7.33503178e-02 -1.68974802e-01 -7.94318393e-02 -1.10089684e+00
-1.55302966e+00 -9.36994910e-01 -1.71305332e-02 -2.43651688e-01
-6.72855854e-01 6.54304802e-01 1.02895290e-01 -3.82458746e-01
5.35699129e-01 -4.36041236e-01 -2.74462938e-01 -1.10231137e+00
6.39115572e-01 6.37198538e-02 7.18648672e-01 -1.08057046e+00
1.54022694e+00 -2.06234074e+00 -8.28992605e-01 5.26949286e-01
2.58150458e-01 -1.51927844e-02 7.34363973e-01 1.13446486e+00
3.75802487e-01 7.68727124e-01 1.00692058e+00 8.57035697e-01
1.51946351e-01 1.80718243e-01 -2.23044693e-01 4.14171517e-02
-9.18029398e-02 4.69224542e-01 -2.63671905e-01 -3.77142839e-02
3.35673392e-01 4.26078051e-01 -2.80487597e-01 -1.94382370e-01
4.19724524e-01 2.05828547e-02 -5.39223969e-01 1.11261380e+00
6.29891872e-01 -1.94110408e-01 3.92270386e-01 4.53121722e-01
-8.03548992e-01 5.16954064e-01 -1.04541576e+00 8.10385197e-02
-3.30113024e-01 5.53151786e-01 4.84146059e-01 -1.19989371e+00
1.31398511e+00 4.83083695e-01 2.87828475e-01 -1.15849328e+00
3.61134768e-01 6.77562475e-01 6.58716559e-01 -6.19207859e-01
4.39742804e-01 -3.00880671e-01 4.22873735e-01 2.95862138e-01
-2.14710191e-01 3.14641833e-01 2.68612206e-01 -1.98224470e-01
6.71121299e-01 -1.77034825e-01 8.20951104e-01 -1.25694707e-01
-3.94465886e-02 3.89638335e-01 4.99369532e-01 3.53926092e-01
-3.54999423e-01 -6.55012310e-01 5.34859538e-01 -6.17296994e-01
-9.51149046e-01 -2.36325622e-01 -2.25266740e-01 1.25160670e+00
7.72058070e-02 -2.20679104e-01 -3.83416384e-01 -5.07142663e-01
3.85349095e-01 1.02186561e+00 -2.30490908e-01 4.54922020e-01
-8.12506750e-02 -4.95211035e-01 -4.51510876e-01 6.05509639e-01
8.91933262e-01 -8.22807968e-01 -8.66765618e-01 2.99520165e-01
5.68450391e-01 -8.87017071e-01 5.18994033e-01 4.46300417e-01
-1.63751495e+00 -8.16987634e-01 6.22515157e-02 -8.61243904e-01
6.46097243e-01 7.28691220e-01 9.41743851e-01 1.04691014e-01
3.72562796e-01 1.62122786e-01 -7.49233007e-01 -1.39111626e+00
-7.11545944e-01 -2.55533099e-01 -1.15000747e-01 -5.87124288e-01
1.45032883e+00 -3.40193957e-01 -3.04558426e-01 5.17347038e-01
-5.27616739e-01 -1.92066118e-01 8.18679869e-01 4.35496986e-01
4.70955431e-01 8.84933531e-01 1.29103959e+00 -8.01883936e-01
9.16709483e-01 -1.31149447e+00 -4.40323621e-01 -1.40298769e-01
-1.62370956e+00 -7.95693994e-01 3.41005772e-01 -1.21260561e-01
-1.12924135e+00 -8.63474548e-01 1.43668830e-01 2.55722314e-01
-2.29842409e-01 1.00351679e+00 1.43324614e-01 -1.12283900e-02
2.82966822e-01 -2.77553827e-01 5.92810690e-01 -3.53298545e-01
-3.78794044e-01 1.21598876e+00 -3.50479335e-01 3.33691984e-01
1.97453156e-01 6.84570298e-02 -2.96593487e-01 -6.59412861e-01
2.57894874e-01 -9.16286469e-01 -4.73531872e-01 -5.69941163e-01
6.46632254e-01 -7.74631917e-01 -8.32575858e-01 -1.70458525e-01
-4.66341734e-01 -5.13713807e-02 7.51476213e-02 1.23022294e+00
-1.46996588e-01 -4.59082097e-01 -8.01956236e-01 -9.14306939e-01
-2.08352968e-01 -7.93609500e-01 -2.89511681e-01 3.36953819e-01
-3.84935319e-01 -1.19118154e+00 -6.42518997e-01 1.18994689e+00
8.32829952e-01 3.43820333e-01 9.93495226e-01 -1.52735293e+00
-6.39762163e-01 -6.51836038e-01 -6.42822757e-02 6.77940428e-01
1.85166925e-01 1.97929755e-01 8.19518268e-02 -4.31397967e-02
7.54694164e-01 2.14524314e-01 -4.07555252e-02 3.21252108e-01
-2.59574801e-02 -7.00375378e-01 -1.05516873e-01 1.21860400e-01
1.77625227e+00 1.22499478e+00 4.79215831e-01 1.80938315e+00
-1.74366295e-01 1.32424402e+00 1.73337567e+00 8.21480513e-01
3.91665518e-01 -8.10893327e-02 2.86462247e-01 4.57770854e-01
5.95955193e-01 -8.92212167e-02 2.36074522e-01 1.11031246e+00
-7.96612024e-01 6.58874512e-01 -1.24170601e+00 3.45298380e-01
-1.58731937e+00 -9.08853173e-01 -5.64246714e-01 1.83327484e+00
-1.15644792e-02 5.23420930e-01 5.84273994e-01 1.44558340e-01
5.10863900e-01 -6.76609516e-01 -1.60646111e-01 -8.74274015e-01
4.13491160e-01 -1.20938644e-01 8.47463012e-01 -9.16685089e-02
-4.72395688e-01 4.19846565e-01 5.94872379e+00 -9.62224677e-02
-1.18007803e+00 1.51825026e-01 8.59909117e-01 1.68289505e-02
-2.78364539e-01 4.57784459e-02 -1.10581064e+00 4.90946621e-01
1.18474233e+00 -4.68568414e-01 1.81455463e-01 1.47058451e+00
8.02382827e-01 -2.25466624e-01 -5.36038160e-01 4.77714509e-01
-2.76194602e-01 -1.37475848e+00 -3.92787576e-01 8.46565604e-01
8.32608938e-01 -1.65295646e-01 8.21270421e-02 4.60685223e-01
-2.36338899e-02 -6.11979663e-01 6.88800573e-01 -7.18797185e-03
-1.98971123e-01 -9.21021879e-01 1.37807083e+00 5.50093591e-01
-4.03573185e-01 -9.28560793e-01 -2.65757293e-01 -8.81862879e-01
-1.18275568e-01 -2.38332734e-01 -1.10864973e+00 -5.78329302e-02
1.07343996e+00 2.38396272e-01 -1.69120282e-01 3.49025369e-01
3.02119851e-01 8.06126714e-01 1.02463588e-01 -4.67823863e-01
4.62056696e-01 -1.04760766e+00 5.34974504e-03 1.07280052e+00
3.27226043e-01 1.96109310e-01 -5.09888351e-01 7.23384678e-01
6.37650311e-01 7.21004128e-01 -8.31809163e-01 -7.08247423e-01
8.03171635e-01 9.27909613e-01 -1.13218594e+00 -2.55445778e-01
-1.28727067e+00 -2.14976985e-02 -4.95732874e-01 1.12748854e-01
-2.68887639e-01 -6.79769456e-01 2.05737919e-01 1.13737726e+00
6.14742935e-01 -3.28489900e-01 -1.06653571e+00 -3.42663884e-01
-1.23980887e-01 -1.17748618e+00 1.13049872e-01 -1.16310306e-01
-9.53664243e-01 6.57733018e-03 -2.66760468e-01 -8.71655464e-01
-3.85738909e-01 -8.86449456e-01 -3.31783623e-01 8.49922776e-01
-9.50349391e-01 -1.00772178e+00 -2.61854809e-02 9.73285288e-02
1.88416079e-01 -8.31030309e-01 5.14844060e-01 -7.10893124e-02
-9.42257568e-02 -1.77921504e-01 8.65482628e-01 -9.15858895e-02
4.81142908e-01 -1.00931561e+00 -1.98519439e-01 6.85751379e-01
-7.08173811e-01 1.16572535e+00 3.46860677e-01 -1.12998676e+00
-1.61474943e+00 -1.20687798e-01 1.37098587e+00 -1.64042011e-01
8.81007850e-01 2.30321512e-01 -5.16182065e-01 9.57682788e-01
3.93545985e-01 -8.68123829e-01 1.17682183e+00 2.68967628e-01
3.23997051e-01 -6.41054511e-01 -1.48970973e+00 2.13431314e-01
1.40998185e-01 -2.10215095e-02 -8.48426580e-01 -2.59570599e-01
2.60958433e-01 4.36202168e-01 -1.54272163e+00 -9.13344417e-03
6.93628073e-01 -1.34988070e+00 7.30967999e-01 -6.58984125e-01
6.19101286e-01 5.25177300e-01 -2.66123414e-01 -5.87110162e-01
-5.46362698e-01 -3.84612143e-01 3.39639261e-02 1.18596029e+00
2.68822908e-01 -1.48699677e+00 7.22667873e-01 1.35808980e+00
1.83269247e-01 -7.10602462e-01 -3.39345545e-01 -4.72572148e-01
-2.27726740e-03 1.22813880e-01 6.11124575e-01 1.24340189e+00
6.27416790e-01 1.50720272e-02 5.93981147e-01 -3.46427888e-01
4.40481663e-01 1.48764431e-01 1.09059083e+00 -9.42896843e-01
-1.08210489e-01 -5.50594509e-01 -4.72370774e-01 -4.62034672e-01
-7.67222285e-01 -4.12863851e-01 -9.31454897e-01 -1.77483380e+00
2.79203430e-02 -2.02510953e-01 -3.59451950e-01 -7.32854679e-02
2.92996347e-01 -5.12225509e-01 7.99439967e-01 7.94431806e-01
8.30250233e-02 -5.10290861e-01 1.27113318e+00 6.16473019e-01
-7.22182930e-01 3.36222887e-01 -1.46833014e+00 8.15221548e-01
8.67058039e-01 -1.54616520e-01 -2.58650273e-01 2.23830014e-01
8.68794858e-01 4.23723191e-01 -2.37212598e-01 -1.76866025e-01
2.76687264e-01 -3.80567640e-01 3.96332055e-01 -4.99098510e-01
-6.29780054e-01 -1.56634963e+00 4.74282235e-01 5.57715595e-01
-2.74439007e-02 7.21815586e-01 -1.72794759e-01 -7.40894154e-02
-1.00470304e+00 -6.45898640e-01 1.34339884e-01 -1.09116919e-01
-6.97741032e-01 -4.40728813e-01 -7.25155115e-01 -5.83014250e-01
1.46274006e+00 -1.22157407e+00 -3.26182991e-01 -4.40831959e-01
-7.53505409e-01 -7.61126205e-02 5.28966844e-01 1.45313188e-01
2.95882314e-01 -9.20291901e-01 -1.64822355e-01 1.97833210e-01
-3.09574455e-01 -2.95317173e-01 -2.91050941e-01 1.14368725e+00
-1.05194855e+00 8.80865037e-01 -1.06825364e+00 2.75733888e-01
-1.18542492e+00 5.43252587e-01 -1.87793121e-01 -1.61180884e-01
-1.30112991e-01 3.32021743e-01 -1.72318965e-01 -7.46888146e-02
7.29298294e-02 -1.90406919e-01 -7.31463253e-01 2.64664084e-01
3.49114805e-01 1.29123187e+00 -1.68667406e-01 -5.26424825e-01
-2.62959488e-02 -2.46983156e-01 -2.21474648e-01 -1.68891400e-01
1.89663565e+00 -5.79906106e-01 -4.63520080e-01 6.20372236e-01
6.24087274e-01 2.49345094e-01 -5.89540720e-01 1.78790584e-01
5.53896487e-01 -1.13531685e+00 1.20072544e-01 -9.39367473e-01
-7.13413417e-01 5.10044575e-01 2.05757484e-01 9.56256390e-01
1.27110553e+00 -1.43558070e-01 1.63369015e-01 2.99195021e-01
2.40224585e-01 -1.99812901e+00 1.00359574e-01 1.80460632e-01
8.94107819e-01 -1.40089476e+00 -2.82246828e-01 -6.52598917e-01
-1.00256729e+00 1.46185684e+00 5.04980445e-01 -1.53781444e-01
1.51834834e+00 4.51304525e-01 2.01268658e-01 -2.26262480e-01
-3.70732933e-01 5.84479868e-01 -5.68774939e-01 5.52817404e-01
1.03673041e+00 4.19642568e-01 -1.17412102e+00 1.54428637e+00
-1.58792719e-01 5.86954474e-01 5.70841193e-01 1.13060582e+00
-5.60016930e-01 -1.11216116e+00 -6.62118554e-01 7.75414348e-01
-1.44719172e+00 3.13157402e-03 -2.66799599e-01 1.59932005e+00
1.91277742e-01 1.23135912e+00 3.67450416e-01 -4.97887164e-01
5.65142632e-01 -1.13251232e-01 -1.39532283e-01 -3.00810546e-01
-1.28350341e+00 3.99011850e-01 3.69682074e-01 1.33065730e-01
-3.08257371e-01 -1.14292264e+00 -1.47173202e+00 -9.59821343e-01
-3.49621683e-01 1.30453795e-01 1.54941797e+00 4.26452100e-01
4.80797410e-01 2.86596954e-01 6.00980639e-01 1.93718165e-01
-1.02870846e+00 -1.30792177e+00 -1.20920610e+00 5.05948246e-01
-3.78748536e-01 -3.29406589e-01 -4.10514593e-01 1.53520539e-01] | [9.05168342590332, 6.154146194458008] |
a8449aff-e2ce-4eb6-8bf4-f689ce47599c | epistemic-uncertainty-weighted-loss-for | 2204.09389 | null | https://arxiv.org/abs/2204.09389v1 | https://arxiv.org/pdf/2204.09389v1.pdf | Epistemic Uncertainty-Weighted Loss for Visual Bias Mitigation | Deep neural networks are highly susceptible to learning biases in visual data. While various methods have been proposed to mitigate such bias, the majority require explicit knowledge of the biases present in the training data in order to mitigate. We argue the relevance of exploring methods which are completely ignorant of the presence of any bias, but are capable of identifying and mitigating them. Furthermore, we propose using Bayesian neural networks with an epistemic uncertainty-weighted loss function to dynamically identify potential bias in individual training samples and to weight them during training. We find a positive correlation between samples subject to bias and higher epistemic uncertainties. Finally, we show the method has potential to mitigate visual bias on a bias benchmark dataset and on a real-world face detection problem, and we consider the merits and weaknesses of our approach. | ['David C Hogg', 'Andrew J Bulpitt', 'Nishant Ravikumar', 'Rebecca S Stone'] | 2022-04-20 | null | null | null | null | ['face-detection'] | ['computer-vision'] | [ 3.42568189e-01 5.15007794e-01 -5.64427860e-02 -7.41388083e-01
-4.64129567e-01 -4.03477401e-01 9.39206481e-01 -1.99524779e-02
-7.41118371e-01 8.32062721e-01 2.01440632e-01 -2.49145269e-01
-3.55152011e-01 -8.69188070e-01 -1.07916760e+00 -7.43445218e-01
1.22770973e-01 2.64422297e-01 1.22361489e-01 1.87654495e-01
4.55561191e-01 6.30954623e-01 -1.65176201e+00 1.65963084e-01
6.88163877e-01 1.08304644e+00 -2.83756793e-01 2.04465121e-01
2.86453307e-01 9.14555609e-01 -7.55612671e-01 -6.46429658e-01
3.65908742e-01 -1.02044769e-01 -5.60010135e-01 -1.22647792e-01
1.01365554e+00 -5.11665761e-01 -2.98738778e-02 1.39445651e+00
7.63070405e-01 -1.81922391e-02 9.45972323e-01 -1.21505380e+00
-5.81097841e-01 6.29296243e-01 -5.65260410e-01 5.99428117e-01
-2.62490779e-01 3.77175629e-01 7.16169119e-01 -7.17553139e-01
5.04139185e-01 1.54088044e+00 6.92407131e-01 8.23656857e-01
-1.35969210e+00 -8.33313763e-01 4.25988883e-01 8.77811015e-02
-9.06190217e-01 -8.33712637e-01 8.60957742e-01 -7.85950601e-01
6.14133120e-01 -1.46007150e-01 4.86773491e-01 1.49900913e+00
1.13617413e-01 6.13641560e-01 1.46766460e+00 -5.98742604e-01
4.34222728e-01 3.02127570e-01 3.83202076e-01 5.96326053e-01
5.76869249e-01 8.13385189e-01 -7.59151220e-01 -3.18449378e-01
4.18237567e-01 -5.09575546e-01 -6.00207001e-02 -4.42534238e-01
-7.05652416e-01 8.24546337e-01 5.97807646e-01 -1.48320228e-01
-4.78194118e-01 5.51618040e-01 2.65099794e-01 -3.65605764e-02
6.53870761e-01 5.22951841e-01 -2.78736353e-01 4.04338777e-01
-9.31822002e-01 3.85205388e-01 3.75909775e-01 5.00182807e-01
6.12750113e-01 1.83028564e-01 -4.28345501e-01 6.71133399e-01
5.78056991e-01 4.86730158e-01 1.61498293e-01 -1.20645237e+00
6.34978861e-02 1.87018469e-01 2.61051178e-01 -1.13440371e+00
-1.62579745e-01 -3.18227559e-01 -3.86895329e-01 8.72525156e-01
6.75581992e-01 -3.80727589e-01 -1.19369352e+00 1.97335434e+00
3.44981700e-01 6.57518627e-03 -1.85495093e-01 8.23725462e-01
5.33570707e-01 1.18475422e-01 3.59068990e-01 7.97193497e-02
1.14278162e+00 -1.50642172e-01 -4.42591757e-01 -5.51647902e-01
3.53157744e-02 -3.54894817e-01 7.90777683e-01 5.52570462e-01
-1.10831153e+00 -1.96994305e-01 -1.18216789e+00 1.96248457e-01
-1.99802369e-01 -8.53340924e-02 7.03139842e-01 1.09320593e+00
-7.76518047e-01 7.25712180e-01 -5.24409652e-01 2.23248564e-02
9.93756533e-01 3.82816136e-01 1.47710312e-02 1.67624116e-01
-1.30398488e+00 1.20973206e+00 3.23177993e-01 3.40462297e-01
-1.40380025e+00 -8.15436721e-01 -5.98880112e-01 -5.73591776e-02
4.64752108e-01 -4.82712269e-01 1.20361066e+00 -1.52970088e+00
-1.03002691e+00 9.20771420e-01 1.57317054e-02 -5.96411288e-01
9.63116229e-01 -1.77163988e-01 -3.12900603e-01 -1.17320701e-01
-2.91306376e-01 9.60661948e-01 1.30810177e+00 -1.56361604e+00
-6.73812509e-01 -3.76003921e-01 3.44636351e-01 -1.62629023e-01
-4.01260376e-01 5.31788357e-02 2.34392956e-01 -5.37580371e-01
-2.14884460e-01 -8.68248880e-01 -4.34452519e-02 2.37393439e-01
-2.82467693e-01 -2.13444844e-01 5.41234553e-01 -2.65463442e-01
7.28371322e-01 -1.97291005e+00 -3.04734230e-01 4.77150798e-01
2.07480907e-01 3.34385931e-01 -1.36518002e-01 -3.22409034e-01
-4.89368550e-02 3.44882578e-01 -8.01992044e-02 1.45328194e-02
9.48400795e-02 1.64857656e-01 -5.28284073e-01 7.13007092e-01
6.05609000e-01 5.28865159e-01 -8.43377709e-01 -3.75148088e-01
1.96446162e-02 6.48372769e-01 -5.32027721e-01 -1.64269939e-01
-3.54169756e-01 4.48993668e-02 -1.96503714e-01 4.63340431e-01
7.49566078e-01 -4.41724956e-02 1.51591122e-01 -2.01352000e-01
1.56894743e-01 3.62161219e-01 -1.16432774e+00 7.74584174e-01
-2.17020959e-01 8.10369194e-01 1.57981053e-01 -6.89827621e-01
6.48167372e-01 9.18999538e-02 -8.35509077e-02 -5.90364814e-01
4.00903612e-01 1.19959943e-01 2.67119348e-01 -2.66458839e-01
2.52869099e-01 -5.38572133e-01 4.35207605e-01 4.41012472e-01
1.47117943e-01 2.16625798e-02 -1.86789967e-02 1.45006571e-02
7.79720664e-01 1.85540114e-02 -4.96387407e-02 -4.99709457e-01
1.24944687e-01 -2.29501769e-01 6.82029068e-01 1.00424576e+00
-5.40811956e-01 3.57133329e-01 8.31166208e-01 -4.48476344e-01
-8.51706982e-01 -9.56898928e-01 -4.17135566e-01 8.59415054e-01
-1.93280831e-01 3.49935740e-01 -6.64599061e-01 -1.07815087e+00
3.85093540e-01 9.83780086e-01 -1.19059217e+00 -4.02058721e-01
-2.15725675e-01 -1.06511605e+00 4.84909385e-01 5.09526312e-01
1.86990619e-01 -1.10326123e+00 -9.79702890e-01 -1.86084673e-01
2.35693827e-01 -6.50884748e-01 5.53292036e-02 3.72039080e-01
-8.94753754e-01 -1.14333570e+00 -4.50495034e-01 -9.24528614e-02
9.14023459e-01 -1.07563429e-01 1.29872918e+00 2.37523913e-01
-2.43498504e-01 6.20339155e-01 2.24649206e-01 -1.02035582e+00
-5.39194643e-01 -4.35374767e-01 2.39263531e-02 -1.71730831e-01
5.96927166e-01 -1.94736466e-01 -6.72853231e-01 2.17595786e-01
-7.93057203e-01 -3.93819124e-01 4.72451091e-01 8.77477646e-01
2.13593692e-01 -3.41219120e-02 6.92987561e-01 -1.22865427e+00
7.39193022e-01 -5.63004851e-01 -1.05166042e+00 1.27256677e-01
-1.20549452e+00 1.24594830e-01 -5.81709929e-02 -5.94866455e-01
-1.50673318e+00 -7.09772334e-02 1.65577397e-01 -1.42133549e-01
-2.37036213e-01 3.21277171e-01 -7.76382610e-02 -2.37522990e-01
1.15444112e+00 -5.27965784e-01 -1.03925746e-02 -1.61009684e-01
2.80189335e-01 3.05887341e-01 8.42791572e-02 -8.75998318e-01
5.11833131e-01 6.63157582e-01 3.29901129e-01 -3.71610820e-01
-1.12770510e+00 3.40982556e-01 -4.07353073e-01 -6.56740785e-01
4.94368970e-01 -8.26052129e-01 -6.27952576e-01 5.01305640e-01
-1.09828353e+00 -3.35017741e-01 -2.18642324e-01 3.88488084e-01
-3.47159415e-01 1.44544125e-01 -2.37390980e-01 -1.12687862e+00
3.83095518e-02 -1.01369655e+00 4.86011058e-01 1.98227882e-01
-2.56387770e-01 -9.43572342e-01 -1.46367401e-01 4.20502335e-01
3.45744073e-01 2.66070962e-01 9.68441546e-01 -5.29098988e-01
-5.25494397e-01 -6.18004203e-02 -3.54076207e-01 8.27121615e-01
-4.27359454e-02 3.75530422e-01 -1.52539468e+00 -2.13360414e-01
-2.28246804e-02 -5.46529472e-01 1.21090806e+00 7.88284063e-01
1.12381196e+00 -2.49544233e-01 -1.88827261e-01 7.70418048e-02
1.24403477e+00 2.58598998e-02 5.97176492e-01 3.51450711e-01
3.76814306e-01 1.17816329e+00 3.83123666e-01 2.36175299e-01
-1.90281179e-02 3.16934377e-01 9.18992996e-01 1.11286901e-01
-2.45764390e-01 -1.21356338e-01 3.09610426e-01 -2.31387243e-01
-5.82636707e-02 -2.09782004e-01 -9.74022806e-01 7.66860008e-01
-1.59349632e+00 -1.00309312e+00 -8.76874626e-02 2.16166949e+00
9.45594966e-01 5.06645560e-01 6.05801977e-02 8.40921849e-02
7.98502088e-01 1.34597421e-01 -7.85781384e-01 -2.57372856e-01
-1.85879692e-01 -2.05615070e-02 7.25871146e-01 5.19003153e-01
-1.05381846e+00 7.19896138e-01 7.75271797e+00 4.67018276e-01
-1.08368349e+00 -5.28555326e-02 9.70159709e-01 -4.71608996e-01
-4.86081004e-01 -1.01947403e-02 -6.79329276e-01 4.57320213e-01
7.81058609e-01 1.43255323e-01 4.09472793e-01 8.21691513e-01
-2.60008369e-02 -4.84115928e-01 -1.19742560e+00 4.13695276e-01
1.81485638e-02 -1.19644988e+00 -2.40045954e-02 7.08417520e-02
7.69082785e-01 1.18778069e-02 4.51655120e-01 -2.88407560e-02
7.92525947e-01 -1.25100553e+00 1.12741065e+00 6.29709303e-01
2.50010371e-01 -9.35510457e-01 5.93267679e-01 8.76172259e-02
8.78324732e-02 -1.60082623e-01 -4.88712698e-01 -1.35348588e-01
-3.52754533e-01 1.15731680e+00 -9.32876885e-01 -1.71227247e-01
8.47993970e-01 3.74072760e-01 -7.18264341e-01 9.21088874e-01
-5.87244809e-01 8.82561028e-01 -2.76292205e-01 -8.83689001e-02
-4.90007550e-02 1.98415622e-01 4.77368891e-01 9.66920853e-01
2.28523090e-01 -3.55781466e-01 -5.94284534e-01 1.10598958e+00
-1.25494570e-01 -3.75841320e-01 -6.42916799e-01 6.31048977e-02
5.11920929e-01 9.96095896e-01 -5.09196341e-01 -2.39184320e-01
-2.02517629e-01 2.74584860e-01 3.91125172e-01 4.45096076e-01
-5.53462148e-01 1.78736169e-02 7.21842289e-01 -7.89285228e-02
1.43068090e-01 3.16828221e-01 -5.91474295e-01 -7.29705334e-01
-1.23791136e-02 -9.83613312e-01 3.64052624e-01 -7.83918440e-01
-1.62858367e+00 1.20588027e-01 1.59526020e-01 -6.42655373e-01
-6.81545436e-02 -1.02559519e+00 -3.37059587e-01 9.80166078e-01
-1.74440682e+00 -8.29377472e-01 -2.04818808e-02 4.21506047e-01
-9.29207951e-02 -1.36820391e-01 3.54006618e-01 9.67555388e-04
-3.64485830e-01 3.98782551e-01 -2.61572003e-01 6.63948059e-02
9.70729947e-01 -1.24039984e+00 1.09475173e-01 8.36750567e-01
-3.84449288e-02 8.95408928e-01 1.07124984e+00 -6.96294844e-01
-8.63690436e-01 -8.95723581e-01 5.53789854e-01 -7.96825886e-01
4.41823363e-01 -2.13151008e-01 -8.53704035e-01 6.69932187e-01
1.65373966e-01 1.16868004e-01 6.87357962e-01 3.69885057e-01
-7.74136186e-01 -1.94420636e-01 -1.60854352e+00 6.03825152e-01
9.26711082e-01 -4.59809691e-01 -8.07174563e-01 9.99754593e-02
3.96967120e-02 -2.31497034e-01 -4.18230087e-01 5.62761724e-01
7.86068857e-01 -1.25756359e+00 9.59676623e-01 -7.68233359e-01
4.71518904e-01 -2.02022761e-01 -2.92088389e-02 -1.52773583e+00
-3.39773148e-01 4.30318993e-03 4.32747416e-02 1.28962302e+00
5.01161397e-01 -7.29859769e-01 9.19500232e-01 9.53064561e-01
3.06913525e-01 -3.32835406e-01 -1.08183241e+00 -6.43738329e-01
4.78884965e-01 -6.02266312e-01 5.82740784e-01 9.18202102e-01
-4.72864419e-01 -1.96244523e-01 -4.42017436e-01 3.47176939e-01
1.09315956e+00 -4.16767806e-01 4.07929420e-01 -1.59249198e+00
6.89104646e-02 -6.94374681e-01 -1.89963043e-01 -2.83060223e-01
3.84647846e-01 -5.72719991e-01 3.57583731e-01 -1.24001765e+00
3.36218834e-01 -3.17853421e-01 -7.10845947e-01 5.33076465e-01
-2.05148414e-01 1.34768963e-01 8.66384618e-03 -4.15849537e-02
-1.81134090e-01 3.31276476e-01 9.18897629e-01 -3.05872321e-01
4.40753490e-01 -1.29349798e-01 -9.40123916e-01 9.76935089e-01
6.78388774e-01 -9.57864761e-01 -4.25022840e-01 -4.63025242e-01
7.70508170e-01 -7.15084612e-01 9.69930053e-01 -8.04840982e-01
5.00108255e-03 -4.16009068e-01 8.21695864e-01 -2.68676370e-01
1.58047184e-01 -9.48004425e-01 -1.70123711e-01 4.90120888e-01
-6.82789683e-01 -4.70536858e-01 3.40668678e-01 7.89379776e-01
1.52125701e-01 -7.27460265e-01 1.09145939e+00 -1.04120933e-01
-5.52350461e-01 -6.38073832e-02 -5.58327973e-01 8.72848779e-02
8.71464014e-01 1.55288026e-01 -6.93207324e-01 -2.15500116e-01
-4.84743297e-01 1.38920844e-01 4.46107417e-01 2.36708805e-01
5.06052434e-01 -1.20271206e+00 -4.70846623e-01 9.16863885e-03
1.15681954e-01 -3.63630176e-01 1.01841867e-01 3.56256694e-01
8.58315965e-04 3.78790945e-02 -4.11247075e-01 -4.13556248e-01
-1.30043328e+00 5.38119614e-01 6.69232190e-01 2.97778815e-01
1.93251316e-02 1.09838462e+00 -1.91021077e-02 -2.84298509e-01
6.17898941e-01 -2.09294349e-01 -8.83184224e-02 3.81144315e-01
4.41621721e-01 5.01825571e-01 7.65569508e-02 -2.95724332e-01
-4.99220848e-01 1.40385404e-01 -3.63272466e-02 -2.33258650e-01
1.22127616e+00 1.38362318e-01 -8.49163160e-02 3.34388584e-01
5.02597153e-01 1.37512669e-01 -1.86513817e+00 -1.54136389e-01
2.63433903e-01 -6.94999337e-01 5.45496225e-01 -1.31697273e+00
-1.23977792e+00 1.10312760e+00 1.04903579e+00 2.06963364e-02
9.03353155e-01 -2.68689930e-01 -1.13610558e-01 4.42546129e-01
2.69156456e-01 -1.53304160e+00 1.57138288e-01 3.01874369e-01
8.77059340e-01 -1.49994993e+00 3.62127990e-01 -9.95767340e-02
-5.31395912e-01 9.61642385e-01 6.21588409e-01 -1.80951610e-01
7.39758134e-01 2.86105663e-01 3.59410346e-01 -3.22846413e-01
-7.82906473e-01 7.31570497e-02 4.10310179e-01 8.73461783e-01
3.91391575e-01 -1.72492743e-01 -1.72405943e-01 2.21623868e-01
2.49118030e-01 2.36184090e-01 5.88907957e-01 8.71939480e-01
-3.20482314e-01 -8.60014498e-01 -5.79058230e-01 6.13363981e-01
-6.53488457e-01 -6.78774044e-02 -6.08042657e-01 5.47526240e-01
3.95148367e-01 9.67048764e-01 -4.59746905e-02 -1.13538401e-02
2.43914068e-01 2.87947178e-01 7.55836368e-01 -3.63481700e-01
-6.99980795e-01 -3.43115419e-01 3.20577025e-01 -3.42684060e-01
-6.46183729e-01 -8.16952109e-01 -7.44826496e-01 -1.33485600e-01
-3.20895970e-01 -3.44405442e-01 7.29940295e-01 8.11876297e-01
2.57853061e-01 5.51956534e-01 1.49032712e-01 -7.03543544e-01
-9.46350276e-01 -8.69915664e-01 -4.03213322e-01 4.49409008e-01
5.97573698e-01 -1.18473065e+00 -7.52635300e-01 -1.20576859e-01] | [8.826953887939453, 4.782235145568848] |
496359b5-cfec-40ed-a8a0-06517443a1c2 | noise2noise-learning-image-restoration | 1803.04189 | null | http://arxiv.org/abs/1803.04189v3 | http://arxiv.org/pdf/1803.04189v3.pdf | Noise2Noise: Learning Image Restoration without Clean Data | We apply basic statistical reasoning to signal reconstruction by machine
learning -- learning to map corrupted observations to clean signals -- with a
simple and powerful conclusion: it is possible to learn to restore images by
only looking at corrupted examples, at performance at and sometimes exceeding
training using clean data, without explicit image priors or likelihood models
of the corruption. In practice, we show that a single model learns photographic
noise removal, denoising synthetic Monte Carlo images, and reconstruction of
undersampled MRI scans -- all corrupted by different processes -- based on
noisy data only. | ['Timo Aila', 'Tero Karras', 'Jacob Munkberg', 'Jaakko Lehtinen', 'Miika Aittala', 'Samuli Laine', 'Jon Hasselgren'] | 2018-03-12 | noise2noise-learning-image-restoration-1 | https://icml.cc/Conferences/2018/Schedule?showEvent=2014 | http://proceedings.mlr.press/v80/lehtinen18a/lehtinen18a.pdf | icml-2018-7 | ['salt-and-pepper-noise-removal'] | ['computer-vision'] | [ 8.56007099e-01 3.78123671e-01 3.91184151e-01 -4.32152390e-01
-1.43403804e+00 -1.17063880e-01 4.96553630e-01 -4.09213424e-01
-4.38568324e-01 1.06006765e+00 4.14607793e-01 -2.18294561e-01
-2.33683258e-01 -4.07216370e-01 -8.82946372e-01 -1.17182708e+00
-3.26419473e-01 2.37040326e-01 -2.86669344e-01 1.67837769e-01
2.67911586e-03 3.44580054e-01 -9.93737400e-01 4.26747292e-01
3.79748791e-01 5.73107064e-01 1.25691965e-01 1.09572935e+00
3.31296235e-01 1.24035001e+00 -6.00340724e-01 -2.44082049e-01
2.69117177e-01 -8.80405128e-01 -4.75819916e-01 3.42841893e-01
2.02788964e-01 -5.82861543e-01 -7.45300114e-01 1.08080828e+00
4.57986683e-01 -2.21074060e-01 8.95913303e-01 -6.18662655e-01
-8.10388863e-01 5.69130361e-01 -5.13985336e-01 1.65643230e-01
3.37659597e-01 4.59550500e-01 2.76919931e-01 -5.76410770e-01
5.43815911e-01 1.14406502e+00 9.56025243e-01 6.98494673e-01
-1.80403233e+00 -2.12713331e-01 -5.20095468e-01 -3.28209624e-02
-9.65169430e-01 -8.36080253e-01 5.57569802e-01 -4.93056774e-01
6.52473271e-01 3.97414565e-01 2.94599086e-01 1.47698629e+00
4.92193431e-01 6.78885520e-01 1.61822808e+00 -4.59512889e-01
2.76448846e-01 -2.90269405e-01 -1.55529737e-01 6.16369665e-01
3.61560464e-01 5.05976200e-01 -4.24459100e-01 -1.49052218e-01
9.93597388e-01 -2.76433360e-02 -6.12523198e-01 -1.54017419e-01
-1.40439630e+00 5.55885434e-01 2.60576814e-01 2.71564633e-01
-5.84204912e-01 6.42512560e-01 4.16915148e-01 6.49778545e-01
3.00793886e-01 4.55686390e-01 -3.03178132e-01 1.27880976e-01
-1.07383156e+00 7.52037810e-03 7.80227959e-01 5.50126076e-01
5.20565271e-01 5.94300270e-01 1.82069153e-01 4.53058004e-01
-4.04530913e-02 1.02637613e+00 4.16336805e-01 -1.41806734e+00
6.12806855e-03 -7.46115744e-01 2.32797712e-01 -5.59435725e-01
-4.46606457e-01 -6.43843710e-01 -1.33334136e+00 6.44465744e-01
3.54234338e-01 -4.49393719e-01 -1.05650866e+00 1.64596832e+00
-3.21036637e-01 8.31412852e-01 2.24948451e-01 9.11273539e-01
5.65142870e-01 5.76952577e-01 -1.07205205e-01 -9.21525061e-01
6.76222682e-01 -3.49934936e-01 -9.95155156e-01 -2.37987369e-01
2.27540970e-01 -5.55723727e-01 7.40880907e-01 9.58621681e-01
-1.51235640e+00 -4.96341795e-01 -1.32563305e+00 2.35171288e-01
4.38182801e-01 -2.19066724e-01 3.76309633e-01 5.68213046e-01
-1.13432038e+00 9.80083525e-01 -9.26890016e-01 -4.56658285e-03
6.14689708e-01 8.56265426e-03 -5.78764260e-01 -4.96136904e-01
-6.79443836e-01 1.18799746e+00 -8.07186365e-02 6.14773110e-02
-1.47291148e+00 -8.20588112e-01 -8.87612939e-01 -2.70782381e-01
-7.59745075e-04 -1.09070981e+00 1.25572407e+00 -1.01369071e+00
-1.27703917e+00 8.09315920e-01 -2.36348823e-01 -6.29937649e-01
6.65823519e-01 -2.66403675e-01 -4.39596355e-01 4.16884840e-01
-1.18619479e-01 1.47829920e-01 1.38428140e+00 -1.93509603e+00
5.21754567e-03 -2.94945747e-01 -4.68699485e-01 -5.66068031e-02
6.69173300e-01 -1.27401277e-01 9.54565182e-02 -7.29728699e-01
5.98212123e-01 -4.95164305e-01 -5.37744880e-01 1.29960002e-02
-4.17353630e-01 7.31885433e-01 3.26877773e-01 -1.10797203e+00
3.74698460e-01 -2.19961429e+00 1.65057659e-01 1.68858722e-01
2.80251086e-01 -6.64296895e-02 -2.69183159e-01 4.75076474e-02
-6.17553413e-01 -5.90559989e-02 -7.38415658e-01 -5.37976027e-01
-3.35701793e-01 4.57293957e-01 -6.35419190e-01 1.19453204e+00
1.63751975e-01 9.53895152e-01 -1.11498356e+00 -4.31802928e-01
5.93084395e-01 6.03291929e-01 -1.61331013e-01 8.73823613e-02
8.99088904e-02 9.54060316e-01 -2.26896703e-01 2.22162604e-01
7.01050937e-01 -1.71465963e-01 6.28285483e-02 -1.11229368e-01
4.80257779e-01 -2.96100855e-01 -1.23314345e+00 1.74406314e+00
-4.54897672e-01 7.52177835e-01 4.64166075e-01 -1.47090816e+00
3.05733651e-01 6.00392163e-01 5.02816498e-01 -6.74162447e-01
4.70127352e-03 2.46132836e-01 -1.68164596e-01 -1.04014564e+00
-4.15783167e-01 -1.00012755e+00 1.82390139e-01 4.22700554e-01
2.52668530e-01 -6.03900850e-01 -4.90765691e-01 -1.18310638e-01
1.49578798e+00 1.14374846e-01 3.81035879e-02 1.96212903e-01
9.27010775e-02 -2.10770637e-01 3.53645414e-01 1.44403541e+00
8.11869800e-02 1.13575733e+00 2.08725259e-01 -3.80344987e-01
-1.24603081e+00 -1.57244301e+00 -3.31293076e-01 2.80367583e-01
-1.46624252e-01 2.18911767e-01 -8.43720675e-01 -2.35252649e-01
-3.46555382e-01 1.02294064e+00 -5.80116272e-01 5.24806753e-02
-6.81396902e-01 -1.15185332e+00 4.09340829e-01 1.24592572e-01
7.04546738e-03 -1.16635835e+00 -4.02152687e-01 2.11747497e-01
-1.19799957e-01 -1.03308749e+00 2.46439442e-01 5.91745555e-01
-1.14974368e+00 -1.00509167e+00 -9.25155640e-01 -6.70646548e-01
7.64967382e-01 1.38278797e-01 1.09712708e+00 2.12512165e-01
-6.47686303e-01 5.29629648e-01 3.35578993e-02 -1.46172076e-01
-9.04411256e-01 -1.04727161e+00 5.97015060e-02 -6.65846393e-02
-3.61165673e-01 -1.02753818e+00 -4.48532432e-01 -6.02333955e-02
-1.01670516e+00 -1.24585934e-01 8.66306961e-01 1.01652455e+00
6.44096851e-01 2.47055694e-01 3.65318239e-01 -8.24608028e-01
3.74206752e-01 -3.23924482e-01 -1.69671014e-01 9.79355350e-02
-7.24682510e-02 1.57367617e-01 7.18236148e-01 -3.08478832e-01
-1.01481283e+00 -9.63080972e-02 -1.65842906e-01 -4.64047402e-01
-6.34492338e-01 2.86733180e-01 -4.59724516e-02 -4.12467010e-02
1.15561938e+00 6.14486694e-01 2.12033466e-01 -4.43744957e-01
6.72445238e-01 3.18872988e-01 1.34231257e+00 -4.82761294e-01
8.07411373e-01 1.01915133e+00 3.40901673e-01 -1.09410179e+00
-8.85827005e-01 -4.02235724e-02 -4.84788537e-01 -9.62502286e-02
7.94006884e-01 -1.10023797e+00 -3.22995484e-01 4.91548598e-01
-9.84522164e-01 -4.80878979e-01 -7.05570459e-01 9.43784893e-01
-1.13194311e+00 5.63358247e-01 -9.59994376e-01 -8.19109023e-01
4.07312326e-02 -1.00090981e+00 9.93572652e-01 -1.99661493e-01
-2.03275885e-02 -1.02407706e+00 1.12200566e-01 2.26656958e-01
3.22684139e-01 5.17723680e-01 9.05466914e-01 -3.83442968e-01
-4.98902172e-01 -1.81510121e-01 -9.09875184e-02 8.39357793e-01
1.10538021e-01 -5.99158764e-01 -1.29437244e+00 -3.44821334e-01
9.69044089e-01 -3.57809871e-01 9.09235477e-01 1.14667094e+00
1.41308105e+00 -2.63793081e-01 -2.08164588e-01 8.36650252e-01
1.72402287e+00 -1.49998173e-01 9.27867770e-01 -1.04317747e-01
2.65272617e-01 2.63824761e-01 -2.12497488e-01 1.99257448e-01
-4.80453551e-01 8.27572420e-02 3.24524730e-01 -2.57841170e-01
-5.85996628e-01 -1.56823378e-02 1.53498277e-01 9.10366178e-01
5.53375445e-02 7.55019560e-02 -4.72320825e-01 6.34686470e-01
-1.54435802e+00 -1.33474004e+00 -2.41232887e-01 2.13694096e+00
8.83535683e-01 3.44497472e-01 -4.21220720e-01 1.97327167e-01
5.63278615e-01 1.46253362e-01 -5.84504187e-01 1.10529639e-01
-3.29409003e-01 5.33066869e-01 6.84557438e-01 8.59970629e-01
-9.29819226e-01 1.63137555e-01 8.74343872e+00 7.05067456e-01
-8.19828868e-01 3.69071305e-01 7.51796305e-01 -2.24961713e-01
-3.78489971e-01 -1.23856496e-03 3.03038865e-01 3.60765696e-01
1.05285108e+00 2.63160050e-01 7.10188091e-01 4.56730962e-01
5.07514417e-01 -2.51326263e-01 -1.18758607e+00 1.35965109e+00
3.36010635e-01 -1.46094763e+00 -1.45123944e-01 -1.61798224e-01
9.03878689e-01 7.61367306e-02 -1.24302775e-01 -5.80692105e-02
5.25213838e-01 -1.36280537e+00 6.13601685e-01 1.07371593e+00
7.41618156e-01 -2.39902496e-01 6.42804027e-01 6.23112679e-01
-1.63274631e-03 4.67420444e-02 -3.96126956e-01 5.72933704e-02
5.06075382e-01 1.01763010e+00 -3.91644776e-01 4.12477255e-01
3.88969630e-01 5.03899574e-01 -8.08108822e-02 1.26537466e+00
-4.06275630e-01 9.47834790e-01 -1.99816540e-01 6.44836128e-01
-1.77822590e-01 -1.85796723e-01 7.70810843e-01 1.38327515e+00
4.21531737e-01 4.92834985e-01 -6.18408658e-02 5.86788297e-01
2.40643322e-01 -6.03155613e-01 -6.69066668e-01 7.11839676e-01
-6.32386804e-02 1.03829408e+00 -5.22802949e-01 -3.64803165e-01
-3.23220104e-01 9.85159755e-01 -2.73921549e-01 7.32350588e-01
-6.68348014e-01 -5.02165481e-02 2.53605157e-01 2.34562561e-01
9.07141566e-02 -1.92888215e-01 -7.44163036e-01 -1.18054175e+00
-1.38965413e-01 -9.60321188e-01 -4.40181196e-02 -1.48616421e+00
-1.65001500e+00 2.81457305e-01 1.95599958e-01 -8.61010790e-01
-3.98604929e-01 -4.93630022e-01 -7.06255913e-01 1.01736391e+00
-1.28016436e+00 -8.37812006e-01 6.38214573e-02 5.65641403e-01
4.27865654e-01 -1.20979384e-01 7.31515110e-01 -4.92406450e-02
9.39396620e-02 -9.35524516e-03 5.17426550e-01 6.36759996e-02
4.21657026e-01 -1.32667398e+00 3.05797517e-01 8.67276013e-01
2.41927519e-01 4.42258507e-01 1.42401528e+00 -5.16402543e-01
-1.42359924e+00 -6.67320609e-01 3.05728465e-01 -3.72000873e-01
6.10999703e-01 1.29784690e-02 -1.13832939e+00 7.65878320e-01
5.33539295e-01 2.77132571e-01 4.69798803e-01 -3.18053335e-01
-6.06982820e-02 -5.94540499e-03 -1.35481703e+00 4.26498443e-01
8.75118971e-01 -6.79342091e-01 -1.00498295e+00 7.55322754e-01
3.91220689e-01 -3.16423744e-01 -5.39993107e-01 2.07779974e-01
2.52604961e-01 -8.07137966e-01 1.23225629e+00 -6.65855408e-01
3.59369516e-01 -2.30937928e-01 -3.09585094e-01 -1.66448689e+00
-2.23158166e-01 -9.15543437e-01 3.30780111e-02 4.46949720e-01
2.48579174e-01 -3.20502311e-01 5.30240893e-01 3.51201266e-01
-2.52394676e-01 -2.04366222e-01 -1.19290590e+00 -7.03531206e-01
8.08050334e-02 -8.65416527e-01 1.39535129e-01 8.95072758e-01
-2.42354125e-01 -4.75371964e-02 -6.84153080e-01 3.89462441e-01
1.41194332e+00 -3.44016820e-01 4.98713195e-01 -7.45224953e-01
-6.62544191e-01 -9.85213220e-02 -3.17829430e-01 -8.94071162e-01
1.44796669e-01 -6.90942049e-01 4.87229079e-01 -1.70063710e+00
3.86367857e-01 2.52566300e-03 -2.21260577e-01 1.36958078e-01
8.37666821e-03 3.91537815e-01 -3.78794372e-02 2.52377540e-01
-4.07033801e-01 1.22690991e-01 1.32284439e+00 -5.28960600e-02
4.31322455e-01 1.26405712e-02 -6.65221810e-01 1.00162470e+00
3.97354394e-01 -6.78143263e-01 -3.17858279e-01 -5.37681937e-01
1.85552880e-01 6.71272755e-01 9.10221219e-01 -1.15256870e+00
-7.77942967e-03 -1.39122486e-01 8.98187935e-01 -1.91564471e-01
3.04675817e-01 -6.74391150e-01 5.75859427e-01 5.56255221e-01
-5.96969903e-01 -4.86856997e-01 -3.48707885e-02 7.42254436e-01
1.08085290e-01 -7.94997036e-01 9.15092945e-01 -6.45038009e-01
-4.07915860e-01 -1.48527905e-01 -4.47194576e-01 8.73551145e-03
5.18951237e-01 -2.43203957e-02 -2.44249165e-01 -8.84908497e-01
-1.46899617e+00 -3.66396487e-01 3.72597069e-01 -1.91070065e-01
8.51302147e-01 -1.08626246e+00 -1.15953636e+00 3.05318743e-01
-5.19397020e-01 -1.48427352e-01 5.11124790e-01 7.71378636e-01
-6.47945285e-01 -4.65466790e-02 2.65071541e-02 -6.67192042e-01
-7.28935122e-01 7.08509982e-01 6.93370223e-01 1.83801390e-02
-1.06143308e+00 5.90774655e-01 2.37999395e-01 -1.46001279e-01
1.04504526e-01 -2.31275409e-01 4.79836941e-01 -6.45139754e-01
7.79708564e-01 1.56906560e-01 5.47950529e-02 -3.83881629e-01
7.47789219e-02 5.54440618e-01 4.15945202e-01 -5.33964694e-01
1.53163016e+00 -2.83073932e-01 -4.85455804e-02 5.29267073e-01
1.26510859e+00 -1.58090442e-01 -1.60433865e+00 -2.29939312e-01
-2.65284777e-01 -5.89726567e-01 2.71839470e-01 -1.02569342e+00
-8.63410234e-01 9.53562975e-01 7.32890427e-01 2.04802334e-01
1.09740138e+00 7.65930722e-03 4.30375993e-01 4.79679286e-01
3.95721614e-01 -8.52978945e-01 2.59707123e-02 -1.07294992e-01
1.01685393e+00 -1.15082204e+00 4.14265186e-01 1.31611601e-01
-3.88334423e-01 9.25210595e-01 -3.19753557e-01 -6.39773071e-01
8.86111617e-01 6.13904178e-01 8.57441053e-02 -1.53516039e-01
-6.20407283e-01 1.86672702e-01 5.44246957e-02 1.02860379e+00
-8.24257806e-02 -8.67168605e-02 2.89663255e-01 2.27140710e-01
4.71931957e-02 3.34056377e-01 7.58066118e-01 8.51731002e-01
-4.77762371e-01 -6.44368529e-01 -8.91084075e-01 6.56947136e-01
-6.32999480e-01 -1.54587939e-01 3.38449001e-01 6.03019238e-01
6.20025881e-02 9.48337138e-01 -2.00144947e-01 1.43215671e-01
1.60946429e-01 -3.78273092e-02 1.00529766e+00 -4.41946238e-01
-8.51281881e-02 5.12782156e-01 -5.12324907e-02 -6.38139129e-01
-6.82548046e-01 -9.30287004e-01 -1.02158022e+00 -1.98854327e-01
-4.03616466e-02 -6.85620010e-02 8.40792835e-01 1.21396387e+00
-2.70402551e-01 5.27462006e-01 6.84425950e-01 -1.16621923e+00
-9.64562297e-01 -8.17344725e-01 -1.01987267e+00 5.50013840e-01
1.07420540e+00 -3.04931961e-02 -8.54666054e-01 5.66253603e-01] | [11.808302879333496, -2.3925271034240723] |
ad4d1149-5dd2-456f-b4bc-2630b9871742 | weakly-supervised-learning-of-human-dynamics | 2007.08969 | null | https://arxiv.org/abs/2007.08969v2 | https://arxiv.org/pdf/2007.08969v2.pdf | Weakly-supervised Learning of Human Dynamics | This paper proposes a weakly-supervised learning framework for dynamics estimation from human motion. Although there are many solutions to capture pure human motion readily available, their data is not sufficient to analyze quality and efficiency of movements. Instead, the forces and moments driving human motion (the dynamics) need to be considered. Since recording dynamics is a laborious task that requires expensive sensors and complex, time-consuming optimization, dynamics data sets are small compared to human motion data sets and are rarely made public. The proposed approach takes advantage of easily obtainable motion data which enables weakly-supervised learning on small dynamics sets and weakly-supervised domain transfer. Our method includes novel neural network (NN) layers for forward and inverse dynamics during end-to-end training. On this basis, a cyclic loss between pure motion data can be minimized, i.e. no ground truth forces and moments are required during training. The proposed method achieves state-of-the-art results in terms of ground reaction force, ground reaction moment and joint torque regression and is able to maintain good performance on substantially reduced sets. | ['Petrissa Zell', 'Bastian Wandt', 'Bodo Rosenhahn'] | 2020-07-17 | null | https://www.ecva.net/papers/eccv_2020/papers_ECCV/html/5301_ECCV_2020_paper.php | https://www.ecva.net/papers/eccv_2020/papers_ECCV/papers/123710069.pdf | eccv-2020-8 | ['human-dynamics'] | ['computer-vision'] | [-1.52027041e-01 5.82046434e-02 -6.71223581e-01 -6.33431301e-02
-5.48819780e-01 -3.50065082e-01 5.43649375e-01 -2.11368397e-01
-7.70679832e-01 8.37598205e-01 7.15913326e-02 3.66823487e-02
1.66461561e-02 -7.11293101e-01 -9.91247892e-01 -7.73714542e-01
-2.68825948e-01 6.05838180e-01 1.79629937e-01 -4.45790350e-01
-1.72219157e-01 5.45901418e-01 -1.39204192e+00 -1.71193987e-01
7.12522507e-01 8.32055211e-01 4.38932151e-01 8.87346387e-01
5.81341326e-01 1.07397556e+00 -2.41417259e-01 8.86089951e-02
2.82102525e-01 -5.24629831e-01 -8.06112528e-01 1.83166534e-01
2.91137815e-01 -5.23516476e-01 -5.78701735e-01 6.98035538e-01
8.16175699e-01 5.77210605e-01 6.46567106e-01 -1.16573107e+00
-2.69645661e-01 3.29747081e-01 -3.09016854e-01 1.50971293e-01
4.84162003e-01 2.97221154e-01 8.53821516e-01 -7.35017717e-01
1.04162872e+00 8.62363338e-01 7.52095938e-01 7.46060550e-01
-1.15399015e+00 -3.71359408e-01 -6.46278933e-02 3.04798245e-01
-1.15769863e+00 -4.53613847e-01 1.09699857e+00 -5.96466303e-01
1.00348759e+00 1.30365580e-01 8.69383931e-01 1.38235128e+00
2.52030969e-01 9.19319510e-01 5.63672841e-01 -2.11108267e-01
4.13078181e-02 -7.50804320e-02 -2.53294021e-01 7.19537973e-01
-1.20388735e-02 2.95863990e-02 -4.00447160e-01 2.40680873e-01
9.63231802e-01 -4.86323051e-02 -2.77640045e-01 -7.06468225e-01
-1.53641021e+00 3.58721197e-01 5.93323588e-01 1.97336867e-01
-5.26010871e-01 3.80776644e-01 5.56297064e-01 2.92099684e-01
3.89256269e-01 1.95809886e-01 -4.15334076e-01 -6.80023551e-01
-8.69951606e-01 7.49211669e-01 6.78551912e-01 7.73494780e-01
5.25750160e-01 2.75269777e-01 3.05427194e-01 5.19142747e-01
-1.08203329e-01 6.94939733e-01 4.22585309e-01 -1.08499885e+00
7.98702717e-01 4.28528249e-01 4.42699850e-01 -1.50528038e+00
-8.16967487e-01 -1.63565323e-01 -1.08483291e+00 1.45709723e-01
7.54652619e-01 -3.03876609e-01 -6.56417549e-01 1.75858748e+00
4.23161924e-01 -4.07980829e-02 -9.12950858e-02 1.38079739e+00
3.71900439e-01 5.70669949e-01 -1.41065210e-01 -1.13466844e-01
8.61344516e-01 -9.52239990e-01 -7.87428677e-01 -2.23173335e-01
8.12815905e-01 -3.42241943e-01 1.27408862e+00 2.21074849e-01
-1.32244039e+00 -7.10106373e-01 -1.04173708e+00 -3.09249878e-01
-7.20515698e-02 3.84563625e-01 5.32784164e-01 -7.96970166e-03
-4.64176625e-01 1.16826057e+00 -1.52674091e+00 -1.26426607e-01
9.91795212e-02 5.42397082e-01 -6.09564424e-01 2.90673673e-01
-1.07990956e+00 1.15848196e+00 1.93277299e-01 4.49562520e-01
-8.97054553e-01 -4.41231787e-01 -8.83958042e-01 -4.05949801e-01
5.42629719e-01 -9.16233301e-01 1.04039943e+00 -5.48955858e-01
-1.95939291e+00 6.36644840e-01 1.50956079e-01 -4.16289300e-01
1.02009141e+00 -6.72326446e-01 4.09249812e-02 1.86789274e-01
-3.25411931e-02 5.03367901e-01 8.44138205e-01 -8.79573941e-01
-3.09400707e-01 -1.02712542e-01 -1.89923737e-02 2.24266335e-01
-3.85335743e-01 -5.79173028e-01 -3.87137383e-01 -8.59977961e-01
2.37474609e-02 -1.44516778e+00 -1.58634260e-01 1.29898936e-01
-2.91114241e-01 4.18244973e-02 6.20472014e-01 -8.30624104e-01
1.21991134e+00 -1.68054295e+00 8.74181390e-01 5.52892163e-02
1.19509526e-01 2.69971400e-01 4.59788367e-02 3.37754130e-01
2.18780443e-01 -2.15875208e-01 -1.48812875e-01 -1.50050119e-01
3.20747006e-03 2.97282219e-01 -1.40449703e-01 6.66254282e-01
6.91595674e-02 1.05345500e+00 -8.97805214e-01 -4.10690457e-01
4.98813599e-01 3.58613104e-01 -7.03805208e-01 4.83006597e-01
-2.63050944e-01 9.86169636e-01 -6.00586653e-01 6.14937901e-01
1.03841633e-01 -3.19536686e-01 2.78765168e-02 -2.88136482e-01
6.72425479e-02 1.98702648e-01 -1.14925277e+00 2.02607918e+00
-6.87128961e-01 5.51373363e-01 -6.71491548e-02 -1.39836526e+00
8.38191152e-01 3.45422775e-01 7.70851910e-01 -4.69791383e-01
4.61462468e-01 4.10939246e-01 9.45876390e-02 -7.97222435e-01
3.19563985e-01 -2.79672295e-02 -2.04790205e-01 4.18609262e-01
1.05266131e-01 -1.60302430e-01 1.88516751e-01 -4.44806606e-01
9.65969145e-01 8.41988862e-01 1.53468385e-01 -2.39516478e-02
6.30312026e-01 3.01412225e-01 6.55406475e-01 1.66237444e-01
-1.44550562e-01 4.23563272e-01 8.19930658e-02 -4.82009202e-01
-1.43088090e+00 -9.98890460e-01 5.31734586e-01 9.45753157e-01
1.97935641e-01 -8.86892229e-02 -7.47767329e-01 -6.30880445e-02
-6.62939474e-02 3.41513939e-02 -4.23343688e-01 -2.29374811e-01
-1.22057390e+00 -3.99745554e-01 4.78892535e-01 8.43664587e-01
3.50286365e-01 -1.15878439e+00 -8.51532817e-01 3.71986151e-01
-4.80526835e-01 -1.35625148e+00 -4.83227998e-01 -2.82915205e-01
-1.14751649e+00 -1.02616632e+00 -1.04129279e+00 -7.92434156e-01
5.81878126e-01 -7.04723746e-02 8.17648590e-01 -1.80251494e-01
-2.68961132e-01 8.66480470e-02 -1.86350375e-01 -7.27656037e-02
-2.02636018e-01 2.34275326e-01 4.63845313e-01 -2.61858135e-01
-2.74262950e-02 -9.07743514e-01 -6.89571857e-01 4.08975750e-01
-4.15180534e-01 8.01837966e-02 6.37181044e-01 9.74718213e-01
8.37851644e-01 -2.25883514e-01 2.52015948e-01 -3.11810732e-01
3.69385630e-01 -8.29093456e-02 -5.21683395e-01 -1.27670124e-01
-3.15626174e-01 -1.56590104e-01 9.08780575e-01 -8.51796508e-01
-1.05096686e+00 2.71662891e-01 8.28891434e-03 -6.81847274e-01
-5.77389002e-02 5.14281273e-01 4.12262864e-02 -5.53863635e-03
8.84533644e-01 1.64877936e-01 3.07485461e-01 -5.81866026e-01
5.80798030e-01 2.79781461e-01 9.43508625e-01 -4.95496511e-01
9.06883478e-01 7.76442647e-01 2.82382399e-01 -8.81817997e-01
-4.66654062e-01 -4.64021653e-01 -1.02703202e+00 -3.03522855e-01
7.76794076e-01 -9.67264414e-01 -1.03278482e+00 6.22399032e-01
-9.55073714e-01 -6.26109540e-01 -2.52887696e-01 9.37821567e-01
-1.13245893e+00 4.05544966e-01 -8.89749527e-01 -7.87149549e-01
-3.32818687e-01 -9.22700226e-01 9.81587529e-01 -2.88389772e-01
-5.88408053e-01 -9.73403931e-01 1.50531664e-01 3.15553010e-01
7.28255734e-02 9.34822440e-01 4.24006730e-01 1.10937938e-01
-4.92287666e-01 -5.46033442e-01 4.31284130e-01 3.56970698e-01
2.74032682e-01 -1.96415946e-01 -4.74059850e-01 -5.42160749e-01
-1.15306951e-01 -7.82385647e-01 4.99907613e-01 5.33249259e-01
8.54372263e-01 -3.51212561e-01 -3.01152319e-01 5.26335001e-01
9.71604943e-01 -2.50057042e-01 4.98909622e-01 4.83344674e-01
1.19022763e+00 9.92336750e-01 8.02044094e-01 3.82827282e-01
4.00640845e-01 9.49464858e-01 7.30947852e-02 -7.02455863e-02
-3.84246260e-02 -3.84678692e-01 5.79761147e-01 1.25333476e+00
-7.60601521e-01 -6.23765737e-02 -1.03622222e+00 6.95439816e-01
-2.08370137e+00 -9.69299555e-01 -3.83287407e-02 1.91581738e+00
7.83442914e-01 2.23275319e-01 4.04821217e-01 4.64947492e-01
3.82300913e-01 1.60534188e-01 -7.80493617e-01 5.57034537e-02
9.84388441e-02 -8.94040987e-02 3.87142718e-01 3.88299614e-01
-1.11015368e+00 8.14435422e-01 5.50033140e+00 6.44441843e-01
-1.32142651e+00 -4.28013839e-02 -7.88526330e-03 -4.05117273e-01
2.03541264e-01 -3.07424217e-01 -5.41820645e-01 3.97604316e-01
9.07785892e-01 -1.72660351e-01 4.01268452e-01 1.03889203e+00
6.88974738e-01 1.20713033e-01 -1.36058021e+00 9.59183276e-01
-2.37926066e-01 -1.31553519e+00 -2.66668081e-01 -4.85328175e-02
8.21480155e-01 -4.86765662e-03 -9.81552377e-02 8.58784765e-02
-8.17818791e-02 -8.48592937e-01 8.69752705e-01 7.41087914e-01
7.47280240e-01 -6.80198908e-01 7.27112949e-01 9.70764041e-01
-1.27476180e+00 -8.99518132e-02 -2.73793578e-01 -4.21498775e-01
5.45286536e-01 4.41559881e-01 -3.19947153e-01 5.93539238e-01
5.72972417e-01 1.02458179e+00 5.19223511e-02 4.68826652e-01
-1.91137612e-01 3.54509205e-01 -4.50287491e-01 -2.19989702e-01
1.54107645e-01 -1.95695207e-01 7.67446697e-01 8.89386117e-01
2.03197792e-01 6.29892796e-02 5.44869125e-01 7.35164464e-01
1.17399514e-01 -7.57547095e-02 -6.63993955e-01 -1.95796162e-01
1.77422259e-02 1.03397381e+00 -5.24680674e-01 -2.40913466e-01
-4.85324077e-02 1.02044094e+00 3.17481309e-01 1.49229690e-01
-8.21496665e-01 -2.34830275e-01 5.74119925e-01 3.05070728e-01
1.09502085e-01 -7.35150099e-01 -3.97983603e-02 -1.37834740e+00
4.91697222e-01 -8.15537393e-01 1.01050608e-01 -4.55303639e-01
-1.11119699e+00 2.48758897e-01 2.37178862e-01 -1.86644638e+00
-1.10978425e+00 -5.89532852e-01 -2.56236613e-01 6.49824917e-01
-1.04868901e+00 -1.00042224e+00 -5.06718338e-01 6.28490388e-01
2.84574538e-01 1.53414905e-01 7.13437438e-01 5.06732821e-01
-3.71065289e-01 4.67245400e-01 -1.22864723e-01 3.07250500e-01
4.88574803e-01 -9.39124286e-01 4.29000378e-01 6.14063621e-01
-9.96181089e-03 4.86012787e-01 8.25795591e-01 -5.35109937e-01
-1.64780891e+00 -8.18041623e-01 8.10719192e-01 -5.15276194e-01
7.93801963e-01 -4.08290207e-01 -9.78834748e-01 4.92463648e-01
-5.16924322e-01 3.79264772e-01 2.33850956e-01 -2.93688446e-01
2.87790835e-01 -6.12419844e-02 -6.29742622e-01 8.06327045e-01
1.40010393e+00 -6.21798456e-01 -6.35656357e-01 3.06350291e-01
4.14570093e-01 -8.72764349e-01 -9.72505510e-01 6.16201341e-01
8.29994857e-01 -5.70315659e-01 1.06935251e+00 -7.02578604e-01
5.31917393e-01 -3.44580024e-01 1.60601035e-01 -1.11481047e+00
-7.63374791e-02 -7.28640199e-01 -8.05809915e-01 5.62127650e-01
1.88572332e-01 -1.29901871e-01 1.26689863e+00 3.07864398e-01
6.50239885e-02 -9.07881141e-01 -8.81936312e-01 -1.20098794e+00
1.52095850e-03 -4.60699111e-01 1.83062837e-01 9.86594617e-01
5.67247719e-02 2.13418335e-01 -1.02660954e+00 -2.51140110e-02
5.19077778e-01 1.95961416e-01 1.46570158e+00 -9.35022175e-01
-4.52616662e-01 -1.98072344e-01 -5.77637017e-01 -1.27644348e+00
3.03923965e-01 -6.81052744e-01 2.98826247e-01 -1.48630571e+00
-3.31810415e-01 -5.34658879e-02 4.25173119e-02 4.01071519e-01
4.69810283e-03 2.87491173e-01 1.34571254e-01 6.11297071e-01
-1.71777412e-01 7.89426863e-01 1.61625516e+00 -9.18670818e-02
-5.08004367e-01 6.47578165e-02 3.09185654e-01 1.06226075e+00
6.76407099e-01 -2.89934874e-01 -5.52045107e-01 -3.94976169e-01
-6.22137748e-02 3.60591710e-01 5.72787762e-01 -1.09874511e+00
2.74071395e-01 -2.78504997e-01 3.18703622e-01 -6.11630440e-01
3.20867777e-01 -6.47023141e-01 3.96088630e-01 7.17352629e-01
-2.21583650e-01 7.10001513e-02 6.71841903e-03 7.05887556e-01
-3.65657002e-01 3.83395284e-01 4.61988211e-01 -8.53881463e-02
-7.96145678e-01 4.65710431e-01 -2.96478033e-01 1.00020524e-02
8.34464371e-01 -4.71770495e-01 2.18459457e-01 -5.22469640e-01
-1.09347606e+00 5.90517633e-02 3.61488730e-01 5.78373373e-01
4.17872369e-01 -1.53222477e+00 -5.02138615e-01 -2.15432584e-01
1.21813249e-02 1.86957613e-01 3.37121308e-01 9.98019040e-01
-6.75651610e-01 4.06129211e-01 -5.81732333e-01 -7.87553132e-01
-1.05946970e+00 3.13994348e-01 1.03598557e-01 -4.63771701e-01
-1.09097636e+00 5.20023882e-01 -2.82468468e-01 -5.22030175e-01
2.15707272e-01 -4.81180102e-01 -7.24544674e-02 -1.63956210e-01
1.87315002e-01 7.62058794e-01 -1.21677548e-01 -9.44011867e-01
-3.01255822e-01 8.59643757e-01 3.99358422e-01 -9.42435190e-02
1.38066363e+00 -1.48215175e-01 1.47659078e-01 7.32747495e-01
1.34777939e+00 -3.26597452e-01 -1.55443120e+00 -7.73419142e-02
4.24934998e-02 -3.02910328e-01 -1.91218719e-01 -2.61739165e-01
-8.04646492e-01 1.01524043e+00 5.07506549e-01 -3.18590820e-01
1.03406429e+00 -3.62698823e-01 1.25562394e+00 9.18715477e-01
5.47388256e-01 -1.56995893e+00 3.46872300e-01 4.69626099e-01
1.04207253e+00 -1.31771207e+00 1.50277317e-01 -4.60579395e-01
-4.62644726e-01 9.53504026e-01 6.70537293e-01 -5.49343526e-01
5.40619969e-01 1.08432218e-01 6.08115755e-02 7.21517727e-02
-5.40134490e-01 -5.82453683e-02 6.72971487e-01 5.01176655e-01
3.53332043e-01 1.18662999e-03 -4.75172400e-01 4.33573514e-01
-3.36421281e-01 3.19479883e-01 2.31399998e-01 1.11921084e+00
-1.53058931e-01 -8.62778842e-01 -1.88563719e-01 2.90941098e-03
-4.07914013e-01 5.75403214e-01 -1.53773963e-01 1.20052290e+00
-7.59064825e-03 6.05278373e-01 -2.97323000e-02 -6.10837042e-01
6.55790269e-01 -1.87608138e-01 7.09745467e-01 -4.04278487e-01
-1.61432981e-01 8.08097515e-03 2.20030695e-01 -9.89684105e-01
-6.81489885e-01 -4.39867556e-01 -1.45701873e+00 -5.92272282e-01
-2.88007557e-01 -2.29299664e-01 4.23680574e-01 8.07419777e-01
1.37914672e-01 4.48028147e-01 4.41905230e-01 -1.62848616e+00
-6.73393905e-01 -1.01383901e+00 -3.70347500e-01 8.79515767e-01
5.27463496e-01 -1.07743967e+00 -1.67921573e-01 4.90918636e-01] | [7.271344184875488, -0.3582984209060669] |
131f0269-46db-42ea-a824-99eaec2c775a | versatile-speech-databases-for-high-quality | null | null | https://aclanthology.org/L12-1014 | https://aclanthology.org/L12-1014.pdf | Versatile Speech Databases for High Quality Synthesis for Basque | This paper presents three new speech databases for standard Basque. They are designed primarily for corpus-based synthesis but each database has its specific purpose: 1) AhoSyn: high quality speech synthesis (recorded also in Spanish), 2) AhoSpeakers: voice conversion and 3) AhoEmo3: emotional speech synthesis. The whole corpus design and the recording process are described with detail. Once the databases were collected all the data was automatically labelled and annotated. Then, an HMM-based TTS voice was built and subjectively evaluated. The results of the evaluation are pretty satisfactory: 3.70 MOS for Basque and 3.44 for Spanish. Therefore, the evaluation assesses the quality of this new speech resource and the validity of the automated processing presented. | ["Inma Hern{\\'a}ez", 'I{\\~n}aki Sainz', 'Jon Sanchez', 'Daniel Erro', 'Ibon Saratxaga', 'Eva Navas', 'Igor Odriozola'] | 2012-05-01 | null | null | null | lrec-2012-5 | ['emotional-speech-synthesis'] | ['speech'] | [-2.59434104e-01 4.58169341e-01 5.43407314e-02 -2.71185964e-01
-9.48498189e-01 -4.57723141e-01 7.28781641e-01 -1.00293141e-02
-2.06240103e-01 1.12488866e+00 4.33247864e-01 -3.12555403e-01
2.91274190e-01 -2.72026271e-01 -8.51873532e-02 -6.94792688e-01
3.99929494e-01 6.58096910e-01 2.51628429e-01 -6.02915168e-01
5.91511317e-02 6.94993615e-01 -1.77164769e+00 4.04764600e-02
5.33318996e-01 9.03843999e-01 4.25869405e-01 1.17158270e+00
-7.09147006e-02 7.39192724e-01 -1.29609573e+00 -2.88031459e-01
-3.65142524e-01 -7.85718203e-01 -1.06513286e+00 2.86693394e-01
-2.53336191e-01 1.04619879e-02 -1.02298945e-01 8.80622983e-01
8.96860719e-01 5.31753838e-01 5.02944291e-01 -9.02670503e-01
-1.90148130e-01 9.02998447e-01 5.45822144e-01 2.60377526e-01
8.34154248e-01 5.61439358e-02 8.02796364e-01 -9.01558816e-01
6.50064170e-01 1.26840389e+00 1.23727545e-01 8.20295811e-01
-9.97572780e-01 -5.78643262e-01 -5.79068005e-01 1.99014872e-01
-1.41384387e+00 -1.37202644e+00 5.16921699e-01 -2.96098650e-01
1.41030729e+00 7.37576902e-01 7.92534769e-01 1.29310787e+00
-5.89561202e-02 3.80240530e-01 1.29982162e+00 -8.75240624e-01
4.84333307e-01 6.49948120e-01 -8.03180709e-02 1.45101845e-01
-4.39168543e-01 3.44778270e-01 -4.74459499e-01 1.01858191e-01
2.74738640e-01 -1.05482507e+00 -3.14946651e-01 5.60348451e-01
-1.05660272e+00 5.54957390e-01 -5.25381804e-01 9.22450542e-01
-4.76686597e-01 -4.15431470e-01 9.29436445e-01 5.65192282e-01
2.51336396e-01 4.04699475e-01 -2.46974334e-01 -7.33051240e-01
-1.09975219e+00 1.99603468e-01 1.15239191e+00 1.25399983e+00
1.82312086e-01 1.02505255e+00 1.03980310e-01 1.37439942e+00
2.66996384e-01 7.76171923e-01 8.11502457e-01 -1.01795709e+00
2.92550981e-01 -2.21361980e-01 6.06617182e-02 -5.41887343e-01
-2.84392804e-01 1.48360580e-01 -3.64321202e-01 3.32810543e-02
7.84415230e-02 -3.97579104e-01 -7.92134881e-01 1.26617682e+00
-8.77527893e-02 -6.93919897e-01 5.70208192e-01 3.49693626e-01
1.17558157e+00 1.38432384e+00 -9.13593918e-02 -8.15287769e-01
1.16388357e+00 -8.83334398e-01 -1.72178626e+00 1.47693411e-01
2.16194227e-01 -1.25807059e+00 1.31541955e+00 6.54998362e-01
-1.63777614e+00 -6.40742421e-01 -1.13962686e+00 6.80253729e-02
-4.55058783e-01 1.14681132e-01 -1.91332087e-01 1.17169297e+00
-1.34107339e+00 1.39659435e-01 -3.31012279e-01 -4.54220414e-01
-6.15002811e-01 1.46804392e-01 -1.33512467e-01 6.81948423e-01
-1.26759851e+00 1.10827374e+00 6.33468211e-01 -1.88998953e-01
-7.34551132e-01 2.20897168e-01 -8.39412034e-01 1.72753006e-01
1.42360643e-01 -5.95383234e-02 1.77114654e+00 -9.50872540e-01
-2.56148648e+00 8.55562449e-01 -2.62144774e-01 -2.58616060e-01
3.73074144e-01 4.76394594e-02 -1.29103327e+00 3.71215701e-01
3.25368457e-02 3.19700718e-01 7.64948070e-01 -1.24956536e+00
-7.65358210e-01 5.70882522e-02 -5.77056587e-01 1.14501223e-01
1.83727816e-01 7.81656981e-01 -3.27704191e-01 -9.83143985e-01
-2.75476187e-01 -7.59199142e-01 2.50373602e-01 -8.61789167e-01
-4.36577290e-01 -2.77287275e-01 4.98496622e-01 -1.20455718e+00
1.55256999e+00 -2.23757935e+00 2.62413323e-02 3.08531106e-01
-4.18717563e-01 5.42123497e-01 2.71789849e-01 5.87855518e-01
-1.96896031e-01 1.36909917e-01 -2.54262201e-02 -2.48351008e-01
9.83038247e-02 3.94092530e-01 -1.41302407e-01 -2.55060382e-02
-3.41838412e-02 4.25229102e-01 -5.89523017e-01 -6.20224595e-01
5.51764786e-01 3.51102084e-01 -2.52784312e-01 4.58885431e-01
1.02037176e-01 3.98668617e-01 1.84801534e-01 5.72152495e-01
9.69983414e-02 7.87733436e-01 -2.06263624e-02 4.63725701e-02
-6.20539963e-01 5.85384667e-01 -1.04025352e+00 1.18581295e+00
-6.96958780e-01 8.26839209e-01 3.32961410e-01 -6.28310800e-01
1.43281102e+00 1.38260520e+00 2.06242114e-01 -6.62962615e-01
4.15112525e-01 8.43369603e-01 4.36291359e-02 -7.75853395e-01
8.56550872e-01 -4.34438527e-01 -4.59268913e-02 -6.28297403e-02
4.48597461e-01 -9.63652015e-01 4.23948288e-01 -1.78862125e-01
3.87750089e-01 -3.19850504e-01 6.15674138e-01 -3.41550559e-01
9.84488249e-01 -1.52432948e-01 3.90126526e-01 -5.16873933e-02
-4.48352605e-01 5.50265670e-01 2.80764461e-01 4.10070568e-01
-1.30042028e+00 -8.13599825e-01 -7.22483844e-02 7.63669908e-01
-4.47275102e-01 -5.14619172e-01 -1.06260920e+00 -9.27630514e-02
-8.17855358e-01 1.26401734e+00 1.80071235e-01 2.30769992e-01
-6.37442052e-01 2.62056123e-02 8.00085723e-01 2.60235697e-01
2.13752136e-01 -1.60005307e+00 -8.41999277e-02 4.67359006e-01
-5.77015460e-01 -1.17248142e+00 -3.98985475e-01 2.89886624e-01
-3.48085672e-01 -4.46669549e-01 -8.59441459e-01 -1.02834344e+00
-1.60684720e-01 -3.29603881e-01 8.71378899e-01 -3.29676837e-01
3.21907878e-01 2.98496246e-01 -5.70605993e-01 -5.44925153e-01
-1.46860242e+00 -5.36786281e-02 2.16398522e-01 -2.89707690e-01
1.47984236e-01 -3.47623497e-01 2.84313262e-01 2.96243876e-01
-4.73592103e-01 -4.46566552e-01 2.63065904e-01 6.93373084e-01
3.13741058e-01 1.72308102e-01 9.26586151e-01 -3.72982085e-01
9.18097615e-01 -1.17853373e-01 -4.19474304e-01 -7.89997354e-02
-3.33714664e-01 -3.41982663e-01 8.59892547e-01 -7.33020455e-02
-1.39597547e+00 -1.65138856e-01 -1.17785943e+00 1.62847772e-01
-6.63211405e-01 1.28604099e-01 -7.43101716e-01 3.82501721e-01
6.25677824e-01 1.15228347e-01 -1.41021997e-01 -5.56786239e-01
9.28262025e-02 1.46934879e+00 7.55120754e-01 -3.16579044e-01
3.10366303e-01 -3.38713437e-01 -6.35819256e-01 -1.62256944e+00
1.12846373e-02 -3.87736380e-01 -4.57644701e-01 -5.39122880e-01
1.03274822e+00 -6.66745007e-01 -5.26897669e-01 5.90683639e-01
-1.24212217e+00 -3.11301261e-01 -6.63286626e-01 7.29777217e-01
-1.02340293e+00 3.10053051e-01 -7.05172539e-01 -1.22625089e+00
-4.24839437e-01 -1.33203065e+00 7.28989720e-01 -1.03934098e-03
-6.80971444e-01 -8.24706733e-01 2.24702537e-01 4.20003891e-01
8.88586864e-02 -2.45818511e-01 7.33056366e-01 -6.82679772e-01
4.35509443e-01 -1.11574657e-01 2.23068923e-01 9.60560083e-01
1.15308218e-01 4.36964154e-01 -1.25573421e+00 8.90873969e-02
1.80480719e-01 -3.86008173e-01 -4.02742736e-02 1.68206349e-01
4.40454185e-01 -3.51656049e-01 3.84405524e-01 5.47007434e-02
8.20988953e-01 1.07547379e+00 6.02846980e-01 1.72546785e-02
-7.37600103e-02 8.46943438e-01 7.04924166e-01 3.52076262e-01
4.41836156e-02 9.55615997e-01 -2.78028101e-01 1.42983034e-01
-4.40969586e-01 -1.07003704e-01 1.01499259e+00 2.00571227e+00
-1.21369503e-01 -5.24733424e-01 -6.95731401e-01 6.59507692e-01
-1.15146518e+00 -9.38292682e-01 -2.18109354e-01 1.83074951e+00
9.15196300e-01 2.13004842e-01 5.28795958e-01 1.02417803e+00
8.91263545e-01 2.23407164e-01 2.74347484e-01 -1.41294444e+00
-4.33428049e-01 5.28647363e-01 -3.81071493e-02 9.86658931e-01
-5.87104142e-01 9.50962603e-01 7.14909363e+00 1.02409971e+00
-1.06506872e+00 1.40319869e-01 1.81053281e-01 1.15830205e-01
-7.73254707e-02 -3.76047850e-01 -6.19855046e-01 5.59584796e-01
1.77344084e+00 -4.27715361e-01 5.54361939e-01 5.57088435e-01
7.17422724e-01 -1.23011291e-01 -7.66473591e-01 1.08872700e+00
5.23889530e-03 -1.05358326e+00 1.29734268e-02 -1.27890170e-01
5.08233547e-01 -1.91658124e-01 -1.39232367e-01 4.34509069e-01
1.37161452e-03 -7.36738861e-01 1.41235137e+00 2.87770271e-01
1.00220096e+00 -1.00248325e+00 7.43897915e-01 1.52966827e-01
-1.04311132e+00 1.95058107e-01 -1.02832794e-01 2.12507218e-01
6.88830554e-01 2.74273008e-01 -1.02281427e+00 5.30835032e-01
4.78858709e-01 6.29379647e-03 -1.22669198e-01 7.04216659e-01
-2.96566218e-01 1.00173950e+00 -2.11889714e-01 -3.16851825e-01
2.96275437e-01 -4.92276773e-02 9.26954508e-01 1.61439753e+00
6.70190334e-01 5.50627150e-02 -1.98122948e-01 4.16191936e-01
3.32090080e-01 7.96687543e-01 -4.41657692e-01 -2.68636256e-01
6.37450457e-01 7.01413691e-01 -5.11393666e-01 -6.31068468e-01
-7.91532174e-02 9.64945793e-01 -4.99126166e-01 3.08190227e-01
-4.84598041e-01 -8.47917080e-01 3.83661807e-01 -1.70191124e-01
4.89449725e-02 -1.30315557e-01 -2.26266891e-01 -8.76219034e-01
-1.87980607e-01 -1.19724536e+00 -2.44387332e-02 -9.70222533e-01
-7.02759266e-01 1.08306825e+00 -1.37842502e-02 -8.01067591e-01
-8.44956100e-01 -4.85746056e-01 -4.59246099e-01 9.27169919e-01
-8.52546573e-01 -5.10821521e-01 1.37356892e-01 4.36585724e-01
1.08607745e+00 -6.51812017e-01 1.18662179e+00 5.97130239e-01
-6.92171276e-01 4.60507244e-01 1.28533334e-01 -2.31363520e-01
4.54859316e-01 -1.25006604e+00 2.02459171e-01 7.25764215e-01
-8.45177919e-02 1.02248406e-02 9.71327305e-01 -4.13599283e-01
-7.17135608e-01 -5.55829346e-01 1.61978018e+00 -4.47858460e-02
6.65562630e-01 -2.18154252e-01 -6.46620274e-01 2.38362059e-01
6.48064315e-01 -8.53462160e-01 6.81939244e-01 -4.44474250e-01
3.42141360e-01 9.78712179e-03 -1.27848363e+00 6.01235926e-01
4.78787363e-01 -6.51242316e-01 -7.78497279e-01 2.25180131e-03
6.85659051e-01 -1.42675117e-01 -1.17146075e+00 -1.77434012e-01
2.33846948e-01 -9.67403948e-01 3.48995715e-01 -7.21028075e-02
2.41213609e-02 -3.85546274e-02 -4.19869691e-01 -1.56172943e+00
9.33383033e-03 -1.37555075e+00 2.74764121e-01 1.54264593e+00
6.45542920e-01 -5.78521788e-01 9.67076570e-02 2.19183996e-01
-6.35517716e-01 1.55482784e-01 -1.02010608e+00 -9.70099986e-01
-1.01658903e-01 -6.67474627e-01 7.15723395e-01 6.64824247e-01
5.07692695e-01 6.66746795e-01 -3.30455124e-01 -1.79902047e-01
-1.14829846e-01 -6.89065039e-01 6.37061715e-01 -7.68767655e-01
-8.88820663e-02 -7.08952963e-01 -2.32249171e-01 -6.25088930e-01
2.12992668e-01 -5.61475933e-01 3.55150819e-01 -1.18477321e+00
-7.37522185e-01 -9.15085524e-02 3.33148450e-01 4.66754846e-02
2.38537729e-01 -1.97228845e-02 3.56105536e-01 -2.82245278e-01
3.06354731e-01 5.23253500e-01 9.12737489e-01 1.65220082e-01
-3.74052435e-01 2.67866850e-01 -3.90994400e-02 5.07016540e-01
1.02972567e+00 -9.13597792e-02 -3.96883339e-01 2.17156962e-01
-2.56368011e-01 4.48365629e-01 -1.29902050e-01 -1.09166765e+00
-1.74672127e-01 -2.47108396e-02 -3.24528784e-01 -7.59547710e-01
6.81372285e-01 -8.58471215e-01 4.50203091e-01 1.86919391e-01
-2.26705730e-01 -2.37767007e-02 2.06046432e-01 -1.37876526e-01
-6.43554509e-01 -6.73472285e-01 1.17221773e+00 1.08550347e-01
-5.95398962e-01 -4.95466739e-01 -1.26310074e+00 6.99768215e-02
8.24849904e-01 -4.97599274e-01 1.92066535e-01 -7.62902141e-01
-1.22341323e+00 -3.73946428e-01 2.38001615e-01 2.28733554e-01
5.15860081e-01 -1.21256351e+00 -6.61027253e-01 3.58019918e-01
4.57480252e-02 -5.47496915e-01 -6.44996613e-02 5.95078766e-01
-7.37662375e-01 6.63558364e-01 -4.24887359e-01 -1.37711838e-01
-1.35263038e+00 3.64052892e-01 3.10874701e-01 2.63183415e-01
-4.00037974e-01 4.49017525e-01 -5.08825600e-01 -2.76296765e-01
4.42160934e-01 -2.92095512e-01 -4.95916188e-01 3.27720433e-01
3.59562814e-01 8.89304996e-01 3.71820927e-01 -1.30363262e+00
-8.44029188e-02 2.35742368e-02 4.96035576e-01 -1.12375736e+00
8.78513575e-01 -4.94945556e-01 -1.43960267e-01 9.84000802e-01
1.13371360e+00 4.58059967e-01 -2.81378984e-01 2.51926988e-01
1.42695755e-01 -1.99410141e-01 5.90188392e-02 -8.94545615e-01
-6.23790979e-01 8.49927962e-01 2.76724637e-01 5.94034553e-01
1.13682055e+00 -2.86197871e-01 1.06010425e+00 2.58669674e-01
3.46266888e-02 -1.71849072e+00 -3.34912866e-01 8.50395739e-01
1.21694827e+00 -5.81219912e-01 -6.07393980e-01 -4.16496605e-01
-8.92685235e-01 1.19805562e+00 1.31694913e-01 4.18648005e-01
7.08254218e-01 4.24731880e-01 3.29988062e-01 1.94821835e-01
-7.54595518e-01 -3.09582174e-01 9.26332027e-02 9.65476334e-01
6.55814648e-01 3.19601059e-01 -6.80992663e-01 6.07086599e-01
-1.07513416e+00 -2.35013947e-01 6.65936232e-01 6.37138486e-01
-6.49817705e-01 -9.63238657e-01 -6.82996094e-01 -1.21043928e-01
-6.34583771e-01 8.35617185e-02 -7.85125136e-01 6.64148808e-01
-1.75822243e-01 1.59444249e+00 -2.01937243e-01 -3.74707907e-01
7.68920660e-01 5.44217467e-01 1.63905352e-01 -6.49660707e-01
-9.31915462e-01 7.35243082e-01 9.15019214e-01 -1.41255975e-01
-3.94542158e-01 -6.37738764e-01 -1.23129106e+00 -4.52225029e-01
-4.32308406e-01 7.65020013e-01 1.01315510e+00 7.85220385e-01
-2.90288866e-01 6.98170543e-01 9.18799341e-01 -8.27623010e-01
-2.65480727e-01 -1.19827676e+00 -8.39956880e-01 -1.29602179e-01
3.25787872e-01 -1.80467740e-01 -4.19624388e-01 3.72771174e-01] | [14.709670066833496, 6.585424900054932] |
053436f3-96ac-4092-bda4-1db0e8e36bfd | multi-agent-reinforcement-learning-methods | 2305.10091 | null | https://arxiv.org/abs/2305.10091v1 | https://arxiv.org/pdf/2305.10091v1.pdf | Multi-Agent Reinforcement Learning: Methods, Applications, Visionary Prospects, and Challenges | Multi-agent reinforcement learning (MARL) is a widely used Artificial Intelligence (AI) technique. However, current studies and applications need to address its scalability, non-stationarity, and trustworthiness. This paper aims to review methods and applications and point out research trends and visionary prospects for the next decade. First, this paper summarizes the basic methods and application scenarios of MARL. Second, this paper outlines the corresponding research methods and their limitations on safety, robustness, generalization, and ethical constraints that need to be addressed in the practical applications of MARL. In particular, we believe that trustworthy MARL will become a hot research topic in the next decade. In addition, we suggest that considering human interaction is essential for the practical application of MARL in various societies. Therefore, this paper also analyzes the challenges while MARL is applied to human-machine interaction. | ['Ying Tang', 'Guanjun Liu', 'Ziyuan Zhou'] | 2023-05-17 | null | null | null | null | ['multi-agent-reinforcement-learning'] | ['methodology'] | [-3.10267478e-01 2.33041532e-02 -5.50315201e-01 4.43937480e-02
1.26659378e-01 -3.57424259e-01 6.66478872e-01 9.91320014e-02
-5.41980803e-01 1.38899243e+00 -4.62510616e-01 -5.95250428e-02
-3.62703562e-01 -7.92079449e-01 -2.80746460e-01 -8.74132037e-01
-2.58778900e-01 1.99402332e-01 -7.79821500e-02 -5.45626163e-01
3.98440570e-01 3.59961212e-01 -1.37827754e+00 -4.34223384e-01
1.05227470e+00 7.49078155e-01 -1.23494618e-01 3.42083782e-01
2.64865547e-01 1.01452398e+00 -8.12012076e-01 -6.31574571e-01
1.20885625e-01 -6.64391041e-01 -6.58927202e-01 -1.82529598e-01
-5.35040140e-01 -5.70343554e-01 1.36415824e-01 1.11469054e+00
4.53039885e-01 1.86573178e-01 6.49016201e-01 -2.16647458e+00
-8.53133202e-01 7.38402307e-01 -5.77133179e-01 -1.90968320e-01
6.03875995e-01 2.39294499e-01 6.78694904e-01 -2.70351708e-01
3.59947592e-01 1.15869224e+00 4.30401832e-01 1.05357683e+00
-3.38950306e-01 -8.59441280e-01 1.98874518e-01 5.93535364e-01
-1.11972713e+00 -2.13298619e-01 8.14898193e-01 -8.87814090e-02
7.38888562e-01 3.17169160e-01 8.72489452e-01 1.12891102e+00
7.50272572e-01 8.85537863e-01 1.31319678e+00 -6.10199213e-01
7.92753398e-01 3.12295854e-01 -1.65920481e-02 5.17276764e-01
8.83465767e-01 4.95032012e-01 -5.02076566e-01 -2.92874098e-01
6.28818393e-01 -2.97747910e-01 3.32468539e-01 -3.53239536e-01
-9.66422558e-01 1.07266438e+00 1.42195374e-01 4.82817411e-01
-6.23501897e-01 1.26860544e-01 2.96672940e-01 1.51652098e-01
1.36110961e-01 3.81914526e-01 -7.97435492e-02 -1.67211503e-01
-3.31059098e-01 2.22852886e-01 8.48799050e-01 6.35524571e-01
2.60548294e-01 4.34098572e-01 3.65719974e-01 4.11701858e-01
6.61682367e-01 6.76705718e-01 4.14293289e-01 -1.38603151e+00
-3.31434906e-01 2.15878978e-01 3.18248957e-01 -9.96417880e-01
-4.77056891e-01 -1.01413876e-01 -7.82102108e-01 5.47834158e-01
5.61913438e-02 -4.91584569e-01 7.59147387e-03 1.35931563e+00
5.60663223e-01 9.92117077e-02 6.93072855e-01 5.81495643e-01
8.12091231e-01 6.57286942e-01 1.36634529e-01 -7.80298293e-01
1.13600087e+00 -9.49345469e-01 -1.18170512e+00 -1.47763789e-01
1.65936738e-01 -4.87809896e-01 5.04156351e-01 4.90576059e-01
-1.12544453e+00 -6.14187904e-02 -1.05773723e+00 6.81391120e-01
-3.13134193e-01 -2.48272136e-01 9.66298938e-01 1.06418037e+00
-6.86361790e-01 2.22709447e-01 -8.96884739e-01 -6.12317860e-01
2.65597273e-02 2.97847122e-01 1.26551405e-01 3.29136401e-01
-1.58185971e+00 1.12463522e+00 3.18499446e-01 1.37371287e-01
-7.26062298e-01 -5.64076006e-03 -9.52022314e-01 -4.65502024e-01
3.08165371e-01 -3.81910682e-01 1.35365152e+00 -8.64197612e-01
-2.05143094e+00 3.91424656e-01 1.38686270e-01 -5.24981022e-01
4.45136040e-01 -6.51287585e-02 -7.04517543e-01 2.93093115e-01
5.90592287e-02 2.06813008e-01 4.79914188e-01 -1.52367616e+00
-9.77419734e-01 -1.27188757e-01 1.02747880e-01 3.09156954e-01
-3.04514587e-01 1.16124839e-01 1.00064717e-01 -4.12502229e-01
-6.41668797e-01 -9.11706686e-01 -2.79016972e-01 -1.51424244e-01
2.17151061e-01 -6.96811974e-01 6.31734252e-01 -3.10951203e-01
1.16411662e+00 -1.97875428e+00 -1.36958912e-01 3.19916695e-01
-3.77558134e-02 3.28391820e-01 -2.31914669e-01 9.04578924e-01
7.49284327e-01 -1.00899432e-02 7.72665516e-02 2.96124548e-01
1.43756866e-01 4.04283643e-01 2.23368760e-02 5.85186303e-01
-4.31103632e-02 8.57300580e-01 -1.24150324e+00 -4.80115086e-01
3.72148752e-01 2.57882684e-01 -2.95775943e-02 2.42108300e-01
7.27703720e-02 4.02775705e-01 -9.19215858e-01 9.81448531e-01
5.22988260e-01 2.30745777e-01 4.16712344e-01 2.57650942e-01
-1.34572923e-01 -3.73203874e-01 -8.06772351e-01 9.22829509e-01
-2.89322138e-01 1.95246115e-01 1.38092726e-01 -1.09580433e+00
9.91272807e-01 5.82702696e-01 6.74528420e-01 -9.39666986e-01
2.35762343e-01 4.70755070e-01 1.34314045e-01 -8.39235961e-01
3.49906027e-01 -8.90005827e-02 -9.51180011e-02 8.34872663e-01
-3.85190099e-01 -3.56562257e-01 2.91119516e-01 -1.02591388e-01
4.16791499e-01 1.23074837e-01 8.43220174e-01 -1.39467373e-01
7.15437472e-01 6.15845434e-02 8.09394360e-01 6.84403181e-01
-9.95312035e-01 -5.67228258e-01 2.43687510e-01 -5.73253751e-01
-5.26714921e-01 -6.92699134e-01 2.37618297e-01 9.37678635e-01
6.55330062e-01 1.85765028e-01 -7.86187649e-01 -7.58397698e-01
2.39447765e-02 1.15169883e+00 -6.18825972e-01 -3.34652185e-01
-3.03266227e-01 -8.57067883e-01 4.60368276e-01 1.81010708e-01
8.19806099e-01 -1.44984794e+00 -1.03673518e+00 3.12831342e-01
-8.73503536e-02 -1.00790453e+00 1.99749961e-01 -3.29921126e-01
-4.97462958e-01 -1.12282825e+00 -5.43622494e-01 -6.67138398e-01
4.88438487e-01 3.52424979e-01 7.97439992e-01 4.03091252e-01
1.63712800e-02 8.39104950e-01 -7.67031550e-01 -8.90965521e-01
-5.89729309e-01 -2.50574023e-01 5.06103694e-01 -2.64291406e-01
5.93339205e-01 -2.83689588e-01 -4.93910104e-01 5.25897980e-01
-6.18797719e-01 -3.46954346e-01 4.66421396e-01 8.36568892e-01
3.30456682e-02 5.81872463e-01 1.27461076e+00 -4.86696243e-01
1.03879440e+00 -4.81379449e-01 -5.96902430e-01 7.77583241e-01
-8.70283246e-01 -4.09168571e-01 3.97050083e-01 -2.59676397e-01
-1.18121207e+00 -3.74219000e-01 7.16779903e-02 3.96640182e-01
-1.51527002e-01 3.98282319e-01 3.94911245e-02 -4.67749178e-01
5.80125749e-01 2.04941362e-01 6.29770637e-01 2.67031580e-01
7.18540773e-02 1.03553319e+00 1.42510061e-03 -7.16947496e-01
6.06731594e-01 3.98458242e-01 2.87737679e-02 -7.57024229e-01
-4.47983563e-01 2.26128936e-01 -1.02371380e-01 -6.76200330e-01
5.65205455e-01 -5.13720691e-01 -1.22583807e+00 4.68339920e-01
-8.68868351e-01 -3.72939557e-01 -1.50610000e-01 8.87623250e-01
-6.18071675e-01 5.54408431e-01 -5.46859205e-01 -1.43412936e+00
-7.05139101e-01 -1.12888813e+00 2.57014573e-01 5.96188664e-01
-1.70063376e-01 -1.14771891e+00 7.43613616e-02 6.00898683e-01
3.19339871e-01 2.44350091e-01 4.18264776e-01 -6.21768236e-01
-1.62743092e-01 -2.00252816e-01 3.83749634e-01 2.16892257e-01
3.20431948e-01 4.53330427e-01 -6.64120972e-01 -4.73650753e-01
2.43764624e-01 -5.50606191e-01 -2.24326123e-02 4.27058160e-01
3.76667082e-01 -5.83660126e-01 -8.69966820e-02 -1.82808980e-01
1.08784962e+00 1.11472964e+00 3.07845682e-01 8.56188297e-01
-1.24265701e-01 9.13815796e-01 1.33787763e+00 7.22658634e-01
6.73914135e-01 8.62698555e-02 5.91598630e-01 1.74044997e-01
4.21065241e-01 5.97516187e-02 7.33091354e-01 9.11230206e-01
-5.58522820e-01 -2.63278335e-01 -7.73163080e-01 1.47843137e-01
-2.26332569e+00 -1.08703923e+00 2.21346587e-01 1.95247531e+00
4.68291372e-01 -5.74776269e-02 3.89826924e-01 7.86919370e-02
7.67353237e-01 -2.84176290e-01 -6.81190014e-01 -8.23000908e-01
-1.38237685e-01 -3.13350886e-01 1.51885301e-01 4.24297869e-01
-8.71024072e-01 8.84186983e-01 7.36673021e+00 7.70178318e-01
-9.09497082e-01 2.46297251e-02 5.26626348e-01 4.22898799e-01
-2.53716409e-01 -1.35456607e-01 -4.48606670e-01 2.38175631e-01
6.43500566e-01 -7.31721342e-01 6.32404387e-01 7.03160703e-01
4.23334539e-01 -5.59530914e-01 -6.91555619e-01 5.41372418e-01
1.86247706e-01 -9.16460991e-01 -1.34050809e-02 -1.18467860e-01
6.57748103e-01 -4.53194469e-01 -1.59284174e-02 2.41991028e-01
5.29211640e-01 -9.07801926e-01 6.16115749e-01 1.58922076e-01
1.70005336e-01 -1.17931151e+00 1.08086109e+00 4.29092407e-01
-7.76833415e-01 -2.21096188e-01 -4.79417622e-01 -5.67787588e-01
2.79604048e-02 2.31027037e-01 -3.65534306e-01 8.71351063e-01
6.37303054e-01 6.52755737e-01 -1.14074074e-01 8.14447582e-01
-5.42062938e-01 2.81866550e-01 -1.52508587e-01 -9.69907582e-01
3.45366746e-01 -5.75437069e-01 2.75717437e-01 5.93371272e-01
2.41299242e-01 1.43576741e-01 1.90325633e-01 4.01465565e-01
4.70768631e-01 1.20276339e-01 -7.89339662e-01 -2.60771424e-01
8.33204687e-01 1.10546267e+00 -6.24062717e-01 -7.21319243e-02
-5.81342697e-01 5.87043285e-01 -5.07356897e-02 3.18047225e-01
-8.92515957e-01 -3.37882519e-01 4.63716626e-01 -4.83693451e-01
-4.74327475e-01 -2.32673600e-01 -2.96770066e-01 -8.28876734e-01
-3.27597916e-01 -1.08394766e+00 4.63055789e-01 -5.37300944e-01
-1.36308610e+00 6.22708261e-01 6.68701902e-02 -1.19594884e+00
-4.31703478e-01 -3.49054098e-01 -5.52316427e-01 6.02101795e-02
-1.57722998e+00 -1.49818122e+00 4.99749705e-02 4.23512161e-01
5.04790664e-01 -6.54924154e-01 9.41277802e-01 -2.65386283e-01
-8.07123721e-01 6.18249595e-01 2.48773128e-01 -1.99715823e-01
5.08784235e-01 -6.73401654e-01 -2.14500129e-01 5.20936131e-01
-4.43587869e-01 3.57417345e-01 8.46928596e-01 -7.38057733e-01
-1.61489916e+00 -7.37225533e-01 7.57551908e-01 3.37241851e-02
8.73240352e-01 1.36238694e-01 -3.02549452e-01 5.13311327e-01
8.51987839e-01 -4.73616093e-01 8.56821716e-01 -1.27368227e-01
3.17698091e-01 -2.90119588e-01 -1.66204512e+00 1.07012200e+00
3.85133654e-01 -1.12242617e-01 -4.48686600e-01 2.14993879e-01
4.20679897e-01 2.51958631e-02 -8.03185999e-01 2.55137384e-01
8.19127977e-01 -8.72519612e-01 7.31988430e-01 -2.77812928e-01
-1.11848591e-02 -1.74612746e-01 1.13564014e-01 -1.26575160e+00
-3.33562225e-01 -9.61182475e-01 6.90053683e-03 1.15804982e+00
1.28193259e-01 -1.04225516e+00 5.98057091e-01 8.50500584e-01
5.03862798e-01 -5.97341359e-01 -8.92395854e-01 -1.07918739e+00
3.38071585e-01 -1.07792124e-01 6.14056587e-01 1.04331517e+00
6.64056301e-01 -6.84391037e-02 -8.12858641e-01 3.51748392e-02
1.08386433e+00 -6.10955879e-02 5.28599381e-01 -9.92662847e-01
1.98830515e-01 -3.62057626e-01 -6.50365129e-02 -5.50125130e-02
4.88617122e-01 -2.65016377e-01 -1.83811590e-01 -1.54509604e+00
8.35391358e-02 -2.80230075e-01 -3.39045823e-01 4.39462185e-01
-4.43444587e-02 4.14794497e-02 3.21210891e-01 1.18017972e-01
-6.99727595e-01 7.19919264e-01 1.27338612e+00 -8.79257247e-02
-2.54896849e-01 2.03863084e-01 -5.67870080e-01 7.90741742e-01
1.78715074e+00 -1.75778434e-01 -4.84042585e-01 -9.64468643e-02
1.91715866e-01 2.41884351e-01 6.33566231e-02 -6.18638456e-01
4.59224194e-01 -1.23536420e+00 -6.15488365e-02 -1.59347087e-01
2.10702017e-01 -1.19049776e+00 1.93299890e-01 1.06531405e+00
-1.32516593e-01 1.34325907e-01 1.74571364e-03 3.12272847e-01
-2.90549919e-02 -5.81414878e-01 7.63828516e-01 -1.82810679e-01
-7.59912670e-01 -1.65286720e-01 -1.02836168e+00 -4.14589614e-01
1.95237446e+00 -2.02053890e-01 -2.13656887e-01 -5.43301105e-01
-2.02319175e-01 5.91212392e-01 2.95354694e-01 3.95756483e-01
9.32230651e-01 -1.27781701e+00 -6.53229117e-01 4.27442007e-02
-7.91652873e-02 -6.93397284e-01 1.67623222e-01 5.57724416e-01
-6.66891515e-01 2.51891255e-01 -7.27764070e-01 1.36406243e-01
-1.23496401e+00 7.73323476e-01 1.45652518e-01 -2.94740987e-03
-2.40693927e-01 3.35197449e-01 -1.24936461e-01 -2.07669228e-01
4.59903657e-01 5.21403551e-01 -5.24259448e-01 -3.23918194e-01
5.62641680e-01 6.63014054e-01 -5.14199138e-01 -5.05865335e-01
-6.40600026e-01 6.23572886e-01 1.41526714e-01 -2.87137121e-01
1.23793292e+00 -4.21679854e-01 -5.02708673e-01 4.84900355e-01
2.16521248e-01 -1.70989811e-01 -9.17786062e-01 1.96294382e-01
5.17652370e-02 -1.86833784e-01 -1.66574001e-01 -9.26136434e-01
-1.09941804e+00 4.72508788e-01 4.93880868e-01 3.71011168e-01
1.11390078e+00 -4.76270139e-01 5.91431797e-01 6.06695235e-01
9.25941408e-01 -1.63938880e+00 -9.98606626e-03 2.40913630e-01
9.67281878e-01 -1.26263273e+00 3.03967297e-01 -2.89308876e-01
-1.07424152e+00 1.13617826e+00 7.86066532e-01 1.19803203e-02
8.70021999e-01 2.29037404e-01 3.31273943e-01 1.56048819e-01
-7.06974268e-01 1.92223303e-02 -3.01596940e-01 1.12923038e+00
3.09559643e-01 2.16903895e-01 -1.00083506e+00 5.01821935e-01
2.50591133e-02 1.79224148e-01 8.60370338e-01 1.34143949e+00
-4.40319598e-01 -1.49158096e+00 -6.50481403e-01 -1.56241516e-02
-3.78864020e-01 4.61916119e-01 -4.70495760e-01 8.88651431e-01
-2.77532376e-02 1.63964033e+00 -3.46310794e-01 -4.17990446e-01
-2.04419764e-03 -5.24481118e-01 3.57032180e-01 1.22552149e-01
-6.31993592e-01 6.79982454e-02 3.48096415e-02 -2.91262060e-01
-9.04935360e-01 -4.62386400e-01 -1.43736839e+00 -7.17138290e-01
-1.93566486e-01 7.17253804e-01 6.74921930e-01 7.09082723e-01
4.81716208e-02 1.77224576e-01 8.25568676e-01 -3.99684846e-01
-5.68504393e-01 -4.54689592e-01 -8.39807272e-01 -9.76997763e-02
9.32395980e-02 -7.87107289e-01 -2.67671496e-01 -3.84281009e-01] | [4.036454200744629, 1.9121806621551514] |
e71d14ad-545b-44f0-b5a8-6e490b7d750c | a-hierarchical-residual-network-with-compact | 2109.13536 | null | https://arxiv.org/abs/2109.13536v1 | https://arxiv.org/pdf/2109.13536v1.pdf | A hierarchical residual network with compact triplet-center loss for sketch recognition | With the widespread use of touch-screen devices, it is more and more convenient for people to draw sketches on screen. This results in the demand for automatically understanding the sketches. Thus, the sketch recognition task becomes more significant than before. To accomplish this task, it is necessary to solve the critical issue of improving the distinction of the sketch features. To this end, we have made efforts in three aspects. First, a novel multi-scale residual block is designed. Compared with the conventional basic residual block, it can better perceive multi-scale information and reduce the number of parameters during training. Second, a hierarchical residual structure is built by stacking multi-scale residual blocks in a specific way. In contrast with the single-level residual structure, the learned features from this structure are more sufficient. Last but not least, the compact triplet-center loss is proposed specifically for the sketch recognition task. It can solve the problem that the triplet-center loss does not fully consider too large intra-class space and too small inter-class space in sketch field. By studying the above modules, a hierarchical residual network as a whole is proposed for sketch recognition and evaluated on Tu-Berlin benchmark thoroughly. The experimental results show that the proposed network outperforms most of baseline methods and it is excellent among non-sequential models at present. | ['Yu Sang', 'Xiaoxiao Zhang', 'Huan He', 'Shihui Zhang', 'Lei Wang'] | 2021-09-28 | null | null | null | null | ['sketch-recognition'] | ['computer-vision'] | [ 5.23812361e-02 -4.77728516e-01 -2.10907698e-01 -3.22560281e-01
-3.99783850e-01 -2.29354918e-01 5.08835614e-01 -2.24196628e-01
-3.07793051e-01 4.19144988e-01 1.46556288e-01 -6.07454181e-02
-1.80820376e-01 -7.80520439e-01 -3.70343596e-01 -7.09480584e-01
4.17224050e-01 8.73179063e-02 3.32415998e-01 -2.20879927e-01
4.87453371e-01 5.98797858e-01 -1.38408494e+00 3.10362726e-01
8.48723590e-01 1.10402274e+00 5.99768579e-01 1.69429898e-01
-4.30595607e-01 4.83381987e-01 -4.89030182e-01 -2.76634037e-01
2.38679096e-01 -3.43593031e-01 -2.14938089e-01 1.40385598e-01
5.28977454e-01 -5.58622777e-01 -4.82914984e-01 9.87349212e-01
5.42051375e-01 2.85051048e-01 5.28283358e-01 -1.10792506e+00
-5.50962806e-01 3.63849998e-01 -7.15598881e-01 -1.12548880e-02
1.81083992e-01 -4.89534286e-04 1.12876868e+00 -1.26327789e+00
3.85683626e-01 1.32505322e+00 4.51479733e-01 3.13409775e-01
-8.85903537e-01 -9.63545978e-01 6.24538660e-01 1.01817682e-01
-1.51038611e+00 -1.15899883e-01 1.24095035e+00 -2.50167191e-01
4.51223612e-01 3.41215283e-01 5.26510000e-01 9.80592906e-01
-1.53981581e-01 1.13183212e+00 1.00215793e+00 -3.23670983e-01
-1.52921513e-01 1.81762919e-01 1.58695683e-01 6.93031490e-01
1.04044326e-01 -2.05111995e-01 -1.82281703e-01 1.05627462e-01
1.25405431e+00 6.22935474e-01 -3.31169814e-01 -5.68581760e-01
-1.18566072e+00 6.25805557e-01 4.38709050e-01 7.61902332e-01
-2.78270483e-01 4.09005284e-02 3.94303799e-01 1.57090604e-01
2.32523248e-01 1.38613418e-01 -2.02948466e-01 -5.89161702e-02
-1.11348546e+00 2.27582213e-02 4.19384480e-01 9.01060343e-01
6.48297310e-01 1.06302604e-01 -2.06306413e-01 1.18094397e+00
2.94787824e-01 2.43220970e-01 3.13207000e-01 -4.04042244e-01
8.84434283e-01 9.29019213e-01 -1.66752234e-01 -1.38880932e+00
-1.84819683e-01 -5.37066340e-01 -1.42191231e+00 2.17385665e-01
3.15721482e-01 1.77807376e-01 -8.22500169e-01 1.55595088e+00
1.02903508e-02 3.11750025e-01 -4.18178558e-01 9.49800432e-01
9.31712687e-01 7.45780408e-01 -3.95628475e-02 -4.52118404e-02
1.26971221e+00 -1.01923442e+00 -7.32377529e-01 -3.83005068e-02
-6.55464604e-02 -8.33744466e-01 1.32494760e+00 4.68029350e-01
-9.23694134e-01 -1.18738592e+00 -1.25057340e+00 -1.14299417e-01
-4.70003694e-01 7.72821248e-01 7.84987271e-01 4.45745856e-01
-6.21162832e-01 5.88202357e-01 -4.09989029e-01 -3.76477689e-01
3.71800989e-01 2.13049918e-01 -4.17752147e-01 -1.79134279e-01
-1.14763796e+00 6.04125142e-01 1.81349382e-01 6.04737699e-01
-9.61168781e-02 -4.93254185e-01 -5.76350272e-01 3.65354717e-01
2.60566473e-01 -2.93622881e-01 6.91791952e-01 -9.31042731e-01
-1.59259033e+00 4.22234565e-01 -3.77384201e-02 2.82619357e-01
8.35260272e-01 -1.26420647e-01 -4.23316628e-01 3.68438810e-02
-2.07220301e-01 5.39971411e-01 1.00995708e+00 -1.33540702e+00
-5.80729365e-01 -3.85523796e-01 -2.14286502e-02 3.04874450e-01
-5.20461977e-01 -2.20313773e-01 -9.08398390e-01 -1.05969536e+00
2.77149737e-01 -6.50824964e-01 -1.10244289e-01 1.57535419e-01
-1.60942912e-01 -5.12709796e-01 8.86377692e-01 -6.67672276e-01
1.59470344e+00 -2.22243285e+00 2.04903796e-01 2.97660768e-01
2.03125596e-01 5.68627179e-01 -1.79001942e-01 4.51426327e-01
-1.40026525e-01 1.06479071e-01 -2.23609835e-01 -2.84200341e-01
3.05157863e-02 5.11711510e-03 -5.01495481e-01 3.57825821e-03
1.18905619e-01 7.35405087e-01 -5.42217731e-01 -4.96058345e-01
5.74302375e-01 4.85213101e-01 -4.61572766e-01 2.47327134e-01
1.46252513e-01 2.67548889e-01 -6.47821188e-01 6.56050026e-01
1.09472215e+00 -2.87671268e-01 1.39902040e-01 -3.93220484e-01
-1.11277439e-01 3.37955542e-02 -1.57219660e+00 1.88602436e+00
-4.23853070e-01 4.06615585e-01 8.79674312e-03 -1.08072519e+00
1.17329419e+00 1.48740634e-01 3.24997306e-01 -7.45479524e-01
-1.26550838e-01 1.46949828e-01 -2.20216393e-01 -3.18049431e-01
3.44326407e-01 -2.10836738e-01 -4.65848483e-02 1.40592769e-01
-1.72106728e-01 3.57869864e-02 -7.79096503e-03 1.14824526e-01
7.64782310e-01 1.03052042e-01 1.66462392e-01 -1.29913926e-01
9.98717546e-01 -6.64974272e-01 6.19061053e-01 4.70573246e-01
3.00737796e-03 7.48854220e-01 4.92989451e-01 -6.52272820e-01
-7.39401221e-01 -8.47534418e-01 1.74543243e-02 7.64012456e-01
3.90481949e-01 -5.14729321e-01 -2.72462606e-01 -9.14279044e-01
2.16481969e-01 2.58452326e-01 -6.43469810e-01 -6.42304644e-02
-6.55884802e-01 -2.91945100e-01 2.69406676e-01 6.60420716e-01
8.67550790e-01 -1.18248832e+00 -1.67277336e-01 1.62900388e-01
1.64235339e-01 -8.12898457e-01 -6.98101938e-01 -1.72084555e-01
-8.45304310e-01 -8.82195175e-01 -1.14374626e+00 -8.66976023e-01
8.12825799e-01 6.85854018e-01 7.45495379e-01 4.59886909e-01
-2.89563924e-01 1.97076380e-01 -3.74355584e-01 -1.16429381e-01
3.16032022e-01 1.74268618e-01 -3.54167104e-01 3.27857405e-01
1.67951971e-01 -7.52814710e-01 -7.13122249e-01 3.55560422e-01
-8.75371933e-01 2.75710553e-01 1.11038983e+00 1.03888261e+00
2.69357324e-01 1.42038763e-01 6.62699401e-01 -6.79389119e-01
6.59433782e-01 -8.60290527e-02 -4.11060601e-01 6.03911877e-01
-4.61246878e-01 4.52887043e-02 9.50394809e-01 -5.13189197e-01
-1.14721489e+00 -5.65651879e-02 -1.93778366e-01 -4.50471073e-01
-1.19831599e-03 3.08346868e-01 -5.56225419e-01 -1.84351485e-02
-9.76209268e-02 4.18143481e-01 -6.53975755e-02 -7.65351593e-01
1.07596576e-01 7.22251117e-01 1.02345925e-02 -5.75425506e-01
9.28218782e-01 1.70967996e-01 3.84344459e-02 -9.72419918e-01
-4.17456836e-01 -3.24479938e-01 -6.76915705e-01 -2.14474782e-01
5.40029585e-01 -6.51245713e-01 -8.86530101e-01 5.55425644e-01
-1.18709075e+00 -2.00510640e-02 -1.86690483e-02 3.04485768e-01
-2.78403740e-02 8.15805674e-01 -3.57804924e-01 -8.19886029e-01
-2.46938974e-01 -1.21087527e+00 1.03616095e+00 4.36876506e-01
2.27107167e-01 -7.53210723e-01 -1.54192984e-01 1.66033208e-01
4.77903605e-01 -8.46396983e-02 9.34261084e-01 -3.07003051e-01
-7.99724162e-01 -3.43415350e-01 -8.78852069e-01 5.24533331e-01
4.16278809e-01 -9.85783804e-03 -6.83687508e-01 -2.55670428e-01
-1.68205202e-01 -2.30023578e-01 1.01266980e+00 -5.38998395e-02
1.50173461e+00 -5.26047163e-02 -1.89423248e-01 5.72842538e-01
1.42976880e+00 3.32348287e-01 7.43229568e-01 -7.37973861e-03
8.81351113e-01 4.11115497e-01 5.96119821e-01 4.09257650e-01
4.37679559e-01 7.58020461e-01 -2.66720373e-02 -3.05901259e-01
-1.15821913e-01 -5.15933394e-01 9.90341678e-02 9.40524638e-01
-1.76979914e-01 -4.35152948e-02 -5.63467026e-01 7.95580372e-02
-1.88867891e+00 -9.80827987e-01 2.17153877e-01 2.31381226e+00
5.73866427e-01 1.36545137e-01 -5.08623198e-03 1.11489408e-01
7.66824961e-01 5.96112370e-01 -3.95778805e-01 -6.28603995e-02
-2.77436338e-02 8.32739398e-02 3.66379693e-02 3.73104841e-01
-9.75157320e-01 9.62502599e-01 5.50730848e+00 1.15003502e+00
-1.32658672e+00 -4.19848800e-01 5.19245207e-01 2.75131106e-01
-3.02593380e-01 7.22881779e-02 -7.52661109e-01 7.40772069e-01
-7.86511078e-02 1.78703547e-01 5.83997548e-01 7.19298959e-01
-7.07380995e-02 -4.51394208e-02 -1.01682317e+00 1.43219960e+00
8.67190138e-02 -1.14440906e+00 4.13398683e-01 -9.34031755e-02
3.13249856e-01 -6.78721130e-01 -5.28432371e-04 5.99096477e-01
-2.77826816e-01 -8.97440374e-01 4.41594750e-01 7.46870100e-01
9.27684784e-01 -8.53783429e-01 6.61030471e-01 4.13302779e-01
-1.70561945e+00 2.31583174e-02 -6.22205138e-01 -8.78586695e-02
-1.91608936e-01 4.90365982e-01 -1.95138201e-01 7.21868157e-01
2.97579229e-01 8.90743852e-01 -6.93153918e-01 1.05651939e+00
-2.12041974e-01 1.29054502e-01 -1.04047932e-01 -1.41732171e-01
2.00001448e-01 -5.25235832e-01 2.82707393e-01 1.23087597e+00
2.95753986e-01 2.48131081e-01 2.61732519e-01 8.15809190e-01
-1.15085877e-01 2.19958782e-01 -5.06524444e-01 -5.85275292e-02
3.96349400e-01 1.41917670e+00 -7.31308877e-01 -4.45747495e-01
-5.55318713e-01 1.16533244e+00 3.05701345e-01 4.29851115e-01
-7.53916442e-01 -7.68062830e-01 3.68118137e-01 -1.08272925e-01
3.68263066e-01 -2.30170250e-01 -2.79637456e-01 -1.44883680e+00
4.68652934e-01 -8.80208194e-01 1.83987111e-01 -5.10432005e-01
-1.41803360e+00 6.03795469e-01 -4.09890831e-01 -1.34461761e+00
3.01236361e-01 -6.56639576e-01 -8.31508100e-01 9.93297696e-01
-1.67629135e+00 -1.31219900e+00 -6.67060792e-01 5.37260950e-01
8.14318657e-01 -1.55497640e-01 6.11495793e-01 6.55384243e-01
-7.91589499e-01 7.89153218e-01 -1.08864672e-01 3.10099512e-01
9.18619096e-01 -1.11493385e+00 3.64023089e-01 7.28524148e-01
9.75303054e-02 9.85493124e-01 9.53334570e-02 -6.67136431e-01
-1.38273048e+00 -6.24103785e-01 6.62369609e-01 -1.41555935e-01
4.05540913e-01 -6.12187803e-01 -1.04671049e+00 2.81018108e-01
1.30411536e-02 1.08489320e-01 2.70727575e-01 1.28783375e-01
-4.49783176e-01 -5.67298472e-01 -8.40583622e-01 5.40403903e-01
1.08476150e+00 -5.87162077e-01 -6.46016657e-01 -2.23608837e-02
2.58172125e-01 -1.28571272e-01 -6.61117852e-01 5.02188087e-01
9.91536438e-01 -9.78068650e-01 1.06498289e+00 -4.02126700e-01
4.08744752e-01 -3.91310751e-01 2.79877894e-02 -9.78783429e-01
-4.67764199e-01 -4.87471491e-01 1.95768178e-01 1.53482199e+00
1.86110493e-02 -5.37173510e-01 9.75315154e-01 6.49805963e-01
1.88007697e-01 -9.45462942e-01 -7.17620313e-01 -7.59049058e-01
-6.34007677e-02 2.78744306e-02 5.26807368e-01 8.92210245e-01
-6.90827370e-02 5.41246653e-01 -5.32560885e-01 -2.14838162e-01
4.54623640e-01 4.57306296e-01 9.36887026e-01 -1.46084893e+00
-1.58438340e-01 -6.96388185e-01 -2.75849998e-01 -1.65627408e+00
1.58379599e-02 -5.60500503e-01 -2.49685019e-01 -1.67193806e+00
2.95312971e-01 -8.31784606e-01 -5.78807533e-01 2.50650793e-01
-5.39101958e-01 3.60111590e-03 5.88074446e-01 2.94085175e-01
-6.44003689e-01 8.55322838e-01 1.47729766e+00 -2.47878969e-01
-9.35053006e-02 1.71062574e-01 -5.48935056e-01 6.24178529e-01
5.47494531e-01 -1.33830532e-02 -4.25569832e-01 -2.47321084e-01
-1.05015924e-02 1.80746928e-01 2.42741868e-01 -8.73074055e-01
5.14642239e-01 -8.67393687e-02 5.56196690e-01 -1.06351280e+00
3.18100363e-01 -1.11022377e+00 -1.44728854e-01 1.69083670e-01
-1.55836806e-01 -1.04318529e-01 1.01292819e-01 5.74755907e-01
-4.99619395e-01 -8.67346674e-02 7.19333053e-01 -1.12674989e-01
-7.02982843e-01 3.82204682e-01 1.89407694e-03 -3.32255363e-01
7.66272426e-01 -3.72171700e-01 2.06455607e-02 -1.90239444e-01
-3.38335246e-01 1.78456500e-01 2.21504062e-01 7.70703554e-01
8.52576077e-01 -1.68115115e+00 -4.03133780e-01 3.94353241e-01
-6.64336095e-03 -5.25931716e-02 7.67002523e-01 6.32877350e-01
-2.77791768e-01 5.21457672e-01 -4.94269878e-01 -3.02016228e-01
-1.15351403e+00 6.22463584e-01 3.90214995e-02 -4.79815841e-01
-7.40501940e-01 8.35182846e-01 5.93552887e-01 -4.84654486e-01
6.14999652e-01 -4.94133741e-01 -3.81110102e-01 2.73600876e-01
6.79482758e-01 3.37217003e-01 -1.99819207e-01 -3.31199527e-01
-3.30989212e-01 9.90022838e-01 -1.62416190e-01 1.22853473e-01
1.45320857e+00 8.50927234e-02 -1.54403374e-01 4.57234442e-01
1.23202300e+00 2.11505771e-01 -1.42934048e+00 -1.90591276e-01
-1.05253600e-01 -6.38268590e-01 -7.04056546e-02 -8.24244559e-01
-1.18622470e+00 1.20004618e+00 4.70691890e-01 1.28723040e-01
1.18427551e+00 -6.46266818e-01 8.34188342e-01 3.93729776e-01
3.48725677e-01 -1.10135233e+00 3.37041587e-01 3.53430510e-01
1.17914522e+00 -1.27594686e+00 9.28600654e-02 -4.45473999e-01
-5.92220247e-01 1.40264058e+00 7.52337575e-01 -4.13490981e-01
5.09641707e-01 4.46562693e-02 -3.32226425e-01 -5.43749258e-02
-3.73207033e-01 4.72288616e-02 6.15693510e-01 5.79462908e-02
4.44316208e-01 -4.06095460e-02 -4.07166660e-01 9.15808797e-01
4.61959451e-01 4.07890379e-02 4.59445175e-03 6.18488312e-01
-3.93918037e-01 -1.47285056e+00 -2.30633169e-01 4.57052886e-01
-8.18199292e-02 -3.45444330e-03 -4.69626009e-01 8.85058641e-01
1.28843144e-01 6.35212839e-01 -2.31629804e-01 -4.90352780e-01
5.05978584e-01 -6.08667433e-02 5.28923333e-01 -3.82116497e-01
-4.20028061e-01 9.44843665e-02 -2.11424783e-01 -4.71601844e-01
-1.79056883e-01 -4.31579143e-01 -8.91568184e-01 -6.50777295e-02
-4.26459402e-01 1.53960526e-01 5.38858533e-01 7.17071533e-01
3.00198764e-01 6.02173448e-01 8.03154826e-01 -9.31758881e-01
-6.37026131e-01 -9.32683885e-01 -6.33512616e-01 2.50039697e-01
2.02374771e-01 -8.24733377e-01 -3.55078340e-01 -4.40780282e-01] | [11.623538970947266, 0.6550179719924927] |
be96cb02-0cf0-4fcd-854e-18d2349cf8d2 | deep-spatio-temporal-forecasting-of | 2106.10940 | null | https://arxiv.org/abs/2106.10940v1 | https://arxiv.org/pdf/2106.10940v1.pdf | Deep Spatio-Temporal Forecasting of Electrical Vehicle Charging Demand | Electric vehicles can offer a low carbon emission solution to reverse rising emission trends. However, this requires that the energy used to meet the demand is green. To meet this requirement, accurate forecasting of the charging demand is vital. Short and long-term charging demand forecasting will allow for better optimisation of the power grid and future infrastructure expansions. In this paper, we propose to use publicly available data to forecast the electric vehicle charging demand. To model the complex spatial-temporal correlations between charging stations, we argue that Temporal Graph Convolution Models are the most suitable to capture the correlations. The proposed Temporal Graph Convolutional Networks provide the most accurate forecasts for short and long-term forecasting compared with other forecasting methods. | ['Francisco C. Pereira', 'Filipe Rodrigues', 'Inon Peled', 'Frederik Boe Hüttel'] | 2021-06-21 | null | null | null | null | ['spatio-temporal-forecasting'] | ['time-series'] | [-4.82802480e-01 -2.80120105e-01 -2.37408176e-01 -4.82847810e-01
6.19823076e-02 -5.56343079e-01 7.59668827e-01 -3.67866084e-02
1.81060538e-01 7.55186915e-01 1.29844129e-01 -8.56034517e-01
-5.29738963e-01 -1.44694507e+00 -2.79895753e-01 -7.21424818e-01
-3.18151504e-01 4.37752932e-01 -2.11094365e-01 -5.43868244e-01
1.68319106e-01 9.22699213e-01 -1.13076437e+00 -4.00626399e-02
1.25125146e+00 1.08183455e+00 6.00380957e-01 4.84185636e-01
-4.08688962e-01 3.96366209e-01 -2.47491494e-01 -1.15770906e-01
-2.64926367e-02 -2.70590127e-01 -4.03779060e-01 -5.29908895e-01
-3.13388795e-01 -1.94373444e-01 -4.21867490e-01 9.87890661e-01
2.91199505e-01 7.45479241e-02 5.82707822e-01 -1.50286841e+00
-5.76151669e-01 7.22086906e-01 -1.47847161e-01 3.96468014e-01
-2.06623897e-01 1.11189045e-01 6.58766270e-01 -4.02927577e-01
1.65043294e-01 6.51517272e-01 6.62048399e-01 2.42962465e-01
-8.47065449e-01 -9.52544749e-01 2.83506572e-01 6.60696208e-01
-1.23167980e+00 -3.29740532e-02 1.07908964e+00 -3.83219153e-01
1.33513939e+00 3.89733523e-01 1.06305397e+00 7.33415127e-01
6.94009840e-01 1.66844115e-01 8.32236230e-01 -3.03168781e-02
8.24879259e-02 -1.79310411e-01 -5.34523055e-02 7.47301504e-02
1.64616570e-01 2.22426936e-01 6.48731142e-02 8.62453580e-02
-1.22434489e-01 2.46757731e-01 -6.08291291e-02 2.48635650e-01
-6.97536349e-01 6.73306167e-01 8.38528156e-01 9.74471271e-01
-5.97017825e-01 5.14274359e-01 2.73422629e-01 -1.84645861e-01
8.46064389e-01 2.03080237e-01 -3.36790740e-01 -8.60490724e-02
-1.25258231e+00 1.56212191e-03 6.22913122e-01 7.14982212e-01
6.16804957e-01 5.52613199e-01 -7.20856851e-03 2.73599237e-01
3.97364348e-01 9.25703764e-01 -4.94191982e-03 -5.28482378e-01
4.06156778e-01 4.91341799e-01 1.94233745e-01 -1.21180105e+00
-1.05039418e+00 -8.24258626e-01 -1.37232852e+00 -2.06643581e-01
-1.19840331e-01 -2.99845964e-01 -6.99926555e-01 1.25750792e+00
-2.83905566e-01 3.88710052e-01 -1.11555815e-01 5.01582325e-01
5.21784246e-01 1.24375355e+00 2.02032730e-01 -3.90333533e-01
8.86428475e-01 -7.07409918e-01 -1.35042155e+00 -1.17333286e-01
7.71390200e-01 -3.60801816e-01 2.56613284e-01 -9.43377614e-02
-7.41315663e-01 -3.44791740e-01 -9.00677264e-01 4.33879375e-01
-9.51329529e-01 4.02760953e-02 8.10419559e-01 3.98861021e-01
-1.28978181e+00 6.20732903e-01 -6.64411247e-01 -2.51626521e-01
3.13757628e-01 2.32854232e-01 3.51397097e-01 7.39710033e-03
-1.64191163e+00 1.20670056e+00 2.95240641e-01 8.01396668e-01
-6.26988590e-01 -8.94784749e-01 -5.53423405e-01 4.72966790e-01
-1.83558136e-01 -2.90472746e-01 1.02612066e+00 -4.38533515e-01
-8.67387176e-01 -6.47806153e-02 -2.07598016e-01 -6.69612944e-01
3.76704365e-01 4.95600581e-01 -9.76072907e-01 -4.63091403e-01
-1.06705919e-01 2.87445914e-02 3.06537777e-01 -1.08914089e+00
-5.60733438e-01 -2.27817386e-01 -4.19816077e-01 -1.80145428e-01
-5.01258194e-01 -3.29717070e-01 2.76558427e-03 -2.62042135e-01
-3.79095495e-01 -1.13756335e+00 -3.14551592e-01 -8.36468279e-01
-2.60783017e-01 -5.74488699e-01 1.33130836e+00 -9.23306048e-01
1.28495824e+00 -2.02327108e+00 -3.82962137e-01 4.18654978e-01
9.71966088e-02 3.07930052e-01 6.76206946e-02 7.99773157e-01
-5.69567569e-02 1.84288844e-01 4.33542207e-02 -1.32521912e-01
2.37578660e-01 5.34564972e-01 -4.14744318e-01 5.01591384e-01
-9.29372013e-02 1.40822423e+00 -9.92871404e-01 2.78350145e-01
4.99325752e-01 8.37960303e-01 6.29063845e-02 3.26780006e-02
-4.77104783e-01 5.80868840e-01 -5.16639888e-01 7.57572651e-02
1.10936666e+00 -2.08171338e-01 4.01770622e-01 -2.39509910e-01
-5.47230303e-01 1.95208326e-01 -4.29132789e-01 1.25848389e+00
-6.78184927e-01 8.61788094e-01 -1.01450026e-01 -1.07645559e+00
1.00902748e+00 1.80373266e-01 8.64734113e-01 -1.41447544e+00
5.79200573e-02 3.54214549e-01 -5.22047095e-02 -2.44269580e-01
4.87601191e-01 -3.85155119e-02 1.39599308e-01 1.39259338e-01
-5.05520225e-01 -3.04634348e-02 2.15101197e-01 2.04381034e-01
6.58351898e-01 -2.88890541e-01 -4.99171793e-01 -6.83235884e-01
3.50409806e-01 1.60869993e-02 5.02035379e-01 3.77904922e-02
2.16846377e-01 -1.09632947e-01 2.92597204e-01 -8.03919494e-01
-9.05023754e-01 -5.86098015e-01 -1.75243486e-02 5.26724041e-01
7.65417814e-02 -1.97409630e-01 -4.48198110e-01 -3.98363590e-01
2.30818689e-01 1.44858539e+00 -2.93653011e-01 -4.50907014e-02
-5.23760498e-01 -8.85022521e-01 1.31004140e-01 6.47700131e-01
5.53645909e-01 -7.05201507e-01 -1.42541453e-01 3.46423358e-01
-2.62706339e-01 -1.12998247e+00 -3.47245276e-01 2.42502972e-01
-5.69543123e-01 -7.92170882e-01 -5.89128971e-01 -4.74670947e-01
5.29453874e-01 3.49533617e-01 1.24194109e+00 2.99024940e-01
8.78187418e-02 7.92145059e-02 -1.25259250e-01 -4.73696381e-01
-2.40712315e-01 2.55968928e-01 -5.81466407e-02 -1.97251305e-01
3.16913635e-01 -7.35073626e-01 -9.06797349e-01 1.94314376e-01
-7.42697835e-01 2.66930133e-01 1.55038550e-01 2.43700534e-01
2.91417360e-01 7.05176115e-01 8.74768734e-01 -5.25414467e-01
7.31314600e-01 -8.97919059e-01 -1.13380659e+00 1.81380078e-01
-1.39086699e+00 -1.10203750e-01 9.62995887e-01 1.74758330e-01
-9.54701602e-01 -1.82194814e-01 -3.19676578e-01 -1.83492571e-01
1.52332753e-01 8.70655537e-01 1.44847453e-01 1.26126066e-01
-2.53110081e-01 2.29928538e-01 -5.77084303e-01 -4.68045026e-01
4.07625854e-01 3.87547582e-01 3.67211431e-01 1.11156508e-01
8.76372755e-01 4.50196266e-01 5.00422418e-01 -5.09410501e-01
-3.25363606e-01 -2.30716512e-01 -5.34050405e-01 -6.61161602e-01
1.09301686e+00 -8.42924893e-01 -7.75052607e-01 1.90530345e-01
-1.43182111e+00 -4.87906754e-01 1.16662458e-01 3.30256522e-01
-5.21428883e-02 -9.78281572e-02 -9.89993569e-03 -1.07633984e+00
-5.36658347e-01 -9.26272511e-01 6.14251852e-01 2.01881900e-01
1.79332182e-01 -1.42604268e+00 1.33650601e-01 -2.03419507e-01
1.24831736e+00 3.56320590e-01 1.00536871e+00 -1.54447779e-01
-5.46894312e-01 -2.54894614e-01 -4.86166030e-01 1.45010412e-01
2.81703442e-01 9.84438285e-02 -7.79261827e-01 -4.76451963e-01
-2.86004990e-01 7.34424472e-01 7.79449582e-01 6.68664098e-01
1.45860922e+00 -4.23747450e-01 -6.13668263e-01 7.77059853e-01
1.54985344e+00 4.90576148e-01 8.92232895e-01 7.34447390e-02
7.87825346e-01 4.92513150e-01 1.64672941e-01 4.33168083e-01
1.04728949e+00 4.34921414e-01 8.68119597e-01 -3.42128098e-01
-9.86791681e-03 -1.65105999e-01 8.51165876e-02 1.27912664e+00
-2.51809377e-02 -5.12931228e-01 -1.37626255e+00 8.07698965e-01
-1.96273780e+00 -1.14138615e+00 -6.68496132e-01 1.74107552e+00
-7.83380494e-02 4.05330732e-02 -1.28468439e-01 -1.56356916e-01
4.94157106e-01 2.14435831e-01 -4.44112062e-01 -7.67732441e-01
-2.08711714e-01 -4.37759459e-02 1.14605761e+00 3.13744038e-01
-4.25206393e-01 4.21983570e-01 6.36524248e+00 5.92578709e-01
-1.24449611e+00 1.50396094e-01 6.44902110e-01 1.69822589e-01
-1.04568768e+00 1.13209352e-01 -6.13866448e-01 8.96957278e-01
1.58227742e+00 -5.56110382e-01 8.05294991e-01 2.80186445e-01
9.41372871e-01 1.42661795e-01 -6.33423090e-01 8.38348389e-01
-3.54225129e-01 -1.79025006e+00 -2.66477942e-01 3.14143389e-01
1.05799925e+00 6.29853249e-01 -1.06393501e-01 1.03517130e-01
3.87017757e-01 -1.22085047e+00 2.70223647e-01 1.18503010e+00
5.97436488e-01 -1.14308882e+00 9.11104500e-01 5.17778754e-01
-1.73847580e+00 -3.13574433e-01 -1.25191167e-01 -5.88507615e-02
4.80147332e-01 1.10669708e+00 -6.19962394e-01 9.89802718e-01
9.22518492e-01 1.02279568e+00 -2.94598669e-01 7.37129569e-01
1.45854324e-01 5.25752604e-01 -2.12625161e-01 -1.63413391e-01
3.20745319e-01 -7.15157032e-01 1.23291768e-01 1.16236138e+00
8.36679339e-01 -1.36299014e-01 4.47438881e-02 9.08269465e-01
-1.22927509e-01 -1.40173778e-01 -9.60233152e-01 -3.39690477e-01
2.29053929e-01 1.29671752e+00 -4.56707895e-01 -1.54686883e-01
-5.74303985e-01 5.08827388e-01 -8.80908817e-02 5.97918510e-01
-1.01604569e+00 -2.40527526e-01 6.39922857e-01 3.51926088e-01
1.97002158e-01 -6.62574589e-01 -3.55244190e-01 -7.17345834e-01
-1.61982343e-01 5.39124645e-02 2.44265169e-01 -1.07076931e+00
-1.21676207e+00 5.13942122e-01 -1.58071611e-02 -1.05588305e+00
-3.41723323e-01 -3.03910911e-01 -1.16122270e+00 1.23902798e+00
-2.55779195e+00 -1.36839807e+00 -4.32608396e-01 4.03249264e-01
3.09146583e-01 6.80122524e-02 7.56036341e-01 4.80211973e-01
-3.52015823e-01 -1.68538213e-01 4.76293445e-01 -1.74023196e-01
-3.75734568e-02 -1.12021792e+00 5.98118246e-01 6.78044915e-01
-2.79686242e-01 -8.09935778e-02 6.34253681e-01 -7.87440360e-01
-1.83384347e+00 -1.44428229e+00 1.27455223e+00 9.43832695e-02
8.71610701e-01 -1.97467774e-01 -8.88606310e-01 4.52986121e-01
8.44081521e-01 1.03754569e-02 4.66514647e-01 -1.14502370e-01
8.54634568e-02 -6.11870766e-01 -9.28433180e-01 2.85565287e-01
6.13471687e-01 -4.52448338e-01 2.33933210e-01 6.02919638e-01
4.44177210e-01 1.82915092e-01 -9.45651293e-01 5.75676024e-01
3.04675996e-01 -5.41318953e-01 3.94905865e-01 -1.80145308e-01
-1.30838335e-01 -3.16487044e-01 -6.90119937e-02 -1.82533562e+00
-6.40580714e-01 -5.66834033e-01 -8.99756923e-02 1.02141702e+00
5.51587880e-01 -9.31093931e-01 5.15825093e-01 7.40822196e-01
-2.13733941e-01 -4.64244485e-01 -1.11417544e+00 -7.95829058e-01
3.78179044e-01 -8.32171261e-01 1.31601655e+00 1.11569810e+00
-3.96703556e-02 -3.07556540e-02 -3.51910532e-01 4.87722278e-01
5.07945538e-01 4.35455859e-01 2.66051054e-01 -1.39943337e+00
4.63196903e-01 -7.12351084e-01 -1.31888986e-01 -5.12503505e-01
5.33450603e-01 -1.30677056e+00 -3.70552570e-01 -2.27710962e+00
5.87038361e-02 -4.44523931e-01 -5.24892628e-01 3.48330230e-01
4.05229211e-01 1.19903259e-01 1.11451484e-01 -1.35702625e-01
-1.25986040e-01 8.39087427e-01 1.03569543e+00 -5.64474642e-01
1.96093351e-01 5.40438305e-05 -2.48806745e-01 2.23184675e-01
1.38957882e+00 -2.59735316e-01 -4.23198283e-01 -7.12869823e-01
9.88119423e-01 3.88985276e-02 1.63688332e-01 -8.03472936e-01
5.08743703e-01 -4.13072258e-01 3.10928285e-01 -1.21237028e+00
-1.12553537e-01 -1.34878421e+00 8.06493044e-01 5.28882086e-01
1.17801942e-01 5.65479279e-01 4.17507797e-01 5.22950232e-01
-2.31718216e-02 2.88396001e-01 3.48471731e-01 2.00878486e-01
-6.17220342e-01 5.85084021e-01 -6.19595110e-01 -6.79906368e-01
1.01203883e+00 1.84277907e-01 -5.03622234e-01 -5.98661721e-01
-6.00795031e-01 9.69658494e-01 7.27534294e-02 6.67405844e-01
4.39254135e-01 -1.65080988e+00 -7.74133563e-01 1.23583995e-01
-1.06822461e-01 -4.30627167e-01 4.56457168e-01 8.44314098e-01
-2.96903104e-01 8.13026369e-01 1.59339264e-01 -4.59798545e-01
-6.17094219e-01 5.50260007e-01 6.12864733e-01 -3.53663474e-01
-5.20792484e-01 1.04042925e-01 -3.49488676e-01 -3.24350059e-01
-2.34151930e-01 -3.91492784e-01 -4.54633981e-01 2.48449937e-01
2.25455806e-01 5.12248099e-01 4.36618447e-01 -7.30673611e-01
-5.41528523e-01 3.41954917e-01 7.21215129e-01 5.96134067e-01
1.74399006e+00 -4.42669213e-01 -5.00565767e-01 4.19556856e-01
1.24085927e+00 -2.52516389e-01 -8.80478919e-01 2.90857941e-01
1.91573888e-01 -2.68771082e-01 8.04711640e-01 -8.85891497e-01
-1.71248448e+00 8.50884855e-01 7.32933760e-01 1.09422898e+00
1.23429298e+00 -3.18038642e-01 1.10891008e+00 -3.03517804e-02
3.57951134e-01 -1.20574880e+00 -8.64429593e-01 5.36281347e-01
8.34670484e-01 -8.57883573e-01 -1.10795423e-01 -4.54985797e-02
-2.08419025e-01 1.45080066e+00 1.72055542e-01 1.37990922e-01
1.11737025e+00 4.55985844e-01 -2.69892514e-01 -4.71403450e-01
-7.43276060e-01 -3.11112791e-01 4.20053184e-01 4.21547711e-01
2.89195538e-01 5.58664680e-01 -2.41626471e-01 3.98793727e-01
-1.82955824e-02 -4.09425199e-02 1.72597274e-01 3.20489287e-01
-1.10187419e-01 -1.01239681e+00 1.18506970e-02 7.99460590e-01
-1.55504480e-01 -9.03912857e-02 9.43293571e-02 2.66963869e-01
3.54398265e-02 1.22344971e+00 3.17405909e-01 -3.77215117e-01
4.45880800e-01 -8.35220050e-03 -5.20739742e-02 8.76988620e-02
-3.87310892e-01 -8.47798511e-02 -2.21696179e-02 -4.76018339e-01
-4.18278426e-01 -2.23030150e-01 -1.39936340e+00 -9.06499207e-01
-3.89628232e-01 3.93160254e-01 1.28961277e+00 9.54739332e-01
4.96025622e-01 1.05842578e+00 1.12813985e+00 -9.70782816e-01
-8.83322507e-02 -8.06227624e-01 -8.26498926e-01 5.19069731e-02
3.64649385e-01 -4.37595189e-01 -5.07418692e-01 -6.34349704e-01] | [6.25131893157959, 2.682584762573242] |
10e6ff4f-82d2-41ec-a2cf-fc21d7abbfe6 | investigating-modulation-spectrum | null | null | https://aclanthology.org/O15-3005 | https://aclanthology.org/O15-3005.pdf | 調變頻譜分解技術於強健語音辨識之研究 (Investigating Modulation Spectrum Factorization Techniques for Robust Speech Recognition) [In Chinese] | null | ['Hsin-Min Wang', 'Kuan-Yu Chen', 'Hsiao-Tsung Hung', 'Ting-Hao Chang', 'Berlin Chen'] | 2015-12-01 | investigating-modulation-spectrum-1 | https://aclanthology.org/O15-3005 | https://aclanthology.org/O15-3005.pdf | roclingijclclp-2015-12 | ['robust-speech-recognition'] | ['speech'] | [-8.63703638e-02 1.71006292e-01 -6.22772932e-01 -4.08054382e-01
-8.41685571e-03 -9.08429027e-01 6.55310392e-01 -6.53472245e-01
-2.85945535e-01 1.06888819e+00 -4.63127941e-02 -1.01159286e+00
-3.91567826e-01 -9.63214397e-01 -4.95059669e-01 -6.31337762e-01
-9.79754329e-01 7.25764990e-01 3.30370307e-01 -6.93831444e-01
7.03166842e-01 7.88774848e-01 -1.68942046e+00 7.18545914e-01
7.04417467e-01 8.52217197e-01 2.49141872e-01 1.14950800e+00
-1.95044339e-01 1.55633950e+00 -7.48382092e-01 -5.46825826e-01
3.13719302e-01 -1.23176083e-01 -7.22945035e-01 -1.01074085e-01
9.28529128e-02 -8.59008506e-02 -2.09758401e-01 9.22211111e-01
5.37373662e-01 4.49454933e-02 1.08379531e+00 -1.42548037e+00
-5.91619551e-01 6.10313773e-01 -4.01565880e-02 1.21627934e-01
1.03678203e+00 -5.39447069e-01 1.19919395e+00 -1.13026452e+00
7.20913768e-01 1.26888943e+00 8.66221786e-01 5.44149756e-01
-1.22286928e+00 -1.94712028e-01 -3.26822817e-01 -9.51717794e-02
-1.46558487e+00 -3.25250506e-01 4.25783843e-02 -2.08119690e-01
1.66093647e+00 1.26596653e+00 1.20609856e+00 1.01401424e+00
1.26658809e+00 8.34431887e-01 1.04267764e+00 -5.13792276e-01
3.35295945e-01 3.66983831e-01 1.54683650e-01 6.33519173e-01
8.40953708e-01 5.26628852e-01 -7.06372619e-01 -9.13127720e-01
9.33553874e-01 -2.94925272e-01 1.71355158e-01 -5.05680561e-01
-9.05919552e-01 6.91228509e-01 1.78732842e-01 3.83959889e-01
-1.39880210e-01 9.89067405e-02 1.26390755e-01 5.30987144e-01
-2.58292928e-02 6.47037446e-01 -9.11868811e-01 -1.33165747e-01
-8.71728659e-01 5.10332465e-01 1.25398111e+00 1.52653182e+00
1.24482810e-01 2.94908643e-01 -9.34252143e-02 3.17179203e-01
8.92314315e-01 1.01808000e+00 4.28362608e-01 -1.36146402e+00
-6.87414408e-02 1.72361732e-01 5.01781464e-01 -8.52631688e-01
-6.33224547e-01 -9.64177120e-03 -8.93263519e-01 4.49267089e-01
3.49161088e-01 4.57367361e-01 -8.02827001e-01 5.07305264e-01
4.33481112e-02 -2.34125629e-01 4.53833073e-01 5.55570945e-02
4.99930978e-01 3.76208365e-01 -1.34477139e-01 -5.73289394e-01
1.06082785e+00 -1.36716676e+00 -1.35299087e+00 2.33215362e-01
9.05734658e-01 -1.07320261e+00 4.35900748e-01 5.33875942e-01
-1.55548143e+00 -1.37560293e-01 -1.08699942e+00 1.78573877e-01
-7.27255583e-01 -3.14239264e-01 8.57801437e-01 1.43120694e+00
-1.60129595e+00 9.73287821e-01 -4.91727620e-01 6.59165755e-02
1.20568443e-02 8.24621081e-01 -2.64718989e-03 4.62812334e-01
-1.33193445e+00 1.08501506e+00 2.22979754e-01 -1.21242590e-01
-1.65216476e-01 -2.13068098e-01 -8.23704481e-01 -5.63443303e-01
-4.78693932e-01 -5.29636025e-01 1.44139910e+00 -2.59346128e-01
-1.65295815e+00 9.71794367e-01 -1.42069459e-01 -1.97814897e-01
6.14786744e-01 -1.28011424e-02 -8.31891418e-01 2.42498964e-01
-1.89849049e-01 5.76383233e-01 9.28263724e-01 -1.35132408e+00
-7.59897232e-01 -1.67359829e-01 -1.23336017e-01 2.66287565e-01
-1.25510961e-01 1.89734384e-01 2.11616129e-01 -1.12999000e-01
3.27147305e-01 -7.20919967e-01 -2.53068686e-01 -5.32041907e-01
-1.46512717e-01 -7.10518599e-01 7.70373225e-01 -4.51523662e-01
1.83705616e+00 -1.67618537e+00 -2.06720144e-01 3.98590982e-01
3.57815564e-01 -1.24705513e-03 2.40583986e-01 1.08380008e+00
-2.76906848e-01 7.32199550e-01 4.11965609e-01 -1.10722095e-01
2.24991128e-01 4.85861301e-01 -4.16602850e-01 3.05609167e-01
-1.29282743e-01 1.15307164e+00 -1.16605783e+00 -5.23096442e-01
4.11106765e-01 9.42391157e-02 -4.58732933e-01 4.79237735e-01
2.99364805e-01 1.47170946e-01 -3.56553018e-01 1.39399457e+00
1.15709066e+00 -1.31984919e-01 1.45911396e-01 5.30878425e-01
-3.79135728e-01 3.55090618e-01 -6.96863770e-01 1.05554795e+00
7.08333924e-02 5.00173986e-01 1.02364108e-01 -7.94621468e-01
3.33247900e-01 8.61540735e-01 4.37155962e-01 -9.67555881e-01
-2.26398129e-02 6.92409754e-01 1.12803578e-01 -6.07703328e-01
7.58228302e-01 3.94563079e-02 -3.64872098e-01 6.40070081e-01
-2.37588286e-01 -6.59476995e-01 9.64643434e-02 3.08100313e-01
6.22585893e-01 -5.18246442e-02 5.62923312e-01 -1.05613089e+00
7.86340594e-01 -1.82965681e-01 -1.81299388e-01 1.03415680e+00
-3.09923887e-01 3.19085121e-01 2.99841821e-01 -6.65102363e-01
-6.45341039e-01 -1.12307119e+00 -4.89381433e-01 1.30636716e+00
3.24267983e-01 -4.39044595e-01 -9.54439282e-01 -2.49762803e-01
1.77620783e-01 6.89606130e-01 -5.90509653e-01 3.84124845e-01
-5.03739953e-01 -8.32535863e-01 7.39044368e-01 3.45434904e-01
-5.07752821e-02 -1.33414865e+00 -6.58416986e-01 1.25490099e-01
-2.20292807e-01 -6.63697243e-01 -6.23428151e-02 4.48765576e-01
-1.35989368e+00 -5.18594682e-01 -6.66252747e-02 -8.20914626e-01
5.87345481e-01 2.46782884e-01 1.27047324e+00 5.39230824e-01
-2.31483161e-01 4.26904231e-01 -1.21292919e-01 -4.95818377e-01
-4.59671497e-01 -8.00336525e-02 5.28869390e-01 -5.87835789e-01
5.19427478e-01 -2.50617653e-01 -7.29350567e-01 5.37953973e-01
-6.88540697e-01 1.62748516e-01 1.79803044e-01 1.04410267e+00
1.35816500e-01 -9.34035778e-02 1.22507080e-01 -6.38007045e-01
8.72274399e-01 -1.69219792e-01 -3.78732830e-01 5.77745810e-02
-6.77108407e-01 -3.74140263e-01 3.21430594e-01 -3.25342178e-01
-1.01981449e+00 -4.87835288e-01 -9.82677937e-02 2.45538145e-01
1.11353043e-02 -1.46784872e-01 6.47139177e-02 -5.24923325e-01
8.02199244e-01 9.25758183e-02 1.99174434e-02 -6.80815242e-03
3.01039815e-01 7.09525108e-01 -6.82967342e-03 -6.68678164e-01
8.44880998e-01 4.91470337e-01 7.98524171e-02 -9.57177758e-01
-1.52186140e-01 -2.60129690e-01 -9.51962709e-01 -6.54426932e-01
6.56643391e-01 -6.78531289e-01 -9.10833478e-01 3.91110867e-01
-9.38691139e-01 -3.38627815e-01 -3.91645581e-01 4.25431967e-01
-1.01278400e+00 2.75717527e-02 -3.90154392e-01 -1.27895141e+00
-5.10977268e-01 -1.02017939e+00 9.43384409e-01 5.30070923e-02
-5.10597289e-01 -1.26927447e+00 5.87685481e-02 2.71537274e-01
1.81734428e-01 -1.73075795e-01 6.90226793e-01 -2.38256708e-01
-4.24233019e-01 -1.53791070e-01 2.34436691e-02 -1.39755070e-01
1.70832314e-02 4.95917559e-01 -9.81751978e-01 -5.31145096e-01
6.65065646e-02 -1.92070693e-01 -1.08835101e-01 6.52520418e-01
5.91872573e-01 -2.29931593e-01 -8.56000841e-01 5.40386558e-01
1.38545322e+00 3.85070026e-01 5.32770038e-01 7.28214979e-01
1.41836226e-01 5.53460240e-01 9.17806149e-01 4.63203549e-01
1.30579369e-02 3.28798652e-01 2.40537539e-01 1.49327129e-01
1.11720070e-01 -1.54819340e-01 3.77893507e-01 1.16112018e+00
-8.18235934e-01 -2.69281328e-01 -5.07867396e-01 4.42987174e-01
-1.72482407e+00 -1.40330648e+00 -4.32368398e-01 6.90478683e-01
6.25676990e-01 1.56016424e-01 -1.48347050e-01 3.35214496e-01
4.99015123e-01 -2.03574806e-01 -1.19133167e-01 -1.06291151e+00
-1.43546045e-01 3.15233678e-01 7.37729073e-01 1.00061214e+00
-7.20721722e-01 1.03317809e+00 1.29781246e+01 1.02230716e+00
2.21112028e-01 1.03134915e-01 5.16071796e-01 3.48020852e-01
-4.36954498e-01 -4.56139445e-02 -1.04416132e+00 2.72933897e-02
1.38140702e+00 -4.30666685e-01 6.85999811e-01 5.44219851e-01
3.44648361e-01 -4.23268199e-01 -1.26188684e+00 5.26221812e-01
9.73738134e-02 -1.40886843e+00 -2.83300440e-04 6.85225725e-01
7.73699820e-01 -5.08050561e-01 6.22419357e-01 3.24184299e-01
6.09259963e-01 -1.14389277e+00 8.60300779e-01 2.53660440e-01
1.03040910e+00 -6.05088234e-01 5.67372203e-01 1.68872893e-01
-1.14389896e+00 -2.20873043e-01 -8.77727985e-01 -1.00755692e+00
3.93533185e-02 -1.81779593e-01 -4.29956943e-01 3.48861217e-01
9.58353162e-01 2.99398601e-01 -3.93658698e-01 9.95779395e-01
-4.78694476e-02 1.04875881e-02 -2.71853864e-01 -4.48467314e-01
4.83122796e-01 -3.54241252e-01 4.66730654e-01 1.00164843e+00
2.48499006e-01 3.51035744e-01 -9.84472036e-02 4.01770771e-01
5.45058846e-01 3.29446048e-02 -1.19659424e+00 -1.78908288e-01
2.83276141e-01 9.16795909e-01 -4.83487815e-01 -4.22520459e-01
-2.00212970e-01 8.62069130e-01 -3.55488248e-02 5.01107454e-01
-6.11489356e-01 -4.35615242e-01 9.72222984e-01 -1.27327025e-01
-1.14700586e-01 -3.48497719e-01 -6.23769283e-01 -7.30352640e-01
-5.89872956e-01 -4.54965204e-01 5.93606755e-02 -5.53365827e-01
-1.39813089e+00 5.79277515e-01 -2.27688253e-02 -1.40553558e+00
-6.99901402e-01 -1.27676582e+00 -4.76714373e-01 4.92853165e-01
-1.11898029e+00 -1.10984349e+00 2.50124663e-01 4.52870727e-01
1.64141744e-01 -5.34416080e-01 1.39563632e+00 3.57715860e-02
1.00637585e-01 9.24474537e-01 6.69434488e-01 -7.37814724e-01
5.56605101e-01 -1.27867436e+00 5.68737745e-01 -1.37897313e-01
-4.31265175e-01 9.05828118e-01 6.28349900e-01 -5.39804697e-01
-1.41196322e+00 -3.66917729e-01 1.08350635e+00 -9.83769417e-01
6.55218959e-01 -3.86345625e-01 4.23767231e-02 7.88592756e-01
7.15902448e-01 -6.18741751e-01 8.21781039e-01 -1.83753878e-01
1.80774391e-01 5.75296998e-01 -1.39248300e+00 6.12354755e-01
1.66275799e+00 -4.63594139e-01 -6.25784039e-01 7.60327101e-01
8.13696027e-01 -6.94087505e-01 -1.30082703e+00 3.34633321e-01
8.65424156e-01 -8.75409484e-01 1.61978090e+00 -1.32660246e+00
-4.43697497e-02 2.81152606e-01 -2.61993498e-01 -9.32519078e-01
-5.96193194e-01 -1.23518765e+00 -5.33532679e-01 -5.83747849e-02
5.96577883e-01 -1.13057327e+00 3.42365682e-01 8.74560475e-01
-2.82833427e-01 -6.42737269e-01 -1.06996536e+00 -1.32016802e+00
-3.35779637e-02 -1.45572275e-01 4.90409225e-01 7.63798356e-01
6.83744550e-01 1.09839931e-01 -6.36873543e-02 -1.00294888e-01
5.33176839e-01 9.55312885e-03 4.41501856e-01 -1.34294486e+00
3.86843324e-01 -5.75816095e-01 -3.07655483e-01 -9.45992947e-01
-8.85957032e-02 -8.12076271e-01 -6.53862000e-01 -1.28511906e+00
-8.32044985e-03 -1.91056758e-01 -1.11109078e-01 -1.65725678e-01
3.67937148e-01 2.13746816e-01 1.20859891e-02 1.03788137e-01
-3.71160030e-01 6.18435517e-02 1.29639816e+00 6.91750320e-05
-1.62315920e-01 4.85058486e-01 -4.81304944e-01 7.84440815e-01
8.58408585e-02 -2.99253196e-01 -6.78878546e-01 6.11881316e-02
6.69384480e-01 4.61409837e-02 3.24159935e-02 -7.42885649e-01
5.37211418e-01 -3.75702560e-01 4.78586555e-01 -1.32223868e+00
1.30741090e-01 -9.61415648e-01 6.95283338e-02 9.40189242e-01
2.74610907e-01 1.20424610e-02 8.66204947e-02 5.39500564e-02
-1.44871444e-01 -5.70943117e-01 9.21121240e-01 -4.07591403e-01
-4.92852688e-01 -5.20386267e-03 -1.03226590e+00 8.97834301e-02
9.92593169e-01 -7.84614205e-01 -3.59281451e-01 -4.20183957e-01
-8.29068601e-01 -1.95836127e-02 6.50830388e-01 3.03609259e-02
7.30431557e-01 -1.51530886e+00 -2.30721906e-01 7.18729138e-01
-3.22939813e-01 -3.74400020e-01 -1.70157343e-01 6.58265352e-01
-1.32361674e+00 1.02442718e+00 -5.41665435e-01 -4.55340147e-01
-1.14228773e+00 4.78126436e-01 4.28307921e-01 -2.41845414e-01
-2.15481281e-01 1.11954463e+00 2.71224789e-02 -8.23025763e-01
1.85185194e-01 -9.21545625e-02 -7.54407048e-01 3.55081353e-03
6.88606799e-01 1.05194807e+00 -2.91290224e-01 -6.04341030e-01
-4.56784427e-01 6.65885091e-01 2.32151806e-01 -2.87484169e-01
9.17833567e-01 -2.19243199e-01 -9.89108324e-01 4.28274393e-01
8.38715494e-01 -1.36269778e-01 -3.69319022e-02 4.16855574e-01
1.25943512e-01 -8.02164078e-01 -4.32406247e-01 -3.55811834e-01
-1.85641110e-01 5.54822803e-01 5.34874737e-01 8.99602413e-01
8.57008278e-01 -3.02566767e-01 8.18335712e-01 9.66778398e-01
5.72402716e-01 -1.68019545e+00 -2.13140488e-01 6.89524531e-01
9.12339568e-01 -9.28350806e-01 5.44190466e-01 -7.27165341e-01
-4.14997995e-01 1.32979155e+00 4.68304873e-01 -1.55325383e-01
1.27306652e+00 5.74917436e-01 1.14069022e-02 -3.70670199e-01
-9.44949508e-01 1.12705544e-01 3.75366658e-01 1.11147714e+00
5.06513238e-01 5.07374525e-01 -9.66500878e-01 3.21953118e-01
-7.28706717e-01 -2.34555230e-01 4.90474731e-01 1.41972518e+00
-6.43810987e-01 -1.20391917e+00 -7.23931909e-01 4.61561680e-01
-5.49773455e-01 -1.16372630e-01 -5.28106689e-01 8.46754074e-01
-5.76629937e-02 1.49448860e+00 -1.97535474e-03 -5.03491640e-01
4.11356747e-01 1.41089618e-01 7.21762300e-01 -1.23501487e-01
-9.04846430e-01 4.15413082e-01 3.84890139e-01 -1.23056793e+00
-8.58632207e-01 -1.05834293e+00 -1.40667629e+00 -1.19437599e+00
-5.12782812e-01 1.89310342e-01 3.83317530e-01 3.90289724e-01
-2.06836104e-01 2.85260603e-02 9.80917513e-01 -1.07949340e+00
-3.90341938e-01 -9.39418614e-01 -1.00026262e+00 -7.84516707e-02
2.89751232e-01 -8.00943017e-01 -7.83523321e-01 2.83909619e-01] | [-7.223554611206055, 3.64066743850708] |
37fc6a65-a18d-41fb-9922-77bc07653d4c | object-semantics-give-us-the-depth-we-need | 2304.12542 | null | https://arxiv.org/abs/2304.12542v1 | https://arxiv.org/pdf/2304.12542v1.pdf | Object Semantics Give Us the Depth We Need: Multi-task Approach to Aerial Depth Completion | Depth completion and object detection are two crucial tasks often used for aerial 3D mapping, path planning, and collision avoidance of Uncrewed Aerial Vehicles (UAVs). Common solutions include using measurements from a LiDAR sensor; however, the generated point cloud is often sparse and irregular and limits the system's capabilities in 3D rendering and safety-critical decision-making. To mitigate this challenge, information from other sensors on the UAV (viz., a camera used for object detection) is utilized to help the depth completion process generate denser 3D models. Performing both aerial depth completion and object detection tasks while fusing the data from the two sensors poses a challenge to resource efficiency. We address this challenge by proposing a novel approach to jointly execute the two tasks in a single pass. The proposed method is based on an encoder-focused multi-task learning model that exposes the two tasks to jointly learned features. We demonstrate how semantic expectations of the objects in the scene learned by the object detection pathway can boost the performance of the depth completion pathway while placing the missing depth values. Experimental results show that the proposed multi-task network outperforms its single-task counterpart, particularly when exposed to defective inputs. | ['Homayoun Najjaran', 'Iraj Mantegh', 'Miodrag Bolic', 'Fardad Dadboud', 'Sara Hatami Gazani'] | 2023-04-25 | null | null | null | null | ['depth-completion'] | ['computer-vision'] | [ 5.01616955e-01 -1.13360379e-02 8.18084851e-02 -2.18061790e-01
-5.61384678e-01 -6.09732330e-01 4.15830761e-01 2.41977319e-01
-4.35304344e-01 3.26123178e-01 -4.31907505e-01 -1.88045427e-01
-1.96452171e-01 -9.03573275e-01 -7.89363265e-01 -5.40840924e-01
-3.25654633e-02 5.68788826e-01 4.24661934e-01 -9.59803835e-02
1.08877748e-01 7.23314464e-01 -1.84211516e+00 1.98308364e-01
8.30987215e-01 1.22498631e+00 8.95924628e-01 6.77346528e-01
8.78057033e-02 4.90724862e-01 -4.35056210e-01 1.02514967e-01
7.76877642e-01 1.82038516e-01 -3.15388620e-01 4.62177336e-01
3.63088191e-01 -8.70221317e-01 -2.45591439e-02 1.07749736e+00
2.79408902e-01 2.51043171e-01 4.95549858e-01 -1.39289975e+00
-5.65128736e-02 -2.02111527e-01 -7.48792589e-01 -4.62012663e-02
3.07295233e-01 3.15276653e-01 6.83439493e-01 -9.75139678e-01
2.91971952e-01 1.18829119e+00 3.84168476e-01 3.45653415e-01
-8.73837531e-01 -5.57456374e-01 5.79340756e-01 -1.78881556e-01
-1.47144699e+00 -3.15949172e-01 7.20522404e-01 -6.20698035e-01
8.43545556e-01 -1.20213687e-01 5.66508591e-01 6.84113026e-01
1.71870559e-01 5.74037135e-01 7.25113392e-01 8.96763057e-03
3.16765040e-01 8.87836441e-02 -5.29761970e-01 9.66495395e-01
4.30678308e-01 3.71166587e-01 -4.64400232e-01 4.29118536e-02
9.10965919e-01 3.04702073e-01 -2.00054988e-01 -5.86930037e-01
-1.20857811e+00 5.08916318e-01 7.13001132e-01 -2.51460820e-01
-6.17374122e-01 7.57065788e-02 9.04335156e-02 2.63435572e-01
6.10955775e-01 2.59789199e-01 -4.35472369e-01 4.32530522e-01
-8.80145073e-01 4.13787872e-01 3.85250509e-01 1.37271774e+00
9.99172866e-01 3.41395549e-02 4.47427109e-02 2.20924675e-01
4.77158129e-01 6.00382447e-01 -1.07109308e-01 -6.89781368e-01
7.59912908e-01 8.94189775e-01 4.74981248e-01 -8.34356546e-01
-4.34637845e-01 -2.60699928e-01 -6.24591768e-01 7.27984130e-01
1.97107866e-01 -1.94709614e-01 -1.07997787e+00 1.42208564e+00
7.43941605e-01 2.87294447e-01 9.07561108e-02 1.27911615e+00
6.31127954e-01 7.17769265e-01 -6.29458427e-02 1.55018419e-01
1.04156518e+00 -7.46935487e-01 -3.15467507e-01 -7.22317934e-01
5.79807222e-01 -5.88038266e-01 6.17277503e-01 3.09779257e-01
-9.25168633e-01 -7.35273957e-01 -1.30062068e+00 -1.40021369e-01
-3.10607433e-01 4.75804895e-01 4.52342808e-01 1.16048574e-01
-8.41743886e-01 4.61626410e-01 -8.36708963e-01 -1.96109295e-01
5.83617926e-01 4.54081893e-01 -3.32740724e-01 -3.68349701e-01
-6.50262594e-01 9.17614877e-01 3.10409099e-01 2.20666528e-01
-1.53594720e+00 -6.33920848e-01 -9.87601519e-01 -1.60534814e-01
7.03041255e-01 -9.34321940e-01 1.08689654e+00 -4.44751501e-01
-1.08172739e+00 7.16170371e-01 -2.49396041e-02 -4.34511632e-01
5.24808466e-01 -4.71543252e-01 2.95215845e-01 1.20440133e-01
3.62425208e-01 1.01543164e+00 1.27539468e+00 -1.33578527e+00
-1.15511858e+00 -8.30141127e-01 3.49359721e-01 6.37840331e-01
8.83659720e-02 -4.34522450e-01 -3.23228151e-01 -2.93991208e-01
4.15904343e-01 -8.17963660e-01 -4.73237008e-01 6.84312165e-01
-3.02600116e-01 3.27958278e-02 1.07437432e+00 -3.05105954e-01
3.65639925e-01 -2.02506280e+00 3.52083385e-01 -4.47036400e-02
2.06542462e-01 1.58029348e-01 -2.80554622e-01 3.58231872e-01
2.46390715e-01 -2.27736086e-01 -1.78087741e-01 -6.92667425e-01
-3.78998727e-01 1.92936584e-01 -4.65952724e-01 5.16975641e-01
5.31936944e-01 5.52055538e-01 -9.27625299e-01 -2.46563807e-01
5.07216275e-01 3.78465980e-01 -4.04484093e-01 5.59249282e-01
-6.25168502e-01 7.78301299e-01 -7.27512777e-01 9.12557065e-01
8.55593324e-01 3.11932508e-02 -1.50867879e-01 -2.03586742e-01
-3.62387359e-01 1.18250713e-01 -1.24216235e+00 2.10127616e+00
-6.14013016e-01 2.22287744e-01 3.87722045e-01 -7.00966835e-01
9.26140189e-01 2.02865869e-01 4.50050175e-01 -3.41741115e-01
1.05034456e-01 9.24411342e-02 -2.63043433e-01 -3.81342560e-01
5.96053720e-01 -3.24062183e-02 -7.93743208e-02 1.97602421e-01
-6.76084757e-02 -8.15553546e-01 -1.93048120e-01 6.97005317e-02
1.02705073e+00 4.87560987e-01 2.31279477e-01 1.41118288e-01
2.74805665e-01 3.97985101e-01 6.56188846e-01 4.73866045e-01
-9.76927578e-02 3.16787928e-01 7.43536651e-02 -6.47685528e-01
-8.80976260e-01 -8.43785882e-01 1.66086793e-01 7.52919972e-01
5.60372055e-01 -1.36905730e-01 -3.75853807e-01 -6.84650004e-01
2.74868548e-01 4.78221357e-01 -4.70349938e-01 -1.92872703e-01
-2.06049189e-01 -2.86239862e-01 1.57784328e-01 5.65075457e-01
4.41665262e-01 -5.47663510e-01 -1.46918237e+00 2.36871332e-01
2.69393455e-02 -1.63093174e+00 -4.68975492e-02 4.64677483e-01
-1.03303528e+00 -1.16819727e+00 -4.64425653e-01 -3.79373044e-01
8.43175232e-01 9.78377640e-01 7.16638029e-01 5.88725917e-02
-4.74357605e-01 4.48244900e-01 -4.05547529e-01 -7.68197060e-01
-1.40998691e-01 -8.21596608e-02 2.02374354e-01 -1.08881138e-01
3.47778462e-02 -4.85020936e-01 -3.86428207e-01 3.73001724e-01
-8.72576356e-01 3.09750527e-01 8.09652388e-01 5.47618151e-01
8.17625642e-01 2.92548269e-01 2.14614332e-01 -4.11666304e-01
9.06224474e-02 -3.92634064e-01 -1.26387167e+00 -1.05829284e-01
-3.25293243e-01 -1.17891625e-01 4.22424555e-01 -2.38532037e-01
-7.51554012e-01 7.20381796e-01 2.80252665e-01 -1.01359880e+00
-3.22842360e-01 1.70377403e-01 -2.33864054e-01 -3.64889115e-01
3.42349589e-01 1.03544362e-01 -5.42708635e-02 -3.07064414e-01
-5.27334213e-03 4.15016115e-01 3.33318084e-01 -2.62986898e-01
1.08292580e+00 6.69632256e-01 4.72089857e-01 -8.38401973e-01
-9.95676041e-01 -5.44505894e-01 -8.33337009e-01 -3.71602982e-01
8.24754834e-01 -1.45795989e+00 -6.22257829e-01 2.99574792e-01
-1.54093623e+00 -2.93784678e-01 -2.96152383e-01 4.61922884e-01
-5.86228430e-01 7.91402161e-02 1.29574105e-01 -1.07773364e+00
-4.06882800e-02 -1.38211179e+00 1.74700046e+00 -2.40116008e-02
3.47004861e-01 -6.21480525e-01 -2.55315542e-01 1.71531707e-01
4.65298109e-02 3.56185794e-01 7.45618045e-01 -2.20399216e-01
-1.25226939e+00 -2.52955616e-01 -3.98405641e-01 1.42129466e-01
1.44831702e-01 -3.18477541e-01 -1.19315064e+00 -3.42510402e-01
7.00418651e-02 -4.07811373e-01 9.07978892e-01 3.45468640e-01
1.10406494e+00 5.61882667e-02 -5.39058208e-01 6.88220561e-01
1.42295980e+00 8.31494927e-02 1.98722873e-02 -3.16403806e-02
8.02829146e-01 9.41430092e-01 1.30705285e+00 8.31906021e-01
5.68971634e-01 6.21616960e-01 1.42833114e+00 -6.52114674e-02
-3.16863433e-02 -3.15346777e-01 2.23931909e-01 5.68314567e-02
-1.81581751e-02 -4.13385063e-01 -8.84791851e-01 4.41457659e-01
-1.81793141e+00 -4.88381892e-01 1.52717996e-02 2.30675602e+00
1.16957150e-01 6.95127770e-02 -3.03711891e-01 1.49613619e-01
4.83132809e-01 2.77204275e-01 -8.63595784e-01 1.79205671e-01
2.33539999e-01 -2.73299992e-01 6.38146758e-01 4.49239403e-01
-1.22638845e+00 1.05801022e+00 5.29198503e+00 2.33625889e-01
-8.62903535e-01 -9.78291407e-02 1.18378326e-02 -3.71169746e-01
-5.61008602e-02 -6.53738063e-03 -1.04771483e+00 8.59052017e-02
4.68443334e-01 1.05740428e-01 2.92367727e-01 9.81985509e-01
2.54451811e-01 -5.57475686e-01 -1.47003698e+00 1.09795892e+00
5.21432534e-02 -1.07929564e+00 2.25167617e-01 2.64232278e-01
5.58978260e-01 8.06912407e-02 7.84251466e-02 -3.73038389e-02
2.37412229e-01 -6.69026673e-01 8.79525781e-01 3.11328024e-01
8.82452071e-01 -6.84685171e-01 5.88701069e-01 8.61451268e-01
-1.50176322e+00 -4.05809462e-01 -6.09595954e-01 -2.32627958e-01
3.09600383e-01 7.50889421e-01 -1.13280833e+00 6.85950994e-01
6.16515636e-01 7.18872130e-01 -2.88383186e-01 9.77529526e-01
-2.39108667e-01 -2.41320938e-01 -3.14243853e-01 2.96934605e-01
4.20808524e-01 -1.77461982e-01 9.31459963e-01 4.57968652e-01
5.50694942e-01 1.74173594e-01 7.40097165e-01 9.12218094e-01
4.77643982e-02 -4.36903834e-01 -1.25135458e+00 9.38284546e-02
4.25845861e-01 1.25371754e+00 -6.14741385e-01 -1.18860923e-01
-4.01339412e-01 1.04933047e+00 2.21290737e-01 1.60035849e-01
-7.50389397e-01 -1.48701057e-01 9.96254563e-01 2.74982482e-01
2.58250296e-01 -6.32101238e-01 -1.88000113e-01 -8.02301347e-01
1.43821448e-01 -3.05773616e-01 4.21941765e-02 -9.21941996e-01
-7.71792173e-01 6.48539126e-01 7.54518881e-02 -1.52674699e+00
-1.73880041e-01 -7.34271646e-01 -3.32633138e-01 9.83502805e-01
-2.00257421e+00 -1.41351366e+00 -9.29594040e-01 5.36586285e-01
9.39092338e-01 -4.29886095e-02 6.62687957e-01 1.18020035e-01
-3.01903069e-01 -1.78769514e-01 -7.85380781e-01 -3.97609562e-01
4.18954462e-01 -9.95256305e-01 4.62630898e-01 1.00466835e+00
-1.84400268e-02 -9.93318632e-02 4.19007272e-01 -9.66677010e-01
-1.69411123e+00 -1.50633073e+00 4.33798462e-01 -3.53081763e-01
1.24964625e-01 -4.37477797e-01 -5.50536990e-01 5.80608487e-01
-3.25153202e-01 1.24618135e-01 2.31980398e-01 -4.36828583e-01
1.98328514e-02 -7.64721259e-02 -1.18767524e+00 2.26688445e-01
1.19203770e+00 -3.39774072e-01 -2.45305225e-01 4.90701020e-01
9.11171019e-01 -7.00757086e-01 -4.24434185e-01 6.79949403e-01
2.37463519e-01 -9.61527526e-01 9.68286097e-01 -3.31289619e-01
2.64883369e-01 -6.07640505e-01 -4.20503736e-01 -1.29980755e+00
-8.32423568e-02 -3.38317931e-01 -1.21406034e-01 6.33735836e-01
1.09927863e-01 -3.30887914e-01 7.56866455e-01 2.73061514e-01
-5.81839681e-01 -5.09048164e-01 -9.16253328e-01 -6.48129523e-01
-5.42016447e-01 -5.86250305e-01 7.03732014e-01 4.53957140e-01
-7.91196048e-01 2.71570385e-01 -2.77803212e-01 9.98176813e-01
8.00830781e-01 3.81052494e-01 1.01425350e+00 -1.56270885e+00
8.50282535e-02 5.99967167e-02 -3.71580839e-01 -1.36231744e+00
6.59125522e-02 -7.64502108e-01 3.65286648e-01 -1.62195909e+00
-3.67347598e-01 -7.37210631e-01 4.23559099e-01 4.26860720e-01
1.33669004e-01 -2.85270084e-02 1.98323175e-01 1.51791185e-01
-4.51645941e-01 7.52365470e-01 1.34110785e+00 -9.04550180e-02
-3.21506262e-01 2.89114028e-01 -3.48004431e-01 9.56750393e-01
4.37806427e-01 -5.23694575e-01 -7.48587251e-01 -1.11893260e+00
1.49918512e-01 3.18192065e-01 7.59174466e-01 -1.33567202e+00
5.96222937e-01 -2.17776939e-01 5.68897963e-01 -1.07766342e+00
9.30210829e-01 -1.31003964e+00 -2.37713709e-01 3.89771461e-01
1.10080317e-01 1.43089831e-01 3.39662582e-01 8.64647031e-01
-4.15925980e-02 -1.17811047e-01 6.14010632e-01 -3.66946131e-01
-8.47158492e-01 7.61399329e-01 -3.19748372e-01 -3.36894423e-01
1.46174693e+00 -3.40353966e-01 1.41443074e-01 -9.34149250e-02
-5.22276044e-01 5.36722600e-01 4.63550597e-01 4.57760215e-01
1.20912170e+00 -9.18997288e-01 -6.45575345e-01 7.56460369e-01
3.25108618e-01 9.55629468e-01 1.75702512e-01 6.36298299e-01
-2.23254085e-01 4.59760100e-01 -2.03778863e-01 -9.10918117e-01
-1.00865507e+00 5.20474494e-01 1.45397097e-01 1.98010020e-02
-5.83256900e-01 8.77619088e-01 4.53734249e-01 -3.95093232e-01
3.90663087e-01 -6.61857963e-01 1.47872325e-02 6.84756637e-02
4.64963108e-01 3.37744981e-01 2.30080664e-01 -3.81641299e-01
-2.68823028e-01 7.16048837e-01 -6.93741534e-03 -4.14470024e-02
1.29364872e+00 -2.79333204e-01 2.34298229e-01 3.05013448e-01
6.00372553e-01 -4.87791300e-01 -1.72379577e+00 -1.94776475e-01
-3.46683592e-01 -8.19822013e-01 4.44778800e-01 -4.74314451e-01
-8.19789410e-01 1.16705155e+00 4.10994291e-01 -8.93497467e-02
1.14415467e+00 -1.66518122e-01 6.28391802e-01 5.48884630e-01
9.85723674e-01 -8.11409593e-01 2.30257109e-01 4.31411058e-01
9.49359536e-01 -1.47773051e+00 2.35911813e-02 -6.39232516e-01
-6.51837945e-01 9.17048156e-01 9.33351696e-01 -6.64642900e-02
3.96125704e-01 3.31095934e-01 -2.30563611e-01 -2.89809585e-01
-8.79206777e-01 -5.28087914e-01 1.27386406e-01 8.34336221e-01
-3.86522382e-01 -1.92025393e-01 7.29870021e-01 3.01034153e-01
2.90796697e-01 -7.04722181e-02 3.84654582e-01 1.16112339e+00
-6.67531431e-01 -8.31910193e-01 -3.16785306e-01 3.60329688e-01
9.02558267e-02 9.37335938e-02 -3.76516223e-01 6.20602608e-01
3.66016686e-01 8.10517192e-01 1.97840542e-01 -5.21232963e-01
4.88931060e-01 -2.34251946e-01 4.81349170e-01 -1.12161112e+00
-2.73517162e-01 -1.17718436e-01 -9.74076539e-02 -8.74552548e-01
-3.55203122e-01 -4.79267657e-01 -1.09724963e+00 1.51987121e-01
-4.09109682e-01 -2.89901406e-01 1.04200697e+00 8.71726394e-01
4.87859607e-01 6.35531843e-01 7.85895348e-01 -1.41874182e+00
-6.56576753e-01 -6.22222006e-01 -4.63452041e-01 -2.53721565e-01
7.50070870e-01 -1.00139570e+00 -1.37153998e-01 -6.06670007e-02] | [7.8760986328125, -2.575617551803589] |
41cab8b3-33e0-48ad-9160-64f508eea672 | using-statistical-and-semantic-models-for | 1805.04579 | null | http://arxiv.org/abs/1805.04579v2 | http://arxiv.org/pdf/1805.04579v2.pdf | Using Statistical and Semantic Models for Multi-Document Summarization | We report a series of experiments with different semantic models on top of
various statistical models for extractive text summarization. Though
statistical models may better capture word co-occurrences and distribution
around the text, they fail to detect the context and the sense of sentences
/words as a whole. Semantic models help us gain better insight into the context
of sentences. We show that how tuning weights between different models can help
us achieve significant results on various benchmarks. Learning pre-trained
vectors used in semantic models further, on given corpus, can give addition
spike in performance. Using weighing techniques in between different
statistical models too further refines our result. For Statistical models, we
have used TF/IDF, TextRAnk, Jaccard/Cosine Similarities. For Semantic Models,
we have used WordNet-based Model and proposed two models based on Glove Vectors
and Facebook's InferSent. We tested our approach on DUC 2004 dataset,
generating 100-word summaries. We have discussed the system, algorithms,
analysis and also proposed and tested possible improvements. ROUGE scores were
used to compare to other summarizers. | ['Anukarsh Singh', 'Divyanshu Daiya', 'Mukesh Jadon'] | 2018-05-11 | using-statistical-and-semantic-models-for-2 | https://aclanthology.org/O18-1018 | https://aclanthology.org/O18-1018.pdf | roclingijclclp-2018-10 | ['extractive-document-summarization'] | ['natural-language-processing'] | [ 2.65391767e-01 1.11786254e-01 1.63719233e-03 -4.99800384e-01
-7.27336824e-01 -5.65396369e-01 9.63130951e-01 9.77090120e-01
-5.96575499e-01 1.12198925e+00 1.45401859e+00 7.94016868e-02
-2.14570850e-01 -7.66907573e-01 -2.69313276e-01 -2.89141268e-01
2.76862495e-02 5.68811357e-01 5.69600642e-01 -7.12756872e-01
1.29715466e+00 -2.38251258e-02 -1.75625253e+00 7.13849366e-01
9.67012584e-01 3.06125343e-01 6.66107178e-01 1.14500415e+00
-8.18721175e-01 8.29398811e-01 -1.23584449e+00 -1.55633455e-02
-3.15465003e-01 -5.01262009e-01 -1.07347584e+00 -1.51221424e-01
4.99537289e-01 1.71975523e-01 -5.30526191e-02 9.45943058e-01
4.66263920e-01 4.28762406e-01 1.11232674e+00 -7.23385215e-01
-4.68526214e-01 1.01992357e+00 -2.26369470e-01 4.20879871e-01
8.98636281e-01 -4.93209362e-01 1.26324105e+00 -6.03520870e-01
7.41583347e-01 1.49120545e+00 5.27593374e-01 4.45905030e-01
-8.01995695e-01 -7.10969865e-02 2.37737540e-02 2.04888090e-01
-8.30006897e-01 -2.66124636e-01 4.51714784e-01 -1.29367754e-01
1.58985925e+00 6.89447343e-01 3.11130166e-01 8.59669089e-01
1.29215226e-01 7.05946147e-01 9.07709718e-01 -7.47358084e-01
2.28785723e-01 3.09245110e-01 8.31990182e-01 4.12163794e-01
5.16465425e-01 -7.46805608e-01 -6.62894011e-01 -3.33063215e-01
-4.45365673e-03 -2.01518610e-01 -1.96280554e-01 1.42822236e-01
-1.07459342e+00 9.41584229e-01 -2.38971077e-02 8.40422630e-01
-4.42175269e-01 5.73651344e-02 8.37871432e-01 1.59837142e-01
6.02877021e-01 8.35065305e-01 -6.09523833e-01 -3.15142602e-01
-1.41319644e+00 5.33659637e-01 1.09828830e+00 8.86373460e-01
5.68925142e-01 -1.48010328e-01 -5.47064483e-01 1.11877465e+00
3.84011343e-02 4.08163637e-01 1.10868216e+00 -6.89394176e-01
5.02064168e-01 6.85590982e-01 1.62581429e-01 -1.17862368e+00
-4.28188354e-01 -2.87479639e-01 -3.98143679e-01 -4.39879775e-01
-1.34808064e-01 -1.77671015e-01 -1.14105415e+00 1.31058371e+00
-2.78833956e-01 2.05485150e-02 5.09259820e-01 5.81355572e-01
1.07030451e+00 9.27257419e-01 8.50204900e-02 -3.38537514e-01
1.49975562e+00 -9.46075559e-01 -9.77169871e-01 -2.06105292e-01
7.96383917e-01 -1.16353559e+00 9.58240509e-01 2.78153062e-01
-9.62109625e-01 -4.21694994e-01 -1.31095946e+00 -1.38263151e-01
-8.37872446e-01 2.88042694e-01 4.53156441e-01 5.77150285e-01
-1.22264159e+00 1.11281097e+00 -5.69278002e-01 -1.25846565e+00
9.38999802e-02 9.95146930e-02 -4.56200764e-02 2.47583315e-01
-1.14599156e+00 1.05514264e+00 6.56645596e-01 -6.11434519e-01
-5.03066838e-01 -1.56699330e-01 -7.50584543e-01 2.08525777e-01
2.95054108e-01 -9.05730605e-01 1.03107274e+00 -6.09350145e-01
-1.11346221e+00 6.00096941e-01 -6.30481899e-01 -7.72638500e-01
-2.12020595e-02 -3.69494826e-01 -3.51184815e-01 2.31629834e-01
3.54115784e-01 5.12188256e-01 4.21019942e-01 -1.14019334e+00
-7.47563541e-01 -2.28875041e-01 -1.77428558e-01 4.09831822e-01
-3.37391406e-01 1.72504678e-01 1.24275550e-01 -7.04917550e-01
-9.37598422e-02 -4.28883940e-01 -1.36997744e-01 -1.06470442e+00
-4.34075326e-01 -5.03038108e-01 5.91449320e-01 -8.46985161e-01
1.72077203e+00 -1.51382101e+00 1.95779260e-02 2.06001122e-02
-8.32439810e-02 4.08006847e-01 -7.06679597e-02 1.12213218e+00
2.46773973e-01 3.09260577e-01 -3.10235888e-01 -3.79662573e-01
-1.25285517e-02 3.46694261e-01 -4.84402716e-01 -1.44114852e-01
-8.21650028e-02 5.35841584e-01 -8.87514830e-01 -7.08109379e-01
2.22111866e-01 1.39418557e-01 -4.39790130e-01 2.00322811e-02
-3.01905185e-01 -5.41619837e-01 -6.27215981e-01 1.62849277e-01
3.26704919e-01 1.26400948e-01 -6.86225519e-02 -9.78029296e-02
9.79820192e-02 7.03022242e-01 -1.01576364e+00 1.95967340e+00
-5.83110809e-01 7.63103724e-01 -4.55228984e-01 -1.34914792e+00
1.10688877e+00 3.27194005e-01 -5.40256947e-02 -3.67679656e-01
1.96448416e-01 2.82400757e-01 -2.87804961e-01 -7.52709270e-01
1.43163955e+00 -1.45060152e-01 -2.22488549e-02 3.49189848e-01
4.04778838e-01 -2.99874961e-01 8.03487420e-01 7.31059492e-01
9.90462482e-01 -9.75369364e-02 5.41062832e-01 -7.62761593e-01
7.09954143e-01 4.34674680e-01 -2.19179522e-02 1.04360914e+00
1.36520684e-01 9.02666450e-01 5.88542700e-01 -8.72796550e-02
-9.69794214e-01 -7.73268759e-01 1.94413126e-01 1.23189139e+00
2.55100466e-02 -1.03340340e+00 -9.83620524e-01 -7.74793029e-01
-1.81091830e-01 1.37228513e+00 -6.34737372e-01 -4.58870316e-03
-5.19306362e-01 -9.29109454e-01 4.67869878e-01 3.09282988e-01
1.24467619e-01 -1.06298828e+00 -4.30562079e-01 4.40299124e-01
-2.27155223e-01 -7.65513420e-01 -3.79025847e-01 1.34117186e-01
-1.23897588e+00 -9.10097957e-01 -6.68092847e-01 -7.36989021e-01
3.13880056e-01 3.06243956e-01 1.28580737e+00 6.63233995e-02
-2.06967920e-01 4.47835177e-01 -8.50153863e-01 -8.04418445e-01
-5.51937938e-01 3.73877555e-01 -8.64950120e-02 -6.19111478e-01
6.17755175e-01 -5.19970238e-01 -5.40754199e-01 -2.48156995e-01
-9.92798388e-01 -2.39368320e-01 3.27483088e-01 7.22864211e-01
-8.78243446e-02 -1.78506687e-01 8.11094403e-01 -1.10346174e+00
1.42478180e+00 -4.01101500e-01 2.99157649e-01 3.97862077e-01
-6.23312294e-01 6.80695951e-01 6.48401856e-01 1.60903782e-02
-9.11600888e-01 -6.56933665e-01 -2.03325778e-01 4.90058959e-01
-1.68480992e-01 4.82816219e-01 2.70190805e-01 7.06843317e-01
9.83726144e-01 4.40262407e-01 -3.23845074e-02 -6.41066909e-01
4.46432114e-01 1.03043675e+00 3.01236123e-01 -4.46672142e-01
2.15133384e-01 3.44805419e-01 -4.00360465e-01 -1.22740185e+00
-9.45930958e-01 -1.01843441e+00 -3.94560605e-01 1.44441754e-01
8.33979130e-01 -7.23815501e-01 -2.26339512e-02 -4.49239686e-02
-1.31222498e+00 4.21150714e-01 -3.09841096e-01 3.56369913e-01
-5.06588757e-01 6.49162769e-01 -1.45030722e-01 -7.80999720e-01
-8.27165842e-01 -6.38265729e-01 1.10247386e+00 3.74341786e-01
-8.79062295e-01 -1.45343184e+00 2.38813549e-01 3.07801664e-01
7.35611439e-01 3.10024135e-02 8.20391059e-01 -1.43485725e+00
1.80469096e-01 -1.93854406e-01 -7.30664209e-02 5.81167161e-01
3.94850105e-01 -4.91860770e-02 -9.18666720e-01 5.40594161e-02
-8.40574577e-02 -8.36751089e-02 1.42588961e+00 5.83852828e-01
8.29315305e-01 -6.44412816e-01 -4.57215399e-01 -2.60449350e-01
1.54034424e+00 -3.61603103e-03 6.81370914e-01 4.14867103e-01
4.05748338e-01 7.07880914e-01 8.03807616e-01 4.74885613e-01
2.29047269e-01 1.11226350e-01 -7.90374205e-02 4.57193792e-01
-3.05256218e-01 -4.11984175e-01 3.28467369e-01 1.25136518e+00
-5.99278137e-02 -5.30108154e-01 -5.57498276e-01 7.87345290e-01
-1.79650211e+00 -1.09868515e+00 -1.39348626e-01 1.92285264e+00
7.88347960e-01 4.04066890e-01 1.46729693e-01 9.73186120e-02
6.54772043e-01 5.97186625e-01 2.12118223e-01 -9.50380266e-01
-2.49712080e-01 4.32144135e-01 6.20534241e-01 9.38136518e-01
-7.00919569e-01 1.14453864e+00 6.29077482e+00 1.25537968e+00
-9.33238924e-01 -4.54557166e-02 2.05218837e-01 5.01383357e-02
-6.18559599e-01 2.28732333e-01 -9.01622593e-01 3.62812459e-01
1.03119433e+00 -6.32530153e-01 6.24840595e-02 7.28232682e-01
4.14280057e-01 -6.12414002e-01 -8.29788983e-01 6.89625025e-01
8.23451519e-01 -1.60995996e+00 6.06752336e-01 -4.52221841e-01
7.98314273e-01 6.11072108e-02 -6.12223208e-01 2.91339934e-01
2.93289244e-01 -1.00063503e+00 4.30825442e-01 5.87160647e-01
-7.07175815e-03 -7.45810270e-01 1.19582176e+00 3.59299958e-01
-6.62168682e-01 3.34036410e-01 -8.46392453e-01 -2.43180558e-01
-6.53723776e-02 3.98021638e-01 -1.20578611e+00 8.19054484e-01
1.81360126e-01 8.59675288e-01 -9.01692748e-01 7.73784816e-01
9.93978903e-02 5.49308062e-01 -1.14085205e-01 -1.00943244e+00
4.92187649e-01 -3.95644531e-02 9.23393846e-01 1.76884139e+00
4.37832564e-01 -8.89690444e-02 -1.38056815e-01 3.94995987e-01
2.17400596e-01 6.20868385e-01 -5.32223105e-01 -9.48110521e-02
4.51915592e-01 1.02433276e+00 -8.50979924e-01 -9.79275405e-01
5.41666523e-02 9.67577517e-01 -1.16522662e-01 2.32268587e-01
-3.25066417e-01 -9.52760994e-01 5.13736069e-01 8.63682404e-02
1.67498365e-01 -1.13611296e-01 -4.42857742e-01 -1.22965896e+00
-1.31673053e-01 -6.66997671e-01 4.58782345e-01 -8.73874247e-01
-9.24775839e-01 6.75466657e-01 3.16702127e-01 -7.73533404e-01
-2.33974829e-01 -4.92141277e-01 -9.44134772e-01 8.13933909e-01
-1.17009270e+00 -5.86113811e-01 -5.20731546e-02 1.42065259e-02
1.28289473e+00 -4.46316987e-01 8.52904737e-01 -2.72244304e-01
-4.53839079e-02 1.73939839e-02 2.86309898e-01 -2.18977660e-01
8.31995487e-01 -1.69712067e+00 3.45643669e-01 7.11400926e-01
4.26325589e-01 8.67390573e-01 1.56859112e+00 -7.29075134e-01
-8.47810030e-01 -8.39189529e-01 1.45137560e+00 -5.80743551e-01
6.21676385e-01 -7.94344097e-02 -1.01584876e+00 2.96483576e-01
8.57713819e-01 -9.89512324e-01 5.43807387e-01 2.36782268e-01
3.98136750e-02 7.61312768e-02 -9.96208787e-01 5.42961717e-01
7.83156037e-01 -2.15959176e-01 -1.54966402e+00 5.56737721e-01
8.95603418e-01 1.39838845e-01 -3.45256865e-01 1.91377406e-03
1.57440767e-01 -1.11842358e+00 8.03437769e-01 -7.49398053e-01
6.45780325e-01 -1.48415297e-01 -2.44791016e-01 -1.84278846e+00
8.40215012e-02 -4.39969927e-01 2.66553968e-01 1.53105259e+00
7.51656353e-01 -4.75684047e-01 6.74376190e-01 1.31220266e-01
-1.98852316e-01 -3.83126050e-01 -4.16539162e-01 -7.06266105e-01
1.51856899e-01 -3.56733769e-01 4.14635509e-01 6.21646643e-01
3.71646166e-01 8.25079024e-01 -1.62428603e-01 -2.82543033e-01
3.07362169e-01 -2.37691239e-01 5.47879815e-01 -1.00597882e+00
3.13638486e-02 -6.18533611e-01 -5.59632301e-01 -8.75460267e-01
6.75175935e-02 -9.18327153e-01 -1.54508948e-01 -2.20055723e+00
3.79369825e-01 1.78071961e-01 -2.82632411e-01 5.73295057e-02
-4.06021297e-01 -1.86898246e-01 2.39281962e-03 -1.74110517e-01
-7.20409274e-01 4.24299270e-01 8.18865001e-01 -1.45161852e-01
-1.28653839e-01 -2.77307898e-01 -9.13618803e-01 6.58253849e-01
9.30739880e-01 -6.49907589e-01 -5.11042714e-01 -2.88962454e-01
1.89117804e-01 -2.46993318e-01 -1.52112255e-02 -1.04336429e+00
2.86599189e-01 -3.26412432e-02 -8.89708661e-03 -7.67664969e-01
7.60405809e-02 -3.24407607e-01 -5.20949721e-01 3.98209661e-01
-7.74244905e-01 1.50638297e-01 1.19107114e-02 5.25440872e-01
-5.08875966e-01 -9.14852142e-01 3.90797257e-01 -4.62103039e-01
-6.36048734e-01 -5.33358634e-01 -5.24828017e-01 3.29919249e-01
4.26175147e-01 -3.34205538e-01 -2.09279269e-01 -5.53541064e-01
-6.33580804e-01 2.60290921e-01 3.13444585e-01 6.09221876e-01
5.35213411e-01 -8.58469546e-01 -9.61296201e-01 -3.58066469e-01
1.52656749e-01 -4.91419941e-01 8.91539603e-02 3.99281502e-01
-8.91748488e-01 8.09557199e-01 -5.10549583e-02 -3.02229881e-01
-1.72446644e+00 3.59878600e-01 -1.74345627e-01 -4.37499434e-01
-3.72151136e-01 5.72249711e-01 -2.27146521e-01 -2.29113668e-01
2.15786129e-01 -7.24164307e-01 -9.15582836e-01 5.52166581e-01
7.68763006e-01 7.20709741e-01 1.30475774e-01 -4.24640775e-01
-2.51635879e-01 6.47874773e-01 -2.68490881e-01 -4.03726578e-01
1.36462653e+00 -3.97534966e-01 -2.95250535e-01 6.77895367e-01
1.37517476e+00 1.98526740e-01 -2.23534822e-01 -1.13914840e-01
7.12620914e-01 -3.08991790e-01 1.34551860e-02 -8.03451657e-01
-2.67355710e-01 8.73852372e-01 2.02135473e-01 6.37456298e-01
9.08919692e-01 2.77682138e-03 8.30978572e-01 6.02798998e-01
1.28204832e-02 -1.56381238e+00 9.10754874e-03 7.56018758e-01
8.66577268e-01 -1.18738675e+00 3.23700637e-01 -2.18172893e-01
-9.10966754e-01 1.37629747e+00 1.06910005e-01 -5.33516765e-01
2.87152767e-01 -5.39064109e-02 -5.81371831e-03 -2.39666179e-01
-9.25914645e-01 -2.70860791e-01 4.13255095e-01 5.20371974e-01
8.20685327e-01 1.33189023e-01 -1.07390940e+00 5.49421728e-01
-6.90693617e-01 -2.04260051e-01 9.41597581e-01 8.50827575e-01
-1.39243484e+00 -1.14011514e+00 -2.28564635e-01 7.46639252e-01
-7.54547417e-01 -3.89075667e-01 -6.50783539e-01 6.57132447e-01
-3.38460505e-01 1.27557433e+00 -4.71556187e-02 -2.18638808e-01
3.40062261e-01 4.11051422e-01 3.40139538e-01 -1.02503562e+00
-6.66340947e-01 -8.96743685e-02 7.34199643e-01 -2.28975147e-01
-4.63373750e-01 -6.87740266e-01 -1.16188192e+00 -8.89483746e-03
-2.86146045e-01 8.50700021e-01 9.22860801e-01 1.03842402e+00
2.20287830e-01 6.04870737e-01 3.91953766e-01 -4.10756379e-01
-6.60168827e-01 -1.70065689e+00 -5.37610888e-01 5.44159770e-01
1.88058168e-01 -1.93364784e-01 -5.45008838e-01 2.01836705e-01] | [12.380186080932617, 9.495598793029785] |
043a10f3-bb6f-495a-814a-e593a4fc72f8 | effective-handwritten-digit-recognition-using | null | null | https://www.researchgate.net/publication/341068954_Effective_Handwritten_Digit_Recognition_using_Deep_Convolution_Neural_Network | http://www.warse.org/IJATCSE/static/pdf/file/ijatcse66922020.pdf | Effective Handwritten Digit Recognition using Deep Convolution Neural Network | This paper proposed a simple neural network approach towards handwritten digit recognition using convolution. With machine learning algorithms like KNN, SVM/SOM, recognizing digits is considered as one of the unsolvable tasks due to its distinctiveness in the style of writing. In this paper, Convolution Neural Networks are implemented with an MNIST dataset of 70000 digits with 250 distinct forms of writings. The proposed method achieved 98.51% accuracy for real-world handwritten digit prediction with less than 0.1 % loss on training with 60000 digits while 10000 under validation. | ['Sriram V.P', 'Yellapragada SS Bharadwaj', 'Rajaram P', 'Sudhakar S', 'Kolla Bhanu Prakash'] | 2020-04-30 | null | null | null | international-journal-of-advanced-trends-in | ['handwritten-digit-recognition'] | ['computer-vision'] | [-1.36172310e-01 -2.89029002e-01 -2.88573429e-02 -4.04419780e-01
5.23409069e-01 -7.42360115e-01 6.85955882e-01 -7.88815841e-02
-6.30416751e-01 1.08752179e+00 -1.30346432e-01 -4.66884702e-01
-2.98253655e-01 -7.88638175e-01 -2.86108673e-01 -4.14652020e-01
2.49950305e-01 3.52024913e-01 -5.18368669e-02 -1.15726940e-01
7.89680660e-01 1.06040895e+00 -1.26530027e+00 4.87468302e-01
3.96073103e-01 1.46110570e+00 -1.48060411e-01 9.44300711e-01
-2.32845694e-01 1.35143304e+00 -8.96215677e-01 -5.95230758e-01
5.35903990e-01 -1.14636041e-01 -5.58392584e-01 9.20900777e-02
4.55467939e-01 -4.55285251e-01 -8.86636198e-01 9.22721803e-01
3.78340453e-01 7.85654262e-02 1.08503902e+00 -8.15404356e-01
-1.23792040e+00 5.75784624e-01 -2.96072572e-01 4.19097096e-01
-1.51615739e-01 -7.57402182e-02 3.63246411e-01 -1.06171834e+00
3.72188777e-01 8.85606050e-01 8.25494945e-01 5.97185016e-01
-8.66026759e-01 -5.17775178e-01 -5.28421581e-01 1.56590343e-01
-1.09085178e+00 -1.38562888e-01 6.85572743e-01 -4.44726616e-01
1.06811750e+00 2.57034779e-01 1.98826566e-01 8.81453514e-01
3.20818186e-01 4.71570432e-01 1.33567059e+00 -5.94719648e-01
3.09727341e-01 4.51741785e-01 7.27702141e-01 4.79103088e-01
4.67406154e-01 2.24939431e-03 -3.29813749e-01 2.09677085e-01
1.05803275e+00 1.07919514e-01 1.06086440e-01 4.29520756e-01
-9.87082958e-01 5.69877028e-01 5.50285697e-01 5.25921047e-01
-2.61866301e-01 6.19176850e-02 6.26694858e-01 5.66949546e-01
-2.85939947e-02 4.45283562e-01 -2.84072608e-01 -2.72923857e-01
-9.10910726e-01 -6.36335611e-02 8.86658907e-01 7.88878739e-01
1.79243967e-01 6.33810401e-01 -5.27237840e-02 1.02879810e+00
-1.01998799e-01 2.77604789e-01 9.65123236e-01 -4.06990945e-01
4.87278700e-01 7.32190609e-01 -3.78968529e-02 -1.23863125e+00
-1.49053991e-01 -4.70795512e-01 -1.48853016e+00 7.64911115e-01
9.15552437e-01 -2.54184514e-01 -1.25574768e+00 7.99268126e-01
-5.95568836e-01 -2.12076455e-01 3.57749075e-01 8.04008543e-01
8.84287953e-01 6.82696402e-01 -2.38711879e-01 3.11469346e-01
1.08765697e+00 -9.44852054e-01 -6.01330459e-01 -2.47028440e-01
-1.86835721e-01 -9.18440044e-01 6.33899868e-01 1.05720663e+00
-8.55768204e-01 -9.77641642e-01 -1.20504761e+00 -2.00632792e-02
-5.79152048e-01 9.39672172e-01 7.24399507e-01 7.81699419e-01
-6.91842258e-01 8.42717707e-01 -3.67949039e-01 -4.21455443e-01
7.50151634e-01 3.43724757e-01 -3.27688783e-01 7.67724309e-03
-7.39210665e-01 8.99959028e-01 1.06792748e+00 1.48801744e-01
-5.06618142e-01 -1.90546468e-01 -1.43821677e-02 1.24246724e-01
-3.27142805e-01 1.40445769e-01 9.81477559e-01 -1.34038198e+00
-1.55126750e+00 7.14288235e-01 2.59964347e-01 -7.67484248e-01
8.99460912e-01 9.04728174e-02 -8.90455067e-01 -8.29435289e-02
-6.12467945e-01 3.20261627e-01 8.49500656e-01 -7.67885804e-01
-4.48862493e-01 -2.75837183e-01 -3.23165804e-01 -2.99677759e-01
-8.26587141e-01 9.97015089e-02 2.15890765e-01 -1.10709012e+00
2.25970045e-01 -4.17315871e-01 2.03309968e-01 2.78453857e-01
-3.01186889e-01 -1.74135417e-01 9.46982265e-01 -9.17906761e-01
1.06512785e+00 -1.95898163e+00 -3.41814131e-01 3.35461766e-01
1.17388360e-01 6.27990782e-01 1.15769967e-01 3.66494000e-01
1.41030475e-02 -3.85745704e-01 8.22599158e-02 1.00623474e-01
-1.89856097e-01 3.00889850e-01 -5.42363644e-01 1.29479900e-01
3.37422907e-01 7.18163133e-01 -4.27743345e-01 -3.83175053e-02
1.68516979e-01 3.93315256e-01 1.21467866e-01 4.83136140e-02
2.50432710e-03 -1.79144427e-01 -1.34368584e-01 9.76935506e-01
7.46523142e-01 -8.61193910e-02 2.91315522e-02 -1.60359159e-01
-9.78482962e-02 -3.99998337e-01 -1.30861855e+00 9.96728003e-01
-1.61399484e-01 1.21618414e+00 -5.25664926e-01 -1.12178850e+00
1.67609346e+00 1.59163609e-01 -4.40190792e-01 -5.03609955e-01
3.49519461e-01 5.51433921e-01 2.91093379e-01 -3.46684396e-01
5.25980473e-01 -2.97291726e-02 2.16877952e-01 3.44242334e-01
3.01962674e-01 3.13783795e-01 1.50576383e-01 -3.18154842e-01
9.20140803e-01 -2.97731280e-01 3.58170688e-01 -2.87895799e-01
4.54603344e-01 1.15035482e-01 5.92008792e-02 9.05501187e-01
-1.29674152e-01 6.02282405e-01 2.24556416e-01 -1.14008224e+00
-1.31191826e+00 -8.09903324e-01 -4.45621073e-01 5.75027227e-01
-5.23332775e-01 5.03857315e-01 -5.12839079e-01 -2.31591552e-01
3.02535176e-01 4.85231608e-01 -7.43947566e-01 8.09000134e-02
-5.09143353e-01 -6.37994409e-01 1.24528193e+00 9.99148250e-01
1.13810933e+00 -1.33181477e+00 -6.43872619e-01 2.89197296e-01
7.99432278e-01 -8.17004263e-01 1.66757494e-01 5.80250621e-01
-1.00693846e+00 -8.99382472e-01 -1.22430336e+00 -1.07757223e+00
8.07189167e-01 -3.97436708e-01 7.54132807e-01 -1.98937550e-01
-5.38007379e-01 -3.22217911e-01 -2.40416005e-01 -5.57769120e-01
-3.68905514e-01 -6.87500089e-02 1.80515081e-01 1.29301650e-02
5.36563814e-01 -4.35545206e-01 -2.92702883e-01 3.07984799e-02
-6.99465871e-01 -3.94149065e-01 8.43401313e-01 1.09641850e+00
-4.87474427e-02 2.66838342e-01 6.72365248e-01 -5.74178398e-01
1.12131655e+00 -2.84668028e-01 -6.14989460e-01 5.31704366e-01
-5.84383190e-01 -1.62221696e-02 1.33042252e+00 -8.26164842e-01
-9.64289129e-01 4.06709425e-02 1.43478438e-01 -1.95022866e-01
-6.20962918e-01 6.16937876e-01 3.76500219e-01 -3.13925564e-01
1.10808420e+00 7.98463881e-01 -7.50571350e-03 -7.19962478e-01
-2.62670755e-01 1.18037641e+00 9.39839482e-01 -4.44990605e-01
5.22960484e-01 4.93897162e-02 1.00438163e-01 -8.31785917e-01
-2.31644407e-01 1.40375018e-01 -8.22694480e-01 -2.47004732e-01
4.27703470e-01 -4.80425298e-01 -9.22739863e-01 1.21459222e+00
-1.16350818e+00 -7.50875659e-03 9.01645198e-02 3.73500645e-01
1.63777471e-02 4.09052312e-01 -9.34833884e-01 -1.06894469e+00
-4.83619422e-01 -6.05914593e-01 -1.39026567e-01 4.02812183e-01
-3.67074162e-01 -9.83994305e-01 -4.37894136e-01 -7.30238184e-02
8.02112699e-01 2.85846740e-01 7.95604289e-01 -9.42073345e-01
-5.72709143e-02 -7.72460163e-01 -5.45680165e-01 9.50556338e-01
4.49383557e-01 3.27386409e-01 -1.11638916e+00 -6.88934177e-02
-9.43551660e-02 -5.02457321e-01 9.27656770e-01 7.18706474e-02
1.49320149e+00 -5.18828273e-01 2.45117247e-01 4.64368641e-01
1.73988509e+00 8.62031400e-01 7.82476783e-01 5.30407727e-01
4.00933892e-01 3.03962622e-02 6.79725185e-02 8.04356456e-01
-4.62066710e-01 -4.49857302e-03 1.53638944e-01 2.28877783e-01
-1.25648454e-01 5.95251881e-02 -1.66418284e-01 5.51053107e-01
-3.82496655e-01 -2.28131726e-01 -1.29193962e+00 3.04706216e-01
-1.40345562e+00 -1.01698053e+00 -7.44588375e-02 1.85565162e+00
8.40470493e-01 5.19797623e-01 7.66827762e-02 8.37156892e-01
5.99137783e-01 -2.15640813e-01 -6.89985037e-01 -7.94149578e-01
-4.44274038e-01 6.89343452e-01 6.55335665e-01 9.41244513e-02
-1.15042269e+00 4.94815052e-01 6.05643797e+00 6.11734092e-01
-1.72940111e+00 -5.66337109e-01 9.89644468e-01 3.26785147e-01
6.27984643e-01 -7.51372337e-01 -7.36028016e-01 8.33352447e-01
6.73733830e-01 1.16682807e-02 5.07217228e-01 1.02063215e+00
-1.26019090e-01 -3.62657495e-02 -8.64891648e-01 1.05567181e+00
9.76057649e-02 -1.42922759e+00 4.27546263e-01 -4.32467550e-01
8.75380516e-01 -3.62662286e-01 3.30404401e-01 -3.89252789e-02
9.08808857e-02 -1.65731013e+00 7.30945110e-01 8.01200509e-01
8.17733347e-01 -8.91602576e-01 8.81652176e-01 4.47490990e-01
-5.48926771e-01 -3.68406355e-01 -6.08054459e-01 -6.80152118e-01
-5.54781377e-01 4.47963506e-01 -7.19960392e-01 1.60512164e-01
6.32097185e-01 7.20047951e-01 -6.65666640e-01 1.09435594e+00
2.63123333e-01 6.84921443e-01 -2.65506864e-01 -6.72234178e-01
5.53975701e-01 3.34797204e-02 -1.85127467e-01 1.30764604e+00
4.88387734e-01 1.66869432e-01 -2.93366730e-01 6.69648588e-01
-2.02755824e-01 -5.97227179e-02 -3.15988362e-01 -5.37158072e-01
5.22407413e-01 1.00674486e+00 -1.02774262e+00 -4.43501592e-01
-9.28672850e-02 1.09244096e+00 8.52598473e-02 3.03697199e-01
-3.06114852e-01 -1.08654737e+00 2.53608137e-01 -1.58158466e-01
3.34942073e-01 -3.28059226e-01 -1.06216586e+00 -8.32069933e-01
5.11482432e-02 -8.21688235e-01 7.46669099e-02 -6.15396440e-01
-1.50935531e+00 8.71831834e-01 -7.65810788e-01 -1.28166711e+00
5.62300254e-03 -1.76400959e+00 -8.13846111e-01 1.25449216e+00
-9.58740294e-01 -8.53232980e-01 -5.84113419e-01 4.37598884e-01
5.65976322e-01 -1.03837645e+00 9.00582075e-01 2.04219088e-01
-4.76887375e-01 8.59620750e-01 7.55412042e-01 8.60827744e-01
5.58080018e-01 -1.15809619e+00 3.55312109e-01 6.04435861e-01
-1.18289903e-01 6.32452548e-01 4.52583641e-01 -4.15225655e-01
-1.09103227e+00 -1.12684917e+00 9.22468901e-01 5.22968359e-02
6.62195742e-01 2.10994557e-02 -8.10787737e-01 4.95241582e-01
2.83994168e-01 1.51820511e-01 5.50195456e-01 -5.01351416e-01
-5.03449917e-01 -1.55597419e-01 -1.40878737e+00 2.65638024e-01
4.60197061e-01 -3.29460680e-01 -5.59275448e-01 1.48196846e-01
-3.28950614e-01 -3.72437090e-01 -9.51216817e-01 -4.32709232e-02
8.34244967e-01 -8.37671041e-01 8.45488846e-01 -8.87025237e-01
1.04524720e+00 -5.44907898e-02 -2.82860518e-01 -9.99451756e-01
-4.63163882e-01 1.14493757e-01 -2.92291731e-01 1.16812932e+00
4.02871251e-01 -6.91577554e-01 9.94439304e-01 6.01759136e-01
2.23572388e-01 -7.54295290e-01 -6.75505161e-01 -1.11218774e+00
2.67698109e-01 1.05682453e-02 2.86321700e-01 1.12570024e+00
-2.71170139e-01 -2.27743909e-01 -5.88711798e-01 -2.66397029e-01
5.54941237e-01 -1.20276988e-01 3.69530648e-01 -1.38341582e+00
-5.26528120e-01 -6.42462373e-01 -9.14924383e-01 -6.15520716e-01
-2.50088811e-01 -7.47960627e-01 -4.07779664e-01 -1.24575341e+00
1.90013479e-02 -1.69074923e-01 -5.86604178e-01 5.77599585e-01
4.06186998e-01 5.82936406e-01 1.06716968e-01 2.24411294e-01
1.61342353e-01 -1.31907254e-01 9.43798721e-01 -5.05581856e-01
1.17534705e-01 1.02704667e-01 -3.30562592e-01 4.98321891e-01
1.16157496e+00 -2.43255850e-02 -2.12270930e-01 -5.51649690e-01
-2.06963032e-01 -2.36312747e-01 3.92787308e-01 -1.48955846e+00
4.63324904e-01 -8.92732441e-02 1.34408498e+00 -5.68449080e-01
2.05123410e-01 -6.83018267e-01 3.04969907e-01 8.99017334e-01
-6.45682156e-01 7.27937222e-02 4.26486731e-01 1.62643030e-01
-3.63876134e-01 -6.35932863e-01 7.82962561e-01 -2.79123098e-01
-1.00900650e+00 -1.63931042e-01 -4.62715387e-01 -4.41812545e-01
1.08933806e+00 -7.51059771e-01 -4.91261333e-01 -3.74130458e-02
-8.63096595e-01 -4.82236683e-01 6.98878989e-02 3.98694128e-01
7.90462732e-01 -1.34509778e+00 -6.72137558e-01 1.22270972e-01
-1.27525359e-01 -5.56605756e-01 -4.98643331e-02 -5.61730331e-03
-1.20592201e+00 7.59140134e-01 -1.08579850e+00 -7.07409903e-02
-1.42122090e+00 2.76023000e-01 5.06335855e-01 1.25819430e-01
-4.32046235e-01 9.44991469e-01 -8.44974577e-01 -2.44488969e-01
5.96318007e-01 -2.61336058e-01 -2.78217375e-01 6.27476498e-02
6.34745657e-01 8.29451323e-01 2.36567467e-01 -2.66765445e-01
-2.16527522e-01 3.34633410e-01 -4.32669133e-01 1.81136847e-01
1.40469718e+00 7.60213196e-01 -1.79343998e-01 5.17509818e-01
9.99711156e-01 -5.82892954e-01 -1.01598096e+00 -2.33128235e-01
2.66667336e-01 -4.54522550e-01 -3.81006151e-02 -1.53213131e+00
-8.87623668e-01 1.00364602e+00 8.76245499e-01 1.91853330e-01
9.74083781e-01 -8.04044783e-01 4.01789844e-01 1.26521695e+00
2.36212626e-01 -1.51652861e+00 -4.46832506e-03 7.11265981e-01
9.59775031e-01 -1.18886876e+00 -1.39815837e-01 2.55326182e-01
-4.62282211e-01 2.17490315e+00 6.89811647e-01 -6.50306821e-01
7.50792384e-01 4.66635704e-01 2.47892186e-01 1.73535153e-01
-4.28121239e-01 6.36890173e-01 2.58272409e-01 4.79371876e-01
5.04700303e-01 1.47642568e-01 -4.10310835e-01 7.02191710e-01
-2.16006100e-01 5.20550847e-01 5.50855875e-01 1.05590749e+00
-4.47851092e-01 -6.58854365e-01 -4.33356076e-01 9.97437716e-01
-5.83341956e-01 -1.66301697e-01 -6.24823093e-01 5.37220359e-01
-9.16976184e-02 5.83165765e-01 2.45405138e-01 -5.34374118e-01
1.56085134e-01 2.04132080e-01 4.42696214e-01 -5.43585531e-02
-7.80056477e-01 -8.67171407e-01 -2.58047312e-01 2.55968690e-01
-6.06316030e-02 -3.65152597e-01 -1.00067759e+00 -6.47931755e-01
1.58471808e-01 -2.56978899e-01 7.89856195e-01 6.66270375e-01
1.41880348e-01 4.76214021e-01 5.76582253e-01 -5.19428611e-01
-1.14481795e+00 -1.18377876e+00 -8.80622089e-01 2.13741943e-01
1.04596660e-01 -2.25991175e-01 -2.24281281e-01 2.53737330e-01] | [11.829987525939941, 2.6613848209381104] |
62d974ca-d290-41eb-b671-e903db9b5d3f | semi-perspective-decoupled-heatmaps-for-3d | 2207.02519 | null | https://arxiv.org/abs/2207.02519v1 | https://arxiv.org/pdf/2207.02519v1.pdf | Semi-Perspective Decoupled Heatmaps for 3D Robot Pose Estimation from Depth Maps | Knowing the exact 3D location of workers and robots in a collaborative environment enables several real applications, such as the detection of unsafe situations or the study of mutual interactions for statistical and social purposes. In this paper, we propose a non-invasive and light-invariant framework based on depth devices and deep neural networks to estimate the 3D pose of robots from an external camera. The method can be applied to any robot without requiring hardware access to the internal states. We introduce a novel representation of the predicted pose, namely Semi-Perspective Decoupled Heatmaps (SPDH), to accurately compute 3D joint locations in world coordinates adapting efficient deep networks designed for the 2D Human Pose Estimation. The proposed approach, which takes as input a depth representation based on XYZ coordinates, can be trained on synthetic depth data and applied to real-world settings without the need for domain adaptation techniques. To this end, we present the SimBa dataset, based on both synthetic and real depth images, and use it for the experimental evaluation. Results show that the proposed approach, made of a specific depth map representation and the SPDH, overcomes the current state of the art. | ['Roberto Vezzani', 'Guido Borghi', 'Stefano Pini', 'Alessandro Simoni'] | 2022-07-06 | null | null | null | null | ['2d-human-pose-estimation'] | ['computer-vision'] | [ 9.01476070e-02 2.51069665e-01 4.74980801e-01 -3.05997938e-01
-1.45647705e-01 -3.75028610e-01 6.04032636e-01 -9.38290805e-02
-8.36824596e-01 4.86927480e-01 -3.49838912e-01 1.03431486e-01
-2.35722110e-01 -7.15925932e-01 -9.04999673e-01 -6.71459198e-01
5.44069894e-02 9.44820702e-01 1.61713228e-01 -1.71092764e-01
1.20622426e-01 8.60436678e-01 -1.55055988e+00 -4.56070155e-01
2.92709023e-01 1.00811028e+00 5.44988155e-01 5.78536093e-01
4.09221739e-01 3.99455190e-01 -7.05445945e-01 -1.22108646e-01
5.61008215e-01 5.25273271e-02 -4.11811888e-01 -6.39567077e-02
1.69462383e-01 -6.49632454e-01 -2.28586763e-01 7.42454052e-01
7.36539662e-01 2.67908096e-01 6.45224810e-01 -1.29949093e+00
-2.20733240e-01 -6.74523562e-02 -4.15922970e-01 -4.29715782e-01
6.56463504e-01 1.90781131e-01 3.72954793e-02 -5.23819566e-01
7.07869172e-01 1.44226849e+00 8.78520012e-01 4.58160073e-01
-9.16173697e-01 -6.37135267e-01 -9.61826369e-02 1.98703125e-01
-1.47688115e+00 4.15945910e-02 9.90539432e-01 -6.46067083e-01
7.87903726e-01 -3.01829696e-01 6.34959340e-01 1.50118160e+00
3.51818681e-01 3.17049712e-01 1.07335210e+00 -4.35061127e-01
5.05283058e-01 1.14063248e-01 -2.89704502e-01 8.59847844e-01
4.01688397e-01 1.05989771e-02 -4.61990565e-01 -9.59551707e-02
1.14419162e+00 3.61997455e-01 -1.70312062e-01 -1.09081721e+00
-1.34192753e+00 6.11065865e-01 6.59818769e-01 9.05206576e-02
-5.70187211e-01 2.10721284e-01 1.97432399e-01 -9.87106562e-02
3.21402460e-01 5.54621994e-01 -4.23981041e-01 -1.56908661e-01
-1.59035563e-01 4.84175831e-01 7.44332314e-01 9.31775391e-01
1.00024104e+00 -5.96050799e-01 8.18186775e-02 2.68917620e-01
3.72415125e-01 7.42367327e-01 5.23240387e-01 -1.30634117e+00
4.60147798e-01 8.21277916e-01 7.09340811e-01 -1.44423854e+00
-1.00870955e+00 -1.38854206e-01 -7.59397745e-01 4.12842244e-01
5.95227659e-01 -2.93239713e-01 -5.90976715e-01 1.73499501e+00
6.70451939e-01 -9.35950875e-02 4.91706990e-02 1.09217012e+00
4.36894268e-01 3.10470641e-01 -3.97789121e-01 1.79737672e-01
1.27721250e+00 -7.40128338e-01 -5.93851447e-01 -3.77730727e-01
5.18107831e-01 -2.22350299e-01 9.51446533e-01 5.55361390e-01
-6.41095519e-01 -6.90272331e-01 -9.64208424e-01 -1.81661472e-01
-5.27198076e-01 4.81688678e-01 2.90816754e-01 4.56470609e-01
-1.05874729e+00 5.79303801e-01 -1.24403048e+00 -8.24333251e-01
-5.11097498e-02 6.09760821e-01 -8.09204578e-01 1.01041310e-01
-1.08029616e+00 1.12277019e+00 4.56144035e-01 5.41106462e-01
-1.06126928e+00 -6.12740591e-03 -1.01494181e+00 -1.99332133e-01
3.66066813e-01 -7.14322865e-01 1.22331107e+00 -3.57758343e-01
-1.91178489e+00 7.44333982e-01 1.44563928e-01 -3.57170939e-01
7.33656645e-01 -6.78340435e-01 3.72901022e-01 2.65811354e-01
7.47620538e-02 5.69833815e-01 6.24139667e-01 -1.24542439e+00
-1.33320794e-01 -9.12801445e-01 2.30899915e-01 5.23275912e-01
-2.85233349e-01 -5.27886927e-01 -3.21258485e-01 -5.97943272e-03
2.52580196e-01 -1.36757207e+00 -2.22100943e-01 3.48321050e-01
-3.96126449e-01 -3.17981653e-02 7.22747564e-01 -4.82731909e-01
1.70788646e-01 -1.87776756e+00 4.07940954e-01 2.34783888e-01
-3.01043671e-02 8.83278996e-03 2.26820603e-01 4.73021626e-01
3.54498625e-01 -6.92099214e-01 -3.04660648e-01 -8.36730182e-01
7.52548575e-02 1.83510900e-01 2.56481081e-01 9.94948566e-01
1.53905936e-02 5.66835582e-01 -9.65523720e-01 -2.22313702e-01
6.04652405e-01 7.70315111e-01 -2.71440357e-01 5.45874536e-01
1.43267781e-01 8.62229824e-01 -4.15489048e-01 9.90448520e-02
7.66216040e-01 1.46577433e-01 2.67889261e-01 -1.69469267e-02
-5.87813035e-02 3.15442905e-02 -1.11154604e+00 2.17760277e+00
-6.48524582e-01 4.24067199e-01 1.59960702e-01 -8.72786045e-01
1.29505968e+00 2.72818476e-01 3.21294755e-01 -4.77981120e-01
4.49759841e-01 8.91492590e-02 -4.94096220e-01 -6.39339566e-01
3.88132364e-01 -2.02566688e-03 -3.73721749e-01 4.14796114e-01
-2.74927751e-03 -2.65834838e-01 -3.11905235e-01 -3.07981580e-01
1.08244407e+00 6.35223150e-01 2.87172973e-01 1.20619871e-01
4.87151831e-01 -2.19543204e-01 2.25270584e-01 5.86088896e-01
-2.35842004e-01 5.49386859e-01 3.51057529e-01 -6.10472977e-01
-8.88783097e-01 -1.11547375e+00 3.19423735e-01 7.16214299e-01
6.16192877e-01 2.05943897e-01 -8.76528859e-01 -6.12060428e-01
3.63129284e-03 3.10280353e-01 -7.90434897e-01 -2.35549331e-01
-7.41080165e-01 -2.41244629e-01 4.29549575e-01 4.02250260e-01
7.23748267e-01 -1.01403844e+00 -1.43753469e+00 -1.57331347e-01
-2.58699864e-01 -1.23919976e+00 1.73063770e-01 2.38086581e-01
-5.94954431e-01 -1.04790461e+00 -8.77287507e-01 -7.19467163e-01
7.95922577e-01 3.04927915e-01 5.43520093e-01 -6.27578199e-01
-1.47469446e-01 6.53273284e-01 -3.13452452e-01 -5.08418560e-01
-8.85522664e-02 2.41019949e-01 4.07265186e-01 9.30508450e-02
3.91266525e-01 -6.85823083e-01 -7.02553868e-01 5.01635969e-01
-4.03173298e-01 -5.25099784e-02 6.90923095e-01 4.53586549e-01
3.99877489e-01 -9.33232680e-02 2.04958573e-01 -6.02156341e-01
5.57734072e-01 -2.85071433e-01 -7.69692004e-01 -5.95139638e-02
-1.40864715e-01 2.37352431e-01 4.61820722e-01 -4.27748471e-01
-1.17015862e+00 7.27296650e-01 3.07601839e-02 -5.49190402e-01
-6.40823841e-01 1.68612655e-02 -1.30824268e-01 -1.19581237e-01
8.82427156e-01 -2.04111524e-02 3.61044675e-01 -7.29722261e-01
2.29904205e-01 7.78281152e-01 7.69361198e-01 -4.68246818e-01
7.80891001e-01 8.49252105e-01 4.18367624e-01 -5.62968016e-01
-3.33910406e-01 -4.62992519e-01 -1.18338382e+00 -2.03627080e-01
9.70046997e-01 -9.57039416e-01 -1.34583342e+00 7.77495801e-01
-1.58315754e+00 -4.43312317e-01 2.88342442e-02 7.25391567e-01
-9.99991715e-01 4.10941958e-01 -3.39818358e-01 -1.03638053e+00
-6.48314729e-02 -1.10929716e+00 1.60798979e+00 8.89163241e-02
-4.61145192e-02 -7.76588440e-01 2.64538229e-01 1.41597375e-01
6.03022575e-02 6.59564316e-01 2.96823740e-01 -3.53215814e-01
-2.85687536e-01 -5.45403838e-01 9.09697488e-02 8.02616477e-02
4.79308292e-02 -6.81128144e-01 -1.05794704e+00 -3.28412890e-01
3.00620168e-01 -4.87575412e-01 1.52848259e-01 1.47979960e-01
8.24720085e-01 -9.27171018e-03 -4.10418034e-01 3.57950509e-01
1.10350585e+00 1.10877514e-01 3.70686948e-01 5.09750962e-01
7.88662374e-01 8.94940913e-01 8.13619733e-01 6.74603343e-01
5.69631159e-01 1.03776312e+00 1.00002182e+00 8.37821737e-02
4.59747225e-01 -2.74138242e-01 3.71195644e-01 3.35640341e-01
-2.17505142e-01 -1.80368945e-01 -9.80991364e-01 3.56574744e-01
-2.00247622e+00 -5.99192262e-01 1.73275381e-01 2.24410391e+00
1.98491946e-01 -5.30833937e-02 1.22727761e-02 1.10502906e-01
8.30887675e-01 -2.50037044e-01 -5.48372328e-01 -2.54900277e-01
5.52286386e-01 -1.75601199e-01 6.45048261e-01 4.05527592e-01
-1.08514535e+00 6.89375818e-01 5.15771055e+00 1.54242560e-01
-1.02533746e+00 2.48208597e-01 3.73791121e-02 -5.50573915e-02
4.83666688e-01 -4.52078849e-01 -6.19868934e-01 1.46864459e-01
6.75430775e-01 5.06540954e-01 4.54135209e-01 1.14745831e+00
2.59229541e-01 -3.43468994e-01 -1.13869715e+00 1.13579464e+00
1.88539326e-01 -5.08899570e-01 -4.84570205e-01 4.24475931e-02
3.47141355e-01 -4.25334573e-02 6.50379509e-02 1.72833815e-01
2.78346669e-02 -5.93448520e-01 8.79532933e-01 5.62753618e-01
5.01841366e-01 -7.61287749e-01 1.07781196e+00 1.03810441e+00
-7.73182750e-01 -1.81225598e-01 -4.71742600e-01 -4.21283215e-01
1.51716202e-01 2.27588207e-01 -1.41295993e+00 5.85717440e-01
8.47617507e-01 4.85983819e-01 -3.69291991e-01 6.74977064e-01
-3.08226913e-01 -2.87722468e-01 -6.06979251e-01 -2.54680574e-01
7.76481703e-02 -1.16208224e-02 3.60914648e-01 7.28183806e-01
6.48082674e-01 -3.57125640e-01 9.62299295e-03 8.04600060e-01
2.35456303e-01 -3.46109599e-01 -1.11536503e+00 4.73117054e-01
3.63575816e-01 1.18645644e+00 -7.64733434e-01 1.46039188e-01
1.36316359e-01 1.17189026e+00 5.30440390e-01 2.55468965e-01
-6.93539917e-01 -6.43579781e-01 5.07120848e-01 -2.02705711e-02
1.71226114e-01 -7.74947107e-01 4.40397635e-02 -9.04183865e-01
1.92121655e-01 -3.27206880e-01 -1.74861878e-01 -9.29276764e-01
-8.24613988e-01 7.03688264e-01 3.98900449e-01 -1.33622229e+00
-6.75330102e-01 -1.04568934e+00 -1.51600689e-01 7.69626915e-01
-1.26069725e+00 -9.87438023e-01 -8.77027631e-01 5.98222852e-01
2.59048909e-01 4.20485921e-02 8.10265422e-01 -4.25014831e-02
-3.65372300e-01 3.25081974e-01 1.24387667e-01 -4.87894751e-02
7.42760479e-01 -1.07633305e+00 4.76171643e-01 4.42025840e-01
-2.74338275e-01 4.86543804e-01 7.83190012e-01 -4.74859089e-01
-1.51818120e+00 -9.44830596e-01 6.35541856e-01 -8.70835781e-01
1.72519078e-03 -9.00335789e-01 -2.92901337e-01 7.49546766e-01
-2.60629326e-01 6.24523088e-02 4.10126895e-02 -2.10027754e-01
8.28084946e-02 -1.98544458e-01 -1.49741483e+00 3.88722748e-01
1.13557386e+00 -4.57263052e-01 -4.37789023e-01 3.94211799e-01
7.13374794e-01 -7.22003996e-01 -6.06164694e-01 2.78285652e-01
7.46470034e-01 -1.23625469e+00 1.00592184e+00 1.25993133e-01
4.51826677e-02 -3.60930264e-01 5.47985360e-02 -1.27751756e+00
-1.84667721e-01 -5.00948131e-01 3.72588709e-02 6.83548868e-01
-2.00451404e-01 -5.79031169e-01 9.99572873e-01 5.18363178e-01
-7.43255913e-02 -3.73976499e-01 -1.12503934e+00 -7.82234311e-01
-5.01955211e-01 -4.04094398e-01 5.35521090e-01 4.41362411e-01
-1.12013362e-01 1.05898403e-01 -4.06865060e-01 5.70937455e-01
5.63755035e-01 -1.73865855e-01 1.29859173e+00 -1.44532955e+00
-1.92599729e-01 1.86337426e-01 -7.87068069e-01 -1.06182742e+00
3.48868817e-01 -2.63379276e-01 5.67397654e-01 -1.50102234e+00
-2.41651237e-01 -1.57527670e-01 1.77063838e-01 2.36460283e-01
3.05216491e-01 3.89086843e-01 -1.47189787e-02 1.56895369e-01
-6.04772508e-01 8.37029397e-01 1.11389840e+00 5.91733009e-02
-1.62433371e-01 6.23037964e-02 -1.64403226e-02 9.88774300e-01
6.87766254e-01 -4.52621847e-01 -5.61523259e-01 -5.74171662e-01
3.49116504e-01 -7.86476284e-02 7.62756109e-01 -1.44573152e+00
2.62396395e-01 3.00541297e-02 3.89387280e-01 -5.02202213e-01
8.81594598e-01 -1.21014977e+00 2.74215966e-01 7.87260294e-01
-2.39971280e-01 2.60538399e-01 2.71773804e-02 8.44130993e-01
2.29660735e-01 -2.70220876e-01 3.87697428e-01 -3.08776528e-01
-4.48857814e-01 -4.68536243e-02 -2.02276021e-01 -5.08087575e-01
1.27534509e+00 -3.13704342e-01 -1.47078514e-01 -5.17117441e-01
-4.01797384e-01 -1.31904855e-02 6.82109535e-01 2.57284790e-01
5.77759743e-01 -1.28632700e+00 -1.95697650e-01 4.38180834e-01
1.29616767e-01 5.89812338e-01 3.02536756e-01 6.98375762e-01
-1.13794649e+00 5.35484850e-01 -4.67999488e-01 -6.04003906e-01
-8.45464528e-01 7.27479219e-01 4.29484814e-01 -1.78553954e-01
-5.25509596e-01 6.28792107e-01 4.01467413e-01 -1.03843284e+00
4.23122406e-01 -4.43552554e-01 -2.42841095e-01 -3.95667315e-01
2.63454407e-01 5.39134920e-01 2.31149402e-02 -7.80367136e-01
-4.18402135e-01 7.56770611e-01 4.14721608e-01 -2.61985123e-01
1.24504888e+00 -3.96611542e-01 7.62774348e-02 6.03607416e-01
1.25195551e+00 -4.25826520e-01 -1.67025208e+00 -1.65973172e-01
-2.62210608e-01 -4.58574176e-01 -3.10192972e-01 -5.03069460e-01
-7.18844831e-01 1.13054407e+00 1.01329672e+00 4.81496714e-02
8.45363319e-01 -2.41711903e-02 6.08588934e-01 1.01670766e+00
9.41672087e-01 -9.16589558e-01 2.03259543e-01 4.64626402e-01
9.25483882e-01 -1.29716361e+00 -2.98320383e-01 -2.12916881e-01
-4.33316439e-01 1.03464544e+00 7.72249937e-01 -2.81273395e-01
4.45217282e-01 -1.42528668e-01 1.04184888e-01 -2.02777609e-01
-2.50513554e-01 -2.71502603e-03 -1.72503710e-01 1.06025660e+00
-8.30774456e-02 -4.38967980e-02 9.21063796e-02 2.67268509e-01
-1.90584928e-01 -5.01126349e-02 3.39642942e-01 9.91079569e-01
-3.40670675e-01 -7.13425100e-01 -6.97449744e-01 -2.74887919e-01
3.43742222e-02 7.49302745e-01 -6.24743462e-01 8.53159785e-01
5.34361422e-01 7.06909478e-01 2.97947913e-01 -6.47947907e-01
5.56754887e-01 -8.62975121e-02 6.89807653e-01 -5.01491070e-01
-4.27975804e-01 -4.64708537e-01 -1.78462937e-01 -7.32474864e-01
-5.09316981e-01 -5.06166577e-01 -1.09958708e+00 -1.73663180e-02
-1.35497823e-01 -1.73295990e-01 1.15182018e+00 9.09980893e-01
5.29400170e-01 6.13113716e-02 5.84158659e-01 -1.93294120e+00
-5.62415302e-01 -1.14458013e+00 -6.21169150e-01 4.91393656e-01
3.46716195e-01 -1.21623647e+00 -1.90095544e-01 -3.22848946e-01] | [6.923571586608887, -1.2598028182983398] |
ce9eb72c-9c39-4e5d-a094-d9c3084a1f6d | super-resolution-of-license-plate-images | 2305.17313 | null | https://arxiv.org/abs/2305.17313v1 | https://arxiv.org/pdf/2305.17313v1.pdf | Super-Resolution of License Plate Images Using Attention Modules and Sub-Pixel Convolution Layers | Recent years have seen significant developments in the field of License Plate Recognition (LPR) through the integration of deep learning techniques and the increasing availability of training data. Nevertheless, reconstructing license plates (LPs) from low-resolution (LR) surveillance footage remains challenging. To address this issue, we introduce a Single-Image Super-Resolution (SISR) approach that integrates attention and transformer modules to enhance the detection of structural and textural features in LR images. Our approach incorporates sub-pixel convolution layers (also known as PixelShuffle) and a loss function that uses an Optical Character Recognition (OCR) model for feature extraction. We trained the proposed architecture on synthetic images created by applying heavy Gaussian noise to high-resolution LP images from two public datasets, followed by bicubic downsampling. As a result, the generated images have a Structural Similarity Index Measure (SSIM) of less than 0.10. Our results show that our approach for reconstructing these low-resolution synthesized images outperforms existing ones in both quantitative and qualitative measures. Our code is publicly available at https://github.com/valfride/lpr-rsr-ext/ | ['David Menotti', 'William Robson Schwartz', 'Jorge de A. Lambert', 'Rayson Laroca', 'Valfride Nascimento'] | 2023-05-27 | null | null | null | null | ['optical-character-recognition', 'image-super-resolution', 'license-plate-recognition'] | ['computer-vision', 'computer-vision', 'computer-vision'] | [ 5.19399464e-01 -3.73198569e-01 1.57769382e-01 -2.45570809e-01
-1.32633460e+00 -6.60419643e-01 5.81382394e-01 -6.31891847e-01
-2.55413026e-01 7.12408185e-01 1.04918592e-01 1.16261523e-02
2.32652783e-01 -9.14344609e-01 -1.15291214e+00 -5.85739434e-01
4.84426022e-01 -3.54780070e-02 4.64694649e-01 -3.02409261e-01
4.53327864e-01 7.24345744e-01 -1.54369903e+00 5.97216129e-01
9.82541680e-01 9.98698831e-01 4.15109783e-01 6.59636080e-01
2.49055997e-01 8.71574938e-01 -4.31958467e-01 -7.02401280e-01
5.67039251e-01 -7.32818544e-02 -5.23242116e-01 2.83159524e-01
7.45076537e-01 -5.35457194e-01 -4.80757415e-01 1.17984068e+00
3.25782865e-01 6.67682439e-02 5.14946818e-01 -6.05445087e-01
-1.19169247e+00 6.30716458e-02 -7.22654998e-01 2.87811846e-01
3.82002354e-01 1.50280803e-01 7.05706596e-01 -1.21332741e+00
7.20504940e-01 1.01395631e+00 9.74937975e-01 4.16086733e-01
-1.27180326e+00 -5.63330114e-01 -3.91570061e-01 1.55973405e-01
-1.53785038e+00 -5.49261332e-01 7.09107637e-01 -3.58002335e-01
7.46776044e-01 2.62131184e-01 5.91346994e-02 1.11185896e+00
1.55610874e-01 4.78736967e-01 1.37841594e+00 -3.03149313e-01
-5.17815202e-02 3.76830697e-01 -3.17274302e-01 6.38648212e-01
3.45253527e-01 4.69780490e-02 -2.70630062e-01 3.37603927e-01
1.22109532e+00 1.42354310e-01 -2.30721563e-01 2.57913113e-01
-9.23630118e-01 4.91348654e-01 3.13430727e-01 1.51954085e-01
-2.28020057e-01 -2.21583858e-01 -1.00957602e-02 9.63167846e-02
6.21039748e-01 3.82407635e-01 -6.52004108e-02 2.72501796e-01
-9.97093379e-01 6.38567358e-02 2.60169446e-01 8.25737357e-01
6.74978316e-01 2.20880747e-01 -3.28943282e-02 1.30181193e+00
4.23485376e-02 5.20981669e-01 4.51858699e-01 -1.05050170e+00
5.30168355e-01 3.60299349e-01 2.92289823e-01 -1.10123909e+00
6.52150065e-02 -3.68329406e-01 -8.73316765e-01 4.09672081e-01
4.61267196e-02 1.22510962e-01 -8.69205654e-01 1.14222395e+00
-1.33519232e-01 2.04159558e-01 1.16462603e-01 1.07274127e+00
7.93736935e-01 8.41824234e-01 -2.93247461e-01 1.20948479e-01
1.31074226e+00 -9.67036068e-01 -4.71296787e-01 -2.33400494e-01
-1.11274622e-01 -8.60948622e-01 1.09103310e+00 2.71301389e-01
-1.15538514e+00 -8.21886957e-01 -1.15132296e+00 -3.28321040e-01
-3.23465168e-01 6.81450486e-01 -5.96266240e-02 4.95115250e-01
-1.15432167e+00 4.06792074e-01 -4.53205854e-01 -7.71680251e-02
6.76715851e-01 1.65472314e-01 -6.66466057e-01 -3.88195813e-01
-9.74313438e-01 7.71434784e-01 5.45432493e-02 1.57212913e-01
-8.46275926e-01 -5.61489582e-01 -6.75845623e-01 -2.21482925e-02
1.72169060e-01 -1.49223670e-01 7.09221363e-01 -1.10536087e+00
-1.62719953e+00 9.99139547e-01 1.93375677e-01 -3.81153733e-01
5.82834661e-01 -2.53439128e-01 -5.70206106e-01 3.24043185e-01
-4.81286906e-02 3.88837457e-01 9.95472789e-01 -1.38756466e+00
-5.23725390e-01 -2.68473506e-01 -3.23252268e-02 2.84273148e-01
-9.30480361e-02 4.61564124e-01 -4.93202001e-01 -8.44241619e-01
-3.32870968e-02 -6.15551472e-01 4.87716123e-02 -1.13333665e-01
-4.04720962e-01 4.22656327e-01 6.35557175e-01 -1.32243013e+00
6.39032245e-01 -2.23581624e+00 -2.17875525e-01 -1.68080479e-01
6.77517131e-02 5.34969091e-01 -2.88776547e-01 7.02901408e-02
-7.28175640e-02 7.94642642e-02 -4.91862625e-01 -4.09055352e-01
-3.50824118e-01 -3.77200156e-01 -4.68023062e-01 6.75729156e-01
4.99779135e-01 7.48865485e-01 -3.54201138e-01 -3.14946383e-01
4.61045563e-01 8.24020207e-01 -3.28630924e-01 2.44941011e-01
2.99669076e-02 3.59225303e-01 -9.74019170e-02 8.17787766e-01
1.19753301e+00 -1.06113009e-01 -1.65011048e-01 -2.77070582e-01
-4.62226868e-01 -6.02560416e-02 -9.63455439e-01 1.17874885e+00
-5.39512336e-01 9.11381662e-01 6.79673776e-02 -6.72719717e-01
1.33946443e+00 4.58828881e-02 2.79695421e-01 -1.07706439e+00
-7.19907954e-02 2.89861023e-01 -5.98829329e-01 -4.48969603e-01
6.63258851e-01 2.26477464e-03 8.45971778e-02 7.68642500e-02
-1.60059243e-01 8.75079557e-02 3.47019881e-02 -1.95579320e-01
8.45813572e-01 1.75214365e-01 -1.23171084e-01 1.31056175e-01
9.60236669e-01 -1.28684178e-01 4.32392508e-01 4.68639731e-01
7.25386068e-02 1.16957033e+00 1.66738987e-01 -4.21981066e-01
-1.69686604e+00 -1.06300497e+00 -2.50001937e-01 8.09074700e-01
1.65644184e-01 2.20619082e-01 -7.91705370e-01 -1.80144206e-01
-3.11030596e-01 4.15251344e-01 -4.34523612e-01 1.09206587e-01
-7.46384501e-01 -6.83902442e-01 6.97736442e-01 4.53162342e-01
9.88041818e-01 -9.89250064e-01 -1.68119505e-01 -1.74750149e-01
-2.68993318e-01 -1.63014996e+00 -4.69362885e-01 -4.76405919e-01
-6.99947357e-01 -8.35073531e-01 -9.97908354e-01 -9.37030435e-01
8.16473722e-01 4.06160414e-01 8.61406922e-01 -2.67657846e-01
-3.40975970e-01 -1.41960802e-02 -2.97605842e-01 1.99255660e-01
-5.91654301e-01 -2.22463325e-01 -1.33033291e-01 4.87247169e-01
1.32360950e-01 -3.70221138e-01 -5.24964154e-01 3.86808872e-01
-1.10240483e+00 2.59596378e-01 1.02096963e+00 5.65373957e-01
7.46129930e-01 1.17223047e-01 3.10332745e-01 -5.83866715e-01
6.14972115e-01 -3.09258938e-01 -1.13953006e+00 2.69288152e-01
-2.20288351e-01 -2.03811139e-01 9.53611791e-01 -1.44029096e-01
-1.39679337e+00 8.49568751e-03 -2.65099227e-01 -5.75510263e-01
-4.05681908e-01 9.62165594e-02 -1.53293639e-01 -2.89179951e-01
5.70367515e-01 8.65073323e-01 -5.13788387e-02 -6.19209468e-01
5.83775043e-02 1.07453537e+00 8.19071114e-01 -1.41608536e-01
9.67277825e-01 6.96762264e-01 -2.50762463e-01 -1.04492188e+00
-5.42767286e-01 -7.37195760e-02 -4.08681065e-01 -2.49970168e-01
8.64138424e-01 -1.09140027e+00 -5.77249765e-01 7.98662663e-01
-9.56439912e-01 -1.63118571e-01 -7.78714567e-02 3.58046561e-01
-5.04783213e-01 4.80873048e-01 -7.99865305e-01 -7.70023167e-01
-2.67610997e-01 -1.16045690e+00 1.25359201e+00 3.26257527e-01
5.32722712e-01 -4.67313647e-01 3.05158254e-02 9.86720979e-01
7.83096671e-01 2.11189970e-01 4.03910160e-01 -3.45975667e-01
-1.04916918e+00 -4.37395722e-01 -7.65281320e-01 8.57021689e-01
-1.11521155e-01 5.14857192e-03 -1.03479755e+00 -1.41737789e-01
-7.00636655e-02 -3.47532213e-01 1.01098418e+00 4.68122900e-01
1.18354416e+00 -5.06681383e-01 2.24494442e-01 8.32180917e-01
1.81788814e+00 -1.89118609e-02 1.10898829e+00 4.62942362e-01
7.06204951e-01 3.90153676e-01 4.72332299e-01 3.35321128e-01
3.14993709e-01 8.74294341e-01 1.89999625e-01 -2.02307418e-01
-4.61837083e-01 -2.07620576e-01 7.92387784e-01 4.35818762e-01
-4.82735395e-01 -8.56812820e-02 -8.71528327e-01 3.20439756e-01
-1.30170333e+00 -1.17807758e+00 7.61626661e-03 2.21412706e+00
6.98757052e-01 -4.06138450e-02 -3.68909165e-02 -1.41804278e-01
1.05480087e+00 1.89386994e-01 -4.26807761e-01 -3.63774329e-01
-5.65831840e-01 -1.22293070e-01 7.77906597e-01 5.27272642e-01
-1.21887553e+00 1.02805293e+00 5.38924122e+00 8.73338163e-01
-1.24768293e+00 4.37265448e-02 9.29473877e-01 3.07203680e-01
-1.99731097e-01 -5.63315868e-01 -7.98558831e-01 5.81336677e-01
7.92045295e-01 7.90168196e-02 7.44075179e-01 6.33749187e-01
2.03373551e-01 9.88514498e-02 -4.60541666e-01 1.00550067e+00
3.96412879e-01 -1.61005414e+00 1.69543371e-01 4.74648215e-02
9.04715180e-01 2.96141446e-01 6.26210749e-01 -1.01215933e-02
4.95610684e-02 -1.25047970e+00 6.46512628e-01 8.25191200e-01
1.42719924e+00 -7.63034523e-01 8.87220562e-01 1.16477273e-01
-1.06945336e+00 -3.39975715e-01 -5.99240482e-01 4.48300809e-01
-2.00537518e-01 3.35328460e-01 -6.70439005e-01 4.80664849e-01
7.23342061e-01 8.19110751e-01 -6.90061808e-01 8.78805935e-01
1.29736841e-01 2.92130232e-01 2.62732245e-02 4.02103394e-01
-7.51457410e-03 -4.56259221e-01 5.85273981e-01 1.21331537e+00
4.39897656e-01 -1.78306267e-01 -3.70864689e-01 1.25569224e+00
-3.56220543e-01 -8.96686316e-02 -5.16970634e-01 5.43021001e-02
3.16945851e-01 1.25719142e+00 -5.13323605e-01 -6.00370392e-02
-6.95497990e-01 1.07028866e+00 -1.24991769e-02 3.26403588e-01
-1.07092643e+00 -3.32256913e-01 4.30638850e-01 3.32844377e-01
5.67153633e-01 -9.14674848e-02 -1.99336380e-01 -1.35783267e+00
2.82647163e-01 -9.71253872e-01 -5.35694137e-02 -9.85184014e-01
-1.19075155e+00 8.81915212e-01 -4.06074166e-01 -1.62548149e+00
8.58848244e-02 -7.27842152e-01 -3.02188754e-01 9.28703487e-01
-1.72175539e+00 -1.27547348e+00 -4.18165058e-01 4.94010329e-01
8.40967357e-01 -4.39253777e-01 5.32043576e-01 5.00717461e-01
-6.43892288e-01 4.84138697e-01 5.00540733e-01 4.96937722e-01
5.14811277e-01 -8.03173542e-01 5.32009065e-01 1.13928699e+00
-2.77436495e-01 2.83740789e-01 5.51082969e-01 -5.60539901e-01
-1.23483062e+00 -1.48873234e+00 4.75219339e-01 -3.90239030e-01
2.66857296e-01 -3.92641842e-01 -9.42729115e-01 6.49283051e-01
-3.43282521e-02 2.04152972e-01 5.04632235e-01 -7.81540513e-01
-4.68140811e-01 -3.43468398e-01 -1.41825092e+00 5.69550812e-01
6.43792808e-01 -5.70291698e-01 -4.19619799e-01 9.41212252e-02
5.92931330e-01 -1.36291876e-01 -9.02642846e-01 3.07853162e-01
5.85538089e-01 -1.18004763e+00 1.23166156e+00 -4.23370264e-02
8.70550752e-01 -5.52236140e-01 -4.07981128e-01 -8.09878170e-01
-3.31667393e-01 -2.02453379e-02 2.18680158e-01 1.09899616e+00
4.78354752e-01 -5.42456269e-01 5.51308155e-01 5.51112235e-01
-1.75141364e-01 -5.22832513e-01 -6.47988081e-01 -7.06863463e-01
-1.59615740e-01 -8.52746740e-02 4.15967643e-01 6.85919046e-01
-6.81508303e-01 -7.89107531e-02 -6.72338963e-01 4.41500902e-01
9.63474751e-01 1.08935945e-01 5.16461074e-01 -9.06483650e-01
-1.88591674e-01 -3.00910562e-01 -4.95317280e-01 -7.00713873e-01
1.35772498e-02 -6.45523906e-01 3.98959331e-02 -1.43156993e+00
3.07731688e-01 -1.28331371e-02 -2.12679029e-01 8.77824724e-02
1.99603483e-01 9.28285360e-01 1.76058203e-01 3.55457395e-01
-7.11309254e-01 3.80254239e-01 1.12408483e+00 -1.04982413e-01
3.10539305e-02 5.25667034e-02 -5.90618491e-01 6.62188172e-01
9.70909655e-01 -1.71481282e-01 1.47375748e-01 -4.18339044e-01
1.92152128e-01 1.42605871e-01 4.59691972e-01 -1.14631617e+00
1.81107163e-01 1.31004229e-01 7.85246909e-01 -4.88718569e-01
5.41764617e-01 -6.53008401e-01 1.61702514e-01 1.79221705e-01
-4.23361003e-01 -1.26905680e-01 2.69840419e-01 4.51751500e-01
-3.65992010e-01 -7.63611170e-03 1.15591145e+00 -2.06245184e-01
-8.22202444e-01 2.17537209e-01 -3.63779336e-01 -2.50755489e-01
8.11540008e-01 -4.91299331e-01 -5.79431832e-01 -1.38641804e-01
-2.45687351e-01 -3.97890359e-01 8.31066847e-01 4.85879242e-01
1.18506062e+00 -1.15995896e+00 -1.17869222e+00 5.02480268e-01
1.59619935e-02 -1.53942719e-01 5.89040518e-01 6.79525793e-01
-1.03271508e+00 4.88788068e-01 -7.26912141e-01 -2.92648494e-01
-1.24640787e+00 3.65636230e-01 5.53781986e-01 -1.07757270e-01
-7.31082737e-01 6.70172870e-01 7.70870000e-02 -5.05402088e-01
-2.22705491e-02 3.48892901e-03 -3.80269825e-01 -2.58810163e-01
8.30425382e-01 5.89438081e-01 1.60045736e-02 -9.86979485e-01
-2.60916203e-01 8.27414870e-01 -1.85163513e-01 -1.04304112e-01
1.84530854e+00 -2.83120811e-01 -1.38754502e-01 2.48867720e-02
1.23095298e+00 1.39061719e-01 -1.52079809e+00 -2.80111104e-01
-2.98155069e-01 -8.64787459e-01 1.40761025e-02 -7.06644118e-01
-1.04808736e+00 5.39631784e-01 6.77087545e-01 -2.30092648e-02
1.31156731e+00 -1.28875881e-01 7.08342195e-01 3.24508339e-01
2.21437603e-01 -1.13188028e+00 4.61360067e-02 3.17471892e-01
1.16885090e+00 -1.33483505e+00 -8.31023753e-02 -3.21892530e-01
-7.92247593e-01 1.10153592e+00 4.92132783e-01 -5.38375258e-01
8.37215707e-02 3.95719558e-01 -6.63762987e-02 2.00706840e-01
-5.04588187e-01 -1.68025736e-02 2.16698512e-01 5.35431266e-01
3.65669042e-01 -1.15767799e-01 1.88598976e-01 4.82710540e-01
3.38025205e-03 7.68522918e-02 8.46195519e-01 4.06473398e-01
-3.80863488e-01 -6.21941984e-01 -7.02674150e-01 3.18640679e-01
-6.69165134e-01 -2.05832630e-01 -2.45217144e-01 4.57801074e-01
9.07862093e-03 6.31142020e-01 2.02201426e-01 -3.97895008e-01
1.82479322e-01 -3.15265328e-01 2.89285392e-01 -3.54130477e-01
-7.30595738e-02 -5.53386398e-02 -9.73417684e-02 -5.34448504e-01
-3.11522484e-01 -5.35968900e-01 -7.35012770e-01 -3.57345670e-01
-8.29234794e-02 -3.63210410e-01 7.32905447e-01 4.74438548e-01
4.17648077e-01 4.98849541e-01 9.46210980e-01 -9.31196690e-01
-3.85118276e-01 -8.33355904e-01 -6.80893540e-01 3.27079237e-01
5.54488361e-01 -2.41848305e-01 -2.73099720e-01 2.93728024e-01] | [11.106231689453125, -2.1841297149658203] |
5a23e2f3-7977-4a0b-92d9-6dae0d9f016e | signnet-single-channel-sign-generation-using | 2212.02848 | null | https://arxiv.org/abs/2212.02848v1 | https://arxiv.org/pdf/2212.02848v1.pdf | SignNet: Single Channel Sign Generation using Metric Embedded Learning | A true interpreting agent not only understands sign language and translates to text, but also understands text and translates to signs. Much of the AI work in sign language translation to date has focused mainly on translating from signs to text. Towards the latter goal, we propose a text-to-sign translation model, SignNet, which exploits the notion of similarity (and dissimilarity) of visual signs in translating. This module presented is only one part of a dual-learning two task process involving text-to-sign (T2S) as well as sign-to-text (S2T). We currently implement SignNet as a single channel architecture so that the output of the T2S task can be fed into S2T in a continuous dual learning framework. By single channel, we refer to a single modality, the body pose joints. In this work, we present SignNet, a T2S task using a novel metric embedding learning process, to preserve the distances between sign embeddings relative to their dissimilarity. We also describe how to choose positive and negative examples of signs for similarity testing. From our analysis, we observe that metric embedding learning-based model perform significantly better than the other models with traditional losses, when evaluated using BLEU scores. In the task of gloss to pose, SignNet performed as well as its state-of-the-art (SoTA) counterparts and outperformed them in the task of text to pose, by showing noteworthy enhancements in BLEU 1 - BLEU 4 scores (BLEU 1: 31->39; ~26% improvement and BLEU 4: 10.43->11.84; ~14\% improvement) when tested on the popular RWTH PHOENIX-Weather-2014T benchmark dataset | ['Ifeoma Nwogu', 'Lipisha Chaudhary', 'Tejaswini Ananthanarayana'] | 2022-12-06 | null | null | null | null | ['sign-language-translation'] | ['computer-vision'] | [ 4.42846656e-01 -9.66297686e-02 -1.28480509e-01 -5.20103395e-01
-1.03665972e+00 -7.36406326e-01 9.07501221e-01 -6.55536890e-01
-5.12923956e-01 5.73060036e-01 6.39728785e-01 -1.03401802e-01
4.49263901e-02 -2.22906411e-01 -8.47675741e-01 -7.49221742e-01
1.02368094e-01 7.19517112e-01 -5.21841720e-02 -3.54043096e-01
-1.43387347e-01 -5.30243339e-03 -1.29468918e+00 3.68981630e-01
6.07706010e-01 8.80009592e-01 -3.96916330e-01 8.86659086e-01
1.44356251e-01 8.32821608e-01 -6.56697750e-01 -6.33895218e-01
3.91335398e-01 -8.86379898e-01 -6.66308820e-01 -3.53076190e-01
1.05671251e+00 -6.00768387e-01 -4.02835190e-01 7.42345989e-01
9.86610472e-01 -1.72297016e-01 8.99797618e-01 -1.28932571e+00
-9.35371697e-01 4.22913849e-01 -3.64913970e-01 -3.04537743e-01
4.80116993e-01 3.78991723e-01 1.46374643e+00 -9.38837111e-01
9.33540046e-01 1.32345200e+00 5.51467657e-01 8.76257360e-01
-9.52870071e-01 -6.50645435e-01 -8.61809030e-02 2.53391176e-01
-9.57417607e-01 -2.79110134e-01 4.60084975e-01 -1.98579594e-01
8.94286692e-01 3.46977711e-01 7.68725574e-01 1.46909189e+00
2.00984534e-02 1.39145267e+00 1.49270761e+00 -3.77446532e-01
-6.12555817e-02 -4.04437155e-01 -1.76003039e-01 7.92911053e-01
-7.65397400e-02 5.05408049e-01 -9.98232782e-01 3.50770205e-01
5.24581373e-01 -4.89366770e-01 -2.54238993e-01 -3.62663984e-01
-1.65788376e+00 5.70150971e-01 5.69592476e-01 2.66360343e-01
-1.38030797e-01 7.79992044e-01 4.09577787e-01 8.20122242e-01
1.11525476e-01 1.53421253e-01 -3.63343656e-01 -5.40541887e-01
-6.02604628e-01 4.17771488e-01 8.05925369e-01 7.86891580e-01
-7.87170511e-03 1.05608262e-01 -5.54553032e-01 6.46326900e-01
5.99510610e-01 1.21750557e+00 4.08590823e-01 -7.15064824e-01
7.85110772e-01 2.86911517e-01 -9.05065332e-03 -3.34492505e-01
-3.77787113e-01 -4.37588602e-01 -6.31471694e-01 5.57668328e-01
6.79854512e-01 -8.95732343e-02 -1.48131156e+00 2.05497241e+00
1.09782897e-01 -8.33226070e-02 1.31100610e-01 1.25914419e+00
8.65916729e-01 3.76000315e-01 5.45909889e-02 3.96999598e-01
1.35231614e+00 -9.96874452e-01 -6.73730671e-01 -1.12144761e-01
6.83077693e-01 -9.04333472e-01 1.23613429e+00 1.43793300e-01
-1.00035310e+00 -3.25131804e-01 -1.06574106e+00 -3.66204441e-01
-4.03878599e-01 3.80671412e-01 4.56824392e-01 4.80046034e-01
-1.16322660e+00 4.91229624e-01 -9.08366084e-01 -6.97685242e-01
2.15900585e-01 3.73501092e-01 -4.22673017e-01 -1.74441293e-01
-1.09074318e+00 1.45116270e+00 -3.81207839e-02 1.28881395e-01
-6.54043317e-01 -5.22755146e-01 -8.72367263e-01 -2.88880944e-01
-8.37168843e-02 -1.11580408e+00 1.31572390e+00 -7.83531189e-01
-1.81187439e+00 1.13778365e+00 -1.92850620e-01 -3.66622627e-01
1.16429651e+00 -4.21143591e-01 -4.39830780e-01 -6.53252304e-02
-1.04457267e-01 1.00913823e+00 9.38252747e-01 -1.17176998e+00
-6.37896180e-01 -3.02629381e-01 -9.47667807e-02 4.17655528e-01
1.07462116e-01 1.37624890e-01 -3.06176394e-01 -8.43479514e-01
2.33954802e-01 -1.21821439e+00 5.36466360e-01 6.47538662e-01
-2.54145533e-01 -2.09585831e-01 7.94547021e-01 -1.01043034e+00
8.19310784e-01 -1.85846412e+00 5.83728790e-01 2.42150053e-01
1.09363846e-01 3.61850351e-01 -4.68163252e-01 2.54536271e-01
8.53926167e-02 -2.01726407e-01 -5.05994081e-01 -5.53845644e-01
6.30800962e-01 5.58831930e-01 -2.15542570e-01 5.27087629e-01
8.30751434e-02 1.40196204e+00 -9.65174377e-01 -3.05526257e-01
3.14734370e-01 6.92745149e-01 -3.12912107e-01 -6.13164678e-02
9.34121944e-03 5.14778316e-01 -1.11410454e-01 9.36254144e-01
3.16313177e-01 1.79060735e-02 -7.03218998e-03 -2.89961815e-01
1.72411382e-01 2.60611534e-01 -6.69623375e-01 1.90151286e+00
-5.87777138e-01 8.72355223e-01 -1.83366880e-01 -5.43392897e-01
6.59334600e-01 3.77595991e-01 4.61815596e-01 -1.03456485e+00
2.29956955e-01 6.68354094e-01 2.56051064e-01 -4.09500599e-01
1.64290488e-01 -3.61209601e-01 -1.64723083e-01 6.70488119e-01
1.67166427e-01 -2.92682111e-01 3.06627303e-01 -1.78266078e-01
1.09413850e+00 6.17235899e-01 -4.06695064e-03 3.36964756e-01
2.38674372e-01 -1.36424586e-01 1.28513947e-01 4.66054499e-01
-2.86714047e-01 7.43012607e-01 3.13398480e-01 -3.02371502e-01
-1.06956911e+00 -1.64557087e+00 6.21784851e-02 1.14547539e+00
-1.13413192e-01 -3.08011435e-02 -5.03143370e-01 -8.72109056e-01
3.11359525e-01 8.05646122e-01 -7.34215319e-01 -1.51200265e-01
-1.02193677e+00 -6.50379956e-01 1.11573148e+00 8.82558167e-01
6.05140388e-01 -1.13450348e+00 -7.29490042e-01 -1.35076001e-01
-2.67647088e-01 -1.00730121e+00 -8.49444389e-01 1.49864972e-01
-6.07906461e-01 -9.43955600e-01 -1.35394287e+00 -1.01872611e+00
5.79234004e-01 -3.65688980e-01 8.28294158e-01 -2.43203819e-01
-3.60938869e-02 3.01739693e-01 -5.66669047e-01 -4.16118860e-01
-6.40691876e-01 -1.66075200e-01 1.89749801e-06 8.55446327e-03
2.61067837e-01 -4.63885278e-01 -4.13869828e-01 4.60266441e-01
-8.69169414e-01 1.12629674e-01 1.05137753e+00 1.17459047e+00
3.09220046e-01 -1.15370929e+00 1.34094030e-01 4.73853573e-02
6.05233729e-01 2.32610941e-01 -3.11260521e-01 5.08236527e-01
-5.80631733e-01 4.16688889e-01 3.34355533e-01 -4.73765910e-01
-6.22121930e-01 -1.10129258e-02 -1.11691974e-01 -2.50761271e-01
1.60775200e-01 2.62476683e-01 -8.84423330e-02 -1.93975717e-01
4.84766394e-01 4.83420193e-01 2.04684630e-01 -4.40944582e-01
7.98242033e-01 6.83024406e-01 7.72853732e-01 -3.19901377e-01
1.22510803e+00 5.29025853e-01 1.65287048e-01 -4.06731755e-01
-5.36602259e-01 -2.74908721e-01 -7.14670539e-01 -3.15022260e-01
8.72615159e-01 -6.67003751e-01 -7.33945727e-01 7.43043721e-01
-1.01050186e+00 -5.25468230e-01 -4.03142512e-01 6.57368481e-01
-8.57076883e-01 2.80095547e-01 -4.34359848e-01 -2.37660721e-01
-4.75301713e-01 -8.95072937e-01 1.43627143e+00 -1.46005973e-01
-5.78253806e-01 -8.90747309e-01 2.91256666e-01 4.74996626e-01
5.12964547e-01 3.92548382e-01 9.27271247e-01 -3.88846874e-01
-3.70177358e-01 -2.48811424e-01 -4.57793802e-01 5.18075049e-01
1.12687439e-01 -4.44867700e-01 -8.80217373e-01 -3.87154549e-01
-6.60434723e-01 -6.07622385e-01 1.07260871e+00 2.04353318e-01
2.10706696e-01 -1.20242707e-01 8.75183865e-02 8.88433874e-01
9.95775759e-01 -1.34591972e-02 6.57219231e-01 2.82236516e-01
7.72852600e-01 2.77295440e-01 3.97282958e-01 6.71645552e-02
5.99265277e-01 1.07457066e+00 1.84469745e-01 -2.55689561e-01
-9.54186201e-01 -6.20842993e-01 8.68584871e-01 9.16423380e-01
-2.99679279e-01 -2.89468646e-01 -6.59798443e-01 5.71964324e-01
-1.85343361e+00 -9.77239609e-01 8.06857869e-02 1.84155476e+00
8.27525318e-01 -7.51957744e-02 3.76657061e-02 2.71648794e-01
2.26667598e-01 3.38063270e-01 -6.88863218e-01 -5.51285923e-01
-3.78283173e-01 4.03322786e-01 6.11818075e-01 4.85453814e-01
-1.03939295e+00 1.04905272e+00 5.81124449e+00 4.36315894e-01
-1.34525156e+00 7.98910186e-02 -9.80806202e-02 -2.74306923e-01
-2.00402841e-01 -2.34161705e-01 -3.31871510e-01 2.96466678e-01
5.76167345e-01 8.75099078e-02 5.79081416e-01 3.33671898e-01
2.20690280e-01 1.93515405e-01 -1.46160734e+00 1.08706987e+00
4.76947337e-01 -8.15161943e-01 3.39824766e-01 -7.52035901e-02
8.13609898e-01 3.76805395e-01 3.06416661e-01 4.05450225e-01
3.49464089e-01 -1.05120540e+00 1.15532660e+00 4.44755703e-01
1.30027485e+00 -1.05204336e-01 7.09709704e-01 -1.44016132e-01
-1.21928108e+00 1.70128614e-01 2.44770765e-01 1.73970193e-01
5.99994481e-01 -3.15145007e-03 -8.44398975e-01 4.58280414e-01
3.78618866e-01 8.48415852e-01 -3.30139875e-01 1.07033384e+00
-8.87979627e-01 4.84062314e-01 -6.22215986e-01 -2.99813747e-01
5.09094894e-01 1.82078481e-02 7.54508078e-01 1.19137168e+00
4.45533305e-01 -4.55728322e-01 -1.21714905e-01 7.57027268e-01
-2.02124819e-01 2.71865148e-02 -3.86385888e-01 2.08014026e-02
-8.06161687e-02 5.29845536e-01 -1.91402838e-01 -3.99284482e-01
-1.90623224e-01 1.55600047e+00 -1.75154135e-01 4.72255677e-01
-1.07848561e+00 -4.44404364e-01 7.97944248e-01 -2.86214143e-01
3.62638324e-01 -1.24568321e-01 -4.27389711e-01 -1.13550329e+00
5.73413551e-01 -9.59512234e-01 2.15797022e-01 -9.20103192e-01
-1.14471900e+00 2.76526034e-01 -1.38185024e-01 -1.66593969e+00
-4.24281329e-01 -8.37020993e-01 -4.83488679e-01 7.62695968e-01
-1.66314411e+00 -1.88080096e+00 -2.45099068e-01 5.47733188e-01
2.16381699e-01 3.73377511e-03 8.70992303e-01 4.31394100e-01
1.18492134e-01 1.24257684e+00 2.77028888e-01 5.57938933e-01
1.08083546e+00 -1.41376162e+00 7.68644273e-01 7.57342935e-01
4.92340773e-01 3.48540574e-01 5.29229283e-01 -4.51882839e-01
-1.29303789e+00 -8.85490417e-01 1.39946020e+00 -6.22429311e-01
7.06077814e-01 -1.60365313e-01 -1.46236166e-01 6.31088912e-01
4.85860258e-02 1.95168760e-02 2.08383679e-01 -2.76835769e-01
-7.13155031e-01 -1.20078377e-01 -1.09187222e+00 8.65400016e-01
1.51627088e+00 -7.66916573e-01 -8.13098013e-01 3.13693434e-01
4.28226292e-01 -7.30317354e-01 -6.95227146e-01 3.39025468e-01
1.35330153e+00 -4.68133658e-01 8.82256627e-01 -7.55107284e-01
4.45088863e-01 -5.37363231e-01 -4.15451080e-01 -1.48970544e+00
1.52334943e-01 -6.98432028e-01 -2.02188119e-01 5.80575228e-01
6.28149927e-01 -6.12643182e-01 6.08810186e-01 6.56588450e-02
-2.29169145e-01 -6.13310158e-01 -1.40746129e+00 -1.09965694e+00
1.74638525e-01 -4.06199872e-01 3.59693885e-01 5.56103706e-01
-1.67213649e-01 1.52608261e-01 -7.27143645e-01 -1.73385471e-01
7.49328256e-01 6.13297336e-02 8.67631853e-01 -7.78104305e-01
-3.67047876e-01 -7.11566865e-01 -6.33673072e-01 -1.33165169e+00
-4.99176309e-02 -1.23960102e+00 1.80808857e-01 -1.85906315e+00
-7.95698017e-02 9.44715664e-02 -3.43567133e-01 8.10140312e-01
1.09373331e-01 4.74757224e-01 7.08974540e-01 1.56387389e-01
-3.95322472e-01 6.43048942e-01 1.62956989e+00 -5.09129822e-01
3.17003280e-02 -9.13256258e-02 -2.11854652e-01 5.11207223e-01
4.25635338e-01 -3.14177394e-01 2.06506271e-02 -8.95718038e-01
2.60982495e-02 -3.23963672e-01 5.56161940e-01 -9.72536922e-01
3.43138948e-02 3.28635603e-01 1.38956890e-01 -5.37972569e-01
5.41688800e-01 -6.21562779e-01 -2.32218251e-01 7.45532811e-01
-4.48964536e-01 1.18363626e-01 -2.22430021e-01 3.05158705e-01
-2.26407513e-01 3.53980184e-01 6.74365044e-01 2.62162179e-01
-8.36551011e-01 2.85141058e-02 -4.86074053e-02 1.30598634e-01
7.07152188e-01 -3.56386364e-01 -2.89204866e-01 -6.88074648e-01
-8.02271426e-01 3.54649842e-01 2.17764586e-01 7.44024694e-01
6.76014602e-01 -1.57703805e+00 -1.02753150e+00 1.48187742e-01
4.46012855e-01 -4.78228241e-01 -2.03429699e-01 1.14033246e+00
-6.13271594e-01 5.21671176e-01 -3.48638952e-01 -6.44903421e-01
-1.43797064e+00 -3.81607950e-01 3.81986141e-01 -3.59897912e-01
-6.87059224e-01 9.29162979e-01 -3.21929008e-01 -7.42673278e-01
5.80415249e-01 -8.43822420e-01 3.14753622e-01 -3.47288721e-03
3.67186487e-01 3.46872300e-01 -6.05609789e-02 -8.85035336e-01
-4.55933630e-01 1.19847417e+00 3.36348593e-01 -5.69741189e-01
1.24401474e+00 1.82202458e-01 1.85525522e-01 1.72239915e-01
1.40387082e+00 -2.75409400e-01 -1.12044215e+00 -2.10641414e-01
-2.50608362e-02 -2.02317566e-01 -2.95795083e-01 -1.60035789e+00
-9.48037446e-01 8.27391028e-01 9.69704866e-01 -7.97119558e-01
9.49369609e-01 4.10410389e-02 1.26324475e+00 5.85121095e-01
3.39285553e-01 -1.09867263e+00 6.67200759e-02 7.80926943e-01
1.21994746e+00 -1.35275340e+00 -1.94956169e-01 2.46716514e-01
-7.99975038e-01 9.65289176e-01 1.29491016e-01 1.46156438e-02
3.35250586e-01 1.04896948e-01 5.69144309e-01 -1.44700065e-01
-4.32304442e-01 -5.45743644e-01 7.19433010e-01 6.48814678e-01
3.85572851e-01 3.15187931e-01 -4.30225372e-01 -4.72282544e-02
-5.49077272e-01 1.48349091e-01 -5.95716909e-02 9.25535321e-01
-1.18620075e-01 -1.27787006e+00 -3.26012611e-01 2.40608662e-01
9.82803181e-02 1.50680344e-03 -7.90856302e-01 1.01668072e+00
1.79951236e-01 5.79929650e-01 -2.48276696e-01 -5.32717824e-01
6.86366916e-01 3.19015414e-01 8.97844136e-01 -7.99002573e-02
-6.45348370e-01 -2.30553165e-01 3.33501995e-01 -6.56022072e-01
-5.59273481e-01 -7.81815052e-01 -1.35158300e+00 -1.93054348e-01
-9.05160513e-03 -5.45604646e-01 8.77172470e-01 1.18314183e+00
-1.38904989e-01 4.80455846e-01 1.69060886e-01 -6.93181157e-01
-1.01430511e+00 -1.10757673e+00 -3.21033865e-01 7.99641430e-01
4.66309458e-01 -6.71715975e-01 -3.29880625e-01 -2.80978624e-04] | [9.210980415344238, -6.5356621742248535] |
9637cb78-b8ad-4103-9695-9ab87da79264 | where-are-we-in-discourse-relation | null | null | https://aclanthology.org/2021.sigdial-1.34 | https://aclanthology.org/2021.sigdial-1.34.pdf | Where Are We in Discourse Relation Recognition? | Discourse parsers recognize the intentional and inferential relationships that organize extended texts. They have had a great influence on a variety of NLP tasks as well as theoretical studies in linguistics and cognitive science. However it is often difficult to achieve good results from current discourse models, largely due to the difficulty of the task, particularly recognizing implicit discourse relations. Recent developments in transformer-based models have shown great promise on these analyses, but challenges still remain. We present a position paper which provides a systematic analysis of the state of the art discourse parsers. We aim to examine the performance of current discourse parsing models via gradual domain shift: within the corpus, on in-domain texts, and on out-of-domain texts, and discuss the differences between the transformer-based models and the previous models in predicting different types of implicit relations both inter- and intra-sentential. We conclude by describing several shortcomings of the existing models and a discussion of how future work should approach this problem. | ['Malihe Alikhani', 'Junyi Jessy Li', 'Katherine Atwell'] | null | null | null | null | sigdial-acl-2021-7 | ['discourse-parsing', 'implicit-relations'] | ['natural-language-processing', 'natural-language-processing'] | [ 2.99752593e-01 9.34676886e-01 -3.72160614e-01 -5.06325245e-01
-5.70673525e-01 -7.09335148e-01 1.18429065e+00 5.17780960e-01
-2.44805947e-01 9.24289703e-01 1.03142989e+00 -7.80940711e-01
-2.07776606e-01 -7.80471981e-01 -1.26033038e-01 -2.69392610e-01
-9.02563557e-02 7.20199764e-01 7.01659441e-01 -5.60875237e-01
7.24781573e-01 -5.58580793e-02 -1.45928764e+00 8.82332683e-01
8.69026065e-01 1.26970857e-01 2.50554770e-01 5.33712924e-01
-6.58810973e-01 1.61919403e+00 -9.96698380e-01 -4.68435019e-01
-5.84616482e-01 -7.43753970e-01 -1.54749322e+00 -3.46082561e-02
1.28999641e-02 -4.56291158e-03 -1.14841558e-01 8.30615222e-01
3.10940146e-01 5.57101481e-02 5.47094345e-01 -7.69556820e-01
-5.81755459e-01 1.13079274e+00 -1.98637873e-01 6.84512675e-01
8.15293372e-01 -4.43450570e-01 1.15226555e+00 -4.24767971e-01
1.23002684e+00 1.60348856e+00 5.59447885e-01 5.98124325e-01
-1.01769602e+00 -1.75755382e-01 4.57405061e-01 6.57044232e-01
-4.75204289e-01 -6.99417055e-01 8.29888642e-01 -6.01199389e-01
1.60766566e+00 4.92347062e-01 3.82754803e-01 8.79170537e-01
-5.30394167e-02 7.93387175e-01 1.28961086e+00 -1.07358670e+00
-6.65006042e-02 1.39152527e-01 6.74112558e-01 1.69259980e-01
-5.29880337e-02 -2.33016059e-01 -7.78657079e-01 -6.00574166e-02
5.26850879e-01 -1.01408875e+00 -2.51676053e-01 4.82608229e-02
-1.05287468e+00 1.04684830e+00 -5.59291169e-02 1.24797821e+00
-1.72887012e-01 -4.33629304e-01 6.70151591e-01 4.09770757e-01
7.53673434e-01 5.10873735e-01 -3.10920447e-01 -5.28215230e-01
-6.32743955e-01 6.27686679e-01 1.22199225e+00 6.59536004e-01
1.84361935e-01 -2.99139649e-01 7.20820576e-02 8.23238611e-01
5.42755127e-01 -1.30269468e-01 3.94233674e-01 -1.07499218e+00
9.06915486e-01 7.75817573e-01 -8.40497296e-03 -1.05921304e+00
-3.91258210e-01 2.73078442e-01 -1.17485978e-01 -4.09833118e-02
7.83352256e-01 2.39387676e-02 -2.06202045e-01 1.60143197e+00
3.16243082e-01 -3.93635154e-01 2.63145804e-01 6.28796279e-01
1.19499254e+00 7.41334081e-01 5.35677850e-01 -6.72584713e-01
1.62957549e+00 -8.05452585e-01 -1.32446718e+00 -4.13894743e-01
1.12520480e+00 -8.82512927e-01 7.83982337e-01 -9.36991796e-02
-1.45567226e+00 -1.25053585e-01 -8.10458183e-01 -4.86037672e-01
-1.88378200e-01 -1.86103284e-01 6.05439067e-01 6.66538298e-01
-7.53165901e-01 3.34894449e-01 -8.22615445e-01 -3.88696730e-01
2.17751592e-01 2.03377977e-01 -6.44136891e-02 3.00876826e-01
-1.43254876e+00 1.50847340e+00 7.24395931e-01 -8.89396295e-02
3.66422087e-02 -4.59592164e-01 -9.74986136e-01 -4.32625450e-02
5.57579458e-01 -2.62574255e-01 1.61096668e+00 -1.12994230e+00
-1.48361063e+00 1.38943708e+00 -5.69942474e-01 -5.86362243e-01
3.06929618e-01 -4.02058482e-01 -3.13315958e-01 -8.32424611e-02
2.93438435e-01 -8.23023841e-02 -3.27699594e-02 -9.43521202e-01
-8.71432602e-01 -4.31262523e-01 3.77400428e-01 6.20279253e-01
-1.24519514e-02 8.31134558e-01 1.43632293e-01 -4.51961547e-01
1.71348378e-01 -6.15530193e-01 1.37388542e-01 -6.17333055e-01
-1.74017653e-01 -9.58201349e-01 9.37540352e-01 -7.31693804e-01
1.75866485e+00 -1.80510128e+00 3.44125807e-01 -4.51010913e-01
1.83961585e-01 5.49660623e-01 4.44781393e-01 8.26896548e-01
-1.52415559e-01 2.89771020e-01 -2.05724031e-01 -1.32071927e-01
1.14523873e-01 4.79629487e-01 -5.34085631e-01 1.25841901e-01
2.22052678e-01 8.15953076e-01 -9.42309856e-01 -5.86052120e-01
2.83046097e-01 6.79967999e-02 -1.75524548e-01 2.68290460e-01
-6.83667243e-01 4.07576978e-01 -6.60637796e-01 7.91988745e-02
1.33339912e-01 -1.04156099e-01 1.05449414e+00 3.68558228e-01
-6.57089710e-01 1.63014412e+00 -8.40152979e-01 1.38564289e+00
7.12301359e-02 1.14994764e+00 2.00847089e-01 -1.42650652e+00
6.97891772e-01 6.95372462e-01 -7.85611495e-02 -7.54022300e-01
1.70152113e-01 2.14416623e-01 6.19022369e-01 -8.56200874e-01
7.54270732e-01 -5.66674232e-01 -8.83115605e-02 7.27695346e-01
-2.87242323e-01 1.01813063e-01 5.64633727e-01 2.99788684e-01
8.41617703e-01 2.93905497e-01 8.53362262e-01 -5.25042832e-01
9.35757995e-01 5.58362305e-01 5.57933688e-01 2.16022432e-01
-6.44372925e-02 1.46498844e-01 1.10432458e+00 -3.25426280e-01
-9.48764384e-01 -4.18638110e-01 -2.81054974e-01 1.35605228e+00
-2.29872912e-02 -6.20793164e-01 -7.43996382e-01 -5.90341628e-01
-4.52669650e-01 1.24029064e+00 -6.85490012e-01 6.10489488e-01
-1.60210562e+00 -9.08146560e-01 6.09889150e-01 5.53331196e-01
3.38695467e-01 -1.32887387e+00 -7.99047232e-01 4.97909129e-01
-4.92817342e-01 -1.13180053e+00 5.97567797e-01 4.05030465e-03
-7.41054654e-01 -1.40282369e+00 1.22868363e-02 -1.02107191e+00
1.69238180e-01 -1.43245962e-02 1.29670942e+00 4.21560109e-01
4.83342469e-01 4.13038023e-02 -6.79796934e-01 -6.30579770e-01
-1.04693019e+00 3.17767859e-01 -5.92751682e-01 -8.26843321e-01
8.50122750e-01 -2.74497181e-01 1.62748575e-01 -2.50906311e-02
-5.76610804e-01 1.29772007e-01 -1.50270194e-01 7.88159311e-01
-2.14312211e-01 -1.38515353e-01 7.22331583e-01 -1.63765657e+00
1.14629519e+00 -6.20466232e-01 -2.84868509e-01 3.25708389e-01
-2.88030714e-01 4.65517975e-02 1.06215790e-01 -1.29905090e-01
-1.79869711e+00 -7.41225123e-01 -4.29552227e-01 7.07269192e-01
-9.62507725e-02 7.99080014e-01 -1.22762091e-01 6.24992371e-01
8.10545027e-01 -3.38453680e-01 7.06794113e-02 -3.74142021e-01
2.93613195e-01 5.89150846e-01 2.66782254e-01 -9.21644807e-01
3.65945250e-01 5.10958470e-02 -3.33890378e-01 -1.26580739e+00
-1.04209256e+00 -4.79384810e-01 -9.42631006e-01 -8.14950019e-02
1.00824547e+00 -5.75368166e-01 -3.49382788e-01 2.16160461e-01
-1.60641134e+00 -5.75053215e-01 -2.84084946e-01 2.28360146e-01
-4.99671280e-01 6.67263925e-01 -8.55194390e-01 -9.32197213e-01
-1.07823938e-01 -8.71064186e-01 5.34093261e-01 1.66027546e-01
-1.01643598e+00 -1.63331819e+00 2.04199299e-01 4.77366447e-01
2.15552077e-01 3.51463050e-01 1.45535946e+00 -1.00609601e+00
-4.15012717e-01 5.31774759e-01 -5.08441404e-02 -1.16528787e-01
4.65470180e-02 -6.38979599e-02 -7.64471889e-01 3.50818306e-01
3.43243837e-01 -4.19512987e-01 4.04358149e-01 2.74408251e-01
2.25117266e-01 -2.38899857e-01 -4.56618607e-01 -2.51511574e-01
9.32811141e-01 4.07382011e-01 8.81585419e-01 9.11715150e-01
1.67433530e-01 1.23438430e+00 8.93384218e-01 -1.04939453e-01
7.59553790e-01 5.02681971e-01 -7.24101067e-02 4.63662356e-01
-3.62549603e-01 -4.04589251e-02 2.22219840e-01 1.03520119e+00
-3.74154657e-01 -2.31511265e-01 -1.15225911e+00 7.66602278e-01
-2.26363897e+00 -1.20517397e+00 -7.97951460e-01 1.62361848e+00
1.17189837e+00 2.44105116e-01 2.51636207e-02 4.32669312e-01
7.09801316e-01 5.47994316e-01 1.47117481e-01 -7.90745378e-01
-2.40087554e-01 1.90116182e-01 -4.53861058e-01 9.19658422e-01
-1.01252580e+00 1.18347466e+00 6.77891016e+00 3.13270092e-01
-8.11079085e-01 2.56608516e-01 2.68416017e-01 4.63195175e-01
-3.93872291e-01 2.70304680e-01 -7.76619077e-01 1.18590377e-01
1.15614200e+00 -3.93101811e-01 -1.81973085e-01 6.13556266e-01
-1.97317395e-02 -4.92108792e-01 -1.29577744e+00 2.88547039e-01
1.48266301e-01 -1.71086752e+00 -3.26798260e-01 -5.08630909e-02
3.31229866e-01 -2.33195424e-01 -4.89983797e-01 4.56630111e-01
1.94022626e-01 -9.03363466e-01 7.11474419e-01 -1.41797930e-01
2.51658171e-01 -1.04471758e-01 7.52528489e-01 7.12758422e-01
-8.22070122e-01 1.15648851e-01 -2.05483884e-01 -8.55303586e-01
4.71218348e-01 8.22160114e-03 -9.07670319e-01 4.50634956e-01
4.11474824e-01 5.62581658e-01 -6.28378838e-02 3.53118241e-01
-7.85434544e-01 8.22361529e-01 -5.17312214e-02 -4.58393931e-01
1.54435351e-01 -1.15769789e-01 8.46961021e-01 1.40152740e+00
-3.03158253e-01 8.42293978e-01 -2.22417973e-02 6.25714004e-01
2.77041525e-01 1.55060187e-01 -4.29979265e-01 -3.64873447e-02
6.65290654e-01 6.69441044e-01 -8.08391154e-01 -4.75518227e-01
-5.45569837e-01 8.07071179e-02 4.90748137e-01 -3.97387519e-02
-3.54071409e-01 1.11312151e-01 2.71734953e-01 3.52234274e-01
9.42901298e-02 -3.53157610e-01 -4.52342510e-01 -1.03699076e+00
4.40868437e-02 -1.17759120e+00 4.81780708e-01 -4.05399621e-01
-1.10467875e+00 4.94522840e-01 5.37268877e-01 -5.84802151e-01
-7.86255836e-01 -6.45007610e-01 -6.33961916e-01 8.68514717e-01
-1.43091381e+00 -1.06561959e+00 1.43943459e-01 1.49170592e-01
7.87184536e-01 1.07690029e-01 1.18376791e+00 3.40737440e-02
-1.42078519e-01 -6.62987530e-02 -2.97938377e-01 5.17799854e-01
6.47085488e-01 -1.26458073e+00 4.84397769e-01 7.58341849e-01
-8.55119750e-02 7.78557479e-01 9.98399556e-01 -5.51778793e-01
-9.22345698e-01 -3.54050189e-01 1.67281985e+00 -7.90014923e-01
1.10851526e+00 -7.32750222e-02 -1.38537073e+00 1.11265600e+00
8.58602047e-01 -6.35223091e-01 8.88426781e-01 7.84238696e-01
-1.61907911e-01 6.87754214e-01 -8.41208041e-01 4.44440663e-01
9.63926613e-01 -4.41204488e-01 -1.91435897e+00 2.77664751e-01
5.30523121e-01 -8.80533338e-01 -9.26151574e-01 1.50682375e-01
3.09704423e-01 -1.12203014e+00 5.90454280e-01 -7.18297482e-01
7.96193838e-01 -7.79603943e-02 4.86572683e-02 -9.28597569e-01
-1.29199028e-01 -5.02409160e-01 -1.59136519e-01 1.46649492e+00
4.49855983e-01 -7.05787480e-01 6.49260938e-01 8.80225301e-01
-4.12250161e-01 -4.04018700e-01 -9.99957800e-01 -3.71888220e-01
6.22485161e-01 -4.44756031e-01 2.09536970e-01 1.18566322e+00
9.70231831e-01 1.13633049e+00 2.75631934e-01 -2.75662392e-01
3.00344706e-01 2.11864725e-01 5.80812216e-01 -1.43934810e+00
-1.53703019e-01 -5.19576311e-01 -1.34416893e-01 -1.07986546e+00
5.03772378e-01 -6.43093050e-01 -1.05988897e-01 -1.84705126e+00
-6.29404560e-02 -6.06683850e-01 5.61326027e-01 2.94503003e-01
-1.53578594e-01 -3.94583344e-01 3.36130321e-01 3.12767953e-01
-4.40358430e-01 3.52443337e-01 9.98379469e-01 1.39766484e-01
-4.81372207e-01 -2.36087352e-01 -8.17818165e-01 1.13744760e+00
7.55305648e-01 -4.05310243e-01 -4.57118273e-01 -6.34526312e-01
2.84497708e-01 3.35522801e-01 -8.52036849e-02 -2.94124573e-01
3.25014532e-01 -2.68029690e-01 -1.62588820e-01 -6.27237022e-01
1.40878707e-01 -2.38275841e-01 -2.38998145e-01 3.02688777e-01
-5.58278024e-01 8.90946537e-02 2.72561580e-01 1.49063647e-01
-4.64064091e-01 -6.48162901e-01 3.67812216e-01 -2.76477575e-01
-9.32432592e-01 -8.36615026e-01 -7.46343791e-01 5.61130166e-01
1.08877265e+00 -2.28031248e-01 -9.36609209e-01 -1.51155978e-01
-8.09135258e-01 9.39406082e-02 2.02325419e-01 5.16562641e-01
2.11000323e-01 -6.92967534e-01 -8.38271618e-01 -2.88743615e-01
-2.76139349e-01 2.58796275e-01 -1.05397500e-01 8.72695088e-01
-5.09190321e-01 7.72289157e-01 -7.23103434e-02 -4.47667420e-01
-1.85576785e+00 2.28920132e-01 7.50535429e-02 -7.37952828e-01
-7.31662273e-01 5.72257578e-01 1.07469529e-01 -2.74699628e-01
-2.42213178e-02 -1.77251667e-01 -9.02845800e-01 2.69259781e-01
6.81704760e-01 3.12149435e-01 -5.92157692e-02 -9.76118863e-01
-4.30446178e-01 2.42530122e-01 -2.27591842e-01 -2.63085723e-01
1.39375234e+00 -3.00812691e-01 -6.55950427e-01 7.57117927e-01
4.89852995e-01 1.23701319e-01 -5.20716310e-01 -2.97036737e-01
6.60430908e-01 -1.20916560e-01 -2.26218387e-01 -7.36138761e-01
5.15469834e-02 9.42750394e-01 -3.56843054e-01 9.97733176e-01
7.28774667e-01 3.29008847e-01 4.20278788e-01 1.23984322e-01
2.53770560e-01 -1.29822123e+00 -1.11895666e-01 1.03346312e+00
9.44469094e-01 -8.32821786e-01 4.41273481e-01 -1.12583435e+00
-6.86821282e-01 1.26165915e+00 4.81028765e-01 -9.26772356e-02
5.05903184e-01 4.69881386e-01 1.79908410e-01 -5.59409499e-01
-1.00844431e+00 -1.15014419e-01 8.22203048e-03 6.87136590e-01
1.35619545e+00 2.12579016e-02 -1.12565327e+00 3.59366685e-01
-4.65635896e-01 -2.22808719e-01 6.68898046e-01 1.19855285e+00
-5.81865132e-01 -1.77994549e+00 -5.76043367e-01 1.60641335e-02
-7.24374473e-01 -7.43460953e-02 -7.46970713e-01 1.38607574e+00
-8.42547566e-02 1.20047307e+00 2.51766086e-01 1.28607199e-01
3.05175960e-01 2.42493272e-01 6.55682743e-01 -1.02462387e+00
-9.08020377e-01 5.10411449e-02 1.32585239e+00 -4.62590232e-02
-1.31908917e+00 -1.21133935e+00 -1.59663522e+00 -4.21040505e-01
-3.75013858e-01 4.52251256e-01 2.68231869e-01 1.52076089e+00
-1.23336188e-01 6.06028855e-01 -1.82472095e-01 -4.69205201e-01
-4.26190734e-01 -1.26279092e+00 -8.96264315e-02 3.10260832e-01
-2.05097981e-02 -6.86583996e-01 -2.00752452e-01 2.36482725e-01] | [10.748438835144043, 9.449780464172363] |
2eab4122-8d3f-4f05-b45c-341a6df838f7 | amr-to-text-generation-with-graph-structure | null | null | https://openreview.net/forum?id=6zZEZD2LsxW | https://openreview.net/pdf?id=6zZEZD2LsxW | AMR-to-text Generation with Graph Structure Reconstruction and Coverage | Generating text from semantic representations such as AMR is a challenging task. Previous research formalizes this task as a graph-to-sequence learning problem and uses various graph neural networks to model the graph structure. Recently, methods based on pre-trained models improve the performance significantly due to pre-trained on a large text corpus. However, these pre-trained model-based methods take linearized AMR graphs as input and may lose the information of graph structure. In addition, these methods don't consider the coverage of the AMR graph. Therefore, some nodes in the graph may be lost or repeated in the generated text. To address these problems, we propose a graph structure and coverage enhanced model for this task. To enhance the information of graph structure, we design two auxiliary objectives, relationship prediction and distance prediction of nodes in AMR graphs. To consider the coverage of AMR graphs, we design a coverage mechanism to solve the problem of information under-translation or over-translation in AMR-to-text generation.
Experimental results on three standard datasets show that our proposed method outperforms the existing methods significantly. | ['Anonymous'] | 2021-09-17 | null | null | null | acl-arr-september-2021-9 | ['graph-to-sequence'] | ['natural-language-processing'] | [ 4.65469152e-01 6.97410583e-01 -4.36764866e-01 -1.64265856e-01
-5.56949437e-01 -2.17848793e-01 5.11167169e-01 9.11381245e-02
1.90496415e-01 9.14751470e-01 4.74757344e-01 -4.08184260e-01
2.41827726e-01 -1.30422342e+00 -7.88270533e-01 -1.55256033e-01
3.79640192e-01 6.60363197e-01 1.73379898e-01 -5.02658546e-01
2.55818784e-01 -2.80898493e-02 -8.80568862e-01 3.12159389e-01
1.30450904e+00 4.85631943e-01 4.91838664e-01 3.16581994e-01
-8.63650739e-01 9.59442496e-01 -8.23828995e-01 -5.58128953e-01
-4.41567600e-02 -7.80078471e-01 -8.87431681e-01 3.07864040e-01
-7.08267167e-02 -5.67648746e-02 -6.78469598e-01 1.30757117e+00
4.25765842e-01 2.02641487e-02 7.63297617e-01 -1.24013507e+00
-1.19631493e+00 1.35640919e+00 -7.02580035e-01 -1.99195500e-02
5.29366493e-01 -3.26062292e-01 1.13540590e+00 -8.22907090e-01
7.65541911e-01 1.47037983e+00 5.28309464e-01 5.06338418e-01
-6.88846648e-01 -5.16249239e-01 3.95695955e-01 1.55409604e-01
-1.41845572e+00 -2.01256827e-01 1.09154141e+00 -1.74668759e-01
9.54883933e-01 1.31300658e-01 6.05452657e-01 9.64889407e-01
7.26461504e-03 7.73410797e-01 6.62647724e-01 -5.08059323e-01
-1.08144775e-01 -6.46695569e-02 -7.37511516e-02 8.58590603e-01
5.94506323e-01 -4.18014914e-01 -1.66898385e-01 3.66526619e-02
7.71175981e-01 -8.79965648e-02 -3.93603116e-01 -1.35330960e-01
-1.02476287e+00 9.77695465e-01 5.98415196e-01 2.64615983e-01
-7.77946711e-02 2.50839442e-03 3.81705493e-01 1.51255503e-01
6.84345245e-01 4.01484966e-01 -2.40738615e-01 2.95022428e-01
-5.71704209e-01 4.77949437e-03 6.98860288e-01 1.58366776e+00
9.14248109e-01 1.66440561e-01 -2.67291039e-01 9.12858665e-01
5.08072436e-01 2.07235649e-01 7.96070099e-01 -1.82952985e-01
1.25895858e+00 1.20607018e+00 -1.39499307e-01 -1.32731366e+00
-2.43018180e-01 -6.17240429e-01 -1.12552178e+00 -7.15674400e-01
-1.59747049e-03 -3.39201957e-01 -1.13227606e+00 1.53163612e+00
3.50102037e-01 2.32645776e-02 3.06049764e-01 7.64671206e-01
1.14046264e+00 1.11406171e+00 -1.92270920e-01 -4.24178571e-01
1.01371634e+00 -1.43666649e+00 -1.08683848e+00 -4.78520960e-01
1.14855468e+00 -7.91327775e-01 9.67873275e-01 -2.27013215e-01
-8.79659772e-01 -6.72941089e-01 -1.07369864e+00 -6.39901310e-02
-4.91831452e-01 2.01228067e-01 3.96824300e-01 4.18396264e-01
-9.50013041e-01 5.78528821e-01 -2.97880411e-01 -2.25095883e-01
2.95448840e-01 1.06292248e-01 -9.74551961e-02 -4.45763320e-01
-1.68286836e+00 7.39627242e-01 9.40350711e-01 3.32457811e-01
-5.46510935e-01 -2.23732963e-01 -1.17361510e+00 1.17662633e-02
6.30216479e-01 -7.84413040e-01 1.01023233e+00 -1.12287092e+00
-1.20277846e+00 4.12788302e-01 -1.18725933e-01 -2.15499431e-01
4.12612170e-01 1.44126996e-01 -5.64437032e-01 -2.11311746e-02
1.72187969e-01 3.15886825e-01 5.72932422e-01 -1.20217621e+00
-3.36144835e-01 -2.74276227e-01 2.21554428e-01 6.58219039e-01
-2.15351015e-01 -4.00809705e-01 -5.50384223e-01 -9.56109703e-01
2.04098389e-01 -6.77606046e-01 -3.52150232e-01 -6.58057749e-01
-6.86177552e-01 -4.16325778e-01 6.86480343e-01 -9.61867511e-01
1.44320452e+00 -1.58916342e+00 2.28843227e-01 1.12090655e-01
3.18641663e-01 2.65662014e-01 -4.48795736e-01 7.89290130e-01
9.32715309e-04 5.08306563e-01 -1.07969634e-01 -4.44231927e-03
-1.90409318e-01 1.42767921e-01 -2.45362997e-01 -5.23018427e-02
1.33479595e-01 1.03841841e+00 -9.71943974e-01 -8.45281601e-01
-1.08628925e-02 2.15416357e-01 -2.03016073e-01 4.70758945e-01
-5.42497218e-01 2.36789331e-01 -8.27325702e-01 4.51094687e-01
7.47248352e-01 -5.29766858e-01 2.66574681e-01 -1.69176549e-01
5.11405766e-01 4.46602255e-01 -9.21892285e-01 1.75724375e+00
-3.91333252e-01 2.00776234e-01 -4.31337088e-01 -1.28621352e+00
1.36443484e+00 2.76409775e-01 1.28573000e-01 -5.93773425e-01
8.55201408e-02 2.36372143e-01 6.50759134e-03 -4.30041611e-01
6.82029426e-01 -4.14369740e-02 1.18781701e-01 4.80201989e-01
-9.78032500e-02 -6.50906935e-02 2.38353744e-01 5.35226405e-01
1.01456726e+00 2.20779821e-01 2.02201113e-01 1.10697150e-01
6.80390179e-01 2.14624912e-01 5.72130322e-01 5.40810227e-01
2.70301729e-01 6.42298341e-01 5.44656813e-01 -2.10694179e-01
-1.00099993e+00 -4.63491112e-01 4.69109863e-01 7.71622241e-01
4.06156391e-01 -6.52067780e-01 -9.90007520e-01 -1.04869092e+00
-4.81532365e-01 9.59961057e-01 -4.55293745e-01 -5.79409897e-01
-7.59930193e-01 -8.14610898e-01 2.63671845e-01 5.05423009e-01
5.28407514e-01 -1.09554982e+00 4.22072828e-01 4.41716284e-01
-7.49912143e-01 -1.22343814e+00 -8.11762154e-01 -6.31618381e-01
-9.90165710e-01 -9.10430491e-01 -7.93296456e-01 -1.12407291e+00
1.17025566e+00 4.57303584e-01 1.17073750e+00 4.18578297e-01
5.17611243e-02 -1.12786770e-01 -8.67152035e-01 -3.28397334e-01
-7.07362354e-01 4.39575046e-01 -4.27237749e-01 -1.94678202e-01
1.73062220e-01 -3.78195077e-01 -4.03244972e-01 3.03777993e-01
-8.35610271e-01 5.54418504e-01 8.03504765e-01 7.58400798e-01
6.84830248e-01 1.99697226e-01 9.71188962e-01 -1.12672591e+00
9.83362973e-01 -5.46245456e-01 -3.90173972e-01 6.27071679e-01
-9.13754284e-01 2.60816485e-01 9.95491982e-01 -4.12863553e-01
-1.11544383e+00 -1.74809426e-01 -3.26102450e-02 -2.91901618e-01
3.66164476e-01 9.75618184e-01 -5.10528147e-01 3.03840399e-01
3.14242363e-01 6.18457973e-01 -1.04286633e-01 -3.42186183e-01
6.05683327e-01 7.84407616e-01 -1.17274383e-02 -3.87719542e-01
7.79182136e-01 -5.14081828e-02 2.14839634e-02 -4.83234435e-01
-9.69404697e-01 -2.48353586e-01 -4.08344328e-01 -5.01670875e-02
7.15929866e-01 -1.00036776e+00 1.17743723e-01 1.21297158e-01
-1.37147844e+00 -4.25083674e-02 -9.49294418e-02 3.44568521e-01
-4.10631865e-01 8.35626900e-01 -5.21215260e-01 -7.63739169e-01
-6.87365234e-01 -9.38477635e-01 1.01973116e+00 2.14821935e-01
1.05380110e-01 -1.12885094e+00 3.30018364e-02 4.21766996e-01
1.38470083e-01 1.70111701e-01 1.14076698e+00 -7.20930517e-01
-6.77848339e-01 -2.74587810e-01 -4.20361251e-01 5.56397699e-02
4.15245682e-01 -2.68777341e-01 -3.71928215e-01 -2.19223812e-01
-2.59899497e-01 -2.90041655e-01 8.19299281e-01 1.16338454e-01
1.18004715e+00 -8.09025884e-01 -5.74578881e-01 2.93115169e-01
1.31324542e+00 1.59491867e-01 7.34221041e-01 5.41463196e-02
1.33051968e+00 7.74111092e-01 7.36941218e-01 4.56525497e-02
6.05576396e-01 5.17216206e-01 3.97556096e-01 -1.37686864e-01
-2.81133592e-01 -1.07640147e+00 3.05125862e-01 1.39018226e+00
-9.01314169e-02 -7.77921677e-01 -6.41981423e-01 4.26832169e-01
-2.12317014e+00 -5.94625890e-01 -4.30629283e-01 1.88132584e+00
7.81875551e-01 1.78032562e-01 -1.66421384e-01 -1.28889158e-01
1.30904996e+00 2.49668106e-01 -3.32321614e-01 -2.53502488e-01
-8.34714100e-02 -2.74075657e-01 3.89988154e-01 4.67016697e-01
-5.48728049e-01 1.26943743e+00 5.76977968e+00 1.07390463e+00
-5.68625987e-01 -1.06694944e-01 6.08129859e-01 5.48422396e-01
-8.08119237e-01 2.41275743e-01 -8.97470176e-01 4.58636612e-01
6.78271711e-01 -5.37614942e-01 4.12735194e-01 8.33815634e-01
1.17541127e-01 5.58225513e-01 -8.00795853e-01 9.40627098e-01
4.43348438e-01 -1.23756731e+00 8.56703460e-01 -3.75488028e-02
7.77131379e-01 -4.05852586e-01 -4.97197419e-01 6.55161858e-01
5.13254642e-01 -1.19641578e+00 3.89014214e-01 3.14826697e-01
7.15939939e-01 -8.24266136e-01 9.78864193e-01 5.13903320e-01
-1.49987888e+00 3.19379419e-01 -1.01801682e+00 -1.47365537e-02
2.24012971e-01 7.51127303e-01 -1.03397048e+00 1.23497021e+00
-1.15216905e-02 8.71686280e-01 -6.88938022e-01 5.74519575e-01
-5.29891849e-01 4.25868839e-01 1.56878546e-01 -3.23911607e-01
2.85309613e-01 -4.85994488e-01 4.63535339e-01 8.88047516e-01
6.68336809e-01 -7.81091535e-03 3.92565995e-01 1.03547800e+00
-4.87036556e-01 6.20377839e-01 -9.73687470e-01 -4.43149656e-01
4.77044880e-01 1.20544434e+00 -6.38376653e-01 -4.96976852e-01
-5.11250317e-01 9.36034262e-01 5.47050536e-01 3.67857307e-01
-7.97502518e-01 -6.44789219e-01 -2.40476102e-01 3.81564528e-01
2.74053421e-02 5.47283469e-03 2.82053854e-02 -1.36042571e+00
1.14018574e-01 -8.77858162e-01 5.44041932e-01 -1.06097794e+00
-1.24560821e+00 7.44980812e-01 -7.50548467e-02 -1.10078573e+00
-3.08129847e-01 -2.05885023e-01 -6.37039125e-01 9.22194362e-01
-1.39890289e+00 -1.40445030e+00 -2.37696812e-01 4.45423394e-01
8.39990973e-01 -2.75397271e-01 5.56709886e-01 8.76830518e-03
-5.77384114e-01 5.40367007e-01 -1.78901702e-01 3.97330910e-01
3.12642038e-01 -1.14547753e+00 6.78967118e-01 8.95715535e-01
1.19502164e-01 4.13206369e-01 4.80259895e-01 -1.21601927e+00
-1.32197750e+00 -1.62198889e+00 9.65399325e-01 -1.79271072e-01
5.17778754e-01 -2.94394910e-01 -1.00943923e+00 9.97276723e-01
3.38570625e-01 -3.79780233e-01 4.24877793e-01 2.45329402e-02
-8.32603425e-02 1.13193765e-01 -8.56330395e-01 7.80757785e-01
1.53776574e+00 -1.37969092e-01 -8.92223239e-01 7.50649273e-01
1.33514798e+00 -3.90543222e-01 -6.29349351e-01 5.03029168e-01
-1.27965793e-01 -3.68326247e-01 7.35353589e-01 -5.54771066e-01
6.41843617e-01 -3.96791667e-01 1.38140112e-01 -1.62784958e+00
-2.32364655e-01 -6.64136350e-01 -2.80536473e-01 1.42080247e+00
6.25280797e-01 -6.43481135e-01 7.61047721e-01 1.02939740e-01
-3.35703433e-01 -6.65052712e-01 -5.30865431e-01 -7.03461647e-01
-2.72184629e-02 2.43741825e-01 8.61225665e-01 1.09496582e+00
7.73453712e-02 1.18739641e+00 -4.94323939e-01 -4.80955914e-02
4.92735535e-01 2.08424300e-01 7.17570186e-01 -1.04764140e+00
-8.09488669e-02 -2.09481254e-01 -1.49669036e-01 -1.18695509e+00
2.36454278e-01 -1.32734811e+00 -7.01013803e-02 -2.44641376e+00
3.12723368e-01 -3.07314336e-01 1.34963900e-01 3.47783834e-01
-5.87296486e-01 -3.82221371e-01 -2.32890978e-01 3.32929075e-01
-5.72271526e-01 8.87449920e-01 1.78909588e+00 -3.22964907e-01
-1.87714472e-01 -2.38235205e-01 -9.23029304e-01 5.46680152e-01
9.40077364e-01 -4.55618650e-01 -9.63708997e-01 -5.86961746e-01
4.01965350e-01 4.17167872e-01 -2.32912213e-01 -5.94969749e-01
-3.65882777e-02 -3.24140489e-01 2.67850906e-01 -7.87756264e-01
-8.09119567e-02 -5.18485427e-01 1.73644945e-01 3.53177309e-01
-4.10419524e-01 2.48589322e-01 -2.15829730e-01 8.19541097e-01
-2.78784454e-01 -4.34375942e-01 4.49762195e-01 -5.48004329e-01
-4.13503617e-01 5.91721177e-01 -2.78324895e-02 3.23676676e-01
7.63421059e-01 -1.42025426e-01 -5.83000600e-01 -6.39697731e-01
-3.90141994e-01 4.45040077e-01 2.13821903e-01 6.20854974e-01
8.57276618e-01 -1.56802392e+00 -8.86089027e-01 -7.66441748e-02
3.01083088e-01 4.96090949e-01 2.28808135e-01 3.56095999e-01
-6.50695741e-01 2.69192249e-01 1.29672050e-01 -1.81388408e-01
-9.36499178e-01 8.45906734e-01 1.86473265e-01 -7.53922522e-01
-6.67689264e-01 4.03975636e-01 2.41816074e-01 -7.19350815e-01
-1.23734131e-01 -1.79414395e-02 -6.38809860e-01 -2.80725390e-01
3.32692444e-01 1.63126737e-01 -2.19748020e-01 -7.10483849e-01
7.29595572e-02 4.97103035e-01 -3.41961652e-01 1.79908961e-01
9.29810047e-01 -4.48767662e-01 -2.47446671e-01 1.48905590e-01
1.07766449e+00 -3.05460002e-02 -5.02538741e-01 -3.69576126e-01
-1.84306860e-01 -2.63780057e-01 -1.79052040e-01 -5.01971126e-01
-1.08687496e+00 8.97654891e-01 -3.61043550e-02 2.66516775e-01
7.97094166e-01 -1.30205005e-01 1.06487513e+00 4.41594929e-01
1.88944981e-01 -1.26715589e+00 4.30886418e-01 3.95017117e-01
1.04326701e+00 -9.96120512e-01 -2.06965599e-02 -9.48949933e-01
-6.45261467e-01 1.07418740e+00 9.60387230e-01 1.00338042e-01
2.73725659e-01 -2.69523889e-01 -1.77700713e-01 9.78697743e-03
-5.31882882e-01 -2.21891567e-01 2.53520042e-01 7.98241913e-01
3.64158094e-01 1.60151124e-02 -6.62658632e-01 5.68581462e-01
-3.13537151e-01 -5.56218624e-02 7.84888625e-01 6.76345825e-01
-6.90841973e-01 -1.15663826e+00 -5.18772900e-02 5.22075653e-01
-3.60577047e-01 -3.29583615e-01 -7.59264171e-01 6.44565642e-01
-2.13629827e-01 1.06659722e+00 -2.25161582e-01 -5.60680568e-01
2.73683965e-01 -2.05630399e-02 3.50988120e-01 -8.55045319e-01
4.53576334e-02 5.39919995e-02 3.98278445e-01 -1.39872301e-02
-2.57264167e-01 -6.14550291e-03 -1.32161844e+00 -2.30633706e-01
-7.86008060e-01 5.33520818e-01 3.81828725e-01 8.71604323e-01
3.80574286e-01 8.02211106e-01 6.65334880e-01 -1.43442020e-01
-5.94902933e-01 -1.37146747e+00 -5.34198225e-01 5.43768942e-01
-2.96083301e-01 -3.28544945e-01 -2.70257801e-01 -1.03401765e-01] | [10.352855682373047, 8.321170806884766] |
c80327ba-a2ea-4e59-99e8-e96fd4467ada | convolutional-and-deep-learning-based | 2306.10084 | null | https://arxiv.org/abs/2306.10084v1 | https://arxiv.org/pdf/2306.10084v1.pdf | Convolutional and Deep Learning based techniques for Time Series Ordinal Classification | Time Series Classification (TSC) covers the supervised learning problem where input data is provided in the form of series of values observed through repeated measurements over time, and whose objective is to predict the category to which they belong. When the class values are ordinal, classifiers that take this into account can perform better than nominal classifiers. Time Series Ordinal Classification (TSOC) is the field covering this gap, yet unexplored in the literature. There are a wide range of time series problems showing an ordered label structure, and TSC techniques that ignore the order relationship discard useful information. Hence, this paper presents a first benchmarking of TSOC methodologies, exploiting the ordering of the target labels to boost the performance of current TSC state-of-the-art. Both convolutional- and deep learning-based methodologies (among the best performing alternatives for nominal TSC) are adapted for TSOC. For the experiments, a selection of 18 ordinal problems from two well-known archives has been made. In this way, this paper contributes to the establishment of the state-of-the-art in TSOC. The results obtained by ordinal versions are found to be significantly better than current nominal TSC techniques in terms of ordinal performance metrics, outlining the importance of considering the ordering of the labels when dealing with this kind of problems. | ['César Hervás-Martínez', 'Anthony Bagnall', 'Pedro Antonio Gutiérrez', 'David Guijo-Rubio', 'Rafael Ayllón-Gavilán'] | 2023-06-16 | null | null | null | null | ['benchmarking', 'benchmarking', 'time-series-classification'] | ['miscellaneous', 'robots', 'time-series'] | [ 2.52460003e-01 -4.32208717e-01 -4.65670675e-01 -4.98277575e-01
-4.41834450e-01 -6.67442083e-01 1.02286923e+00 8.01392555e-01
-5.65080047e-01 7.11615324e-01 -9.72842723e-02 -2.35793710e-01
-8.07717085e-01 -6.10911429e-01 -3.87752444e-01 -7.61697471e-01
-9.84789312e-01 5.53679168e-01 -6.20238371e-02 -3.42618972e-01
3.22847575e-01 7.77360141e-01 -2.36555910e+00 7.31384993e-01
5.23187816e-01 1.66692257e+00 -4.68185306e-01 2.69418955e-01
-1.85146391e-01 8.02089751e-01 -8.15311730e-01 8.09099525e-02
1.13469921e-01 -1.07400365e-01 -6.81560814e-01 -5.03172986e-02
4.04519022e-01 7.68781975e-02 8.04803446e-02 5.61357021e-01
1.31731391e-01 1.15607589e-01 9.13296342e-01 -1.47262669e+00
-2.11089358e-01 5.87089419e-01 -8.67561847e-02 3.78572524e-01
2.85515845e-01 -4.35592271e-02 1.31517613e+00 -5.40293396e-01
3.32316905e-01 7.85713911e-01 8.10578465e-01 -5.77185825e-02
-1.21077931e+00 -4.25102651e-01 6.86682761e-02 6.89274132e-01
-1.01116800e+00 -1.21430352e-01 9.71762061e-01 -6.75028741e-01
1.17640483e+00 4.42254007e-01 5.99713504e-01 1.19549370e+00
9.97218266e-02 5.17812312e-01 1.48015761e+00 -5.64512610e-01
6.36941373e-01 -1.72094673e-01 4.10357326e-01 -5.41298650e-04
2.20278483e-02 2.68509150e-01 -4.29084539e-01 -1.45096883e-01
8.41090456e-02 1.10848332e-02 1.58406928e-01 -2.80625165e-01
-1.28113616e+00 7.81062543e-01 2.95638055e-01 1.07110584e+00
-3.75779420e-01 6.43809363e-02 1.18512011e+00 8.42204630e-01
8.42795491e-01 5.65883219e-01 -7.29128361e-01 -3.67919445e-01
-1.06493044e+00 3.68902057e-01 7.37785220e-01 4.28519219e-01
3.82391304e-01 4.09366675e-02 -1.76464796e-01 7.38260567e-01
-3.45691532e-01 -6.46907017e-02 8.48712623e-01 -6.45337045e-01
3.44827980e-01 7.91872919e-01 6.22337647e-02 -6.96843803e-01
-6.15633607e-01 -7.36389339e-01 -6.95877433e-01 2.52914757e-01
5.69542527e-01 2.81991005e-01 -6.62868023e-01 1.51050603e+00
5.04055172e-02 1.94689274e-01 -8.22475106e-02 5.57394743e-01
6.27644062e-01 6.73040092e-01 -2.16983527e-01 -6.91530764e-01
1.08260334e+00 -4.54317957e-01 -6.84592485e-01 3.18636954e-01
7.66120911e-01 -3.76215905e-01 1.09685004e+00 7.96644509e-01
-4.56642717e-01 -6.61577106e-01 -1.17351913e+00 3.21727574e-01
-9.79736209e-01 1.44374967e-01 7.10789740e-01 3.74867588e-01
-9.90236282e-01 1.07425272e+00 -6.29916728e-01 -4.15956706e-01
1.26969531e-01 5.00732183e-01 -2.74552584e-01 3.83450210e-01
-1.55496407e+00 8.50790501e-01 7.18234003e-01 1.77203402e-01
-5.67419827e-01 -6.47121131e-01 -4.92665917e-01 -1.08544968e-01
1.52786553e-01 3.38271230e-01 1.37711811e+00 -1.08741498e+00
-1.16051054e+00 9.55851078e-01 2.42823035e-01 -8.87450457e-01
8.45375419e-01 3.70310172e-02 -7.73895621e-01 -4.89531606e-02
1.55191869e-01 1.63198143e-01 8.93742323e-01 -7.00392842e-01
-8.47510457e-01 -2.65720993e-01 1.74652450e-02 -3.06005627e-01
-6.81050956e-01 3.47983576e-02 2.22109854e-01 -5.56802213e-01
-2.49679625e-01 -8.13573301e-01 1.74694970e-01 -3.42734873e-01
1.59491077e-01 -9.59176421e-01 9.22166705e-01 -2.04754725e-01
1.53489447e+00 -2.21824098e+00 6.17923439e-02 -3.29849161e-02
-2.53489222e-02 3.67119424e-02 1.48126125e-01 9.81574893e-01
-6.26176715e-01 -4.45757657e-02 -5.22860050e-01 -3.66865903e-01
2.25608096e-01 3.52724046e-01 -6.49405837e-01 7.61581123e-01
3.96235920e-02 6.16750181e-01 -8.19046199e-01 -3.36177677e-01
5.50676346e-01 -7.73712918e-02 2.10826308e-01 1.25962093e-01
-5.24956048e-01 3.06581229e-01 -2.00332314e-01 5.34311056e-01
2.03141958e-01 1.04996778e-01 -2.15096585e-02 -1.59434065e-01
-7.83134997e-01 1.73321739e-01 -7.68004060e-01 1.14330387e+00
-3.67394656e-01 8.76092196e-01 -8.51442516e-01 -1.59056294e+00
1.10926306e+00 4.88657922e-01 1.14962173e+00 -7.86198199e-01
1.58792049e-01 7.81803191e-01 1.17263101e-01 -5.32503247e-01
2.92439789e-01 -7.85450563e-02 -4.38620687e-01 3.16575736e-01
8.72970447e-02 -1.12011656e-01 6.25673652e-01 -6.22292042e-01
9.21572745e-01 1.14986479e-01 4.63768005e-01 -4.08094019e-01
6.44529045e-01 2.27836102e-01 2.10670903e-01 4.95082200e-01
-1.92142785e-01 4.06813294e-01 8.16992640e-01 -1.02008104e+00
-1.25142264e+00 -6.87744856e-01 -5.96915722e-01 1.06472480e+00
-5.50568223e-01 -1.36875212e-01 -3.89613330e-01 -4.87873673e-01
2.08589002e-01 7.21797168e-01 -1.17976415e+00 1.18550450e-01
-6.63447440e-01 -8.70920479e-01 5.17093837e-01 3.23125154e-01
9.17087123e-02 -1.32969058e+00 -8.32029045e-01 4.22505975e-01
8.29431508e-03 -1.11506879e+00 3.27533603e-01 6.47865832e-01
-1.21076703e+00 -1.03375387e+00 -4.61352199e-01 -4.26104993e-01
-6.33958355e-02 -3.18548024e-01 1.10670197e+00 -2.69830674e-01
-6.35280758e-02 1.23195194e-01 -9.89918888e-01 -6.98053420e-01
-2.84896016e-01 4.46193248e-01 1.06248677e-01 4.62033808e-01
4.77957487e-01 -7.13462830e-01 -1.51450515e-01 1.31895483e-01
-1.03507614e+00 -4.96805698e-01 1.49562731e-01 6.69495106e-01
3.55724066e-01 4.13066387e-01 6.96817577e-01 -5.23127258e-01
6.40803218e-01 -6.22474551e-01 -5.78988910e-01 4.20703031e-02
-5.69538891e-01 -1.18203409e-01 1.20412147e+00 -6.42427206e-01
-2.89012969e-01 -2.51731575e-01 1.48403831e-02 -4.49312776e-01
-3.38943005e-01 8.74058425e-01 4.12780404e-01 1.72524303e-01
7.50879586e-01 2.33715549e-01 -2.54374355e-01 -6.73584223e-01
-1.28105015e-01 7.99163759e-01 4.76337105e-01 -7.36530423e-01
2.37786502e-01 5.71495473e-01 3.80956352e-01 -8.18931162e-01
-1.02245688e+00 -6.05146527e-01 -1.06895375e+00 -6.36411846e-01
4.52653915e-01 -3.94230425e-01 -4.28964704e-01 5.82791209e-01
-8.37363839e-01 -2.06230402e-01 -5.74390829e-01 4.49524403e-01
-7.18391180e-01 2.29194462e-02 -3.44374537e-01 -1.11531079e+00
-1.54173553e-01 -8.90236318e-01 1.06027508e+00 -3.65946591e-01
-4.13525790e-01 -1.07313037e+00 2.11736318e-02 -1.58117414e-01
4.16406780e-01 1.04353809e+00 1.10076475e+00 -1.01179922e+00
1.60143182e-01 -6.03406727e-01 1.55508369e-01 2.60881573e-01
1.33923694e-01 1.14457145e-01 -1.17112410e+00 -4.39743817e-01
1.24842949e-01 -2.16071531e-01 7.55022466e-01 1.82397798e-01
1.45440876e+00 5.02262712e-02 1.36543930e-01 2.57000744e-01
1.45590723e+00 4.22931820e-01 2.71577805e-01 8.94327044e-01
2.67787993e-01 1.02554798e+00 9.07562375e-01 7.34856069e-01
-3.26593518e-02 9.39412355e-01 8.56877029e-01 1.46750823e-01
2.54967719e-01 3.12178940e-01 3.74048412e-01 8.74986291e-01
-1.86355591e-01 -1.24102458e-01 -1.08687150e+00 8.23881149e-01
-1.95177341e+00 -9.38011706e-01 -6.02985919e-01 2.27679014e+00
5.67393661e-01 5.69646716e-01 3.53513360e-01 1.09313631e+00
4.80155796e-01 5.18514037e-01 -5.21882176e-01 -6.52592003e-01
-9.75104868e-02 3.67136598e-01 2.93726295e-01 -1.00061327e-01
-1.57435739e+00 2.48063013e-01 5.58649349e+00 8.76500905e-01
-1.47589767e+00 5.08168526e-02 5.12886345e-01 -1.33688867e-01
2.62748420e-01 -3.18107575e-01 -3.68207246e-01 6.81990623e-01
1.59101939e+00 -5.48149124e-02 2.80267894e-01 6.07905865e-01
3.56641561e-01 2.10833564e-01 -1.61028492e+00 1.06081617e+00
-5.08396607e-03 -1.05997574e+00 -1.39062285e-01 3.00282408e-02
6.35703266e-01 2.24355161e-01 2.33457714e-01 5.83400548e-01
-4.13879901e-01 -9.89256978e-01 1.16372800e+00 5.47435343e-01
8.64567757e-01 -6.17305458e-01 1.07969475e+00 3.16688180e-01
-1.36972630e+00 -6.76830709e-01 2.70304084e-02 -4.08449799e-01
-1.10161550e-01 7.99768448e-01 -3.50690275e-01 9.08128381e-01
8.89571846e-01 1.13584888e+00 -6.24735057e-01 1.01185489e+00
2.34221295e-01 7.96621144e-01 -4.14945573e-01 -2.20704228e-01
6.21792138e-01 -4.33409102e-02 4.49789971e-01 1.24137712e+00
4.54999000e-01 -4.88842547e-01 5.48987165e-02 3.60159606e-01
3.74999344e-01 3.13591897e-01 -6.52943552e-01 -1.44582182e-01
3.11796159e-01 7.26456702e-01 -7.00234652e-01 -3.38934898e-01
-3.63909215e-01 4.61660892e-01 2.45744064e-01 1.74116865e-02
-6.91309869e-01 -2.09570900e-01 3.17002326e-01 9.79897976e-02
8.51743147e-02 -3.26869786e-01 -4.49053228e-01 -9.07301724e-01
3.23128641e-01 -8.01377177e-01 7.07585752e-01 -5.11211336e-01
-1.43677199e+00 7.17810988e-01 5.34938693e-01 -1.97528458e+00
-4.58093405e-01 -9.30528879e-01 -4.18601453e-01 4.32668149e-01
-1.39106119e+00 -1.03750420e+00 -2.29790226e-01 2.46464804e-01
8.68009806e-01 -2.52383411e-01 7.78160334e-01 4.95274067e-01
-2.49824643e-01 2.44158000e-01 4.57969725e-01 -2.12896526e-01
4.26060677e-01 -1.43846834e+00 3.00579935e-01 4.57737625e-01
-3.80413756e-02 1.22266725e-01 9.24738169e-01 -1.70719057e-01
-1.09949601e+00 -1.08727133e+00 1.16695476e+00 -3.74733865e-01
1.14320111e+00 -3.80452037e-01 -9.72390711e-01 2.38472059e-01
-5.61711267e-02 1.14585347e-01 5.17201364e-01 7.09914789e-02
-3.06249708e-01 -4.89050657e-01 -8.43761086e-01 3.03189903e-01
6.76318586e-01 -6.76876783e-01 -8.37688625e-01 3.61872405e-01
5.28250456e-01 -2.31538806e-02 -1.18024063e+00 6.34937346e-01
7.35245168e-01 -1.06323326e+00 7.91113675e-01 -5.03974795e-01
4.77751940e-01 -1.57076910e-01 -2.48368949e-01 -1.24703765e+00
5.08773513e-02 -3.03758472e-01 -3.32583278e-01 9.67214584e-01
1.40592024e-01 -7.72077382e-01 4.30397779e-01 -7.01160505e-02
-3.09194535e-01 -8.97974312e-01 -1.31461418e+00 -1.21190572e+00
2.19345421e-01 -9.56411242e-01 9.50698912e-01 1.24776256e+00
-3.43193277e-03 -1.01097241e-01 -2.07396880e-01 -3.53630066e-01
6.52279794e-01 3.19260597e-01 3.31712037e-01 -1.88083982e+00
1.47006735e-01 -7.26179719e-01 -5.82468212e-01 -1.04614265e-01
2.36916438e-01 -8.89695227e-01 -3.29630762e-01 -1.12868154e+00
-5.15703559e-01 -3.08625102e-01 -8.37648988e-01 3.62602323e-01
4.40351784e-01 2.27981329e-01 1.60486139e-02 5.95182002e-01
-4.14076418e-01 3.22037905e-01 6.40528440e-01 -1.90527201e-01
-1.04565077e-01 9.53747630e-02 1.47642419e-01 5.21100461e-01
1.06066418e+00 -3.90208662e-01 -3.83843392e-01 -8.69807750e-02
2.36390039e-01 1.03809685e-01 3.52547884e-01 -1.26883483e+00
3.64117441e-04 -1.42792314e-01 1.16838060e-01 -8.85939240e-01
2.66702414e-01 -1.13258970e+00 2.88125128e-01 5.36851466e-01
-6.54165626e-01 2.78868943e-01 9.56881046e-02 2.51645088e-01
-5.92132688e-01 -2.93521076e-01 6.77782834e-01 -1.43880127e-02
-1.04249156e+00 3.44316036e-01 -3.70148659e-01 -2.20159009e-01
1.05501294e+00 -3.90273511e-01 -1.62208289e-01 -9.92496237e-02
-7.35196590e-01 -2.92544644e-02 7.18973344e-03 6.27545893e-01
1.27827942e-01 -1.39112365e+00 -8.24298799e-01 -4.59199399e-02
7.65919507e-01 -5.44236302e-01 4.12632450e-02 9.93968904e-01
-4.72903371e-01 7.02205479e-01 -4.75932837e-01 -7.59739459e-01
-1.03997934e+00 6.58426702e-01 4.81435061e-01 -4.21083003e-01
-6.10244989e-01 1.59912258e-01 -4.79477704e-01 -4.88184154e-01
3.90624881e-01 -7.24728763e-01 -8.51860702e-01 8.71191740e-01
2.34129429e-01 5.66506565e-01 7.19907999e-01 -6.42576277e-01
-3.46330881e-01 5.21657825e-01 3.15885007e-01 -1.45463124e-01
1.63527381e+00 3.16458941e-01 -4.29870754e-01 1.43564951e+00
1.57447827e+00 -6.34303749e-01 -8.31282020e-01 -1.26865566e-01
7.06241190e-01 -2.95238197e-01 -1.34134874e-01 -6.61211371e-01
-4.66386586e-01 8.85423243e-01 8.44315350e-01 1.00876749e+00
1.35980225e+00 -3.15478325e-01 3.54041219e-01 1.77947178e-01
7.67547607e-01 -1.21060944e+00 -1.29882842e-01 7.11148798e-01
1.17627287e+00 -1.12829185e+00 -4.39566001e-03 -5.46789058e-02
-1.22666813e-01 1.52803588e+00 1.37849346e-01 -2.54039913e-01
5.81869066e-01 5.74589446e-02 -2.92699188e-01 -1.68559566e-01
-9.99379575e-01 -3.55203003e-01 4.32683021e-01 3.88520330e-01
6.60738051e-01 3.89649540e-01 -7.88067997e-01 2.86535889e-01
-3.18145514e-01 5.74160814e-02 2.68722266e-01 8.18909049e-01
-3.29884589e-01 -1.08180642e+00 -4.27330285e-01 8.76000345e-01
-6.75457537e-01 2.27781564e-01 -4.46075559e-01 9.23125386e-01
4.85945702e-01 1.02115440e+00 1.92407578e-01 -3.09146702e-01
6.62565649e-01 8.77475291e-02 1.77618206e-01 -2.73965240e-01
-1.14821684e+00 -2.18969792e-01 2.80067861e-01 -4.26095515e-01
-5.79899371e-01 -9.92748559e-01 -8.13243449e-01 -8.48012343e-02
9.04052034e-02 4.02512372e-01 8.13860893e-01 1.10320127e+00
-3.97160828e-01 7.11746693e-01 9.19015288e-01 -1.02392673e+00
-6.37376308e-01 -1.08470690e+00 -6.39495015e-01 6.91687047e-01
7.31807411e-01 -8.83734524e-01 -5.98717630e-01 -7.78387114e-02] | [7.257309913635254, 3.300638437271118] |
c4a5a19f-96dc-461a-9ac0-80ef7db6d073 | a-survey-and-implementation-of-performance | 2011.05847 | null | https://arxiv.org/abs/2011.05847v1 | https://arxiv.org/pdf/2011.05847v1.pdf | A Survey and Implementation of Performance Metrics for Self-Organized Maps | Self-Organizing Map algorithms have been used for almost 40 years across various application domains such as biology, geology, healthcare, industry and humanities as an interpretable tool to explore, cluster and visualize high-dimensional data sets. In every application, practitioners need to know whether they can \textit{trust} the resulting mapping, and perform model selection to tune algorithm parameters (e.g. the map size). Quantitative evaluation of self-organizing maps (SOM) is a subset of clustering validation, which is a challenging problem as such. Clustering model selection is typically achieved by using clustering validity indices. While they also apply to self-organized clustering models, they ignore the topology of the map, only answering the question: do the SOM code vectors approximate well the data distribution? Evaluating SOM models brings in the additional challenge of assessing their topology: does the mapping preserve neighborhood relationships between the map and the original data? The problem of assessing the performance of SOM models has already been tackled quite thoroughly in literature, giving birth to a family of quality indices incorporating neighborhood constraints, called \textit{topographic} indices. Commonly used examples of such metrics are the topographic error, neighborhood preservation or the topographic product. However, open-source implementations are almost impossible to find. This is the issue we try to solve in this work: after a survey of existing SOM performance metrics, we implemented them in Python and widely used numerical libraries, and provide them as an open-source library, SOMperf. This paper introduces each metric available in our module along with usage examples. | ['Jérôme Lacaille', 'Hanane Azzag', 'Mustapha Lebbah', 'Florent Forest'] | 2020-11-11 | null | null | null | null | ['self-organized-clustering'] | ['miscellaneous'] | [-9.50371325e-02 -6.62362799e-02 1.06655531e-01 -2.76985705e-01
-1.11151129e-01 -9.30540085e-01 6.45909011e-01 5.90393841e-01
-4.73335177e-01 5.83145320e-01 7.73573071e-02 -1.55443892e-01
-8.25483859e-01 -8.47785652e-01 -2.35099122e-01 -9.85332072e-01
-2.85136104e-01 8.33310068e-01 4.82160389e-01 -2.69722700e-01
6.50420487e-01 6.86364889e-01 -2.12706685e+00 5.08380607e-02
9.49196160e-01 8.36951137e-01 3.24354589e-01 4.17370170e-01
-2.60395437e-01 3.25778335e-01 -4.19253349e-01 1.24327041e-01
4.57033813e-02 -3.66545200e-01 -7.45608330e-01 -7.50861317e-02
-1.13857873e-01 3.51417601e-01 2.30506271e-01 1.07986271e+00
4.20016408e-01 2.47785211e-01 8.48454237e-01 -1.18757713e+00
-3.07532430e-01 3.50768983e-01 -1.52612358e-01 3.28276545e-01
1.35202214e-01 -2.46507376e-01 6.38148427e-01 -9.75708902e-01
7.72241533e-01 8.73127639e-01 5.30408442e-01 4.66846377e-02
-1.44262600e+00 -6.68939278e-02 -3.43511939e-01 4.15577114e-01
-1.72633636e+00 -2.98719168e-01 7.08038330e-01 -7.37037718e-01
5.09589553e-01 7.79156387e-01 5.71716011e-01 4.78931099e-01
2.85647027e-02 1.77552968e-01 1.45906997e+00 -4.64732409e-01
7.47357905e-01 4.63503212e-01 1.35078475e-01 1.42061457e-01
3.40078741e-01 -1.91494226e-01 -4.56895500e-01 -6.66672513e-02
6.70767844e-01 -1.29416212e-01 -4.75019850e-02 -8.64832699e-01
-1.24717116e+00 7.10578918e-01 4.59413022e-01 7.43811369e-01
-2.53929973e-01 -3.00618470e-01 1.38677314e-01 3.24289292e-01
3.63881022e-01 6.78840101e-01 -5.96715845e-02 -2.88041562e-01
-1.16279030e+00 2.60915428e-01 6.85767710e-01 6.70719445e-01
9.68881905e-01 -2.39842951e-01 3.02459657e-01 1.00286055e+00
8.20026249e-02 9.48965549e-02 6.27708316e-01 -8.65287662e-01
-6.07222952e-02 9.34202731e-01 -1.97405554e-02 -1.32729232e+00
-8.42760682e-01 -3.54858696e-01 -9.50962186e-01 4.61600691e-01
5.14799833e-01 1.95001394e-01 -5.85395157e-01 1.14726281e+00
5.08975625e-01 -3.19105595e-01 -1.51006073e-01 1.04722703e+00
6.84942007e-01 4.37140971e-01 -2.22756252e-01 -2.34810218e-01
1.12608874e+00 -3.79903346e-01 -5.23655713e-01 4.94834967e-02
4.37432885e-01 -6.94334865e-01 9.39603031e-01 3.49163085e-01
-1.02131343e+00 -3.02041411e-01 -1.14682806e+00 1.35051534e-01
-9.53287244e-01 -9.08724666e-02 3.96750838e-01 4.46859628e-01
-1.35802913e+00 8.01637709e-01 -1.00864303e+00 -8.63266885e-01
4.36340347e-02 3.70763272e-01 -4.95119303e-01 4.01423544e-01
-9.46458459e-01 1.18484616e+00 6.24011338e-01 5.25846183e-02
-1.04300462e-01 -3.58796120e-01 -5.49904048e-01 9.85723585e-02
-9.40747093e-03 -2.65563011e-01 5.43619156e-01 -6.90112174e-01
-1.20673358e+00 9.76733625e-01 -9.73996613e-03 -2.34713033e-01
5.39815545e-01 5.06771624e-01 -3.85815650e-01 2.06028298e-02
9.09520090e-02 4.91693437e-01 2.85362303e-01 -1.40077233e+00
-4.90793914e-01 -5.56742609e-01 -1.39578477e-01 3.38849276e-01
-6.64713323e-01 -1.89755335e-02 -2.29477346e-01 -5.22530615e-01
6.88146472e-01 -7.22447813e-01 -2.54708558e-01 -4.85691242e-02
-3.04670662e-01 7.29640992e-03 6.53023362e-01 -4.60429043e-01
1.40803802e+00 -2.23475432e+00 2.40438238e-01 7.45456159e-01
1.44419730e-01 -9.56898183e-02 3.95400465e-01 7.13938415e-01
-2.10490096e-02 1.15597084e-01 -6.49290621e-01 1.51124299e-01
1.73691884e-01 1.52653724e-01 6.32976219e-02 5.24863005e-01
1.94559917e-02 4.76127058e-01 -7.40287066e-01 -7.16342986e-01
4.78310883e-01 6.48790419e-01 -1.46610141e-01 -2.22839504e-01
2.57283717e-01 3.46837819e-01 -1.50944024e-01 3.46354038e-01
6.12959027e-01 -1.05571285e-01 1.55690551e-01 1.38897642e-01
-7.05520749e-01 4.73747477e-02 -1.57429779e+00 1.30758548e+00
-1.22553669e-02 7.03075290e-01 -5.90725504e-02 -1.13136268e+00
1.33459771e+00 1.33592576e-01 6.96913838e-01 -7.91834116e-01
-1.00788958e-01 3.92537475e-01 5.39695285e-02 -4.13242280e-01
4.78104204e-01 -6.39608353e-02 2.40203932e-01 4.93564844e-01
-2.06568539e-02 1.19709060e-01 5.51992893e-01 -2.51869977e-01
1.03691673e+00 -7.28025213e-02 4.66203898e-01 -9.84394312e-01
5.61611295e-01 4.44145292e-01 1.45562142e-01 3.85224938e-01
4.63887639e-02 8.28661203e-01 6.01278007e-01 -5.31345606e-01
-1.16720676e+00 -1.18370426e+00 -7.78132975e-01 9.93581653e-01
2.75721043e-01 -2.16034874e-01 -9.66547549e-01 -8.05400386e-02
-1.21632144e-01 4.07450408e-01 -7.55570590e-01 1.46691263e-01
-4.18871611e-01 -9.18762803e-01 2.42831707e-01 2.71720558e-01
2.80221611e-01 -1.19006419e+00 -9.10017848e-01 1.48988679e-01
-9.04075131e-02 -4.11612779e-01 2.70137578e-01 4.66479152e-01
-1.03948784e+00 -9.16688919e-01 -6.36603415e-01 -7.61575341e-01
8.53154361e-01 3.51135917e-02 9.86589074e-01 1.31143302e-01
-2.12309390e-01 6.15534708e-02 -3.89390379e-01 -3.55978578e-01
-2.10421398e-01 3.97379965e-01 1.25835180e-01 -1.06449127e-01
3.97881269e-01 -9.41969275e-01 -5.95832646e-01 9.11723256e-01
-1.20576668e+00 -1.29987508e-01 3.42609704e-01 5.03829956e-01
9.12363589e-01 3.94913435e-01 4.84214246e-01 -7.08256245e-01
6.77129209e-01 -4.59335625e-01 -5.20840824e-01 1.97751254e-01
-6.54144049e-01 1.04441196e-01 6.02649808e-01 -3.60603705e-02
-4.03908074e-01 1.29050881e-01 -1.77761521e-02 -2.77116545e-03
-3.06078166e-01 8.09745610e-01 -2.32832313e-01 -9.16877463e-02
9.39845204e-01 3.34509581e-01 1.87713519e-01 -4.37661856e-01
2.68353000e-02 6.40582919e-01 6.39363468e-01 -3.65492076e-01
7.07788706e-01 6.38000011e-01 1.30152673e-01 -1.02283645e+00
-1.08977906e-01 -7.68845737e-01 -1.09126031e+00 -4.63211983e-01
5.57036757e-01 -1.97253332e-01 -6.11379325e-01 -1.38431400e-01
-8.03781807e-01 -2.47862022e-02 -2.65030175e-01 2.96568155e-01
-7.87925601e-01 6.78461865e-02 1.48493245e-01 -7.64308810e-01
-4.66425642e-02 -1.06288517e+00 6.80281460e-01 -1.67425454e-03
-3.56264114e-01 -1.29070795e+00 2.76580960e-01 -9.59985927e-02
6.77941442e-01 5.89770377e-01 1.03278506e+00 -5.39442420e-01
-1.08161598e-01 -2.38825500e-01 1.23165641e-02 -3.04432027e-02
1.36889264e-01 3.26829441e-02 -9.80998755e-01 -1.80676237e-01
6.78644553e-02 3.87973160e-01 7.89436042e-01 4.16564226e-01
9.93142903e-01 -2.22543016e-01 -3.14735234e-01 4.77125138e-01
1.51237619e+00 2.51752347e-01 7.56300509e-01 7.21284628e-01
3.77670825e-01 1.14350402e+00 4.81091112e-01 2.89082438e-01
2.08095476e-01 8.52289498e-01 3.87998253e-01 -2.48620957e-02
2.72608608e-01 9.97463092e-02 -1.03482135e-01 8.49508345e-01
-4.67194557e-01 5.51422089e-02 -1.35936117e+00 4.02731538e-01
-2.01557684e+00 -9.83627617e-01 -2.94471174e-01 2.61323571e+00
4.28423434e-01 2.86230762e-02 4.20771241e-01 6.60475552e-01
8.59188020e-01 -2.18976125e-01 -2.75005013e-01 -2.80882746e-01
-3.01224709e-01 -7.20546842e-02 3.67549807e-01 5.48992932e-01
-1.21714258e+00 4.64995772e-01 6.13112068e+00 5.78988194e-01
-1.24887919e+00 -1.24177955e-01 4.29158121e-01 1.27411401e-02
-1.77549794e-01 8.08928087e-02 -3.82258236e-01 5.80512106e-01
7.88521886e-01 -3.15377742e-01 4.03628230e-01 7.77501404e-01
5.19868076e-01 -2.56128699e-01 -9.78272557e-01 8.76130104e-01
-8.92703980e-02 -1.22055471e+00 -1.93978995e-01 3.70003521e-01
5.38778961e-01 -2.53683507e-01 -1.49671016e-02 -2.62499362e-01
-8.80671293e-02 -1.19185746e+00 6.35023952e-01 6.81078017e-01
6.41052008e-01 -8.35135162e-01 7.69357324e-01 1.97448030e-01
-1.12745893e+00 5.74186165e-03 -5.50853312e-01 -6.46641552e-02
-4.88422550e-02 6.86914384e-01 -5.88856876e-01 4.93966609e-01
9.83183086e-01 4.13606524e-01 -9.48316157e-01 1.48910582e+00
3.51177335e-01 3.46818775e-01 -5.23150504e-01 -2.35600725e-01
1.82073548e-01 -4.03124839e-01 6.04540586e-01 1.10642600e+00
5.38599908e-01 -1.93980694e-01 -2.61992097e-01 9.42615926e-01
5.21718860e-01 5.11246085e-01 -4.90589589e-01 9.67676342e-02
5.35963893e-01 1.29583716e+00 -1.39269722e+00 -2.59318743e-02
1.73652425e-01 5.67690849e-01 1.30031362e-01 2.03051791e-01
-3.28291208e-01 -5.85133851e-01 5.34077227e-01 7.64133096e-01
-1.95547063e-02 -2.88620949e-01 -7.54977286e-01 -6.02967322e-01
2.46167347e-01 -4.51768428e-01 3.24529648e-01 -6.69903576e-01
-1.08474767e+00 7.06415117e-01 2.60693640e-01 -1.56700218e+00
-3.56847614e-01 -6.78296208e-01 -6.17235184e-01 6.36919320e-01
-9.92326677e-01 -6.15708768e-01 -5.25463879e-01 3.84162307e-01
-8.24704394e-03 -8.16930011e-02 9.12001789e-01 3.64350885e-01
-2.95864552e-01 8.96600038e-02 7.30278790e-01 -1.68750212e-01
5.53460896e-01 -1.60776699e+00 1.00672066e-01 4.64072317e-01
1.14982240e-01 7.21606016e-01 1.06105101e+00 -3.01136196e-01
-9.99758005e-01 -7.55812407e-01 9.44839478e-01 -4.93308932e-01
6.96584105e-01 -4.32530373e-01 -1.09949315e+00 3.69133838e-02
-1.31986856e-01 -7.92342201e-02 5.85276663e-01 -4.79304306e-02
1.40123650e-01 -1.64594010e-01 -1.07626200e+00 5.06730378e-01
8.54339778e-01 -1.34952828e-01 -2.46350661e-01 1.82113677e-01
-7.74375536e-03 -9.75019857e-02 -1.11569285e+00 2.72235185e-01
5.59774816e-01 -1.42244649e+00 8.88682902e-01 -2.03112900e-01
2.62018234e-01 -7.17958808e-01 -1.72776897e-02 -1.36402678e+00
-3.84970754e-01 -1.88209504e-01 3.39424282e-01 1.24645102e+00
4.45678174e-01 -6.75359786e-01 7.95339227e-01 4.09020931e-01
-1.26376703e-01 -7.08102524e-01 -1.08475184e+00 -7.42430508e-01
1.49204269e-01 -1.72514468e-01 6.76710069e-01 1.16647243e+00
4.89531547e-01 3.79545637e-03 2.34242186e-01 3.44032384e-02
6.05663538e-01 -6.69102669e-02 4.88594234e-01 -1.84879780e+00
2.41122693e-01 -8.89878273e-01 -1.07591057e+00 -2.51573205e-01
-2.77562261e-01 -9.60621536e-01 3.35868597e-02 -1.94424999e+00
1.03281885e-02 -6.24401987e-01 -3.17708641e-01 3.76335084e-01
2.20707864e-01 4.69005346e-01 -5.54739609e-02 6.51267529e-01
-4.83429253e-01 1.58746645e-01 8.72055709e-01 4.31926586e-02
-4.27302003e-01 -2.06256509e-01 -5.17340183e-01 5.60740173e-01
8.40563118e-01 -4.47610497e-01 -3.33542675e-01 -1.54047519e-01
4.70611483e-01 -2.96954781e-01 5.13397694e-01 -1.46274233e+00
5.66009998e-01 -1.19843349e-01 4.06803280e-01 -6.02719605e-01
1.54396251e-01 -9.51832473e-01 5.65231264e-01 1.74111798e-01
-2.12774158e-01 2.87549943e-01 3.77182551e-02 1.78478852e-01
-5.17070889e-01 -3.07227641e-01 7.70498455e-01 -4.47762012e-02
-7.10515559e-01 -9.46632102e-02 -4.00214195e-01 -2.54894167e-01
1.04585338e+00 -7.37840950e-01 -3.08993518e-01 -1.97150126e-01
-9.75547194e-01 1.99453399e-01 9.17026937e-01 2.90182233e-01
3.95599425e-01 -1.42722213e+00 -5.70724666e-01 9.74086151e-02
3.89974654e-01 1.20609719e-02 9.61081609e-02 1.00297332e+00
-9.67203259e-01 4.60108489e-01 -5.96618652e-01 -8.18515062e-01
-1.03746462e+00 4.70338255e-01 3.03702384e-01 1.02034964e-01
-4.20858443e-01 1.84564516e-01 -2.96279564e-02 -5.67239106e-01
1.19281396e-01 -5.28523661e-02 -5.48837662e-01 3.18185776e-01
5.50706565e-01 7.45662153e-01 5.10478914e-01 -7.59644806e-01
-4.96219486e-01 6.00128949e-01 5.89575589e-01 -2.22910583e-01
1.55243862e+00 -1.68665439e-01 -5.45935929e-01 8.46720457e-01
9.80877459e-01 -3.75553936e-01 -8.17058146e-01 2.38019824e-01
5.24100602e-01 -2.56867975e-01 -8.11049268e-02 -6.02358580e-01
-6.20851755e-01 8.59880865e-01 7.91570127e-01 8.68946135e-01
1.05807066e+00 2.43216231e-02 -1.24073297e-01 2.96364456e-01
2.78338939e-01 -1.30496490e+00 -2.96219498e-01 3.82465363e-01
9.94785190e-01 -9.50608075e-01 9.99351591e-02 -3.04632366e-01
-4.65639234e-01 1.17891204e+00 1.49139956e-01 -9.64393094e-02
7.86349058e-01 2.62368441e-01 2.02917635e-01 -3.85739326e-01
-1.99029073e-01 -2.84452230e-01 3.62248451e-01 6.68484569e-01
5.16328275e-01 6.20337687e-02 -4.69693720e-01 2.02566758e-01
-7.85433888e-01 -2.20821023e-01 3.02147627e-01 7.72062778e-01
-8.13420534e-01 -8.71603847e-01 -7.60858178e-01 5.81311464e-01
-2.75152568e-02 1.77974671e-01 -7.49302149e-01 6.62483096e-01
2.85899360e-02 8.19207311e-01 3.50687712e-01 -4.39995259e-01
3.68932277e-01 2.07632240e-02 1.00754626e-01 -2.50013381e-01
-4.90982741e-01 -9.31216180e-02 -2.64468968e-01 -1.76564008e-01
-5.89350998e-01 -7.50905097e-01 -1.25748467e+00 -4.71705765e-01
-8.01064223e-02 4.91752148e-01 1.01645088e+00 7.83619583e-01
2.50458717e-01 2.27098197e-01 5.38358271e-01 -9.30570781e-01
2.62197051e-02 -9.11186516e-01 -8.34778607e-01 3.51693749e-01
-5.64003587e-02 -9.20744598e-01 -4.13881272e-01 9.34468433e-02] | [7.670138359069824, 4.505418300628662] |
0b4ef7a4-c11e-46b7-8a83-fb16c66da3c5 | pointgpt-auto-regressively-generative-pre | 2305.11487 | null | https://arxiv.org/abs/2305.11487v2 | https://arxiv.org/pdf/2305.11487v2.pdf | PointGPT: Auto-regressively Generative Pre-training from Point Clouds | Large language models (LLMs) based on the generative pre-training transformer (GPT) have demonstrated remarkable effectiveness across a diverse range of downstream tasks. Inspired by the advancements of the GPT, we present PointGPT, a novel approach that extends the concept of GPT to point clouds, addressing the challenges associated with disorder properties, low information density, and task gaps. Specifically, a point cloud auto-regressive generation task is proposed to pre-train transformer models. Our method partitions the input point cloud into multiple point patches and arranges them in an ordered sequence based on their spatial proximity. Then, an extractor-generator based transformer decoder, with a dual masking strategy, learns latent representations conditioned on the preceding point patches, aiming to predict the next one in an auto-regressive manner. Our scalable approach allows for learning high-capacity models that generalize well, achieving state-of-the-art performance on various downstream tasks. In particular, our approach achieves classification accuracies of 94.9% on the ModelNet40 dataset and 93.4% on the ScanObjectNN dataset, outperforming all other transformer models. Furthermore, our method also attains new state-of-the-art accuracies on all four few-shot learning benchmarks. | ['Yufeng Yue', 'Li Yuan', 'Kai Yu', 'Yi Yang', 'Meiling Wang', 'Guangyan Chen'] | 2023-05-19 | null | null | null | null | ['3d-point-cloud-classification', '3d-part-segmentation', 'few-shot-3d-point-cloud-classification'] | ['computer-vision', 'computer-vision', 'computer-vision'] | [ 2.79486865e-01 9.96960029e-02 -3.53101157e-02 -1.18829548e-01
-1.34629357e+00 -3.38969231e-01 9.61381257e-01 3.49935405e-02
5.00907451e-02 4.94593263e-01 1.02872699e-01 -2.59033054e-01
2.33836211e-02 -1.16573739e+00 -1.09690118e+00 -8.88868093e-01
1.37635171e-01 8.82931888e-01 4.00932491e-01 -1.36111379e-01
3.66114199e-01 3.87432039e-01 -1.63394320e+00 5.10453463e-01
1.00634646e+00 9.58284616e-01 6.54627919e-01 3.84261012e-01
-4.21315759e-01 5.85880816e-01 -3.21473479e-01 -4.51774955e-01
2.35343464e-02 -1.32477909e-01 -5.52267432e-01 3.76957841e-03
2.91894197e-01 -1.26206642e-03 -4.73632634e-01 5.81968784e-01
4.32811946e-01 4.06014062e-02 9.82627153e-01 -1.00461102e+00
-1.05162024e+00 6.25977159e-01 -3.83074224e-01 3.31711099e-02
-7.58517385e-02 2.77151406e-01 1.14613688e+00 -1.43600798e+00
3.86231720e-01 1.10095012e+00 7.20839024e-01 4.88255173e-01
-1.37307334e+00 -8.84995461e-01 8.43313262e-02 7.41161630e-02
-1.59804845e+00 -3.17649484e-01 5.64874530e-01 -4.83695418e-01
1.56320238e+00 -1.43447742e-01 6.23251438e-01 1.42374420e+00
4.12429571e-01 7.54524827e-01 7.19174862e-01 -1.80180311e-01
3.86605680e-01 -2.17513561e-01 -3.31021905e-01 6.14690304e-01
-9.75507945e-02 2.02839263e-03 -7.66527653e-01 -2.20747516e-01
7.41180897e-01 1.43549830e-01 1.67700052e-01 -2.63032883e-01
-1.27044916e+00 8.83386910e-01 6.32187843e-01 1.74586639e-01
-4.31436419e-01 1.65303573e-01 1.40110597e-01 -1.22526281e-01
7.53947496e-01 3.28100413e-01 -2.53648341e-01 4.95523661e-02
-1.18159127e+00 1.91368490e-01 5.49679160e-01 1.36466062e+00
6.31832719e-01 3.63883078e-02 -5.94767690e-01 9.28992152e-01
2.61644006e-01 4.77905661e-01 3.84981543e-01 -3.06106597e-01
9.31558371e-01 4.56888735e-01 -3.39139104e-01 -5.52196622e-01
-7.14486986e-02 -7.41157651e-01 -1.11939287e+00 -2.20818996e-01
-3.08692455e-01 3.06936681e-01 -1.27474141e+00 1.58493364e+00
5.05355969e-02 6.41925991e-01 2.62308925e-01 4.06666517e-01
8.33355427e-01 1.13568556e+00 3.47524226e-01 8.09385031e-02
1.08566427e+00 -9.49745655e-01 -1.18472271e-01 -3.61360848e-01
4.51678991e-01 -5.63991308e-01 1.04221976e+00 3.61679614e-01
-9.26658928e-01 -6.75674796e-01 -9.51667428e-01 -2.06581384e-01
-3.80581021e-01 3.96560021e-02 5.83136201e-01 8.02948475e-02
-1.08418894e+00 7.09645629e-01 -9.82115149e-01 -2.23845601e-01
7.73723960e-01 1.79977924e-01 -3.94574106e-02 -2.24584013e-01
-8.67176175e-01 6.41585827e-01 2.92754114e-01 -2.41605580e-01
-1.21038485e+00 -1.09835505e+00 -8.21879566e-01 3.41858923e-01
3.77113000e-04 -9.92155790e-01 1.12636566e+00 -4.94082533e-02
-1.31397378e+00 7.02233136e-01 -3.08843136e-01 -7.49375343e-01
2.21468687e-01 -1.05615295e-01 -2.05042124e-01 -1.47012081e-02
4.23905134e-01 1.04900372e+00 9.29582953e-01 -1.12193537e+00
-5.25741398e-01 -2.67885715e-01 -4.74565566e-01 1.98938936e-01
-2.28082806e-01 -3.92963707e-01 -6.75423503e-01 -7.31943488e-01
1.77941248e-01 -1.00544238e+00 -3.97675902e-01 -3.43913615e-01
-6.01751804e-01 -5.60969293e-01 7.57869601e-01 -2.79009581e-01
1.05478919e+00 -2.12708759e+00 2.36004159e-01 -1.32688493e-01
2.63730764e-01 1.02294572e-01 -3.17856729e-01 7.27072954e-01
-6.68592155e-02 1.11033857e-01 -4.06096578e-01 -6.77665114e-01
4.83446792e-02 2.20896989e-01 -8.58983696e-01 1.10809907e-01
6.06050253e-01 1.29961848e+00 -8.57337713e-01 -4.20973659e-01
3.60156059e-01 5.40676355e-01 -5.90172291e-01 1.50284544e-01
-4.26354796e-01 4.05941248e-01 -4.50569153e-01 6.55904531e-01
4.68004763e-01 -6.37707889e-01 -3.59172791e-01 4.48167324e-02
-1.19810164e-01 5.57644725e-01 -3.39405537e-01 2.02884841e+00
-6.87172949e-01 4.80897516e-01 -7.04814494e-01 -9.72311258e-01
1.18237400e+00 2.47388825e-01 5.72448611e-01 -7.12773681e-01
-2.49158472e-01 1.68515086e-01 -2.36145958e-01 -2.65589565e-01
4.17399466e-01 -3.99989456e-01 -3.32285374e-01 2.85182774e-01
4.81373847e-01 -4.21050727e-01 -4.88299206e-02 3.09815705e-01
1.25517035e+00 1.19363189e-01 3.77196781e-02 -5.63736483e-02
9.89994183e-02 -8.50438923e-02 1.95034772e-01 7.57677615e-01
5.43209791e-01 9.90728557e-01 3.05642664e-01 -2.37903669e-01
-1.26612949e+00 -1.40417266e+00 -7.73941129e-02 7.79935122e-01
-8.15757662e-02 -7.21544087e-01 -4.00534630e-01 -4.78629708e-01
2.61741653e-02 1.18887317e+00 -4.79430288e-01 -5.21294296e-01
-3.26819897e-01 -8.64559293e-01 6.34087920e-01 7.37254381e-01
3.96433830e-01 -9.60742593e-01 -2.50791848e-01 3.63856375e-01
-2.65937090e-01 -1.24424148e+00 -1.78959250e-01 2.84738958e-01
-1.07303667e+00 -5.67451477e-01 -6.56969190e-01 -8.01625133e-01
5.74506998e-01 2.95606703e-01 1.29315472e+00 -2.63402224e-01
-1.16655044e-01 5.06533906e-02 -3.77176642e-01 -3.42297226e-01
-3.39583814e-01 5.31698704e-01 -1.34484529e-01 -9.42450911e-02
2.94029295e-01 -8.34536374e-01 -3.71679753e-01 1.74548656e-01
-7.87149012e-01 4.01403368e-01 9.60058689e-01 7.12330103e-01
1.09538507e+00 -8.95729214e-02 5.64639330e-01 -5.44132948e-01
3.98939788e-01 -8.27199697e-01 -3.46098065e-01 1.84178784e-01
-6.18843973e-01 2.87088156e-01 5.89350462e-01 -4.98342633e-01
-9.42767978e-01 4.01242338e-02 -4.57352340e-01 -8.41160774e-01
-1.47174709e-02 2.71512479e-01 -1.15911171e-01 9.13011432e-02
5.83566606e-01 8.00696671e-01 -3.96895409e-01 -5.02136350e-01
4.51663196e-01 4.92511600e-01 4.92528856e-01 -6.31132483e-01
9.55993295e-01 3.74760062e-01 2.31716763e-02 -8.76480460e-01
-9.18496192e-01 -3.35117459e-01 -6.56634510e-01 1.00176141e-01
9.03043270e-01 -1.13130593e+00 -3.48178625e-01 3.80798101e-01
-1.32508910e+00 -3.69068652e-01 -4.36304569e-01 1.70860648e-01
-7.52870083e-01 3.49485949e-02 -5.52353799e-01 -5.62644184e-01
-5.86484671e-01 -1.15020823e+00 1.69277096e+00 -6.45695627e-02
4.43318859e-02 -7.03463435e-01 2.86029913e-02 1.72416106e-01
3.95473957e-01 1.15259783e-03 1.16758966e+00 -7.30937839e-01
-9.91535544e-01 -1.60252541e-01 -2.22525060e-01 2.14355573e-01
-1.06095701e-01 -3.31232011e-01 -9.36143756e-01 -9.81748849e-02
-7.21881688e-02 -1.97434545e-01 1.17513621e+00 2.18584329e-01
1.58957601e+00 -2.41825357e-01 -7.98591018e-01 7.40550041e-01
1.39829266e+00 -6.67318255e-02 8.81338179e-01 -3.48702185e-02
8.38949442e-01 1.05679706e-01 3.79284620e-01 3.00120056e-01
4.60897177e-01 6.92817271e-01 4.91954952e-01 1.69616476e-01
-2.46265113e-01 -8.94927621e-01 2.04191685e-01 8.96610022e-01
3.60032506e-02 -4.39823389e-01 -1.05853868e+00 5.94375730e-01
-1.78074646e+00 -8.95447671e-01 -1.37547553e-01 2.08096886e+00
5.85800409e-01 2.20237255e-01 -3.49696517e-01 -2.24213466e-01
4.80571210e-01 3.05759102e-01 -6.47559583e-01 7.71967769e-02
-8.34057480e-02 5.77082634e-01 5.12946546e-01 6.92611113e-02
-1.02005076e+00 1.30685687e+00 6.23454857e+00 1.15883946e+00
-1.00450444e+00 3.51202846e-01 6.26331806e-01 -2.80352920e-01
-4.61813807e-01 3.13238241e-04 -1.13799787e+00 5.59601307e-01
1.11371469e+00 -1.03439979e-01 3.29484403e-01 9.42439020e-01
4.98266630e-02 3.54285777e-01 -1.06505513e+00 9.94442821e-01
1.50364608e-01 -1.76868176e+00 3.81880224e-01 2.98219323e-01
7.62276053e-01 6.36733890e-01 2.66240537e-01 8.00530970e-01
2.16036737e-01 -1.05486655e+00 7.20734179e-01 5.88448286e-01
1.02707338e+00 -7.63527870e-01 3.40100884e-01 7.08313107e-01
-1.30957878e+00 -2.48152902e-03 -5.88878155e-01 1.85047835e-01
3.41586232e-01 6.83189154e-01 -1.06120193e+00 5.47608554e-01
6.65724158e-01 8.53368521e-01 -4.91218328e-01 9.69779491e-01
-3.11657488e-01 6.69196606e-01 -3.31514716e-01 9.47473198e-02
3.62265617e-01 -5.56723773e-02 5.21828651e-01 1.21356845e+00
7.62412786e-01 8.80676955e-02 -2.48748567e-02 1.27863979e+00
-1.86920673e-01 -1.02764890e-01 -8.24375987e-01 -7.10971355e-02
5.59401095e-01 1.11250389e+00 -5.92713475e-01 -3.36136401e-01
-2.90960878e-01 1.00609887e+00 4.75352287e-01 1.83947504e-01
-9.70184803e-01 -1.97026908e-01 5.76557398e-01 3.48584950e-01
7.40584016e-01 -3.48667860e-01 -3.05359662e-01 -1.14380658e+00
5.81896678e-02 -4.42911655e-01 1.24835104e-01 -9.39325690e-01
-1.51161981e+00 8.38744700e-01 2.83629112e-02 -1.45371628e+00
-4.08878893e-01 -3.20482790e-01 -7.55389690e-01 9.30527568e-01
-1.38528872e+00 -1.56891823e+00 -2.77802765e-01 3.78411472e-01
1.01829445e+00 -2.56432593e-01 8.63365650e-01 2.53732741e-01
-2.22278446e-01 3.88182610e-01 -7.29915500e-02 -8.95657465e-02
3.29790980e-01 -9.82007027e-01 1.10282457e+00 7.21241236e-01
4.78083581e-01 5.24722338e-01 3.05950880e-01 -8.03810000e-01
-1.32429004e+00 -1.56439877e+00 1.03078437e+00 -7.22374558e-01
6.16979182e-01 -7.54382968e-01 -1.12444854e+00 6.88842773e-01
-4.47559282e-02 -1.14195585e-01 5.09926677e-01 5.91314957e-02
-4.11126494e-01 2.43838038e-02 -7.20622301e-01 4.73060012e-01
1.44886017e+00 -5.68537593e-01 -6.05468690e-01 6.59825087e-01
1.06152678e+00 -4.94581521e-01 -7.11591005e-01 5.10589957e-01
1.51070371e-01 -8.07627380e-01 1.15309620e+00 -5.15714228e-01
7.46798754e-01 5.02863973e-02 -1.98877156e-01 -1.32873106e+00
-6.13629103e-01 -3.86394382e-01 -3.78229737e-01 1.27979875e+00
6.03745818e-01 -5.13991117e-01 8.25062633e-01 1.26093715e-01
-6.04493082e-01 -1.10236979e+00 -1.02226281e+00 -8.26198936e-01
1.96599066e-01 -8.13419878e-01 6.98895752e-01 4.09778833e-01
-3.91253382e-01 7.75092840e-01 -2.92571664e-01 3.21433634e-01
6.29891038e-01 2.38231376e-01 6.24425769e-01 -1.26475275e+00
-3.04891378e-01 -3.78660917e-01 -2.42489636e-01 -1.39772344e+00
3.07540834e-01 -1.25671792e+00 2.32088074e-01 -1.78125656e+00
1.68272555e-01 -7.51173615e-01 -8.68456066e-02 5.55982709e-01
-1.79864280e-02 1.50424570e-01 2.29630411e-01 5.01042485e-01
-5.74949086e-01 9.98138011e-01 1.01214767e+00 -2.87854016e-01
-4.05361876e-02 3.25686812e-01 -6.97666287e-01 3.80193651e-01
5.88557363e-01 -6.59300864e-01 -4.94532794e-01 -6.32666707e-01
6.79900274e-02 -2.30669789e-02 5.67424595e-01 -1.34503543e+00
4.37652320e-01 8.19840282e-02 4.00851250e-01 -1.10133994e+00
7.02928424e-01 -5.43570459e-01 2.62209326e-01 2.20201150e-01
-1.94017470e-01 9.27525293e-03 2.31985062e-01 8.59070659e-01
-7.32505918e-02 3.51569057e-02 5.98218679e-01 -2.01065212e-01
-5.49203634e-01 7.87016511e-01 -1.34927183e-01 -1.03823408e-01
9.12049413e-01 -1.27434134e-02 -1.79286033e-01 -5.53874373e-02
-5.48317850e-01 7.00308532e-02 2.94714332e-01 8.51023018e-01
8.51533175e-01 -1.35521746e+00 -7.86995292e-01 4.85420436e-01
4.68854547e-01 5.16951323e-01 3.24446619e-01 7.48668194e-01
-6.84194863e-02 7.32250869e-01 1.64029092e-01 -9.82012510e-01
-6.72481477e-01 6.22218847e-01 1.49262518e-01 -4.34775323e-01
-1.05827153e+00 9.05333579e-01 4.61014509e-01 -3.27810973e-01
-1.32296681e-01 -4.45198834e-01 1.60543382e-01 -1.78541079e-01
3.35049927e-01 -5.46140186e-02 3.64872515e-01 -5.05412221e-01
-1.39508903e-01 5.58489501e-01 -7.33763799e-02 -9.89880264e-02
1.60577977e+00 2.09654152e-01 1.94793999e-01 5.10954022e-01
1.20573986e+00 -2.91833103e-01 -1.26535988e+00 -3.96055400e-01
-1.33317515e-01 -2.45610014e-01 -1.12587571e-01 -4.98487175e-01
-8.15360844e-01 1.08916140e+00 1.69774950e-01 1.60074204e-01
8.91703010e-01 6.08104825e-01 9.78106618e-01 2.12648809e-01
6.41692877e-01 -6.57152772e-01 2.10446119e-01 6.78313434e-01
9.88250136e-01 -1.05334401e+00 -2.85898775e-01 -4.52164471e-01
-7.67754555e-01 8.24776053e-01 4.42979544e-01 -2.75760055e-01
5.46926618e-01 2.78933823e-01 -5.98675132e-01 -3.44842315e-01
-1.29131055e+00 -2.27359012e-01 4.01889384e-01 5.68376541e-01
1.02944091e-01 4.68551591e-02 1.64844915e-01 7.06430972e-01
-3.30337346e-01 -3.93124707e-02 -1.33345842e-01 6.00858867e-01
-4.76384759e-01 -9.06310439e-01 3.55239213e-02 6.37287617e-01
5.06019406e-03 -4.98016149e-01 -1.94324821e-01 5.53728819e-01
2.55653299e-02 5.94911158e-01 3.50613743e-01 -6.69484019e-01
2.94416249e-01 1.26522541e-01 3.37683111e-01 -1.04420877e+00
-2.77929723e-01 3.13181654e-02 -2.26514161e-01 -4.39820528e-01
1.12556249e-01 -5.85954309e-01 -1.15789187e+00 -2.65574995e-02
-9.90216061e-02 6.79471642e-02 6.64007843e-01 1.05513716e+00
8.32798839e-01 6.59726322e-01 4.26715612e-01 -1.20446897e+00
-6.36431277e-01 -1.13306212e+00 -1.77712202e-01 7.27260411e-02
1.37222081e-01 -7.30244517e-01 -1.35863721e-01 -1.15687527e-01] | [8.12630558013916, -3.405186414718628] |
e226137d-0047-48cc-8eec-367ad5f338de | 3d-lip-event-detection-via-interframe-motion | 2111.09485 | null | https://arxiv.org/abs/2111.09485v1 | https://arxiv.org/pdf/2111.09485v1.pdf | 3D Lip Event Detection via Interframe Motion Divergence at Multiple Temporal Resolutions | The lip is a dominant dynamic facial unit when a person is speaking. Detecting lip events is beneficial to speech analysis and support for the hearing impaired. This paper proposes a 3D lip event detection pipeline that automatically determines the lip events from a 3D speaking lip sequence. We define a motion divergence measure using 3D lip landmarks to quantify the interframe dynamics of a 3D speaking lip. Then, we cast the interframe motion detection in a multi-temporal-resolution framework that allows the detection to be applicable to different speaking speeds. The experiments on the S3DFM Dataset investigate the overall 3D lip dynamics based on the proposed motion divergence. The proposed 3D pipeline is able to detect opening and closing lip events across 100 sequences, achieving a state-of-the-art performance. | ['Robert B. Fisher', 'Jie Zhang'] | 2021-11-18 | null | null | null | null | ['motion-detection'] | ['computer-vision'] | [-3.08269173e-01 -4.78135943e-01 -3.64069849e-01 -7.62521625e-02
-9.88253832e-01 -5.78586459e-01 6.47926629e-01 -2.56950974e-01
-3.05266351e-01 -2.02541798e-02 6.12953305e-01 7.05679655e-02
2.62586772e-01 -2.27882430e-01 -1.65668517e-01 -7.59224594e-01
-5.05471416e-02 4.61917259e-02 5.09706020e-01 1.64454430e-01
4.87104148e-01 9.91955817e-01 -2.14272118e+00 5.36325693e-01
1.55755177e-01 9.41657364e-01 3.23734790e-01 1.06142449e+00
-3.68081182e-01 -6.05423674e-02 -4.28652078e-01 1.29445896e-01
-5.58790155e-02 -2.44468212e-01 -4.78749454e-01 3.90983652e-03
4.89200830e-01 -3.40940326e-01 1.83597639e-01 8.34925056e-01
1.10082936e+00 -9.62784737e-02 7.03945816e-01 -1.32194662e+00
3.08954448e-01 9.13500488e-02 -4.35277224e-01 4.90003228e-01
8.89305770e-01 2.51997441e-01 7.16056705e-01 -1.21031117e+00
8.15320313e-01 1.65831554e+00 5.30900478e-01 7.62874484e-01
-1.02552748e+00 -6.12344265e-01 -7.43513256e-02 5.60893774e-01
-1.50287616e+00 -1.19023788e+00 9.74857211e-01 -6.25684142e-01
9.25042331e-01 1.74234167e-01 7.57649720e-01 8.93660009e-01
6.34995773e-02 9.76803422e-01 7.30753064e-01 -4.36964333e-01
4.78722379e-02 -2.18945414e-01 1.51978433e-01 4.77303594e-01
-4.20403004e-01 -4.44346517e-02 -1.17421937e+00 -8.92361775e-02
1.82601228e-01 -4.97637331e-01 -3.05335879e-01 -4.98853140e-02
-1.03703177e+00 4.51050848e-01 -3.98746818e-01 1.73542917e-01
-3.60711992e-01 -7.62861297e-02 6.96758628e-01 -1.14328317e-01
6.26813889e-01 -7.97096550e-01 -2.22616732e-01 -7.00524151e-01
-1.08512449e+00 1.63192615e-01 6.63497686e-01 5.11357546e-01
1.76412225e-01 -2.98067808e-01 -5.62017709e-02 8.94010365e-01
8.69907141e-01 4.68353838e-01 5.03376544e-01 -7.92674124e-01
1.54303327e-01 4.53039348e-01 -2.86800545e-02 -3.83010030e-01
-4.57643896e-01 5.79541981e-01 -3.16630125e-01 6.73065305e-01
4.04102474e-01 1.47727191e-01 -7.46989310e-01 1.74556041e+00
8.55576694e-01 2.81110972e-01 -1.05102010e-01 7.32280850e-01
9.42736030e-01 4.70584810e-01 -6.86589926e-02 -6.54385209e-01
1.53798866e+00 -3.56569350e-01 -1.08981001e+00 3.82016718e-01
2.04164743e-01 -1.11953175e+00 1.02608442e+00 2.32900128e-01
-1.34407520e+00 -5.98582029e-01 -5.93413770e-01 -1.19137824e-01
-1.52287304e-01 6.61134198e-02 4.20272425e-02 7.62998223e-01
-1.20654976e+00 7.07105920e-02 -9.15110648e-01 -5.09719849e-01
2.34954312e-01 1.97420105e-01 -3.55430633e-01 4.46154267e-01
-8.36370587e-01 6.26588404e-01 -1.38941661e-01 -5.85536249e-02
-7.73570895e-01 -6.92412019e-01 -9.31907952e-01 -4.07921463e-01
-3.15777451e-01 -5.04064023e-01 1.45789671e+00 -3.65953475e-01
-1.91945541e+00 1.50236964e+00 -1.16464913e+00 -1.54122487e-01
9.21595752e-01 -1.53369740e-01 -4.37541634e-01 4.49671328e-01
-5.31667881e-02 6.12894058e-01 1.22903085e+00 -1.01228738e+00
-6.90558374e-01 -3.56082499e-01 -8.47171068e-01 3.20231944e-01
1.51030079e-01 6.88649893e-01 -5.60146570e-01 -5.51875591e-01
3.17872941e-01 -7.57383525e-01 8.33068311e-01 3.79627734e-01
-2.26571962e-01 -8.80119026e-01 1.31796265e+00 -6.81915820e-01
1.05972326e+00 -2.33596039e+00 -2.15508565e-01 -3.06545079e-01
-1.63673162e-01 2.86210895e-01 1.63184792e-01 1.41659334e-01
-3.58941928e-02 1.16979197e-01 5.42301126e-02 -9.16065097e-01
-9.44897011e-02 -3.06217015e-01 -1.50475413e-01 9.16223228e-01
-7.28053227e-02 5.01743913e-01 -6.25306368e-01 -7.70334244e-01
4.23724145e-01 8.96740675e-01 -3.38284910e-01 2.82357424e-01
1.41599000e-01 1.47600085e-01 -2.31418218e-02 9.89901125e-01
8.60968292e-01 5.96223235e-01 -2.51792848e-01 -3.76620770e-01
-4.72300500e-01 5.43275177e-01 -1.27224994e+00 1.80789030e+00
-3.98076147e-01 1.14586103e+00 5.68592608e-01 -3.47833395e-01
8.18517625e-01 6.84742987e-01 5.23043036e-01 -1.16001286e-01
4.54358421e-02 2.32577145e-01 -1.51050255e-01 -9.21383262e-01
2.31665671e-02 -1.45385712e-01 2.91875333e-01 3.42739820e-01
-6.12275861e-02 -2.50983745e-01 -1.02277517e-01 -3.18800688e-01
6.65981829e-01 1.46944329e-01 2.28656530e-01 -3.69656175e-01
9.60449040e-01 -8.03165972e-01 2.95011818e-01 2.52045766e-02
-9.48319614e-01 5.73817849e-01 3.32123607e-01 -4.26950157e-02
-6.11436367e-01 -1.59473372e+00 -4.22397554e-01 9.66630995e-01
3.25896069e-02 -2.74443060e-01 -8.78483057e-01 -3.42182815e-01
5.92651665e-02 4.12061721e-01 -4.39284533e-01 1.88219458e-01
-5.08822381e-01 -8.55850205e-02 6.33300424e-01 1.26014471e-01
2.92050481e-01 -1.31776226e+00 -8.73444140e-01 1.22628910e-02
-5.16340137e-01 -1.14395618e+00 -1.02260709e+00 -3.73514891e-01
-4.44253892e-01 -8.22172761e-01 -8.17746580e-01 -1.16474175e+00
6.94205314e-02 1.21446751e-01 6.25348628e-01 -4.83156413e-01
-6.49341941e-01 5.09820998e-01 1.36133105e-01 -4.12447900e-01
-8.99310052e-01 -4.17171389e-01 6.45975351e-01 3.09381694e-01
3.16050410e-01 -6.01762533e-01 -6.76175594e-01 5.56573927e-01
-3.90491724e-01 -1.63733602e-01 -2.24673122e-01 3.14446270e-01
5.59576154e-01 -1.71070963e-01 6.18323445e-01 4.89524841e-01
6.95794821e-01 -1.31229118e-01 -5.70683599e-01 9.34269130e-02
-1.69276878e-01 -1.35779127e-01 -9.85983014e-02 -8.46162796e-01
-1.04402089e+00 1.10706903e-01 -4.02175486e-01 -5.34776628e-01
-4.61300641e-01 -2.87732005e-01 -4.95425940e-01 1.09326825e-01
2.47316524e-01 2.56818712e-01 4.16080981e-01 -6.80236280e-01
3.59287024e-01 1.20923555e+00 6.55108631e-01 -4.20897491e-02
7.51278475e-02 8.97461236e-01 -6.82464093e-02 -1.65450799e+00
-1.23029776e-01 -1.05630970e+00 -7.51277328e-01 -7.65036762e-01
9.70257580e-01 -6.51554346e-01 -1.56181550e+00 1.05571949e+00
-1.44529676e+00 -3.79531682e-01 -9.87115651e-02 4.98155862e-01
-9.96409655e-01 5.21104157e-01 -5.43673217e-01 -1.31981933e+00
-5.44801950e-01 -1.32771850e+00 1.42091358e+00 2.18148947e-01
-4.34172750e-01 -7.65721560e-01 3.03708076e-01 2.61894912e-01
-1.08062252e-02 -3.89282778e-02 4.48416770e-01 -1.51762009e-01
-1.11499883e-01 -1.63270161e-02 1.46835715e-01 2.14999661e-01
5.62562346e-01 4.91374314e-01 -1.40530598e+00 6.37566522e-02
-4.53250743e-02 5.78573020e-03 8.18285882e-01 9.77646828e-01
5.73845267e-01 -2.26198554e-01 -4.23110455e-01 4.94401157e-01
7.68482327e-01 2.21794531e-01 3.80477995e-01 -1.99496001e-01
-2.24411488e-03 9.19988096e-01 5.99763095e-01 5.95389545e-01
1.95858449e-01 1.02342772e+00 3.53310615e-01 2.33247355e-01
-7.79564023e-01 -2.58462220e-01 1.06824780e+00 4.97543663e-01
1.31390020e-01 -7.11428970e-02 -1.02310109e+00 9.05000389e-01
-1.43054068e+00 -1.16202986e+00 7.23879784e-02 2.29015136e+00
8.47315967e-01 1.51508167e-01 4.82730538e-01 5.43024421e-01
1.15315914e+00 1.37437180e-01 -4.60296214e-01 -5.70508957e-01
-8.14866796e-02 9.03964639e-02 -8.66196454e-02 1.06909275e+00
-9.53693330e-01 1.09890723e+00 6.54014778e+00 7.92233706e-01
-1.51042891e+00 9.60669070e-02 5.16745411e-02 -2.08071038e-01
3.83037329e-02 -5.59857368e-01 -1.43001068e+00 4.30072933e-01
9.50349152e-01 -2.81082213e-01 -2.11045500e-02 4.16441023e-01
1.20993018e+00 -3.29218894e-01 -9.97660518e-01 1.39266813e+00
2.14084029e-01 -1.16886008e+00 -8.96124542e-02 8.37532505e-02
1.38042169e-02 2.10848659e-01 -5.44905551e-02 -4.58741874e-01
-4.40854639e-01 -8.64967346e-01 9.43195403e-01 5.99074304e-01
1.13542783e+00 -7.19136715e-01 3.48093323e-02 4.76498634e-01
-1.71445394e+00 2.69158512e-01 4.22161341e-01 3.99176180e-01
6.02901220e-01 1.64183959e-01 -1.62859464e+00 -9.79372635e-02
9.14566576e-01 4.77307439e-01 -1.07521778e-02 1.12710583e+00
-1.69659644e-01 5.12263179e-01 -6.39495373e-01 -4.12388369e-02
-3.70906979e-01 2.95488745e-01 1.30476308e+00 1.59880877e+00
3.66545439e-01 -2.54241228e-01 -2.11574882e-01 7.20753729e-01
2.18884215e-01 7.17033446e-02 -7.54610896e-01 2.74634361e-01
4.72798198e-01 1.11475623e+00 -5.89000344e-01 8.55918378e-02
-7.87618104e-03 1.10588431e+00 -5.02796113e-01 1.43317029e-01
-4.22160059e-01 -1.77936643e-01 1.49482524e+00 4.55052078e-01
6.03453666e-02 -3.25736612e-01 6.18266538e-02 -7.97075450e-01
1.55750200e-01 -4.50303733e-01 2.04472527e-01 -9.22020495e-01
-7.56208122e-01 3.54197323e-01 -4.53176573e-02 -1.08237386e+00
-6.89905822e-01 -6.33231759e-01 -7.16284215e-01 1.05626559e+00
-1.62083054e+00 -1.03291392e+00 -3.01520914e-01 9.70971286e-01
1.01358283e+00 9.00442153e-02 8.20150971e-01 8.99761841e-02
-2.48416007e-01 7.06998944e-01 -4.96730238e-01 5.80173358e-02
8.34692657e-01 -7.70469785e-01 5.48969030e-01 7.70174146e-01
-5.26092313e-02 2.54657924e-01 6.77642167e-01 -5.60307264e-01
-1.13336825e+00 -5.85173547e-01 1.21435642e+00 -2.56018400e-01
4.38259512e-01 -5.62775195e-01 -7.47263551e-01 1.11271292e-01
5.54615026e-03 2.94689685e-01 5.49333096e-01 -4.65275288e-01
-2.16830313e-01 -1.22555353e-01 -1.39384878e+00 5.09291768e-01
1.16810942e+00 -9.86972749e-01 -5.21336317e-01 4.70602959e-02
4.46746171e-01 -1.63367122e-01 -7.60890424e-01 3.48701298e-01
1.04915965e+00 -1.05735219e+00 9.74272549e-01 -1.77964181e-01
-4.81399029e-01 -4.10507023e-01 -5.59463724e-02 -7.01995850e-01
4.15641963e-01 -1.26581562e+00 -4.08159435e-01 1.47180903e+00
2.66180150e-02 -4.20625955e-01 7.38679588e-01 -4.98697497e-02
4.65820692e-02 -3.66174728e-01 -1.68788409e+00 -7.69369721e-01
-2.38644689e-01 -8.86684299e-01 8.70250165e-02 3.03607613e-01
1.30885586e-01 -1.01128310e-01 2.63787419e-01 2.13954732e-01
7.45896459e-01 -1.15006104e-01 6.23966277e-01 -1.21692288e+00
4.46334988e-01 -8.61678541e-01 -7.47425377e-01 -1.05265200e+00
4.56973940e-01 -7.48452842e-01 3.16622376e-01 -1.22562063e+00
-4.83867079e-02 2.52912156e-02 9.24534723e-02 1.37833774e-01
2.26537511e-01 -1.23933412e-01 5.52709913e-03 1.43418297e-01
2.45617107e-01 2.46733367e-01 9.01677251e-01 1.81213748e-02
-6.00609601e-01 4.64911044e-01 3.41255277e-01 1.04184949e+00
6.05679452e-01 -1.18650228e-01 -1.49653822e-01 2.93098778e-01
-3.79847497e-01 3.61711383e-01 3.48220557e-01 -7.99436867e-01
3.07546884e-01 8.15233886e-02 -1.42583072e-01 -1.26929224e+00
6.74318969e-01 -3.53253305e-01 -2.46507525e-01 4.97473836e-01
-1.83022916e-01 -5.40202409e-02 5.34132063e-01 1.22952946e-01
-1.25541449e-01 2.93044392e-02 1.31937778e+00 2.71998525e-01
-6.00982785e-01 3.25055897e-01 -5.68377852e-01 1.01417050e-01
1.00708747e+00 -3.52641523e-01 2.81408697e-01 -4.22044545e-01
-8.61826777e-01 -4.23493564e-01 3.66163760e-01 5.52534580e-01
8.79633427e-01 -1.10292923e+00 -7.89389789e-01 5.31392395e-01
-1.01762451e-01 -1.91033661e-01 7.70761520e-02 7.92435706e-01
-3.36765677e-01 1.12286694e-01 -6.77572191e-03 -1.29901671e+00
-2.12754655e+00 2.89089590e-01 7.62288272e-01 3.08495134e-01
-7.31411576e-01 9.48778689e-01 -1.32830618e-02 1.57335147e-01
6.73607528e-01 -5.41901529e-01 -1.84946597e-01 6.52741909e-01
7.36844659e-01 6.34121656e-01 9.42662358e-02 -9.84563589e-01
-7.44026065e-01 1.05286264e+00 3.46942008e-01 -7.73396909e-01
8.08728158e-01 -6.07510924e-01 8.21416155e-02 8.39942455e-01
1.50825298e+00 5.45880556e-01 -1.23720205e+00 2.48902783e-01
-4.98214774e-02 -1.61654159e-01 1.05934016e-01 -4.01144952e-01
-5.09230077e-01 1.31880879e+00 1.02997577e+00 1.41938746e-01
1.20465839e+00 2.38258228e-01 6.84799373e-01 -3.75998139e-01
-6.95443302e-02 -9.02600765e-01 5.12923673e-02 3.36015880e-01
1.16588056e+00 -1.03436530e+00 -5.42083204e-01 -4.61907655e-01
-2.69661933e-01 1.42212069e+00 -6.71186820e-02 5.00395417e-01
1.07339191e+00 7.06378520e-01 2.67298192e-01 3.31303269e-01
-5.79309702e-01 -5.00794709e-01 4.73463833e-01 7.98987567e-01
4.08689380e-01 -4.91227284e-02 -1.59835115e-01 1.33209035e-01
-9.01076645e-02 9.67850015e-02 1.05865546e-01 5.93782008e-01
-4.81079280e-01 -8.13764870e-01 -4.52767462e-01 -4.31600511e-01
-4.37726617e-01 -3.68145965e-02 -2.56011546e-01 6.14207447e-01
2.27003217e-01 1.06761098e+00 2.19801590e-01 -1.19868480e-01
3.39356303e-01 7.25251496e-01 4.57317680e-01 -1.47201061e-01
-2.47765630e-01 4.97026414e-01 -1.12969220e-01 -6.78398252e-01
-4.97366875e-01 -1.21486878e+00 -1.62017715e+00 -1.20737627e-01
-8.11200067e-02 -4.22813147e-01 1.08438802e+00 5.16985953e-01
4.81275201e-01 -6.69007450e-02 7.52997100e-01 -1.05420959e+00
-2.63881922e-01 -1.09150720e+00 -4.60871935e-01 3.44532430e-01
1.07764077e+00 -7.20776975e-01 -9.29079950e-01 3.02895159e-01] | [14.339941024780273, 5.006660461425781] |
a5c87ef4-e54f-48ab-a8fa-460d29eb6aaf | an-application-of-a-runtime-epistemic | 2209.13043 | null | https://arxiv.org/abs/2209.13043v1 | https://arxiv.org/pdf/2209.13043v1.pdf | An Application of a Runtime Epistemic Probabilistic Event Calculus to Decision-making in e-Health Systems | We present and discuss a runtime architecture that integrates sensorial data and classifiers with a logic-based decision-making system in the context of an e-Health system for the rehabilitation of children with neuromotor disorders. In this application, children perform a rehabilitation task in the form of games. The main aim of the system is to derive a set of parameters the child's current level of cognitive and behavioral performance (e.g., engagement, attention, task accuracy) from the available sensors and classifiers (e.g., eye trackers, motion sensors, emotion recognition techniques) and take decisions accordingly. These decisions are typically aimed at improving the child's performance by triggering appropriate re-engagement stimuli when their attention is low, by changing the game or making it more difficult when the child is losing interest in the task as it is too easy. Alongside state-of-the-art techniques for emotion recognition and head pose estimation, we use a runtime variant of a probabilistic and epistemic logic programming dialect of the Event Calculus, known as the Epistemic Probabilistic Event Calculus. In particular, the probabilistic component of this symbolic framework allows for a natural interface with the machine learning techniques. We overview the architecture and its components, and show some of its characteristics through a discussion of a running example and experiments. Under consideration for publication in Theory and Practice of Logic Programming (TPLP). | ['Silvia Rossi', 'Marco Grazioso', 'Salim Malek', 'Luca Raggioli', "Fabio Aurelio D'Asaro"] | 2022-09-26 | null | null | null | null | ['head-pose-estimation'] | ['computer-vision'] | [ 1.14925854e-01 6.26929045e-01 -1.62299007e-01 -4.41075057e-01
-3.35521966e-01 -3.41878474e-01 3.46551418e-01 4.96492445e-01
-6.70217395e-01 6.23674035e-01 1.11504748e-01 3.78851593e-02
-6.13469839e-01 -7.97617972e-01 -6.79715753e-01 -6.81661785e-01
-2.04008422e-03 5.60073435e-01 4.96127337e-01 -6.92221895e-03
-2.46047112e-03 5.12002647e-01 -2.62833953e+00 2.51765430e-01
6.54470444e-01 9.68358099e-01 -2.74379514e-02 7.49343455e-01
4.49013077e-02 7.66700804e-01 -4.08953190e-01 -2.40941107e-01
-4.59684819e-01 -5.79360574e-02 -6.94344163e-01 -3.91577154e-01
-2.07161888e-01 -1.61709219e-01 1.29067391e-01 1.12592256e+00
5.53789437e-01 1.79461658e-01 5.47604561e-01 -1.43072283e+00
1.95639700e-01 7.76300251e-01 1.94099069e-01 -1.02550849e-01
8.86444807e-01 7.20700622e-02 3.36203545e-01 -1.87963396e-01
2.64638990e-01 9.02407408e-01 2.24076852e-01 8.89678299e-01
-1.13662374e+00 -2.74727732e-01 1.41711816e-01 6.76904678e-01
-1.40460563e+00 -5.30719340e-01 6.68926835e-01 -5.65013945e-01
1.00757790e+00 4.85535979e-01 1.27615142e+00 1.10509801e+00
1.00724734e-01 6.98440015e-01 1.11132741e+00 -7.33189642e-01
8.93607736e-01 2.59859979e-01 -2.02846620e-02 3.03836733e-01
1.44567207e-01 3.30212742e-01 -9.33547080e-01 -9.64964032e-02
1.86924338e-01 -5.57593584e-01 -9.76889357e-02 -3.96996588e-01
-6.16940260e-01 5.27512610e-01 -1.90732822e-01 4.41251338e-01
-6.75137281e-01 3.10553998e-01 4.01743829e-01 2.25188192e-02
1.88994288e-01 -3.68845388e-02 -4.55692887e-01 -8.08425784e-01
-8.32913816e-01 5.45444310e-01 9.32181180e-01 7.31562853e-01
1.60397708e-01 -3.48203361e-01 -3.54628831e-01 5.42549968e-01
6.43480480e-01 2.78571546e-01 4.66733098e-01 -9.51962471e-01
-2.50044554e-01 7.09061563e-01 -1.91405360e-02 -5.20465016e-01
-7.31474102e-01 2.58706123e-01 4.73601036e-02 7.17268944e-01
3.92041534e-01 -2.35624030e-01 -4.65930432e-01 2.06699634e+00
7.88190365e-01 2.54708111e-01 1.38766263e-02 6.22037411e-01
9.44246590e-01 3.25327739e-02 4.41286922e-01 -4.36189681e-01
1.53058076e+00 -6.72338903e-02 -9.40019786e-01 -9.37141106e-02
5.02846241e-01 -2.84270287e-01 5.40325820e-01 8.77116382e-01
-1.38513315e+00 1.64127380e-01 -8.15648139e-01 3.03056180e-01
-4.94941443e-01 -1.54479489e-01 7.50333011e-01 1.04313266e+00
-1.00710845e+00 5.00825644e-01 -1.16754377e+00 -4.78003621e-01
8.51268694e-03 6.85984254e-01 -2.39862099e-01 4.73433584e-01
-1.31884086e+00 1.27897644e+00 6.44741178e-01 6.63095415e-02
-6.12831473e-01 -4.54593182e-01 -9.36481416e-01 4.77847233e-02
1.80141851e-01 -3.04886043e-01 1.41524231e+00 -8.66889238e-01
-2.25707626e+00 1.21605408e+00 4.00520653e-01 -2.20256090e-01
4.67714876e-01 -8.81270245e-02 -2.90208936e-01 1.86698452e-01
-2.86369741e-01 6.89065635e-01 6.67513669e-01 -2.56707042e-01
-7.34931469e-01 -6.86009943e-01 1.65434033e-01 3.37115556e-01
-7.75620416e-02 6.29141271e-01 -1.80906594e-01 -1.37987986e-01
-6.36648312e-02 -8.13075364e-01 5.21208793e-02 -1.74943328e-01
-1.26970217e-01 -4.97484893e-01 1.43740237e-01 -2.94455916e-01
1.22358394e+00 -1.96957028e+00 3.62955093e-01 3.08702111e-01
-1.10757321e-01 9.18554589e-02 5.29913127e-01 3.15161608e-02
-3.77079308e-01 -4.34952110e-01 1.73920199e-01 4.24757414e-02
3.06722581e-01 2.04373896e-01 3.46377850e-01 6.71792209e-01
-1.20419040e-01 4.53876972e-01 -7.57287800e-01 -5.65506577e-01
6.37817383e-01 4.86783355e-01 -5.97077608e-01 3.55975330e-01
-4.75694835e-01 4.00973082e-01 -4.44234520e-01 4.26787883e-01
1.48299202e-01 6.73557520e-01 1.68651909e-01 1.99577555e-01
-5.23100853e-01 6.10271931e-01 -1.50269842e+00 1.44046736e+00
-2.25024104e-01 3.10064673e-01 1.54801562e-01 -8.35921824e-01
5.00855148e-01 7.21270204e-01 5.17928064e-01 -3.10042351e-01
6.09512269e-01 3.77297215e-02 -9.71555263e-02 -1.11370766e+00
-3.48533951e-02 1.50851399e-01 -8.33763778e-02 3.50401670e-01
-1.15924738e-01 -2.67850041e-01 3.55144106e-02 -4.02844101e-01
1.21700823e+00 6.04404867e-01 5.77934444e-01 -2.87458122e-01
6.10282004e-01 -2.26376578e-01 4.17845070e-01 3.24919105e-01
-3.09555948e-01 2.93068923e-02 7.71543205e-01 -7.24195465e-02
-3.37462455e-01 -8.16969395e-01 -2.84572542e-01 1.35317969e+00
-4.55268115e-01 -1.70486420e-01 -1.29735005e+00 -1.85422137e-01
-2.62064993e-01 1.06474090e+00 -7.11740851e-01 -5.66722155e-01
-1.84900582e-01 -4.87845868e-01 5.10277629e-01 4.26389456e-01
-1.24684535e-01 -1.47501326e+00 -1.50871909e+00 1.59511149e-01
8.25262591e-02 -6.52444065e-01 2.30031773e-01 2.71367460e-01
-5.76735497e-01 -9.68085408e-01 -1.16992854e-01 -2.68306702e-01
2.68893301e-01 -8.36605906e-01 8.14741850e-01 -2.11918592e-01
-1.40601844e-01 8.49104643e-01 -4.38866317e-01 -8.73698115e-01
-2.90780902e-01 -2.89356828e-01 2.51062751e-01 -1.75673500e-01
5.39655983e-01 -7.97981143e-01 -4.72819388e-01 -2.89476544e-01
-7.45583057e-01 8.32960233e-02 -1.03463180e-01 3.16036314e-01
4.90127027e-01 1.69294596e-01 5.51750176e-02 -3.30494165e-01
5.35757780e-01 -4.63939637e-01 -9.12665188e-01 1.91323698e-01
-6.02507770e-01 3.59487534e-02 -2.85199899e-02 -8.22623849e-01
-8.06288183e-01 1.99266300e-01 -2.62719035e-01 -1.14734851e-01
-6.53153837e-01 5.24858892e-01 -4.15414184e-01 -5.16555309e-02
5.34366429e-01 -2.03931272e-01 -4.47114222e-02 -1.19782001e-01
9.66571271e-02 8.93937230e-01 6.74233496e-01 -7.71439254e-01
-2.74982572e-01 6.94560334e-02 1.13980435e-01 -3.32784742e-01
-5.23400366e-01 6.37855157e-02 -3.60526234e-01 -8.52555633e-01
1.06564224e+00 -3.92716289e-01 -1.55919504e+00 6.96566880e-01
-9.79188085e-01 -4.05476868e-01 -5.28572202e-01 7.45323956e-01
-9.77363944e-01 -3.83454233e-01 -7.22153112e-02 -1.32825911e+00
-3.43241513e-01 -1.21007478e+00 9.35451567e-01 6.06383383e-01
-3.69021356e-01 -7.67471433e-01 1.88426405e-01 2.71533221e-01
8.11711401e-02 4.88535762e-01 7.55678117e-01 -9.06336963e-01
-6.18039966e-02 -3.96848768e-01 4.82889533e-01 1.44943491e-01
-4.33838487e-01 3.68482023e-01 -1.08685517e+00 2.41272241e-01
1.33626074e-01 -1.43083796e-01 -1.67911828e-01 4.80956584e-01
1.01722074e+00 4.75154147e-02 -1.93997115e-01 2.99827754e-01
1.03328562e+00 3.63901466e-01 6.04542077e-01 3.31577986e-01
1.80844888e-01 1.04181385e+00 7.61448741e-01 5.32214940e-01
7.56145716e-01 1.27482677e+00 7.75529504e-01 7.05658317e-01
1.67140946e-01 7.55715370e-02 5.60329437e-01 9.16117132e-02
-2.82978237e-01 2.35940978e-01 -1.09276378e+00 4.47073400e-01
-1.99814701e+00 -7.89633453e-01 -2.22954273e-01 2.59412718e+00
1.02232432e+00 2.22076759e-01 4.29556638e-01 5.41475356e-01
7.56688595e-01 -4.63870645e-01 -3.04883540e-01 -9.47291315e-01
5.60553908e-01 5.40353596e-01 2.23209649e-01 3.24996442e-01
-6.14063203e-01 8.12475920e-01 6.03048706e+00 7.30380297e-01
-9.31726992e-01 2.65114039e-01 2.01359197e-01 -3.77830893e-01
-1.40047483e-02 -2.52664387e-01 -6.91068232e-01 6.54093921e-01
1.32004070e+00 1.10919870e-01 7.59610295e-01 7.68378019e-01
2.34533623e-01 -7.50530183e-01 -1.49134755e+00 9.79332149e-01
-1.87930781e-02 -6.74132407e-01 -9.53918457e-01 -3.69997352e-01
-2.04597369e-01 -8.32762718e-02 -3.97690505e-01 2.44094744e-01
5.05667143e-02 -8.44151258e-01 1.37131202e+00 9.84204710e-01
7.25652218e-01 -8.12669396e-01 5.09271443e-01 5.32781601e-01
-5.80763340e-01 -5.46115600e-02 3.69133264e-01 -2.84036785e-01
-5.70145398e-02 4.77101177e-01 -3.47607881e-01 -3.71485092e-02
1.01386523e+00 -1.80742875e-01 -1.81495935e-01 1.07209110e+00
-8.03231239e-01 4.25245553e-01 -9.36519265e-01 -5.16410708e-01
-2.97845423e-01 1.48340508e-01 8.80252779e-01 9.12291586e-01
1.74205750e-01 4.54287469e-01 -3.72777551e-01 8.83679628e-01
6.53088212e-01 1.07661910e-01 -4.48006660e-01 3.50834191e-01
4.88347322e-01 1.15873003e+00 -5.64950824e-01 -7.73451477e-02
-7.38566741e-02 2.87277162e-01 3.18715990e-01 -1.89303219e-01
-8.18940699e-01 -2.24622265e-01 6.72710061e-01 1.58704117e-01
-1.02091283e-01 4.83356446e-01 -3.23451430e-01 -6.52762294e-01
1.14610428e-02 -7.46863246e-01 3.42791826e-01 -1.12308216e+00
-4.63088065e-01 2.01740026e-01 6.15983546e-01 -7.18226492e-01
-7.73669779e-01 -6.80698514e-01 -6.51240587e-01 6.53318882e-01
-7.49283850e-01 -8.63956571e-01 -1.41254589e-01 6.09121442e-01
-5.12421690e-02 2.90472835e-01 1.01888931e+00 9.15826783e-02
-5.95940948e-01 4.05814260e-01 -6.99208379e-01 -3.64700109e-01
2.44141936e-01 -1.34390223e+00 -4.66058791e-01 4.98457044e-01
-4.23822224e-01 -6.42544925e-02 1.18214798e+00 -4.53862816e-01
-1.26362944e+00 -3.42249364e-01 8.61435413e-01 -2.66352147e-01
5.38381696e-01 -3.59230727e-01 -6.21455848e-01 4.19069916e-01
-1.66652426e-01 -1.43185645e-01 5.53394437e-01 2.32136417e-02
2.52599657e-01 -2.33653858e-01 -1.68736637e+00 6.00873947e-01
9.28993821e-01 -3.70777965e-01 -7.60185182e-01 2.81659901e-01
2.93239743e-01 -8.51349175e-01 -9.53385413e-01 5.95761299e-01
7.89631665e-01 -1.10021949e+00 6.66442811e-01 -3.88128757e-01
-1.05821267e-01 -2.04179913e-01 1.82547327e-03 -1.07540524e+00
1.28908724e-01 -7.21564293e-01 -3.16166162e-01 1.20877314e+00
8.24567080e-02 -5.90098441e-01 6.39775097e-01 1.39406729e+00
1.03201970e-01 -7.35340774e-01 -1.38716233e+00 -1.24868065e-01
-1.89168945e-01 -1.21029747e+00 7.99549997e-01 4.13982093e-01
8.56528461e-01 -1.77642137e-01 2.50071406e-01 3.07021230e-01
5.08087933e-01 -3.34441274e-01 2.23159701e-01 -1.24259341e+00
-1.81376919e-01 -5.71167409e-01 -1.07258177e+00 7.00262487e-02
2.22031638e-01 -5.92382550e-01 2.64515787e-01 -1.15369642e+00
-2.44148582e-01 -1.43986300e-01 -1.18608087e-01 7.86917627e-01
2.05764249e-01 -3.43519479e-01 -1.85976475e-01 -5.81738353e-01
-5.51613688e-01 5.44272140e-02 3.69179815e-01 3.71979147e-01
-5.16431808e-01 4.48219299e-01 -3.26173723e-01 1.09033394e+00
5.96989274e-01 -6.30330801e-01 -2.32554734e-01 3.39088682e-03
9.46946859e-01 4.70992386e-01 2.73052514e-01 -1.09753001e+00
6.00474358e-01 -3.73931944e-01 6.34105280e-02 -4.10325915e-01
3.34721327e-01 -1.13062024e+00 4.54121500e-01 4.07731980e-01
-4.19571668e-01 -3.32846671e-01 2.37695217e-01 -9.54259560e-02
1.76937267e-01 -6.97639763e-01 7.24702895e-01 3.74197364e-01
-3.24583203e-01 -1.57736197e-01 -9.28206205e-01 -2.63433725e-01
1.49652553e+00 -2.48108521e-01 -7.73328766e-02 -1.82520807e-01
-1.32946849e+00 2.28565797e-01 2.13445678e-01 2.55408406e-01
2.96587527e-01 -1.04363596e+00 -3.28405976e-01 2.74114460e-01
2.29055673e-01 3.30101675e-03 1.25695542e-01 1.22989595e+00
-3.49578857e-01 2.07229614e-01 -2.70767331e-01 -5.04979014e-01
-1.45418572e+00 3.50068241e-01 6.00928545e-01 3.80474739e-02
-4.08059597e-01 6.86403096e-01 -4.91118073e-01 -1.39828309e-01
5.90957105e-01 -5.38185835e-01 -8.23428214e-01 2.56117195e-01
7.45948434e-01 6.16215229e-01 5.04594624e-01 -4.52808022e-01
-6.70224190e-01 1.79632843e-01 5.20553946e-01 -4.73198205e-01
1.47636795e+00 -2.50434019e-02 -3.93005639e-01 5.90296566e-01
1.16128303e-01 -1.10535115e-01 -8.84564519e-01 1.86505958e-01
5.33228554e-02 9.81186144e-03 4.07729924e-01 -1.02740479e+00
-8.21247578e-01 4.16271478e-01 9.71112251e-01 5.44336021e-01
1.55779564e+00 2.79974580e-01 -1.33224189e-01 8.79350379e-02
6.33644402e-01 -1.46341622e+00 -7.31576264e-01 3.27910870e-01
8.84503603e-01 -6.82255566e-01 -5.96168786e-02 -2.42931724e-01
-2.51012057e-01 9.86050367e-01 5.38499653e-01 1.43136248e-01
7.79989779e-01 7.42997169e-01 -3.20518702e-01 -4.98047709e-01
-1.05570590e+00 -5.16345561e-01 2.87529111e-01 7.09875524e-01
2.32639924e-01 6.42045498e-01 -6.24466062e-01 1.09233558e+00
-3.81701529e-01 3.34011257e-01 3.28645110e-01 9.50962007e-01
-4.16261405e-01 -9.61854100e-01 -7.66680539e-01 3.32881361e-01
-6.79363608e-01 1.85720414e-01 -1.72970027e-01 4.96483237e-01
6.60257161e-01 1.00093091e+00 3.96539690e-03 -1.99242771e-01
6.79447830e-01 4.59796906e-01 9.68941212e-01 -8.23052049e-01
-7.55907893e-01 -1.43167391e-01 3.99243474e-01 -9.34286535e-01
-6.27959430e-01 -1.21523356e+00 -1.46113515e+00 -7.73544237e-02
-3.85714889e-01 -1.48946509e-01 1.38425660e+00 1.26564765e+00
-1.57336548e-01 6.03616297e-01 1.07172944e-01 -9.21477616e-01
-8.27696547e-02 -8.07456493e-01 -4.82497841e-01 -2.11916581e-01
-6.23240508e-02 -1.05706346e+00 -2.60318905e-01 -2.88952470e-01] | [13.146673202514648, 3.106675148010254] |
8d7b9c7a-1ebd-4bea-8cfa-ae1e21231fb6 | meta-learning-for-vision-and-language-cross | 2305.14843 | null | https://arxiv.org/abs/2305.14843v1 | https://arxiv.org/pdf/2305.14843v1.pdf | Meta-Learning For Vision-and-Language Cross-lingual Transfer | Current pre-trained vison-language models (PVLMs) achieve excellent performance on a range of multi-modal datasets. Recent work has aimed at building multilingual models, and a range of novel multilingual multi-modal datasets have been proposed. Current PVLMs typically perform poorly on these datasets when used for multi-modal zero-shot or few-shot cross-lingual transfer, especially for low-resource languages. To alleviate this problem, we propose a novel meta-learning fine-tuning framework. Our framework makes current PVLMs rapidly adaptive to new languages in vision-language scenarios by designing MAML in a cross-lingual multi-modal manner. Experiments show that our method boosts the performance of current state-of-the-art PVLMs in both zero-shot and few-shot cross-lingual transfer on a range of vision-language understanding tasks and datasets (XVNLI, xGQA, MaRVL, xFlicker&Co | ['Frank Keller', 'Hanxu Hu'] | 2023-05-24 | null | null | null | null | ['cross-lingual-transfer'] | ['natural-language-processing'] | [-2.18797803e-01 -4.58043188e-01 -5.02817631e-01 -4.70531017e-01
-1.52212250e+00 -4.93671924e-01 1.05382490e+00 -3.91772389e-01
-7.83442557e-01 5.74597180e-01 1.72094330e-01 -3.04533720e-01
6.06688797e-01 -4.35661316e-01 -9.14626896e-01 -3.83553386e-01
4.85582888e-01 8.30302298e-01 2.05608562e-01 -3.83231670e-01
-9.06183347e-02 -8.33816901e-02 -1.47370493e+00 6.92743480e-01
8.61912966e-01 3.40815365e-01 6.60412908e-01 5.98970234e-01
-5.63753188e-01 4.90761250e-01 -9.81431901e-02 -5.46708167e-01
1.45498365e-01 -1.82442054e-01 -7.71934330e-01 -2.21011728e-01
1.17590129e+00 -9.47624967e-02 -1.88616998e-02 9.89439309e-01
7.76763737e-01 -4.63169478e-02 7.75544107e-01 -1.24717104e+00
-1.33283031e+00 6.87537670e-01 -9.22377706e-01 -2.54752152e-02
1.20086797e-01 3.69527370e-01 8.89785588e-01 -1.16691101e+00
7.07062364e-01 1.77570736e+00 7.91514874e-01 9.18303251e-01
-1.28980041e+00 -9.03405607e-01 1.14348605e-01 4.57642376e-01
-1.22551703e+00 -6.30099535e-01 3.15206289e-01 -3.64365041e-01
1.27797389e+00 -3.39293480e-01 2.12124616e-01 1.48537266e+00
3.90273809e-01 1.18033886e+00 1.43947816e+00 -7.15333879e-01
-3.75921935e-01 4.66519237e-01 -2.62016326e-01 8.52457821e-01
-4.04336333e-01 6.95891455e-02 -7.71768510e-01 3.23375702e-01
3.93484771e-01 -4.74360853e-01 7.45706856e-02 -4.64702696e-01
-1.56777680e+00 1.11398304e+00 2.01155365e-01 4.16466057e-01
1.73901897e-02 2.85803020e-01 8.48717630e-01 4.96970743e-01
6.91100478e-01 -3.58172767e-02 -7.09404707e-01 -7.49413520e-02
-9.19757783e-01 -1.05282456e-01 3.04070950e-01 1.09566832e+00
9.78065193e-01 2.20330834e-01 -4.02517200e-01 1.30399203e+00
4.39434975e-01 9.85599697e-01 6.84817791e-01 -5.68078995e-01
6.63919091e-01 1.94178998e-01 -4.45698023e-01 -1.94843620e-01
-3.83951575e-01 8.32997113e-02 -5.92637837e-01 1.66852474e-01
3.21298465e-02 2.97854468e-02 -1.41365886e+00 1.91600490e+00
2.00907186e-01 8.13581496e-02 3.48559618e-01 5.01601100e-01
1.07463276e+00 9.39338923e-01 6.74253225e-01 -7.94842094e-02
1.37622678e+00 -1.45823252e+00 -5.62467396e-01 -4.74283755e-01
8.64438236e-01 -1.04943860e+00 1.66441298e+00 5.63602485e-02
-8.20581496e-01 -9.96916652e-01 -6.97649837e-01 -4.77386981e-01
-7.80093074e-01 8.25155079e-02 5.40505230e-01 5.71081161e-01
-1.36093020e+00 -2.78841287e-01 -6.36731386e-01 -8.23119879e-01
2.71579325e-01 -6.21245392e-02 -4.40453708e-01 -6.12615108e-01
-1.26236713e+00 1.37913334e+00 5.85710049e-01 -3.79983306e-01
-1.34278011e+00 -8.93166363e-01 -1.05062330e+00 -4.36273336e-01
1.17446057e-01 -7.97673941e-01 1.17979085e+00 -1.06172025e+00
-1.29792988e+00 1.35302496e+00 -1.33905560e-01 -5.12278199e-01
3.76367956e-01 -2.36107945e-01 -6.60451412e-01 -3.82029712e-02
4.96487230e-01 1.57450557e+00 9.94993269e-01 -1.24763179e+00
-8.69344950e-01 -1.34116724e-01 -3.18210274e-02 5.05108893e-01
-4.90985274e-01 7.71401227e-02 -7.62734771e-01 -4.40147758e-01
-7.41850734e-01 -1.00986874e+00 -4.81145736e-03 -1.69018596e-01
-1.53520927e-02 -5.65377772e-01 9.62749302e-01 -4.57295001e-01
6.87257528e-01 -1.80807912e+00 2.59830981e-01 -7.10412383e-01
-4.31098282e-01 5.65452278e-01 -8.31543565e-01 5.21948993e-01
2.25056693e-01 -2.27884784e-01 -2.76917145e-02 -6.66802347e-01
5.98989874e-02 4.48104143e-01 -3.03428292e-01 2.00111732e-01
-4.61942926e-02 1.32010126e+00 -8.22791874e-01 -8.86259675e-01
6.21424496e-01 6.67922080e-01 -2.36504212e-01 9.20130238e-02
-4.68838573e-01 4.83326197e-01 -6.35549575e-02 7.74156332e-01
5.15335619e-01 -1.01277597e-01 -1.90218493e-01 -4.60241020e-01
-2.84863085e-01 -2.34698713e-01 -3.79920870e-01 2.57582235e+00
-1.00413036e+00 6.19039595e-01 -1.75423563e-01 -7.57075727e-01
5.24754703e-01 3.87489676e-01 1.53200135e-01 -1.06111383e+00
-1.01587884e-01 1.43542588e-01 -2.23582789e-01 -4.64134455e-01
6.73535705e-01 -4.42463696e-01 -2.19763219e-01 3.91941220e-01
7.40318656e-01 -1.41021445e-01 3.27308625e-01 3.42953324e-01
2.59288788e-01 4.76055831e-01 1.80936158e-01 -3.25813651e-01
6.36756539e-01 2.03158632e-01 7.63936266e-02 8.20027769e-01
-4.31606054e-01 3.79297674e-01 -5.00453174e-01 -2.23272666e-01
-8.82605731e-01 -1.18134534e+00 -2.01742783e-01 1.86376917e+00
-2.09660187e-01 -2.16491088e-01 -6.40443027e-01 -6.41930699e-01
-3.12253162e-02 9.96743441e-01 -5.00683188e-01 -1.37096271e-01
-8.14564973e-02 -7.17872024e-01 7.65359819e-01 4.84970778e-01
4.16725427e-01 -1.22704411e+00 -2.63231993e-01 2.63747633e-01
-3.23116004e-01 -1.51123810e+00 -6.22373641e-01 3.56432833e-02
-5.39289236e-01 -6.16765797e-01 -1.01162446e+00 -1.22909987e+00
2.23453611e-01 4.79720205e-01 1.37299180e+00 -7.05622077e-01
-4.24306214e-01 1.07569885e+00 -2.78486967e-01 -5.10202110e-01
-5.68477750e-01 2.16008142e-01 3.67305219e-01 -3.24095488e-02
8.08890224e-01 -1.71417315e-02 5.27218021e-02 8.35472643e-02
-6.91448808e-01 2.63878375e-01 6.27820373e-01 8.42682242e-01
7.89472103e-01 -6.02943480e-01 8.16074193e-01 -7.39649296e-01
6.37431324e-01 -2.66645163e-01 -6.50933087e-01 8.18304837e-01
-4.91204619e-01 1.45625904e-01 4.59333122e-01 -6.99623406e-01
-1.36335266e+00 -5.93094267e-02 -4.63061035e-02 -7.40158975e-01
-1.45736471e-01 5.77719092e-01 1.63907912e-02 -3.55396807e-01
5.80818474e-01 4.14891422e-01 -2.64740407e-01 -3.22695851e-01
1.30389798e+00 7.12281168e-01 6.68676615e-01 -7.50387907e-01
6.96975887e-01 4.69102412e-01 -4.48173255e-01 -9.56572711e-01
-9.94058669e-01 -5.00777483e-01 -8.58696163e-01 -1.91344306e-01
1.28746891e+00 -1.62350154e+00 -1.82383195e-01 6.43260658e-01
-8.99257362e-01 -5.42298436e-01 -8.45735893e-03 4.80815738e-01
-5.84657311e-01 1.79945663e-01 -5.70009708e-01 -4.67170954e-01
-6.35509789e-01 -1.36735582e+00 1.32412267e+00 2.37067476e-01
1.19251482e-01 -1.45585239e+00 5.36183417e-01 6.19763196e-01
5.39576173e-01 -3.89007121e-01 9.38179612e-01 -2.63422936e-01
-4.02889937e-01 3.64821255e-01 -3.82564396e-01 3.35124314e-01
-1.32608125e-02 -1.85884744e-01 -1.16247332e+00 -6.76751733e-01
-5.50495267e-01 -1.40654147e+00 1.14951122e+00 5.20438969e-01
6.11759663e-01 2.61615276e-01 -3.35527599e-01 8.32514346e-01
1.46205783e+00 -1.38828218e-01 2.93277830e-01 4.52380061e-01
1.10232651e+00 3.13495815e-01 7.28990197e-01 -2.27117136e-01
1.17514372e+00 7.10594475e-01 1.71487033e-01 -2.85501122e-01
-4.63757902e-01 -2.69596994e-01 7.51109898e-01 1.34418583e+00
1.80869535e-01 -2.86707040e-02 -9.58811045e-01 1.01223516e+00
-1.82471502e+00 -6.81336641e-01 2.62096435e-01 1.83022761e+00
1.13545156e+00 -2.14307249e-01 -8.73212740e-02 -1.01581800e+00
3.65836293e-01 3.59242022e-01 -5.82808018e-01 -6.40498221e-01
-4.05278713e-01 7.44289905e-02 4.54906911e-01 5.06201327e-01
-1.28179228e+00 1.86162150e+00 6.17042971e+00 1.11407936e+00
-1.34248996e+00 6.46975279e-01 2.77465045e-01 -1.56198546e-01
-2.24803582e-01 -2.20138207e-01 -1.21858013e+00 9.50907916e-03
1.15755832e+00 -9.19188410e-02 3.00309032e-01 7.92265296e-01
-3.27437222e-01 9.35066044e-02 -1.03731775e+00 1.33759403e+00
7.00545967e-01 -1.18374908e+00 4.23253924e-01 -2.35063612e-01
9.66883063e-01 1.09560287e+00 4.45306093e-01 9.83064175e-01
5.95682979e-01 -9.96748984e-01 4.71227646e-01 2.48957932e-01
1.30043364e+00 -7.98700690e-01 2.00266451e-01 2.87173808e-01
-1.27299535e+00 4.57225144e-01 -6.22981727e-01 5.60717523e-01
4.89899546e-01 -9.33964476e-02 -7.80263543e-01 5.29036701e-01
8.59170020e-01 8.18509579e-01 -7.13375986e-01 3.82570416e-01
-9.10862535e-02 3.65963489e-01 -3.01245190e-02 4.32770461e-01
7.50391245e-01 2.32785475e-03 3.35929722e-01 1.59619212e+00
1.98349908e-01 -5.43143690e-01 4.91291732e-01 5.58432221e-01
-1.36088908e-01 3.42578590e-01 -6.73501909e-01 -1.29378363e-01
1.20660022e-01 1.41484618e+00 -8.21251199e-02 -5.84733486e-01
-1.01176834e+00 1.03027058e+00 6.65826201e-01 4.02959377e-01
-7.78509915e-01 1.54469937e-01 5.71062267e-01 -4.95089620e-01
2.55084693e-01 -2.41426215e-01 2.19608232e-01 -1.53135371e+00
-4.94001657e-01 -1.12883961e+00 5.80780804e-01 -9.23720121e-01
-1.68817902e+00 7.07128823e-01 7.26630092e-02 -1.07881391e+00
-3.79157513e-01 -7.43135810e-01 -1.86533809e-01 9.03153360e-01
-2.02064705e+00 -2.30013847e+00 1.25622630e-01 1.24313152e+00
1.12783647e+00 -8.17959726e-01 1.08636808e+00 5.03766954e-01
-1.06810398e-01 8.01231921e-01 1.01246551e-01 -1.20357551e-01
1.52714503e+00 -1.02524209e+00 5.55644512e-01 8.29983652e-01
6.31909609e-01 3.32855463e-01 4.54804450e-01 -5.14011204e-01
-1.61788332e+00 -1.32169378e+00 8.39856207e-01 -6.57095134e-01
8.28142524e-01 -2.78516889e-01 -1.05300570e+00 1.09710848e+00
8.81623983e-01 8.07571504e-03 6.07376933e-01 4.04465705e-01
-8.29293311e-01 3.15279625e-02 -8.67061973e-01 7.52269149e-01
6.90859675e-01 -1.08900380e+00 -7.02973008e-01 4.99534786e-01
8.41262460e-01 -1.92877084e-01 -7.74774492e-01 4.52961743e-01
5.95019162e-01 -5.98690808e-01 1.19642484e+00 -7.21471012e-01
3.02509964e-01 -1.65460736e-01 -3.76368493e-01 -1.64192259e+00
-3.57583642e-01 -1.74327195e-01 -1.42322956e-02 1.24442291e+00
4.42392707e-01 -3.83895040e-01 9.98435393e-02 -5.91467656e-02
-3.19227815e-01 -2.40149543e-01 -1.06218767e+00 -7.51140952e-01
4.65818018e-01 -6.65205061e-01 -1.15657188e-01 1.34908986e+00
-3.81671220e-01 8.03345442e-01 -8.95976961e-01 -3.62345055e-02
1.00202620e+00 3.36425714e-02 1.01528931e+00 -6.39692485e-01
-3.12067986e-01 -2.94416159e-01 -7.77391493e-02 -8.29381883e-01
5.83808541e-01 -1.27531528e+00 -4.47301902e-02 -1.67732966e+00
4.81423944e-01 -1.88860536e-01 -5.60801625e-01 8.43470693e-01
-2.40051001e-01 3.85992169e-01 4.42164242e-01 1.27855167e-01
-8.23964715e-01 6.63438022e-01 1.16977406e+00 -5.32760978e-01
8.08634236e-03 -5.94360173e-01 -4.29403573e-01 7.61427820e-01
3.59166920e-01 -2.96101481e-01 -5.54676056e-01 -1.13857388e+00
-1.41779244e-01 -4.12056357e-01 2.79070809e-02 -8.01210225e-01
2.14505658e-01 -1.49544731e-01 1.70697600e-01 -9.62377787e-01
5.23109972e-01 -4.77707893e-01 -2.14289188e-01 2.04998031e-01
-2.50255346e-01 1.38312308e-02 7.20833898e-01 4.36705619e-01
-2.61090875e-01 1.85410708e-01 1.03574634e+00 -3.47805947e-01
-1.55288839e+00 4.38791096e-01 -1.31496266e-01 3.24512780e-01
8.68019283e-01 2.52238363e-01 -5.73542655e-01 -2.11621761e-01
-4.75557774e-01 6.53115451e-01 3.95054728e-01 1.24821019e+00
4.31780577e-01 -1.53898072e+00 -1.05331194e+00 -4.87257577e-02
8.94387066e-01 -5.35705090e-01 5.42296886e-01 8.39702368e-01
-1.29755929e-01 8.02818775e-01 -6.25528097e-01 -1.09554052e+00
-1.33302462e+00 6.23049796e-01 2.35859439e-01 -3.44689906e-01
-5.04581630e-01 1.14858878e+00 6.35757446e-01 -9.90079820e-01
1.05876178e-01 2.58903317e-02 -1.92776293e-01 3.94278944e-01
6.53500080e-01 -1.44006550e-01 -2.34349132e-01 -1.15476644e+00
-2.59252429e-01 9.19382036e-01 -5.12995660e-01 -3.91393393e-01
1.02805865e+00 -4.70152885e-01 1.27807343e-02 1.05439913e+00
1.32039249e+00 -4.64199066e-01 -1.12313557e+00 -7.39178717e-01
-2.34342843e-01 -7.88958147e-02 1.74929887e-01 -9.42255199e-01
-9.75398421e-01 1.30615461e+00 9.76497352e-01 -6.27637088e-01
9.51630831e-01 2.06562236e-01 9.38550055e-01 3.29519033e-01
7.63784766e-01 -1.26068425e+00 2.18777172e-02 8.85591805e-01
7.05878496e-01 -1.85853243e+00 -2.02949315e-01 4.42098796e-01
-1.16055334e+00 7.86455095e-01 8.60343754e-01 3.69870454e-01
3.85652691e-01 -2.92038801e-03 7.00389802e-01 -2.64278650e-02
-9.56849992e-01 -4.53883827e-01 6.28497839e-01 8.05480182e-01
5.89711547e-01 3.25589001e-01 3.06865154e-03 -5.05824387e-02
2.29921155e-02 -6.00887723e-02 1.13937169e-01 7.63461173e-01
-4.61822689e-01 -1.19737673e+00 -1.47605836e-01 -4.52196598e-02
-2.78124928e-01 -6.41856372e-01 -6.59855455e-02 8.95368576e-01
8.65165591e-02 6.47053838e-01 -7.52475485e-02 -3.86773311e-02
9.89552513e-02 3.90901953e-01 9.29979384e-01 -7.56512523e-01
-4.63265389e-01 3.52558494e-01 4.68972474e-02 -5.43267131e-01
-8.77098143e-01 -5.26860297e-01 -9.28246737e-01 1.25766188e-01
-4.87441197e-02 -3.96557778e-01 7.96411455e-01 1.18987322e+00
1.92416251e-01 5.02085090e-01 -1.12888791e-01 -8.37803721e-01
-3.18280995e-01 -1.14573669e+00 -2.51301855e-01 3.19753408e-01
2.11223647e-01 -6.37920201e-01 1.19606517e-01 2.95641482e-01] | [11.164384841918945, 1.5667983293533325] |
10ff6ee7-43bc-4e06-9059-0832ced71910 | accuracy-of-automatic-cross-corpus-emotion | null | null | https://aclanthology.org/L16-1634 | https://aclanthology.org/L16-1634.pdf | Accuracy of Automatic Cross-Corpus Emotion Labeling for Conversational Speech Corpus Commonization | There exists a major incompatibility in emotion labeling framework among emotional speech corpora, that is, category-based and dimension-based. Commonizing these requires inter-corpus emotion labeling according to both frameworks, but doing this by human annotators is too costly for most cases. This paper examines the possibility of automatic cross-corpus emotion labeling. In order to evaluate the effectiveness of the automatic labeling, a comprehensive emotion annotation for two conversational corpora, UUDB and OGVC, was performed. With a state-of-the-art machine learning technique, dimensional and categorical emotion estimation models were trained and tested against the two corpora. For the emotion dimension estimation, the automatic cross-corpus emotion labeling for the different corpus was effective for the dimensions of aroused-sleepy, dominant-submissive and interested-indifferent, showing only slight performance degradation against the result for the same corpus. On the other hand, the performance for the emotion category estimation was not sufficient. | ['Atsushi Nagaoka', 'Yoshiko Arimoto', 'Hiroki Mori'] | 2016-05-01 | accuracy-of-automatic-cross-corpus-emotion-1 | https://aclanthology.org/L16-1634 | https://aclanthology.org/L16-1634.pdf | lrec-2016-5 | ['cross-corpus'] | ['computer-vision'] | [-2.64297724e-01 3.83112907e-01 1.74888179e-01 -6.38273239e-01
-4.21551198e-01 -5.96592009e-01 6.11102104e-01 2.74935722e-01
-4.32178050e-01 6.94197774e-01 3.34283829e-01 -6.82015298e-03
-1.31823659e-01 -2.28522554e-01 2.60718703e-01 -6.27169549e-01
1.71383649e-01 5.62194288e-01 -2.01214254e-01 -3.30515504e-01
2.20927507e-01 3.81974667e-01 -1.90185606e+00 4.23561066e-01
6.59781337e-01 1.03646266e+00 -1.68995231e-01 5.90737522e-01
-7.71340251e-01 7.56999612e-01 -6.81643128e-01 -6.33152664e-01
-1.69884428e-01 -5.59940100e-01 -9.53729808e-01 2.46143177e-01
-1.56652287e-01 2.46302858e-01 3.57491374e-01 9.61201489e-01
6.06679380e-01 2.85702068e-02 8.15480411e-01 -1.39202249e+00
-1.52879179e-01 5.61283469e-01 -2.85879046e-01 -1.62118778e-01
4.83816534e-01 -4.50249314e-01 8.23747694e-01 -3.51023823e-01
7.94014335e-01 1.16609013e+00 3.83163989e-01 6.05325222e-01
-7.94625640e-01 -5.62226176e-01 4.50912490e-02 8.73310193e-02
-1.09491861e+00 -4.00426477e-01 1.13658297e+00 -6.62572682e-01
1.05761671e+00 3.80942762e-01 7.58044660e-01 1.15031826e+00
-1.97594687e-01 3.42549056e-01 1.49292910e+00 -6.46091342e-01
2.73050010e-01 9.19711947e-01 4.16016102e-01 1.88448399e-01
-2.80202717e-01 -2.39843234e-01 -2.55144835e-01 -2.38938645e-01
-8.67055580e-02 -6.78392470e-01 -2.01491609e-01 -2.59025216e-01
-7.10773408e-01 8.68576050e-01 -4.88247603e-01 1.10555220e+00
-3.93584549e-01 -5.93338966e-01 1.17524350e+00 5.26777208e-01
8.69806826e-01 4.87874240e-01 -6.66805804e-01 -7.64179707e-01
-8.21144700e-01 -7.67063126e-02 1.19293022e+00 4.72194701e-01
5.59358597e-01 2.50720531e-02 1.01543292e-01 1.12628841e+00
2.88616121e-01 -2.68376665e-03 7.68449783e-01 -6.62863076e-01
1.33835092e-01 7.12860644e-01 2.24594469e-03 -1.19824028e+00
-6.86478972e-01 -2.42096618e-01 -5.44290960e-01 4.82034162e-02
3.05140674e-01 -4.39923018e-01 -4.64496940e-01 1.79721105e+00
5.18626451e-01 -3.47884119e-01 6.96891129e-01 8.21401715e-01
9.44782615e-01 6.25875831e-01 4.31145459e-01 -7.89830267e-01
1.44585919e+00 -8.02477777e-01 -1.28312767e+00 -3.01443525e-02
9.47788477e-01 -9.27761734e-01 1.19625270e+00 4.65577662e-01
-8.15719783e-01 -3.87953758e-01 -1.04998744e+00 2.70895660e-01
-7.12763011e-01 2.24140584e-01 6.11595809e-01 1.01714730e+00
-7.77537048e-01 3.29231173e-02 -3.19812626e-01 -6.27267778e-01
-3.30416560e-01 4.64009978e-02 -7.03700304e-01 4.92163241e-01
-1.30109763e+00 1.06803000e+00 5.83324015e-01 -1.68634191e-01
-2.05951750e-01 -1.59149185e-01 -7.43876755e-01 2.21332312e-02
-7.70481229e-02 1.91935763e-01 9.84030485e-01 -1.69605434e+00
-1.68801081e+00 1.23768377e+00 2.30986595e-01 -2.36494675e-01
4.52635556e-01 7.93668553e-02 -1.01120007e+00 1.46783009e-01
1.35718100e-02 4.77998942e-01 4.71844256e-01 -1.37958825e+00
-5.07297993e-01 -4.11017954e-01 -2.76201099e-01 3.10298204e-01
-4.27655816e-01 3.49323124e-01 -2.07946882e-01 -2.95801759e-01
1.11621678e-01 -8.57021093e-01 3.40863466e-01 -6.90781474e-01
-1.47721887e-01 -4.78734642e-01 9.97250617e-01 -7.68882692e-01
1.41637373e+00 -2.40885377e+00 -1.12223066e-02 2.30114296e-01
-1.73152983e-01 6.65933713e-02 2.42521763e-01 4.45394039e-01
-3.84377271e-01 1.44026712e-01 3.14059369e-02 -2.25244179e-01
2.29178444e-01 4.23608392e-01 1.08464755e-01 4.34295148e-01
-2.22217977e-01 1.84485335e-02 -6.51993811e-01 -8.90681982e-01
1.44315585e-01 3.83825362e-01 -3.60765964e-01 4.55864817e-01
-4.37681973e-02 3.66936713e-01 -2.88906932e-01 4.43270355e-01
4.83657509e-01 4.89987493e-01 6.76911950e-01 -4.45035964e-01
-2.82579929e-01 3.53562902e-03 -1.00004339e+00 1.42812157e+00
-6.64195061e-01 6.10305369e-01 2.54874080e-01 -1.16550279e+00
1.29398692e+00 7.91754961e-01 6.49412870e-01 -6.47251308e-01
5.71506381e-01 3.73550773e-01 3.59814733e-01 -1.01048470e+00
5.97032666e-01 -4.91645068e-01 -5.30331016e-01 3.75899166e-01
2.80100256e-01 -2.53908783e-01 2.89669305e-01 -8.45015049e-02
3.80320072e-01 8.66013244e-02 3.72925431e-01 -3.03059697e-01
8.29256058e-01 -1.04532681e-01 4.36097831e-01 -9.71418247e-02
-5.30234635e-01 1.07780680e-01 8.66850793e-01 -2.71125942e-01
-7.65508115e-01 -2.95153975e-01 -1.58646762e-01 1.15317547e+00
-2.03706399e-01 -2.10737243e-01 -1.04330802e+00 -6.74073517e-01
-6.02534771e-01 9.88682389e-01 -5.35310924e-01 8.05458426e-02
6.63335472e-02 -6.38207376e-01 6.49193764e-01 -7.81390145e-02
4.53996539e-01 -1.15396607e+00 -7.53598392e-01 5.80568798e-02
-3.72809529e-01 -1.21158850e+00 1.63848758e-01 4.33505535e-01
-2.59312034e-01 -7.60331869e-01 -2.36041784e-01 -6.87034428e-01
1.25949726e-01 -4.20297444e-01 1.17344999e+00 -1.29787028e-01
1.33323362e-02 3.95832330e-01 -8.17814767e-01 -5.02173126e-01
-7.13212550e-01 -3.45153920e-02 -7.87556767e-02 1.35103419e-01
5.34308970e-01 -5.94058871e-01 -8.43951479e-02 1.81711435e-01
-7.41455853e-01 -2.29065731e-01 2.06005737e-01 4.45454925e-01
1.55981511e-01 3.65732424e-02 7.33627379e-01 -8.53748441e-01
1.03914118e+00 -5.32414079e-01 -6.31834641e-02 1.97095200e-01
-5.29208362e-01 -1.26773670e-01 3.16620469e-01 -4.65867937e-01
-1.38774002e+00 -3.00638899e-02 -4.52984452e-01 -7.32379630e-02
-5.73157609e-01 4.95542318e-01 -3.25808376e-01 1.77446216e-01
4.77116823e-01 -1.97801009e-01 -3.66190188e-02 -2.01134279e-01
2.23292887e-01 1.24896181e+00 3.33433896e-01 -5.12941301e-01
-9.86200348e-02 5.99262342e-02 -4.80450839e-01 -9.50011909e-01
-5.46722949e-01 -5.16155779e-01 -4.30291623e-01 -5.64119637e-01
1.30580866e+00 -6.69872880e-01 -5.44484198e-01 3.29498410e-01
-1.19705439e+00 -1.00694589e-01 -2.26786733e-02 6.90146267e-01
-5.08019030e-01 3.99488747e-01 -3.90696526e-01 -1.21072698e+00
-4.32012081e-01 -1.06666028e+00 7.27033317e-01 5.45082986e-02
-6.72276199e-01 -1.02279699e+00 2.58943558e-01 4.53811795e-01
1.56162247e-01 3.88717920e-01 9.90349531e-01 -1.18502736e+00
8.98486018e-01 -3.55450362e-01 6.41592517e-02 5.50651729e-01
5.72856553e-02 2.93490678e-01 -1.14160395e+00 1.97364748e-01
3.12087506e-01 -6.99231565e-01 4.32214402e-02 -9.63610560e-02
6.71732426e-01 -2.95127816e-02 3.04936455e-03 3.29847820e-03
1.24333656e+00 6.89101636e-01 6.36831999e-01 1.85574099e-01
2.49796242e-01 1.22117150e+00 9.15449142e-01 5.35705984e-01
2.27220356e-01 6.46307588e-01 1.36315137e-01 -5.82400411e-02
3.41796398e-01 2.86056250e-01 1.63856119e-01 1.06855643e+00
1.77993223e-01 -4.14397210e-01 -8.34931195e-01 4.80318874e-01
-1.58371019e+00 -9.47736859e-01 -1.42237008e-01 1.71198297e+00
6.03949845e-01 3.92805450e-02 2.13112816e-01 4.56056714e-01
7.17678905e-01 1.30034745e-01 3.53931971e-02 -1.42336321e+00
-7.62894824e-02 -1.54667854e-01 -1.88356489e-01 4.64486659e-01
-9.26275969e-01 7.65699804e-01 6.03794241e+00 7.28220165e-01
-1.20476127e+00 3.69962513e-01 8.16617072e-01 -5.98653872e-03
-1.65527225e-01 -1.89056411e-01 -2.43527576e-01 4.99341279e-01
1.15609062e+00 1.61949486e-01 1.18224435e-01 1.03650606e+00
2.11825281e-01 -3.02219987e-01 -6.36017859e-01 1.15369284e+00
3.45167965e-01 -5.97598255e-01 -5.42959154e-01 -1.10138156e-01
4.01780516e-01 -3.22125018e-01 -5.60351312e-01 6.59762025e-01
-2.31382787e-01 -5.52551091e-01 6.03748083e-01 3.08617592e-01
5.78237355e-01 -9.08644617e-01 1.13759124e+00 1.84342578e-01
-7.74454832e-01 2.14769199e-01 -1.21464923e-01 -1.29459992e-01
2.11148039e-01 5.14110625e-01 -3.31164002e-01 5.05044222e-01
6.42717659e-01 3.33266929e-02 -1.44239292e-01 1.48536578e-01
9.26895142e-02 4.58194822e-01 -1.26994044e-01 -2.26947963e-01
3.10520649e-01 -5.14041126e-01 3.69785011e-01 1.46998560e+00
2.32008949e-01 2.04440095e-02 1.36273682e-01 2.75189251e-01
3.49084198e-01 7.51415551e-01 -4.44155574e-01 -3.12187552e-01
3.39646846e-01 1.72104573e+00 -9.03541744e-01 -4.81910288e-01
-3.36217314e-01 1.04195333e+00 3.35239500e-01 -2.39733569e-02
-9.46451187e-01 -3.51599902e-01 4.06855464e-01 -3.25732321e-01
-1.07826740e-01 2.52624393e-01 -3.40199023e-02 -8.12666059e-01
-2.04084039e-01 -9.44549978e-01 3.81023765e-01 -7.63951242e-01
-1.19208431e+00 1.13755083e+00 1.40606850e-01 -9.65572715e-01
-5.50642669e-01 -5.57235062e-01 -4.36416745e-01 6.91399157e-01
-7.95863211e-01 -1.04087591e+00 -3.16773325e-01 3.87645066e-01
3.54992390e-01 -1.87099710e-01 1.12655151e+00 5.51420569e-01
-5.40333927e-01 4.58254606e-01 -3.00576389e-01 -4.33942452e-02
7.75872469e-01 -1.11736774e+00 -6.45725846e-01 3.65189940e-01
-2.03338429e-01 -3.87506722e-03 1.07000244e+00 -4.51754689e-01
-7.86057591e-01 -3.87445390e-01 1.13962889e+00 1.42373994e-01
6.57761633e-01 -2.92592585e-01 -8.66122901e-01 1.80861920e-01
5.83315909e-01 -4.37407047e-01 1.14459765e+00 4.03303891e-01
-3.00133854e-01 1.65187776e-01 -1.44380319e+00 3.99976134e-01
4.86905843e-01 -5.22856355e-01 -6.89715862e-01 2.76814073e-01
4.92744029e-01 -1.35519251e-01 -1.17247844e+00 1.33324191e-01
6.79158509e-01 -1.46711302e+00 3.91231865e-01 -4.31111455e-01
2.16563538e-01 -7.48194754e-02 -4.10763353e-01 -1.28937161e+00
9.60987359e-02 -2.02326536e-01 5.46221435e-01 1.85109890e+00
5.52335858e-01 -6.36964738e-01 4.57405508e-01 7.16058195e-01
-2.46934429e-01 -6.02322340e-01 -8.43909800e-01 -4.99123275e-01
1.71621516e-02 -6.69912875e-01 4.78519022e-01 1.26703525e+00
7.35745788e-01 7.23742843e-01 -3.14476013e-01 -2.37202957e-01
-1.98073924e-01 -7.45086595e-02 6.32887721e-01 -1.28336859e+00
-1.94274634e-01 -4.79042411e-01 -5.07675827e-01 1.13564283e-01
6.79029047e-01 -5.72580814e-01 -2.85982471e-02 -1.35240889e+00
-1.40181497e-01 -2.20050946e-01 5.18107191e-02 3.04126263e-01
2.39729896e-01 1.08838916e-01 1.56312972e-01 -5.46214990e-02
-3.43352497e-01 3.67265195e-01 7.22381771e-01 2.16949850e-01
-2.86712915e-01 -3.26281101e-01 -5.44865906e-01 8.40862632e-01
1.10008311e+00 -3.67331475e-01 -5.82573950e-01 2.15412512e-01
2.46727377e-01 2.13261232e-01 -2.35819250e-01 -1.04251146e+00
-3.07280511e-01 9.81672630e-02 -7.82384351e-02 -4.38904792e-01
5.18724442e-01 -1.06245267e+00 3.70430589e-01 1.89673081e-01
-3.42279762e-01 1.02815054e-01 2.50644088e-01 1.94996610e-01
-5.05334735e-01 -5.05187571e-01 8.27427268e-01 -7.16308355e-02
-6.44145250e-01 -4.24446762e-01 -6.08464062e-01 1.21001050e-01
1.18084872e+00 -3.21980357e-01 -9.79719609e-02 -5.27823389e-01
-8.94000113e-01 -2.59825081e-01 2.69699872e-01 2.64172167e-01
-1.86948366e-02 -1.04046071e+00 -4.09581065e-01 -1.59754127e-01
2.79157579e-01 -6.97620153e-01 5.02338171e-01 9.50224638e-01
-2.83855349e-01 1.43513799e-01 -4.79797214e-01 -2.26088151e-01
-1.57311869e+00 7.05776572e-01 4.17972654e-01 -4.18348700e-01
1.71912715e-01 3.98216009e-01 -1.85751185e-01 -5.86378932e-01
1.66717112e-01 -8.06870162e-02 -6.98436499e-01 7.10034430e-01
1.39969647e-01 4.11774576e-01 1.78718299e-01 -1.13059533e+00
-3.25342923e-01 4.16829348e-01 2.76654094e-01 -5.89697003e-01
9.00323153e-01 -3.23804408e-01 -3.65587562e-01 7.69541919e-01
1.40178406e+00 1.64266005e-01 -5.13916254e-01 4.78998482e-01
1.60187244e-01 -6.06671050e-02 1.65764958e-01 -9.89635944e-01
-7.18474269e-01 7.63092220e-01 8.24597776e-01 7.65407622e-01
1.21596682e+00 -1.81631058e-01 2.24114597e-01 1.71006814e-01
1.58711653e-02 -1.66578805e+00 -2.00858131e-01 4.61717576e-01
8.51534963e-01 -1.14526272e+00 -3.04507434e-01 -5.32778203e-01
-1.29815364e+00 1.15126431e+00 6.81272089e-01 3.39127690e-01
7.29565740e-01 3.19063693e-01 5.24830461e-01 -4.20303613e-01
-6.30516768e-01 -2.11905226e-01 2.29112715e-01 5.60098171e-01
1.06271803e+00 3.87711793e-01 -9.43016946e-01 6.95629895e-01
-2.66405821e-01 -2.65267998e-01 3.17111552e-01 6.64858937e-01
-3.09562683e-01 -1.12791336e+00 -3.88089567e-01 2.50139296e-01
-7.38148212e-01 1.41842350e-01 -7.33283401e-01 1.09626484e+00
3.48559171e-01 1.34274435e+00 3.14464062e-01 -5.00111461e-01
2.22044766e-01 7.04066515e-01 1.07239634e-01 -2.70696104e-01
-9.04284418e-01 2.86534429e-01 8.65576982e-01 -1.62646368e-01
-9.96384382e-01 -5.82431257e-01 -1.27179325e+00 -1.01483678e-02
-3.56352508e-01 7.26140261e-01 1.09567153e+00 1.05587995e+00
2.18179569e-01 3.40652078e-01 5.46666741e-01 -7.34508634e-01
-6.03842698e-02 -1.38913071e+00 -7.20502198e-01 6.76117480e-01
-3.56922567e-01 -7.87025988e-01 -6.80638373e-01 -1.21170960e-01] | [13.005096435546875, 6.164022922515869] |
7097374d-824a-4726-8e16-782149839824 | transformers-meet-stochastic-block-models | 2210.15541 | null | https://arxiv.org/abs/2210.15541v1 | https://arxiv.org/pdf/2210.15541v1.pdf | Transformers meet Stochastic Block Models: Attention with Data-Adaptive Sparsity and Cost | To overcome the quadratic cost of self-attention, recent works have proposed various sparse attention modules, most of which fall under one of two groups: 1) sparse attention under a hand-crafted patterns and 2) full attention followed by a sparse variant of softmax such as $\alpha$-entmax. Unfortunately, the first group lacks adaptability to data while the second still requires quadratic cost in training. In this work, we propose SBM-Transformer, a model that resolves both problems by endowing each attention head with a mixed-membership Stochastic Block Model (SBM). Then, each attention head data-adaptively samples a bipartite graph, the adjacency of which is used as an attention mask for each input. During backpropagation, a straight-through estimator is used to flow gradients beyond the discrete sampling step and adjust the probabilities of sampled edges based on the predictive loss. The forward and backward cost are thus linear to the number of edges, which each attention head can also choose flexibly based on the input. By assessing the distribution of graphs, we theoretically show that SBM-Transformer is a universal approximator for arbitrary sequence-to-sequence functions in expectation. Empirical evaluations under the LRA and GLUE benchmarks demonstrate that our model outperforms previous efficient variants as well as the original Transformer with full attention. Our implementation can be found in https://github.com/sc782/SBM-Transformer . | ['Seunghoon Hong', 'Honglak Lee', 'Moontae Lee', 'Jinwoo Kim', 'Seonwoo Min', 'Sungjun Cho'] | 2022-10-27 | null | null | null | null | ['stochastic-block-model'] | ['graphs'] | [ 2.21937656e-01 2.36272112e-01 -2.37656862e-01 -3.58631968e-01
-7.03718603e-01 -2.56291628e-01 2.45134622e-01 -1.07226864e-01
-4.32014525e-01 5.87366581e-01 -1.89087056e-02 -3.67747962e-01
-3.70909576e-03 -9.00974214e-01 -1.13523996e+00 -7.10303128e-01
-8.76844674e-02 6.43627763e-01 1.76120862e-01 -1.65348321e-01
1.24968074e-01 1.65617257e-01 -1.26449764e+00 1.41017213e-01
1.02919590e+00 1.08032537e+00 6.09674096e-01 6.73921704e-01
-2.79174238e-01 7.48281181e-01 -3.28200132e-01 -5.58738172e-01
2.36236513e-01 -5.72898567e-01 -7.99026906e-01 -1.44836336e-01
1.78224504e-01 -7.11101592e-02 -6.22152150e-01 1.29487956e+00
4.73287821e-01 3.36632222e-01 4.62455392e-01 -1.09852040e+00
-9.53354120e-01 1.14098620e+00 -7.53497839e-01 4.51173455e-01
1.22072510e-01 3.34331065e-01 1.38495791e+00 -9.37515020e-01
1.95752412e-01 1.31740713e+00 6.24408066e-01 4.33737367e-01
-1.32263565e+00 -6.86736345e-01 5.68491399e-01 4.46674019e-01
-1.41126525e+00 -2.27492765e-01 7.61504710e-01 -2.17140481e-01
1.04076874e+00 2.36740246e-01 5.56782544e-01 9.10751581e-01
-5.41435145e-02 9.75630879e-01 6.89356327e-01 -3.96177649e-01
1.85941249e-01 1.86741091e-02 3.06626976e-01 9.01995301e-01
3.83410044e-02 -1.54103041e-01 -3.77647042e-01 -9.69311595e-03
6.70080125e-01 5.76324463e-02 -4.58426863e-01 -1.30955130e-02
-8.14893544e-01 8.54901016e-01 8.64090145e-01 1.79281175e-01
-4.69245493e-01 3.43461722e-01 1.90358102e-01 2.19975665e-01
4.00273830e-01 2.41090536e-01 -4.59047437e-01 1.13170452e-01
-9.29226577e-01 -7.73304254e-02 6.32822335e-01 1.19245148e+00
1.02313375e+00 1.12601124e-01 -4.81290936e-01 9.10303593e-01
3.83927077e-01 2.90219605e-01 5.95092118e-01 -6.76999986e-01
6.09191656e-01 5.48985362e-01 -1.98807910e-01 -9.24332738e-01
-3.04682553e-01 -7.15572834e-01 -1.17912316e+00 -1.03286847e-01
3.26336235e-01 -2.46670142e-01 -1.03488529e+00 2.01164103e+00
1.22745432e-01 3.04014534e-01 -4.84747529e-01 9.27581549e-01
6.62329495e-01 7.63914108e-01 -3.28040607e-02 -2.59029232e-02
1.23297310e+00 -1.28765476e+00 -4.76749003e-01 -4.93362576e-01
2.38778472e-01 -3.63649398e-01 1.18973553e+00 1.75866440e-01
-1.40582168e+00 -5.01922905e-01 -9.26563203e-01 -1.24376267e-01
-1.75909519e-01 8.65782611e-03 5.41343510e-01 4.10934508e-01
-1.24787188e+00 6.63614273e-01 -6.85334742e-01 5.20336144e-02
6.52820885e-01 5.29467046e-01 7.49172270e-03 -1.06080145e-01
-1.37710142e+00 6.67100430e-01 3.01433712e-01 2.89026320e-01
-9.19150889e-01 -6.03501439e-01 -9.36785758e-01 6.70640409e-01
3.88188094e-01 -8.23686600e-01 1.02450788e+00 -1.22214150e+00
-1.55285990e+00 6.04106843e-01 -3.74699593e-01 -7.10778117e-01
3.84105414e-01 -8.12620521e-02 1.30466357e-01 -8.14291835e-02
1.47828877e-01 6.36827469e-01 8.96796644e-01 -1.02563560e+00
-5.24408638e-01 -2.01617450e-01 2.48836100e-01 2.34290808e-01
-3.55512112e-01 -8.03814083e-02 -8.29531431e-01 -7.20631957e-01
-1.73519924e-01 -8.65999043e-01 -3.80344838e-01 -2.99014151e-01
-5.85260868e-01 -2.29143664e-01 4.14722353e-01 -5.72575688e-01
1.52725005e+00 -2.10469151e+00 4.25929487e-01 2.60933608e-01
1.40449762e-01 1.97115496e-01 -2.73975700e-01 3.10113788e-01
-1.00904107e-01 2.30454221e-01 -5.80011368e-01 -5.53553700e-01
8.37125480e-02 1.24878794e-01 -9.88226980e-02 3.65671009e-01
2.03060970e-01 1.02327204e+00 -9.72650170e-01 -2.93459117e-01
-6.68981597e-02 4.65746433e-01 -7.88248241e-01 2.61685789e-01
-2.69176155e-01 7.05047846e-02 -2.08644167e-01 4.11406785e-01
6.18635178e-01 -8.06372106e-01 2.28534013e-01 4.67182919e-02
2.22493604e-01 3.58591408e-01 -1.12507737e+00 1.51205778e+00
-5.77908993e-01 4.88841325e-01 2.21034482e-01 -1.09863949e+00
7.12241292e-01 6.04257137e-02 3.94479819e-02 -4.16635484e-01
2.92221665e-01 4.90104370e-02 7.52630383e-02 -2.11411357e-01
3.43875200e-01 8.14905316e-02 1.26024991e-01 3.05243075e-01
3.25715929e-01 4.64482874e-01 3.10487717e-01 3.61000687e-01
1.10017693e+00 -4.90233041e-02 2.02263117e-01 -2.64620215e-01
5.49394608e-01 -6.33689344e-01 6.04951203e-01 9.80433345e-01
-4.53472584e-02 6.77895248e-01 7.53541112e-01 -2.25558862e-01
-7.58213103e-01 -8.04526508e-01 2.40931153e-01 1.39414120e+00
1.27946153e-01 -2.66930699e-01 -8.87160599e-01 -8.28785121e-01
4.11608582e-03 7.33429193e-01 -8.22084308e-01 -2.25306779e-01
-4.91610497e-01 -7.45877743e-01 2.00384438e-01 6.53859794e-01
2.57347435e-01 -1.36719251e+00 -2.33512461e-01 2.44006932e-01
-1.20721310e-01 -7.64466703e-01 -1.13369715e+00 5.03874362e-01
-5.97948372e-01 -8.62356544e-01 -7.60793805e-01 -8.88655365e-01
8.84931684e-01 1.35732174e-01 1.13417327e+00 2.75946826e-01
-7.56048709e-02 4.73941565e-02 -2.56343067e-01 -1.21258870e-01
-1.12318538e-01 5.30771911e-01 -3.61348450e-01 3.99227470e-01
1.52810782e-01 -8.13481808e-01 -7.01224208e-01 5.49540967e-02
-7.21221626e-01 5.43551035e-02 7.66159594e-01 1.14273632e+00
6.59802794e-01 -2.14524254e-01 5.78744829e-01 -1.01745105e+00
5.18155992e-01 -8.82396519e-01 -6.35897398e-01 1.50097966e-01
-4.76906151e-01 2.57714808e-01 7.88807094e-01 -4.30745482e-01
-7.48108983e-01 3.77141833e-02 -3.46595615e-01 -6.36297226e-01
2.10950837e-01 6.68344796e-01 -3.35353792e-01 7.82235265e-02
2.34234929e-01 4.45086449e-01 -4.45032641e-02 -3.70374918e-01
3.81555259e-01 4.12948281e-01 3.46102536e-01 -4.85287905e-01
5.57278395e-01 9.92238969e-02 -2.25907579e-01 -7.06265807e-01
-9.34145391e-01 -2.48840019e-01 -2.04175472e-01 -8.57714638e-02
6.38044059e-01 -6.57577455e-01 -8.19872916e-01 5.25213480e-01
-1.02129960e+00 -7.38523006e-01 -3.32665861e-01 2.63247192e-01
-4.37237829e-01 3.62250566e-01 -8.15970123e-01 -8.58835936e-01
-5.46882868e-01 -1.28484392e+00 7.88176179e-01 2.05968559e-01
8.09490588e-03 -9.81512189e-01 -4.03468937e-01 8.17091018e-02
5.14421701e-01 -1.82478085e-01 8.99486482e-01 -5.70169270e-01
-7.08078265e-01 -1.45264089e-01 -2.39431351e-01 4.32964891e-01
-8.04200470e-02 -1.15365274e-01 -7.25813568e-01 -4.21301514e-01
-1.01248570e-01 -3.04675043e-01 1.19224334e+00 7.03919828e-01
1.49518394e+00 -6.16674840e-01 -2.70106822e-01 9.71831441e-01
1.40281630e+00 1.82168779e-03 5.95997870e-01 1.18680090e-01
7.73783207e-01 1.21105246e-01 8.42168257e-02 3.34322691e-01
5.03106713e-01 3.86808604e-01 6.86752081e-01 -3.47365774e-02
-6.12808764e-02 -4.89658087e-01 3.51582617e-01 9.69674289e-01
5.70263341e-02 -5.29444218e-01 -6.96397841e-01 5.59691608e-01
-1.78352165e+00 -1.04893112e+00 1.54458761e-01 2.09302020e+00
9.34500515e-01 3.86357814e-01 1.03389636e-01 5.73990196e-02
9.59893107e-01 3.02454203e-01 -6.56254411e-01 -2.58943945e-01
-3.58958989e-02 3.23015511e-01 6.24427915e-01 7.97248483e-01
-9.46595311e-01 8.32876444e-01 5.65417242e+00 1.04828024e+00
-1.01487899e+00 5.27787544e-02 8.14796329e-01 -1.19076714e-01
-5.42685211e-01 -2.57810764e-02 -8.82472217e-01 8.95775139e-01
8.70925963e-01 -1.65003940e-01 8.17156792e-01 7.98778534e-01
-1.37414455e-01 2.12688848e-01 -8.85181189e-01 6.06354773e-01
-3.43733281e-02 -1.27716339e+00 5.75821623e-02 -1.73400566e-01
6.44407749e-01 2.97457814e-01 3.92558984e-02 5.12794793e-01
5.22937417e-01 -9.64573503e-01 6.84768796e-01 5.07856369e-01
7.38883853e-01 -7.73336709e-01 5.81680834e-01 4.72872913e-01
-1.25784254e+00 -2.72845805e-01 -3.29088688e-01 -6.37495518e-02
1.53215349e-01 6.14040554e-01 -6.03567064e-01 4.40515548e-01
7.16417015e-01 6.67868435e-01 -2.73447543e-01 9.83125687e-01
-4.28067267e-01 8.71890008e-01 -3.68514806e-01 -2.03253984e-01
3.11895281e-01 -2.95671016e-01 6.64916933e-01 1.41343987e+00
2.82851368e-01 -6.43115044e-02 1.77988946e-01 9.57627177e-01
-4.46298361e-01 -1.71433482e-02 -1.65555149e-01 1.51966795e-01
6.43192112e-01 1.17094266e+00 -6.31634831e-01 -3.65555972e-01
-4.62856501e-01 1.00348067e+00 8.66309524e-01 5.61554015e-01
-9.55629647e-01 -5.67873299e-01 4.37103540e-01 1.24050744e-01
8.49600792e-01 6.72762245e-02 -2.81108230e-01 -1.07966137e+00
-3.65747623e-02 -8.73370528e-01 6.05214477e-01 -6.57460153e-01
-1.44497263e+00 8.53425205e-01 -3.29540223e-01 -7.08100021e-01
6.03065044e-02 -2.75790781e-01 -6.36804700e-01 1.15329301e+00
-1.45881200e+00 -9.63741899e-01 -2.86958516e-01 4.45758462e-01
6.25147641e-01 -1.59979425e-02 5.25302827e-01 4.57083821e-01
-9.06716764e-01 1.06020582e+00 -4.06254269e-02 1.52060449e-01
3.26815784e-01 -1.42621827e+00 6.44310355e-01 8.34040284e-01
1.11232363e-01 6.02200925e-01 5.02445936e-01 -3.01732212e-01
-1.21393120e+00 -1.09127724e+00 7.61235476e-01 -5.60792573e-02
7.13497758e-01 -4.49602008e-01 -1.20075381e+00 9.61563110e-01
3.71424288e-01 1.46649316e-01 4.02106553e-01 1.31570920e-01
-2.70733565e-01 -2.84423202e-01 -8.09602439e-01 4.60317582e-01
1.06852853e+00 -2.52409607e-01 -3.48411798e-01 1.81857511e-01
8.70593369e-01 -5.19916773e-01 -6.16475642e-01 1.69361293e-01
2.59876102e-01 -9.05157447e-01 7.34766185e-01 -5.71127594e-01
5.06839156e-01 -1.58370435e-01 4.25838586e-03 -1.40418291e+00
-7.79304266e-01 -8.46611023e-01 -4.16801959e-01 1.16310346e+00
6.01229429e-01 -8.77058029e-01 7.22459793e-01 2.53420472e-01
-3.92241657e-01 -1.21157563e+00 -7.83129871e-01 -6.65956378e-01
-3.67913134e-02 -2.39954859e-01 6.81826651e-01 7.63369977e-01
-1.07794203e-01 5.32258153e-01 -3.46032560e-01 2.20250636e-01
5.69621384e-01 5.73394783e-02 4.63136584e-01 -7.96764970e-01
-7.51919866e-01 -8.48965585e-01 -1.47460029e-01 -1.47786033e+00
2.14373633e-01 -1.23758042e+00 1.94805130e-01 -1.46921372e+00
3.78037065e-01 -4.18783545e-01 -5.32636464e-01 5.91700077e-01
-7.26289392e-01 -1.02893479e-01 1.03172995e-01 2.65005492e-02
-6.04346156e-01 6.98802173e-01 1.10048890e+00 -1.92802534e-01
-1.21498488e-01 1.60161763e-01 -9.62899625e-01 5.62577426e-01
8.18805337e-01 -5.12665570e-01 -3.63607943e-01 -5.15879810e-01
2.57657886e-01 -9.84204933e-04 1.89902052e-01 -8.02282035e-01
3.71557415e-01 -2.19948161e-02 9.06282663e-02 -3.85094047e-01
5.71938679e-02 -5.52662075e-01 -4.28892076e-02 3.93891841e-01
-5.42580605e-01 8.87226034e-03 1.05110362e-01 6.86635017e-01
-1.75969198e-01 -3.44921261e-01 8.78128290e-01 -1.47291914e-01
-4.04915333e-01 7.37203360e-01 -1.17139161e-01 4.38083559e-01
7.10994244e-01 1.50635779e-01 -9.94378999e-02 -4.90429282e-01
-6.99665129e-01 5.51621139e-01 1.78990513e-01 1.55993059e-01
4.52375740e-01 -1.29717314e+00 -7.48116255e-01 2.92734265e-01
-3.08124930e-01 2.25848705e-01 5.19055724e-01 8.99131238e-01
-1.83097079e-01 1.49748027e-01 8.02475885e-02 -4.28214192e-01
-8.40280294e-01 8.66725683e-01 3.91621560e-01 -4.89599556e-01
-4.82015103e-01 1.27702796e+00 3.37982208e-01 -1.53180391e-01
5.03316283e-01 -1.29844829e-01 -2.84316372e-02 -1.05081774e-01
4.36867297e-01 2.83973604e-01 -1.36156470e-01 -3.62491637e-01
-3.32225055e-01 2.31538594e-01 -5.69130965e-02 2.29327366e-01
1.28807127e+00 -1.22615434e-01 -1.72302455e-01 3.50111395e-01
1.23460066e+00 -1.88369200e-01 -1.49290144e+00 -3.33307654e-01
-2.69882053e-01 -2.29783207e-01 1.14586169e-03 -6.20849848e-01
-1.45484984e+00 9.03664589e-01 6.36808649e-02 5.50967515e-01
1.28667498e+00 -5.48258871e-02 7.93831110e-01 6.81309104e-02
1.05624393e-01 -7.13636160e-01 -1.27950892e-01 7.01364398e-01
9.46959794e-01 -1.00971484e+00 -3.52858633e-01 -2.15442210e-01
-6.00528538e-01 7.69165337e-01 7.08016634e-01 -2.33085811e-01
6.96160495e-01 3.84917945e-01 -3.55820447e-01 -1.89584941e-02
-9.57110345e-01 -2.26192743e-01 2.38823354e-01 3.50199968e-01
4.55079168e-01 2.52928082e-02 -2.26227432e-01 8.31125557e-01
4.92422888e-03 -7.72127435e-02 1.72590390e-01 4.58193213e-01
-4.43766028e-01 -8.60107183e-01 -5.23137525e-02 6.66565895e-01
-4.59388316e-01 -4.56310958e-01 -2.94192266e-02 5.58667421e-01
-5.27460277e-02 5.42847037e-01 1.76999107e-01 -3.28458458e-01
3.23714286e-01 -1.32698752e-02 3.66040170e-01 -5.71909010e-01
-6.82336569e-01 -1.82237387e-01 -1.55373141e-01 -3.99856478e-01
5.82601987e-02 -5.49928963e-01 -1.10042834e+00 -2.57308394e-01
-3.99701625e-01 1.46902099e-01 1.78470328e-01 7.12210536e-01
5.21291733e-01 8.17945480e-01 6.46969914e-01 -9.79827225e-01
-8.52429509e-01 -1.13885021e+00 -4.06968623e-01 2.99318999e-01
4.16774929e-01 -4.33128685e-01 -5.57562292e-01 -1.34101182e-01] | [8.674904823303223, 4.669475078582764] |
2c37b16f-c332-48f3-9997-4545faf10d0d | efficient-approximate-quantum-state | 2009.07601 | null | https://arxiv.org/abs/2009.07601v3 | https://arxiv.org/pdf/2009.07601v3.pdf | Efficient Quantum State Sample Tomography with Basis-dependent Neural-networks | We use a meta-learning neural-network approach to analyse data from a measured quantum state. Once our neural network has been trained it can be used to efficiently sample measurements of the state in measurement bases not contained in the training data. These samples can be used calculate expectation values and other useful quantities. We refer to this process as "state sample tomography". We encode the state's measurement outcome distributions using an efficiently parameterized generative neural network. This allows each stage in the tomography process to be performed efficiently even for large systems. Our scheme is demonstrated on recent IBM Quantum devices, producing a model for a 6-qubit state's measurement outcomes with a predictive accuracy (classical fidelity) > 95% for all test cases using only 100 random measurement settings as opposed to the 729 settings required for standard full tomography using local measurements. This reduction in the required number of measurements scales favourably, with training data in 200 measurement settings yielding a predictive accuracy > 92% for a 10 qubit state where 59,049 settings are typically required for full local measurement-based quantum state tomography. A reduction in number of measurements by a factor, in this case, of almost 600 could allow for estimations of expectation values and state fidelities in practicable times on current quantum devices. | ['Alistair W. R. Smith', 'M. S. Kim', 'Johnnie Gray'] | 2020-09-16 | null | null | null | null | ['quantum-state-tomography'] | ['medical'] | [ 5.63943624e-01 4.19279262e-02 2.32997134e-01 -5.68821311e-01
-1.30327511e+00 -5.10914028e-01 6.99627340e-01 -9.37550701e-03
-8.72354090e-01 1.22043860e+00 -2.44898811e-01 -8.33860397e-01
-1.08958788e-01 -1.13899148e+00 -6.46005094e-01 -1.17214906e+00
7.55558610e-02 8.19913566e-01 -2.36172393e-01 -6.66419715e-02
5.51897347e-01 5.85789025e-01 -1.28460348e+00 8.32757056e-02
4.33495015e-01 6.77882135e-01 -1.38096645e-01 1.11371648e+00
1.28783762e-01 6.69230521e-01 -9.56138551e-01 -1.25320137e-01
2.15248942e-01 -7.53543079e-01 -9.44087446e-01 -5.13434887e-01
7.53146932e-02 -5.52237868e-01 -6.56481743e-01 1.36751497e+00
7.49360979e-01 9.63417143e-02 6.50654256e-01 -6.92110002e-01
-2.86864311e-01 9.03617203e-01 2.31297806e-01 1.99558184e-01
3.27364326e-01 4.52226520e-01 9.72302496e-01 -2.61252403e-01
6.34603441e-01 7.85546660e-01 3.51128966e-01 4.67381805e-01
-1.79747653e+00 -6.77255929e-01 -1.25368571e+00 1.36141106e-01
-1.30763483e+00 -7.47531056e-01 3.84825915e-01 -2.47106068e-02
1.57404780e+00 1.74449340e-01 8.35586250e-01 9.92827833e-01
5.36456883e-01 -1.33924931e-01 1.62795293e+00 -7.69183874e-01
6.04228258e-01 1.93796277e-01 6.95865303e-02 4.46785003e-01
1.26309603e-01 6.29525006e-01 -4.49659862e-02 -2.51701981e-01
8.28397810e-01 -3.65429312e-01 8.97228792e-02 1.07780233e-01
-1.34059143e+00 8.91287148e-01 4.99782801e-01 4.20961112e-01
-4.03101593e-01 5.51122785e-01 5.65000415e-01 8.96929562e-01
1.52821802e-02 8.09563816e-01 -1.69519126e-01 -4.99898463e-01
-1.00939965e+00 6.17078133e-02 9.78748977e-01 7.64525533e-01
1.17778349e+00 2.21161589e-01 2.31516603e-02 2.95408934e-01
-2.15175971e-01 9.33138549e-01 2.52537459e-01 -1.12283790e+00
3.50555032e-01 1.02337390e-01 2.96396166e-01 -1.31910577e-01
-4.18952763e-01 -8.41094628e-02 -9.21437860e-01 1.68304220e-01
4.41662550e-01 -3.50960761e-01 -9.75661933e-01 1.45130563e+00
1.49230644e-01 -1.45560667e-01 2.30368197e-01 3.96477908e-01
5.14649749e-01 9.95745599e-01 -5.92893600e-01 -3.18775266e-01
9.67612803e-01 -2.61639059e-01 -4.38725263e-01 1.38654828e-01
1.04386747e+00 -6.72193229e-01 7.40160942e-01 4.38163608e-01
-1.11115813e+00 -6.24762774e-01 -1.28423703e+00 -1.10503398e-02
-3.32613528e-01 -2.28548110e-01 7.35442758e-01 1.14712584e+00
-1.33137548e+00 1.25813913e+00 -8.46948385e-01 -1.67163573e-02
5.28127700e-02 9.33888972e-01 -5.10298491e-01 -3.17473076e-02
-1.10734558e+00 1.05135000e+00 1.04611719e+00 1.58952415e-01
-9.19297159e-01 -1.32276580e-01 -6.66784942e-01 4.21360396e-02
-3.45210508e-02 -5.63836515e-01 1.21181238e+00 -1.44470513e-01
-1.93834305e+00 6.70920312e-01 -1.95546791e-01 -4.56635624e-01
-1.75430059e-01 5.14100969e-01 -6.52736902e-01 4.90129218e-02
-1.67238027e-01 2.47443169e-01 4.38296318e-01 -5.71399629e-01
-7.15820938e-02 -7.59973601e-02 1.61844611e-01 -3.42798263e-01
3.34351838e-01 -6.37555271e-02 6.14346899e-02 7.34345317e-01
4.73537803e-01 -1.29261672e+00 -2.42626384e-01 -9.16257977e-01
-5.09919286e-01 -2.34448221e-02 1.71884328e-01 -3.84842396e-01
8.34598839e-01 -1.72048128e+00 4.48923893e-02 6.45654500e-01
2.99722105e-02 1.89072058e-01 -1.80348352e-01 8.44989955e-01
-1.81366235e-01 -2.04906806e-01 -1.06572136e-01 -2.91054189e-01
2.12520406e-01 1.66830376e-01 -7.19545856e-02 5.76067567e-01
1.53372198e-01 1.18902636e+00 -8.66327405e-01 4.15576063e-02
4.62114692e-01 2.25922823e-01 -5.52574396e-01 2.36135423e-01
1.19909182e-01 6.84801280e-01 -1.57392293e-01 5.14599264e-01
5.37398160e-01 -2.38999516e-01 3.27906698e-01 -6.27091974e-02
-2.20858585e-02 9.57000375e-01 -1.08368480e+00 1.63258398e+00
-5.38078129e-01 7.14743435e-01 -1.70318782e-01 -1.09695876e+00
1.08600533e+00 3.78669947e-01 2.11996526e-01 -9.11286414e-01
4.78933245e-01 5.35794497e-01 6.43659890e-01 -2.49577925e-01
6.81336641e-01 -7.91148365e-01 -4.27378267e-01 8.60124588e-01
4.49094176e-01 -8.48670602e-01 1.24893010e-01 2.74596095e-01
1.33797133e+00 2.30457727e-02 3.20158958e-01 -2.66586661e-01
3.21933746e-01 -1.80423021e-01 -2.51906123e-02 1.18859684e+00
-3.43155205e-01 2.90065464e-02 5.20984590e-01 -5.61336160e-01
-1.82876134e+00 -1.03573287e+00 -5.50232649e-01 5.19995451e-01
-2.25425676e-01 -5.96856236e-01 -7.47893631e-01 -1.96085796e-02
-4.34661567e-01 8.45409095e-01 -4.54123408e-01 -4.83838350e-01
-3.65937263e-01 -8.74313295e-01 6.62364304e-01 3.16156536e-01
3.69147211e-01 -1.65047467e+00 -2.12587774e-01 4.84552264e-01
1.94006830e-01 -9.12786067e-01 5.05485475e-01 8.78789961e-01
-9.68790054e-01 -6.17920995e-01 -7.98886418e-02 -2.05117121e-01
5.29616892e-01 -5.83119810e-01 1.05299997e+00 -2.72103399e-01
-2.78290749e-01 -1.20331243e-01 -4.36698161e-02 -4.08971421e-02
-1.24139309e+00 -2.53908020e-02 2.95212388e-01 -6.85829818e-01
6.26002908e-01 -6.67746127e-01 -3.13472271e-01 -4.12982315e-01
-7.33235359e-01 -3.56619090e-01 6.23014688e-01 1.29478049e+00
1.95396766e-01 6.63298965e-02 1.14565581e-01 -8.20034862e-01
5.57596684e-01 -2.89670527e-01 -8.95832479e-01 -7.80634582e-02
-4.73338336e-01 5.91501474e-01 8.04166198e-01 -1.04595780e-01
-6.60964668e-01 -3.29630047e-01 -4.71044153e-01 1.36190221e-01
-3.31873626e-01 4.31919843e-01 3.18017900e-01 -2.73942530e-01
8.84381235e-01 5.05532682e-01 -3.51109356e-02 7.14363018e-03
2.18670532e-01 8.84994626e-01 4.06410992e-01 -7.97329426e-01
8.17064822e-01 6.42543808e-02 5.03661633e-01 -6.92580760e-01
-5.81950188e-01 -3.76755714e-01 -9.41303551e-01 2.59358495e-01
5.50228059e-01 -5.83307624e-01 -1.29745185e+00 3.88846725e-01
-8.04548562e-01 -2.96617985e-01 -4.37720418e-01 7.05795050e-01
-7.34842122e-01 2.23316818e-01 -8.51312876e-01 -1.11035526e+00
-4.38921511e-01 -1.48856962e+00 1.13912475e+00 9.56069529e-02
-6.69438839e-02 -1.06615818e+00 2.99276441e-01 2.16884032e-01
6.02549016e-01 1.87469795e-01 8.22709262e-01 -4.85408813e-01
-8.07639778e-01 -6.53671145e-01 -5.89969046e-02 6.46749020e-01
-1.71478596e-02 -2.48193994e-01 -1.16209841e+00 -7.93954670e-01
6.64199740e-02 -8.30891252e-01 7.21441627e-01 1.62130788e-01
8.97112072e-01 1.70668125e-01 -4.74084578e-02 5.76598227e-01
1.50722277e+00 2.67748892e-01 1.03058684e+00 7.44721368e-02
4.59254414e-01 -8.95641893e-02 2.16011852e-01 3.10268402e-01
-2.85373509e-01 3.55195373e-01 1.02957912e-01 3.12555999e-01
4.80286658e-01 -6.85715079e-02 3.25725615e-01 1.20823109e+00
-2.56727338e-01 -1.23591479e-02 -7.91260421e-01 1.68191954e-01
-1.03611839e+00 -1.28319430e+00 -6.12503439e-02 2.60083747e+00
7.98194647e-01 3.89106065e-01 -1.27020046e-01 3.09790820e-01
4.95615482e-01 -8.26746575e-04 -4.50996727e-01 -1.10209477e+00
4.14245129e-01 1.24356115e+00 5.01799226e-01 7.04438269e-01
-7.67049372e-01 8.56704175e-01 7.25152731e+00 6.44089580e-01
-1.11020577e+00 5.34537062e-02 2.20746666e-01 -2.31340721e-01
-2.24854887e-01 4.73713279e-01 -7.02374756e-01 4.57595110e-01
2.12382841e+00 1.66879907e-01 8.99803758e-01 4.11797225e-01
-2.25142181e-01 -3.89395177e-01 -1.23443127e+00 1.20208430e+00
-4.61843938e-01 -1.44086647e+00 -3.57256114e-01 5.59906721e-01
7.44858325e-01 4.89953965e-01 -2.54425108e-01 8.47364128e-01
2.85715431e-01 -1.27926719e+00 1.75540194e-01 5.46091616e-01
1.32996988e+00 -8.99485171e-01 9.56702948e-01 6.93419218e-01
-5.28743863e-01 2.10316405e-02 -8.65592182e-01 -3.12543809e-01
2.00641587e-01 6.27733648e-01 -1.32701218e+00 4.75613922e-01
1.38099983e-01 -1.78793054e-02 -1.34778723e-01 6.20835483e-01
-3.66606712e-02 7.80272305e-01 -7.67657697e-01 -5.21633744e-01
6.07964635e-01 -6.42381370e-01 1.80555269e-01 8.77856553e-01
5.21858275e-01 2.34911945e-02 -2.40828991e-01 1.20600116e+00
-8.77687931e-02 -3.42448354e-01 -7.10114658e-01 -3.57732624e-01
7.71723628e-01 1.17868459e+00 -4.48727578e-01 -6.38900101e-01
1.04972899e-01 1.04348671e+00 3.38290006e-01 1.34518102e-01
-6.13900363e-01 -6.14933848e-01 2.46779155e-02 -3.03511888e-01
1.57270998e-01 -3.35158765e-01 2.86318392e-01 -1.15233886e+00
-9.19677466e-02 -7.76155412e-01 -3.76935601e-01 -6.66637599e-01
-8.66523623e-01 5.84138870e-01 9.00515392e-02 -8.20263207e-01
-9.30522501e-01 -8.61601830e-01 -6.20457053e-01 1.51452732e+00
-8.84788036e-01 -7.35923111e-01 2.24456280e-01 2.37768039e-01
-4.14570719e-01 -2.87806779e-01 1.60921836e+00 -1.42126143e-01
-2.86114633e-01 4.66473639e-01 7.79947162e-01 1.40486106e-01
3.31668437e-01 -1.52261639e+00 7.58369744e-01 5.04301667e-01
5.05722046e-01 1.14550924e+00 6.47271574e-01 -3.17244023e-01
-1.62161660e+00 -5.05565882e-01 7.22400010e-01 -5.58059573e-01
6.81094050e-01 -5.08346736e-01 -6.70606315e-01 7.46781528e-01
2.05197871e-01 1.99128494e-01 7.41759598e-01 5.16448796e-01
-2.01043516e-01 2.31969301e-02 -1.32326865e+00 1.96331546e-01
6.33404195e-01 -1.32526422e+00 -3.71953517e-01 3.98532957e-01
1.34687603e-01 -5.40237129e-01 -1.33723104e+00 8.75098854e-02
6.30122185e-01 -1.28372276e+00 7.18768895e-01 -5.43346345e-01
5.84853366e-02 -1.63302049e-01 -2.50678062e-01 -1.26205075e+00
-5.11408389e-01 -7.59397388e-01 -1.43260777e-01 5.99274635e-01
3.52279246e-01 -9.46005046e-01 7.79762268e-01 5.12962580e-01
1.17924027e-01 -4.76970136e-01 -1.30423832e+00 -6.23993516e-01
3.89572263e-01 -5.78139722e-01 7.66047835e-01 7.01872766e-01
-1.46718659e-02 5.27507603e-01 -2.99526751e-01 6.32272586e-02
6.31225169e-01 5.39192706e-02 8.17510545e-01 -8.77295017e-01
-8.76059353e-01 -1.93919614e-01 -9.06106889e-01 -7.45013297e-01
6.09411784e-02 -1.14522254e+00 -5.39332591e-02 -1.05883193e+00
4.56372082e-01 -4.84136254e-01 -4.22971964e-01 1.22643337e-01
1.32298544e-01 4.59470302e-01 -9.28211361e-02 1.29345879e-01
-2.55393624e-01 3.47754091e-01 1.02247119e+00 3.39768740e-04
2.60271300e-02 -1.66181833e-01 -2.14684367e-01 1.47602230e-01
7.29360402e-01 -5.96197009e-01 -1.43409073e-01 1.10796645e-01
4.12526518e-01 3.78232837e-01 4.53812659e-01 -1.40735757e+00
1.11352727e-01 2.23401070e-01 7.25637197e-01 -5.91872454e-01
7.50665188e-01 -3.11782479e-01 4.52906072e-01 5.57296157e-01
-1.22691467e-01 -2.06396729e-01 3.79899405e-02 1.93378165e-01
-7.52801597e-02 -7.37085700e-01 7.29010940e-01 -3.60159487e-01
-3.93203080e-01 1.41109014e-02 -1.39497831e-01 -4.40006196e-01
6.80226207e-01 -2.77760237e-01 -2.45175064e-01 -3.73263478e-01
-8.48951161e-01 -5.37750602e-01 4.67252076e-01 -4.05064940e-01
2.62717158e-01 -1.23642671e+00 -3.82777482e-01 6.71938181e-01
2.48339679e-03 -2.71691591e-01 3.13470095e-01 7.73023427e-01
-8.33934367e-01 6.49483621e-01 -1.88117415e-01 -7.64593720e-01
-8.07414830e-01 3.15468997e-01 4.76465970e-01 -2.60349840e-01
-4.90313321e-01 7.33772933e-01 -8.51464272e-01 -8.57849956e-01
-5.29361367e-01 -4.31592822e-01 5.71235359e-01 -5.58795273e-01
5.31811178e-01 1.24839261e-01 2.72157729e-01 -4.34598267e-01
-3.25294794e-03 9.90941674e-02 -1.59430891e-01 -4.83168721e-01
1.32018030e+00 4.74039942e-01 -3.52273703e-01 7.22051382e-01
1.56856918e+00 -1.74674645e-01 -7.98909724e-01 -2.29847699e-01
-2.64892578e-01 -1.21873118e-01 3.73585790e-01 -7.86836147e-01
-2.69693851e-01 1.10024703e+00 6.63559735e-01 6.31908953e-01
7.79909134e-01 -7.17574432e-02 7.53151655e-01 1.24180567e+00
1.06023669e+00 -9.45753276e-01 -4.87511009e-01 7.40435898e-01
1.90774992e-01 -1.25907791e+00 3.19243334e-02 4.47200328e-01
-1.03753246e-02 1.28887856e+00 -1.10626876e-01 -4.90981519e-01
2.74648190e-01 4.38411266e-01 -1.64682478e-01 -3.64234954e-01
-5.38021326e-01 1.80835247e-01 1.08783953e-01 4.19609040e-01
4.66189653e-01 5.02751470e-01 3.04147393e-01 -3.83416891e-01
-6.45803988e-01 -2.18425348e-01 7.73481429e-01 8.93553495e-01
-5.85665524e-01 -1.47907209e+00 -4.51016575e-01 6.95438147e-01
-6.87711909e-02 -3.69177431e-01 6.19667359e-02 6.98815823e-01
-1.07288584e-01 7.28345096e-01 1.69008404e-01 -4.90816414e-01
-5.28238863e-02 5.10338783e-01 1.15518725e+00 -6.82212472e-01
-4.38911825e-01 -1.65075928e-01 2.37234935e-01 -6.72177315e-01
-4.30362910e-01 -5.04176378e-01 -1.12482166e+00 -8.64954710e-01
-7.47029245e-01 3.10398489e-01 9.65862334e-01 1.20036769e+00
1.72055379e-01 4.26812738e-01 6.35490239e-01 -1.03974521e+00
-1.13204539e+00 -1.38327754e+00 -9.78798807e-01 1.71126693e-01
3.19484144e-01 -3.29781979e-01 -3.43569815e-01 -4.89712030e-01] | [5.5823469161987305, 4.9362077713012695] |
4cb5e7f6-d827-4707-9081-a7fc5a7a16f9 | active-learning-on-a-programmable-photonic | 2208.02104 | null | https://arxiv.org/abs/2208.02104v1 | https://arxiv.org/pdf/2208.02104v1.pdf | Active Learning on a Programmable Photonic Quantum Processor | Training a quantum machine learning model generally requires a large labeled dataset, which incurs high labeling and computational costs. To reduce such costs, a selective training strategy, called active learning (AL), chooses only a subset of the original dataset to learn while maintaining the trained model's performance. Here, we design and implement two AL-enpowered variational quantum classifiers, to investigate the potential applications and effectiveness of AL in quantum machine learning. Firstly, we build a programmable free-space photonic quantum processor, which enables the programmed implementation of various hybrid quantum-classical computing algorithms. Then, we code the designed variational quantum classifier with AL into the quantum processor, and execute comparative tests for the classifiers with and without the AL strategy. The results validate the great advantage of AL in quantum machine learning, as it saves at most $85\%$ labeling efforts and $91.6\%$ percent computational efforts compared to the training without AL on a data classification task. Our results inspire AL's further applications in large-scale quantum machine learning to drastically reduce training data and speed up training, underpinning the exploration of practical quantum advantages in quantum physics or real-world applications. | ['He-Liang Huang', 'Wan-su Bao', 'Shuo Zhang', 'Yun-Fei Niu', 'Xiao-Yue Xu', 'Chen Ding'] | 2022-08-03 | null | null | null | null | ['machine-learning', 'machine-learning'] | ['methodology', 'miscellaneous'] | [ 4.24170494e-01 1.26233190e-01 -4.22153145e-01 -3.58301401e-01
-1.11677206e+00 -4.81195331e-01 3.24722469e-01 -1.47783486e-02
-6.77155077e-01 8.72559249e-01 -6.01441443e-01 -4.97078866e-01
1.09748796e-01 -1.32575488e+00 -6.91433311e-01 -1.35689592e+00
1.98605910e-01 5.36968112e-01 7.75186718e-02 -1.29157543e-01
4.43726540e-01 -1.22343644e-03 -1.62192774e+00 1.39280960e-01
9.97775793e-01 9.52708244e-01 -2.90633608e-02 3.19858044e-01
-2.11059395e-03 7.15717733e-01 -2.18026578e-01 -4.64352608e-01
3.24174196e-01 -5.30888498e-01 -7.08925784e-01 -5.64317644e-01
8.82352442e-02 -3.02698147e-02 -5.58121920e-01 1.41203022e+00
5.73026121e-01 2.53456831e-01 5.99441707e-01 -9.09601688e-01
-4.82276320e-01 7.09979951e-01 -8.52186605e-03 1.87782213e-01
-9.29631814e-02 5.52738190e-01 1.22535157e+00 -3.25370431e-01
6.65266454e-01 5.37297189e-01 2.86618799e-01 9.84853804e-01
-1.60573375e+00 -1.01471865e+00 -9.37120140e-01 5.61833799e-01
-1.46607208e+00 -6.99603736e-01 6.75481141e-01 -1.97391614e-01
1.28595483e+00 1.04419887e-01 6.55825078e-01 9.03834522e-01
3.87569577e-01 4.05644268e-01 1.55587327e+00 -8.40011060e-01
7.78193593e-01 2.26670966e-01 6.08953357e-01 1.11371207e+00
1.63313448e-01 8.64671946e-01 -8.36985111e-01 -3.03478986e-01
1.17098153e-01 -2.72630185e-01 4.17579450e-02 -1.75147250e-01
-1.11506820e+00 9.77330565e-01 6.34696901e-01 -1.24178521e-01
-9.09593254e-02 5.19358277e-01 2.52387673e-01 2.96987981e-01
2.91372269e-01 6.50198281e-01 -1.44106105e-01 -2.48137221e-01
-7.13382125e-01 -5.55097349e-02 4.77615744e-01 8.25317323e-01
1.39426792e+00 -3.01503599e-01 -4.26556505e-02 3.12397212e-01
1.54071420e-01 7.72221804e-01 6.74537048e-02 -9.32335317e-01
9.24374461e-02 4.31323588e-01 -5.86546510e-02 -1.50346691e-02
-3.97360593e-01 -2.24241212e-01 -4.93707836e-01 8.04529265e-02
-5.90211228e-02 1.02474373e-02 -7.55093038e-01 1.64621496e+00
3.53190243e-01 2.49844596e-01 3.48043919e-01 6.63606644e-01
7.70861864e-01 6.86563909e-01 7.14708269e-02 -6.46766722e-01
1.09073853e+00 -8.10503483e-01 -5.00623167e-01 1.82553411e-01
1.28798735e+00 -5.61405599e-01 1.00848007e+00 2.93860942e-01
-5.77722669e-01 -4.92758930e-01 -1.38294935e+00 -1.70769021e-01
-2.54756421e-01 -1.26159132e-01 1.32227218e+00 1.21087956e+00
-7.17682123e-01 1.02970231e+00 -1.07805026e+00 1.91443369e-01
5.70714712e-01 9.54310834e-01 3.07538696e-02 1.36918779e-02
-1.25803888e+00 6.71827555e-01 7.13741660e-01 -1.29145026e-01
-7.75619626e-01 -3.94680887e-01 -3.49856317e-01 -1.21454477e-01
1.49356037e-01 -7.40699530e-01 1.20766473e+00 -3.47050488e-01
-1.98481977e+00 9.35835481e-01 -1.60050109e-01 -3.68215770e-01
-1.21846460e-01 4.07698303e-01 -4.63465035e-01 -3.42905782e-02
-1.04706675e-01 5.59759915e-01 6.39039814e-01 -4.17261809e-01
-3.44704777e-01 -4.45431054e-01 2.05120847e-01 -9.87960324e-02
-3.03294271e-01 -3.83154124e-01 1.21267296e-01 2.58238465e-01
4.23774689e-01 -1.45303285e+00 -4.87478077e-02 -5.42233944e-01
-2.31340319e-01 -5.96906722e-01 6.48350060e-01 4.58715618e-01
9.82044160e-01 -2.21339059e+00 1.10727109e-01 9.08122510e-02
3.60421598e-01 1.53013214e-01 1.35899112e-01 2.88296700e-01
2.90641397e-01 -1.56513050e-01 -1.53266922e-01 -8.30002129e-02
7.94965178e-02 2.63264179e-01 -3.99724931e-01 5.43002784e-01
-3.47318612e-02 1.29151857e+00 -9.68653381e-01 -3.34827185e-01
2.28540272e-01 1.46919116e-03 -9.83585656e-01 -8.55143592e-02
-4.77317780e-01 7.67319322e-01 -5.84836423e-01 5.70727587e-01
6.40457511e-01 -6.86979234e-01 2.18180448e-01 -2.56056964e-01
-2.22201094e-01 8.53202522e-01 -6.78432584e-01 1.74410248e+00
-3.15357119e-01 5.52402020e-01 -2.25877210e-01 -1.16441011e+00
6.14714384e-01 3.71510349e-02 2.96209484e-01 -1.29562819e+00
3.73394638e-01 6.85101688e-01 4.33727026e-01 -5.23894250e-01
4.14660126e-01 -4.44815099e-01 -2.32280001e-01 7.87192881e-01
2.97914743e-01 -3.77318770e-01 1.12623863e-01 9.98971537e-02
1.07610571e+00 1.06739447e-01 9.93321612e-02 -4.82427597e-01
4.28468078e-01 2.97598362e-01 4.89359856e-01 9.36716259e-01
-6.90015137e-01 -2.17891559e-01 2.04997420e-01 -4.34305817e-01
-9.74175751e-01 -9.98531222e-01 -7.34519720e-01 1.13958442e+00
6.51179969e-01 -6.97971046e-01 -6.99611366e-01 -6.39219344e-01
-3.42520237e-01 8.06216598e-01 -2.47116372e-01 -5.91748357e-01
-4.37334776e-01 -1.59875464e+00 3.93894345e-01 -1.47266746e-01
7.34044492e-01 -8.67577374e-01 -4.44154739e-01 -7.17198998e-02
4.08879295e-02 -9.39799547e-01 2.41279215e-01 7.05389202e-01
-9.96080101e-01 -8.87528181e-01 4.15206790e-01 -6.05722606e-01
4.45626199e-01 1.09070487e-01 7.90781021e-01 -8.22574347e-02
-3.21322650e-01 -8.01025555e-02 -4.69565690e-01 -2.20472172e-01
-6.10453606e-01 2.04567954e-01 5.35080373e-01 -3.76931041e-01
9.68237817e-01 -4.73738223e-01 -6.70586705e-01 -8.18414763e-02
-3.10735732e-01 2.08319411e-01 6.24463797e-01 1.23107147e+00
6.88728750e-01 1.71121508e-01 1.79310918e-01 -1.18018579e+00
2.74505839e-02 -7.20661879e-02 -9.28988576e-01 1.74054101e-01
-8.38997960e-01 5.19073725e-01 6.55406594e-01 -1.35006607e-02
-6.53249443e-01 1.03192985e-01 -3.05830449e-01 1.49907768e-01
1.84252277e-01 1.80398405e-01 3.57004970e-01 -6.37449145e-01
1.13762355e+00 3.57795507e-01 -2.56559014e-01 4.95859236e-02
3.93976957e-01 1.03913617e+00 -1.17905624e-01 -6.76845431e-01
7.39134610e-01 4.28893089e-01 4.90760624e-01 -1.04464924e+00
-1.05847716e+00 -1.65840015e-01 -6.61692202e-01 1.26300558e-01
1.00694215e+00 -8.91779423e-01 -1.13214111e+00 2.32867107e-01
-8.23066533e-01 -3.71720791e-01 -2.71836638e-01 8.86043429e-01
-5.86592615e-01 1.83811739e-01 -6.34219766e-01 -8.82665694e-01
-5.48409581e-01 -1.70750725e+00 9.03360724e-01 3.45375806e-01
2.10535124e-01 -5.95869303e-01 4.12144549e-02 8.47292185e-01
2.15730608e-01 -2.64382243e-01 1.26554215e+00 -3.34267139e-01
-1.32662630e+00 2.31113192e-03 -6.01450540e-02 2.54628032e-01
-3.57165992e-01 -2.57732272e-01 -1.50573206e+00 -4.93955344e-01
-6.94856700e-03 -1.01870263e+00 1.16186702e+00 -7.80029446e-02
1.21209335e+00 2.06238553e-01 -3.80501091e-01 8.40687096e-01
1.32854819e+00 3.23835552e-01 6.53250575e-01 -1.19192913e-01
5.19810259e-01 4.29924764e-02 3.82287502e-01 -2.43524127e-02
-1.66682243e-01 5.88796794e-01 2.81069875e-01 4.46438909e-01
1.86849654e-01 -4.70078662e-02 4.45242852e-01 1.48262262e+00
-2.07011998e-01 3.37968141e-01 -8.45087409e-01 -4.37831938e-01
-1.41732073e+00 -1.05035460e+00 -2.01885089e-01 2.38025308e+00
8.55155647e-01 3.33645344e-01 -3.64203215e-01 1.71968922e-01
4.95845169e-01 2.28093080e-02 -6.87551677e-01 -3.64106447e-01
9.23044235e-02 1.09401703e+00 6.03360713e-01 3.17590743e-01
-1.25988746e+00 1.20176661e+00 6.29416370e+00 1.42951262e+00
-1.48899686e+00 6.19363308e-01 3.00040722e-01 -3.90629247e-02
-2.28371203e-01 5.74766755e-01 -8.19287121e-01 4.55170721e-01
1.39784253e+00 1.19401671e-01 7.68594027e-01 8.99811924e-01
-2.65614659e-01 -1.33198708e-01 -1.30465579e+00 1.28864491e+00
-5.13617933e-01 -1.96853113e+00 -2.14060158e-01 2.89785534e-01
8.69365513e-01 6.25105083e-01 7.61760548e-02 7.64656365e-01
-7.75716603e-02 -8.54628980e-01 5.76420486e-01 2.15717390e-01
1.17338479e+00 -6.44744217e-01 6.07422471e-01 7.34056056e-01
-5.94959795e-01 -2.17636928e-01 -7.26565838e-01 -4.74621773e-01
-2.57263064e-01 4.12508279e-01 -5.37529469e-01 2.33696371e-01
3.18664461e-01 3.82922322e-01 -3.66225928e-01 3.83406997e-01
-2.01634809e-01 8.49762499e-01 -3.33679378e-01 -7.79409349e-01
1.31657422e-01 -8.71144056e-01 1.85703725e-01 6.72826350e-01
9.08890143e-02 5.15846431e-01 8.44941661e-03 1.29355204e+00
-3.93822372e-01 -2.65101474e-02 -5.38399339e-01 -6.66027248e-01
7.51543164e-01 1.18846166e+00 -7.74614513e-01 -3.18650603e-01
-3.34610999e-01 8.86778235e-01 4.72953200e-01 -1.08555466e-01
-9.02086973e-01 -4.47245508e-01 1.88289657e-01 -3.17489296e-01
-1.72826156e-01 -3.78044337e-01 -5.26339650e-01 -1.41483402e+00
-4.58464265e-01 -2.38157615e-01 1.07486732e-02 -3.32293451e-01
-9.72722769e-01 3.41343850e-01 -5.48576474e-01 -1.17589474e+00
1.05710790e-01 -9.31788206e-01 -4.63818908e-01 6.91807449e-01
-1.18252397e+00 -8.45375836e-01 -1.13707565e-01 1.80103257e-01
-2.28145391e-01 -3.99075985e-01 1.30077612e+00 3.03739786e-01
-9.03478622e-01 6.57980025e-01 6.75501466e-01 2.32498739e-02
3.50176901e-01 -1.02684426e+00 2.30719864e-01 4.33731169e-01
6.03072882e-01 1.05732358e+00 4.68314499e-01 -1.65381908e-01
-2.04616976e+00 -8.52296889e-01 6.62393332e-01 -7.63837993e-01
6.48449779e-01 -7.49890685e-01 -2.81606466e-01 4.63848531e-01
-9.57766771e-02 1.01456285e-01 1.12380862e+00 3.74158800e-01
-4.41477120e-01 -1.22901484e-01 -1.13709342e+00 2.96302974e-01
1.05829370e+00 -1.28414452e+00 -3.18838537e-01 8.55249882e-01
6.85382009e-01 -2.51897752e-01 -7.28708863e-01 4.09105062e-01
4.89463538e-01 -1.15039980e+00 7.44944155e-01 -5.24718940e-01
3.68145645e-01 -1.00102149e-01 -3.63793731e-01 -8.37456763e-01
-3.21812063e-01 -7.64307737e-01 -8.90465900e-02 4.59901422e-01
6.64348125e-01 -8.99101555e-01 1.19962299e+00 4.29631829e-01
-3.08469117e-01 -6.47060752e-01 -1.36778784e+00 -6.04872644e-01
4.06586319e-01 -6.79692149e-01 1.33112758e-01 9.95428383e-01
2.79039770e-01 6.88453972e-01 7.83791915e-02 -1.00066692e-01
9.98873532e-01 4.20631856e-01 5.46380103e-01 -9.14286733e-01
-5.57633758e-01 -1.93341374e-01 -6.36368632e-01 -9.59279716e-01
3.26469988e-01 -1.51224911e+00 -1.28155053e-01 -6.39237642e-01
5.72086096e-01 -1.01097834e+00 -4.54806209e-01 1.34084225e-01
-7.35541806e-02 6.50545537e-01 -1.85319945e-01 5.61179876e-01
-7.21014440e-01 7.50162780e-01 1.08964419e+00 -3.46951962e-01
-8.47691819e-02 -1.06878482e-01 -1.82233617e-01 2.12492198e-01
5.20450771e-01 -7.94334769e-01 -2.73620665e-01 -1.73026353e-01
5.34755290e-01 -1.80778459e-01 2.65553981e-01 -1.15816331e+00
2.58176357e-01 -1.79213226e-01 8.48327950e-02 -3.59453142e-01
5.73668659e-01 -2.17585891e-01 -1.70792237e-01 8.47838581e-01
-1.73015103e-01 -7.24091589e-01 -3.70037645e-01 5.14075518e-01
1.81909904e-01 -4.40937370e-01 1.28231227e+00 -1.18717268e-01
-8.10792804e-01 3.83781075e-01 4.20026928e-02 -1.37386406e-02
1.20564663e+00 7.44540393e-02 -7.53382146e-01 4.96729314e-01
-6.64246738e-01 -1.67468622e-01 6.22466445e-01 -2.82188505e-01
1.23081543e-01 -1.11901534e+00 -1.35957778e-01 4.84781504e-01
4.29549545e-01 -3.12593699e-01 3.50482762e-01 1.09240115e+00
-9.95840490e-01 7.77250946e-01 -1.50868827e-02 -1.02279973e+00
-7.31975675e-01 6.59476817e-01 3.18496644e-01 3.06887299e-01
-5.86514056e-01 1.12884343e+00 -2.68570781e-01 -6.76980257e-01
-3.09417784e-01 5.86126857e-02 4.13796961e-01 -2.40272358e-01
1.94732055e-01 3.75521302e-01 3.76173407e-01 -5.67400038e-01
-2.49906704e-01 4.32574719e-01 1.46658123e-01 1.52311369e-03
9.87660289e-01 2.74225593e-01 -4.05816495e-01 4.95143682e-01
1.42212927e+00 -4.16923910e-02 -7.43194818e-01 -4.16720092e-01
-2.31927447e-02 -4.40596268e-02 5.19341469e-01 -3.84582460e-01
-5.65566480e-01 1.27975869e+00 9.73278940e-01 1.36525854e-01
6.95245504e-01 6.36228472e-02 9.19539332e-01 1.16302621e+00
1.27125156e+00 -1.36744797e+00 -2.01634869e-01 5.50394654e-01
-2.38180816e-01 -1.52250004e+00 4.73206416e-02 -3.80254745e-01
-1.16675533e-01 1.20206416e+00 3.31071943e-01 -1.64856330e-01
6.62416101e-01 -7.11463392e-02 -4.93202284e-02 -4.66825873e-01
-7.42272258e-01 -3.10228109e-01 -5.28995320e-02 2.48801485e-01
4.28251296e-01 6.34604871e-01 -3.96947891e-01 1.62866876e-01
-5.50074816e-01 -7.58801699e-02 5.37955463e-01 9.91874456e-01
-5.86361945e-01 -1.42611539e+00 4.48751822e-02 6.80060208e-01
1.89317875e-02 -2.57844746e-01 -1.55032426e-01 4.70381856e-01
6.86157048e-01 9.04584050e-01 1.12709530e-01 -9.66623962e-01
-2.39609599e-01 5.35710633e-01 1.15596080e+00 -9.05032277e-01
-2.16852233e-01 -3.16293359e-01 -1.32572949e-01 -5.23754537e-01
-5.26750207e-01 -4.85025883e-01 -1.56212938e+00 -6.81926847e-01
-8.82464290e-01 5.79290152e-01 6.88118935e-01 1.12951100e+00
3.60346198e-01 2.05136672e-01 7.81232893e-01 -5.84701478e-01
-1.16265750e+00 -7.55269408e-01 -6.80899501e-01 1.74158707e-01
1.80471435e-01 -8.88788223e-01 -4.25044984e-01 -4.26905543e-01] | [5.5508012771606445, 4.9886298179626465] |
aa23a0f1-fdea-4e6d-80b5-594b78ce3899 | remote-sensing-scene-classification-with | 2302.14256 | null | https://arxiv.org/abs/2302.14256v2 | https://arxiv.org/pdf/2302.14256v2.pdf | Remote Sensing Scene Classification with Masked Image Modeling (MIM) | Remote sensing scene classification has been extensively studied for its critical roles in geological survey, oil exploration, traffic management, earthquake prediction, wildfire monitoring, and intelligence monitoring. In the past, the Machine Learning (ML) methods for performing the task mainly used the backbones pretrained in the manner of supervised learning (SL). As Masked Image Modeling (MIM), a self-supervised learning (SSL) technique, has been shown as a better way for learning visual feature representation, it presents a new opportunity for improving ML performance on the scene classification task. This research aims to explore the potential of MIM pretrained backbones on four well-known classification datasets: Merced, AID, NWPU-RESISC45, and Optimal-31. Compared to the published benchmarks, we show that the MIM pretrained Vision Transformer (ViTs) backbones outperform other alternatives (up to 18% on top 1 accuracy) and that the MIM technique can learn better feature representation than the supervised learning counterparts (up to 5% on top 1 accuracy). Moreover, we show that the general-purpose MIM-pretrained ViTs can achieve competitive performance as the specially designed yet complicated Transformer for Remote Sensing (TRS) framework. Our experiment results also provide a performance baseline for future studies. | ['Alex Tien', 'Liya Wang'] | 2023-02-28 | null | null | null | null | ['scene-classification', 'earthquake-prediction'] | ['computer-vision', 'computer-vision'] | [ 4.29324657e-01 -1.39266238e-01 -9.38057601e-02 -3.79141837e-01
-8.68841708e-01 -1.24932647e-01 8.42930496e-01 -1.22965708e-01
-4.14232582e-01 4.77130532e-01 8.46386515e-03 -6.50914729e-01
-5.38528487e-02 -9.30694699e-01 -6.07322574e-01 -9.06036615e-01
-4.57789987e-01 1.57328665e-01 2.94980019e-01 -3.00964117e-01
2.19559282e-01 6.57673717e-01 -1.77848065e+00 4.57899898e-01
7.80664444e-01 1.17702460e+00 5.06145120e-01 4.30687398e-01
-1.63234863e-02 1.09056258e+00 -1.11856133e-01 -2.71328427e-02
4.84720141e-01 -2.91544851e-02 -7.65442908e-01 1.48353338e-01
5.70623696e-01 4.77192318e-03 -1.76837042e-01 9.00474370e-01
4.34176683e-01 -1.23152867e-01 8.12244117e-01 -1.06572664e+00
-4.97121096e-01 5.25057435e-01 -8.13909054e-01 2.86881447e-01
-2.49706313e-01 1.76531568e-01 9.30758774e-01 -1.30800724e+00
4.46298331e-01 1.06110477e+00 6.82323515e-01 3.22482735e-01
-9.53919113e-01 -6.11851335e-01 1.93152145e-01 5.86057425e-01
-1.42814291e+00 -2.77006775e-01 8.24204743e-01 -6.01016939e-01
8.45387936e-01 3.18768591e-01 5.00937879e-01 7.44315684e-01
4.57713893e-03 1.10888481e+00 1.59946966e+00 -6.10839486e-01
2.60670274e-01 1.33651435e-01 3.27061713e-01 9.22964454e-01
4.03290018e-02 1.91363245e-01 -5.15447438e-01 1.27121791e-01
7.54611135e-01 1.72937348e-01 -3.70247841e-01 -2.99821287e-01
-1.21721041e+00 8.64377499e-01 9.13818419e-01 4.03924167e-01
-3.41185480e-01 4.40832376e-02 1.85507655e-01 3.95270437e-01
6.06588423e-01 3.67596626e-01 -5.18236816e-01 2.22913235e-01
-1.03051484e+00 -1.48840621e-01 3.97801816e-01 5.20773292e-01
1.09342051e+00 4.88388002e-01 -1.82609662e-01 8.83628845e-01
2.00214669e-01 7.30904341e-01 3.52352917e-01 -4.57632124e-01
4.90472436e-01 6.66095197e-01 -1.50107130e-01 -1.03745699e+00
-4.91093040e-01 -7.60662794e-01 -1.25319099e+00 3.82383049e-01
1.91880271e-01 1.72410116e-01 -1.08455014e+00 1.15320337e+00
-5.23968227e-02 3.18500251e-01 2.54198194e-01 7.08870411e-01
1.03279150e+00 8.89207661e-01 -1.57816301e-03 1.39212012e-02
1.09328604e+00 -1.02582467e+00 -1.30932122e-01 -4.03871804e-01
6.08522713e-01 -4.80373889e-01 1.28444958e+00 5.40164709e-01
-3.44705611e-01 -7.70700753e-01 -1.11820602e+00 3.05269390e-01
-4.99317408e-01 4.49985504e-01 7.97688186e-01 6.50725842e-01
-8.80255878e-01 5.47280312e-01 -7.13769078e-01 -5.31652749e-01
6.58404469e-01 -1.65649235e-01 -4.46444929e-01 -2.39305928e-01
-7.36103475e-01 8.09457242e-01 2.10918948e-01 3.49689484e-01
-1.16889405e+00 -8.49845827e-01 -9.55983758e-01 5.96391112e-02
2.03975797e-01 -1.17024742e-01 6.79098368e-01 -7.43232250e-01
-1.24537838e+00 1.10668337e+00 2.63467163e-01 -7.64447808e-01
5.05356014e-01 -1.23259723e-01 -3.75627249e-01 1.30968839e-01
7.85555318e-02 7.94556141e-01 9.78605092e-01 -1.48008418e+00
-4.92926240e-01 -3.36727023e-01 -1.54153287e-01 -1.30595919e-02
-3.72513950e-01 -1.93591177e-01 -8.19788873e-02 -8.49709034e-01
3.49223107e-01 -6.82648897e-01 -3.59595269e-01 1.32858917e-01
-4.61020440e-01 5.17207794e-02 1.12443626e+00 -5.78410268e-01
7.25763261e-01 -2.28622222e+00 -1.92805812e-01 3.35975200e-01
1.52826868e-02 6.21641278e-01 -1.84358045e-01 4.99776602e-01
-1.85086131e-01 -5.18366918e-02 -5.22362590e-01 -3.21817458e-01
-2.36722857e-01 2.75178820e-01 -4.59894508e-01 5.07442534e-01
1.90265253e-01 1.07747602e+00 -5.95207930e-01 -4.39899445e-01
5.20512164e-01 2.61178553e-01 -1.47842586e-01 1.69783726e-01
-1.17882185e-01 6.23113096e-01 -2.27269918e-01 9.00120795e-01
7.05033898e-01 -3.64468902e-01 -4.16100062e-02 -3.71451825e-01
-2.83325493e-01 -2.54318982e-01 -9.96236563e-01 1.36830688e+00
-4.06834602e-01 7.27849841e-01 -2.79007435e-01 -1.43404734e+00
1.28714943e+00 -1.75320685e-01 2.30566904e-01 -1.02639043e+00
-3.05402964e-01 1.05669945e-01 -3.66242051e-01 -6.67577922e-01
2.63344437e-01 -1.26621291e-01 5.04964404e-02 8.09334368e-02
-1.36705548e-01 7.08501339e-02 -1.49277464e-01 -3.39174308e-02
9.61978853e-01 6.54788390e-02 3.31117690e-01 -3.23962599e-01
6.10607982e-01 4.05032247e-01 3.53673458e-01 8.95958543e-01
-1.62617080e-02 5.91723323e-01 -6.80143619e-03 -7.13494718e-01
-6.92107081e-01 -9.39364135e-01 -3.22555363e-01 1.04467273e+00
1.32757556e-02 -7.84723014e-02 -3.63124490e-01 -7.09327817e-01
-7.69698992e-02 4.98231173e-01 -5.14650702e-01 4.53409879e-03
-3.95849705e-01 -9.68998253e-01 5.59015572e-01 6.64924622e-01
1.22315657e+00 -1.07916582e+00 -6.87738240e-01 3.59648629e-03
-9.47324559e-02 -1.32484818e+00 2.63832659e-01 2.06871942e-01
-8.79533529e-01 -9.56842601e-01 -8.19592357e-01 -8.01236093e-01
3.83491039e-01 6.65164530e-01 9.04142857e-01 1.57244250e-01
-3.88472259e-01 3.76396924e-01 -4.72158492e-01 -4.29240644e-01
-1.14143409e-01 -5.34886345e-02 -3.31565648e-01 4.44145441e-01
3.49542759e-02 -6.37515843e-01 -4.96977478e-01 1.51927650e-01
-8.30260038e-01 4.34112132e-01 8.62076938e-01 9.03750300e-01
3.94579738e-01 6.49354830e-02 4.93410975e-01 -1.07780516e+00
-1.27378926e-01 -3.51897210e-01 -6.17891371e-01 4.74877775e-01
-6.13325000e-01 -7.61014689e-03 4.92850721e-01 -9.06426087e-02
-1.11723042e+00 1.93170607e-01 -8.36593285e-02 -3.54554921e-01
-3.80538940e-01 9.27365363e-01 5.67825697e-03 -4.64299023e-01
8.48002791e-01 6.22248352e-01 -1.35003254e-01 -8.50694597e-01
5.09591177e-02 8.45939159e-01 6.91425979e-01 -3.23929369e-01
1.13762200e+00 7.24402606e-01 2.19791427e-01 -1.29352343e+00
-1.04234564e+00 -6.52158439e-01 -6.03570998e-01 -2.88646251e-01
7.43288338e-01 -1.27766335e+00 -2.03257874e-01 6.56557739e-01
-6.60452545e-01 -5.49226403e-01 -9.39125344e-02 3.59414876e-01
-3.19315940e-01 5.98510027e-01 -3.22230339e-01 -1.02756691e+00
-4.81656700e-01 -8.36610079e-01 1.24357307e+00 2.28558369e-02
5.02228260e-01 -8.37549567e-01 -1.14540882e-01 5.85260212e-01
7.80084848e-01 4.43861634e-01 1.09664547e+00 -5.35743415e-01
-7.07595646e-01 5.37430234e-02 -4.80321407e-01 5.61693609e-01
-9.80291665e-02 -2.24273980e-01 -1.29173553e+00 -3.12741816e-01
-2.29296863e-01 -3.17268997e-01 1.41513395e+00 1.90319315e-01
1.18142664e+00 -7.90953934e-02 -1.38172403e-01 7.08186626e-01
1.77247953e+00 -6.53454959e-02 1.03673816e+00 6.19580269e-01
8.75582218e-01 5.62368870e-01 6.29791677e-01 4.42857027e-01
4.88922775e-01 5.90319335e-01 6.93597853e-01 -5.77188969e-01
-2.42439717e-01 -1.92175526e-02 5.16693532e-01 5.34361482e-01
-3.69138956e-01 2.72153348e-01 -1.22137201e+00 4.06418443e-01
-1.97614491e+00 -1.13563049e+00 -4.21611011e-01 2.23776460e+00
4.02502716e-01 5.46392146e-03 -1.07754193e-01 6.18613243e-01
3.18118900e-01 2.88213491e-01 -2.06153169e-01 1.77366510e-01
-3.98051322e-01 4.66878891e-01 6.08574688e-01 1.68564335e-01
-1.50815487e+00 9.89588916e-01 5.93845606e+00 1.07316995e+00
-1.52435398e+00 9.17263851e-02 6.47529721e-01 5.81967831e-01
7.20814764e-02 -1.04318485e-01 -7.14555323e-01 1.98160991e-01
4.89173621e-01 4.55260277e-01 -2.00266708e-02 1.04823720e+00
2.81313777e-01 -3.20284814e-01 -7.54826486e-01 1.22260821e+00
3.83382314e-03 -1.65115952e+00 2.85595417e-01 -2.38974988e-02
7.15470493e-01 3.38265687e-01 1.57137364e-01 5.14908910e-01
-1.11261822e-01 -1.04778993e+00 7.09956288e-01 5.03075123e-01
7.19543815e-01 -4.42801267e-01 8.46968234e-01 4.44799185e-01
-1.35386407e+00 -1.86785653e-01 -5.67568898e-01 -2.47640714e-01
-1.94498569e-01 7.26569951e-01 -7.88314819e-01 8.45899343e-01
8.48320782e-01 1.14594185e+00 -1.10315943e+00 1.21366012e+00
-2.21555419e-02 1.03126228e+00 -7.18333721e-02 2.40868971e-01
4.35339451e-01 -1.39673248e-01 4.18425232e-01 1.28114557e+00
1.12761006e-01 -2.10015103e-01 4.90761846e-01 5.37900150e-01
3.23649436e-01 1.12195201e-01 -7.74661601e-01 3.21788579e-01
1.11678606e-02 1.25188875e+00 -6.39616013e-01 -4.15013194e-01
-3.20220679e-01 7.37705529e-01 2.40001976e-01 3.97274166e-01
-6.85049772e-01 -5.70260622e-02 3.67565423e-01 9.39964205e-02
5.28587580e-01 -1.75928652e-01 -3.57948363e-01 -1.25000632e+00
1.73471551e-02 -7.33709276e-01 5.41049242e-01 -9.20999944e-01
-1.27446759e+00 6.32008851e-01 1.99045707e-02 -1.68420947e+00
1.36047572e-01 -8.53646159e-01 -7.53247678e-01 4.12268698e-01
-2.38405895e+00 -1.61091208e+00 -6.89680040e-01 7.71465182e-01
6.29227042e-01 -4.56259936e-01 7.65702069e-01 2.11626321e-01
-6.74005389e-01 2.89602757e-01 3.92060094e-02 2.72996187e-01
2.78212756e-01 -1.16582263e+00 8.01835805e-02 9.95416939e-01
4.59622502e-01 -6.93004131e-02 3.94558400e-01 -2.28695318e-01
-1.48912466e+00 -1.60508382e+00 6.67784035e-01 1.02857299e-01
4.81139898e-01 -1.81960389e-01 -1.02036631e+00 5.27070880e-01
1.57522280e-02 2.78558314e-01 4.72466439e-01 -4.03325081e-01
-6.91754222e-01 -5.33441842e-01 -1.12954533e+00 3.87836695e-01
9.75535810e-01 -4.85732496e-01 -4.75914985e-01 4.51862872e-01
2.90478855e-01 3.69109847e-02 -6.41883910e-01 6.96863472e-01
2.04628378e-01 -1.16849005e+00 1.20590854e+00 -3.49542975e-01
5.17007649e-01 -4.95910048e-01 -4.75536495e-01 -9.99884784e-01
-2.91767687e-01 -2.41169125e-01 1.87026024e-01 1.02281773e+00
4.01153982e-01 -5.87313175e-01 7.28142202e-01 -7.94206485e-02
-2.56522775e-01 -6.53998017e-01 -8.31366062e-01 -9.75578249e-01
-6.76793680e-02 -6.69939697e-01 2.72963256e-01 1.02914298e+00
-4.62374330e-01 2.91407347e-01 -5.56495428e-01 4.25051540e-01
9.35392976e-01 4.56509471e-01 8.49852681e-01 -1.48384917e+00
-3.21236670e-01 -3.49205762e-01 -6.93003297e-01 -9.76398408e-01
-6.36417046e-02 -1.06720865e+00 -1.17588095e-01 -1.50624919e+00
2.62227029e-01 -6.75394535e-01 -2.65188396e-01 7.90667415e-01
9.12966440e-04 5.70208907e-01 3.27498078e-01 2.95962602e-01
-4.65639979e-01 5.63478589e-01 1.04259562e+00 -4.65692461e-01
1.79612711e-01 3.32122929e-02 -3.69606823e-01 7.39937544e-01
7.54015446e-01 -2.87385166e-01 -8.24908242e-02 -3.52418065e-01
-2.42533325e-03 -1.66682467e-01 7.79866993e-01 -1.45886254e+00
2.64044851e-01 -9.42239538e-02 2.84625918e-01 -7.42660642e-01
2.63650894e-01 -8.78361106e-01 1.00816764e-01 7.14139581e-01
-4.24770787e-02 -2.82916516e-01 3.31024155e-02 5.54944694e-01
-4.06298637e-01 -8.14322233e-02 8.96538198e-01 -3.12577695e-01
-1.40481162e+00 1.87922895e-01 -4.70407248e-01 -2.14783072e-01
9.00921106e-01 -4.20142978e-01 -3.73546302e-01 -2.26238996e-01
-5.73623359e-01 4.63469736e-02 -3.79979126e-02 1.89960510e-01
8.44760776e-01 -1.06138599e+00 -1.02193069e+00 3.45914692e-01
5.70150793e-01 -5.75729869e-02 2.06164941e-01 9.36972916e-01
-5.49588323e-01 2.24661201e-01 -2.27692679e-01 -9.74134862e-01
-1.15355599e+00 2.35499993e-01 3.54235113e-01 -4.71685648e-01
-7.96606481e-01 6.92199349e-01 2.36812994e-01 -5.21572411e-01
2.91022986e-01 -3.53778630e-01 -4.24487889e-01 -1.40405804e-01
5.87559760e-01 4.68214244e-01 3.66080791e-01 -5.98227680e-01
-3.48622411e-01 8.68678868e-01 2.13601723e-01 1.64181754e-01
1.84097803e+00 1.65527299e-01 1.44072883e-02 4.30809528e-01
1.11132550e+00 -3.31432819e-01 -9.55594540e-01 -6.50733471e-01
2.93221682e-01 -3.61964345e-01 2.88913578e-01 -6.51467562e-01
-1.36779857e+00 9.91029143e-01 7.84592092e-01 1.43232420e-01
1.21665776e+00 -5.30950865e-03 2.95994997e-01 8.01897466e-01
7.56729126e-01 -6.07899308e-01 1.31649032e-01 5.97995996e-01
1.05404305e+00 -1.48517919e+00 -1.62498847e-01 -5.51404893e-01
-8.24496210e-01 1.15794194e+00 3.96350563e-01 -1.19840480e-01
9.25976038e-01 2.29978025e-01 1.29460916e-01 -4.18499470e-01
-4.26858097e-01 -6.12793922e-01 3.22986215e-01 7.73397565e-01
4.79942970e-02 9.99095589e-02 3.15191954e-01 2.41061121e-01
5.44417165e-02 -1.51666388e-01 2.80986875e-01 9.92415726e-01
-7.35966742e-01 -7.76690364e-01 -4.83949780e-01 5.59087932e-01
-9.96214822e-02 -3.16222876e-01 -1.14492610e-01 6.72930598e-01
-4.28954838e-03 8.87301981e-01 -3.31481434e-02 -4.19939548e-01
2.99031377e-01 -8.92535448e-02 2.37190843e-01 -5.67703605e-01
-6.42911255e-01 -2.59872168e-01 3.20657864e-02 -5.24125278e-01
-8.67908061e-01 -4.88570184e-01 -9.39221919e-01 2.52507515e-02
-2.27643296e-01 -6.54182285e-02 4.19976205e-01 1.22140586e+00
1.80077642e-01 1.59359768e-01 9.43807125e-01 -9.95374501e-01
-4.90803450e-01 -9.96743202e-01 -6.77308798e-01 3.04875582e-01
2.06769437e-01 -6.91074729e-01 -3.08891684e-01 1.43856511e-01] | [9.64865493774414, -1.3016871213912964] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.