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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
db4dec29-c984-4fee-906d-37fecea605cc | robust-unsupervised-graph-representation | 2201.08557 | null | https://arxiv.org/abs/2201.08557v2 | https://arxiv.org/pdf/2201.08557v2.pdf | Toward Enhanced Robustness in Unsupervised Graph Representation Learning: A Graph Information Bottleneck Perspective | Recent studies have revealed that GNNs are vulnerable to adversarial attacks. Most existing robust graph learning methods measure model robustness based on label information, rendering them infeasible when label information is not available. A straightforward direction is to employ the widely used Infomax technique from typical Unsupervised Graph Representation Learning (UGRL) to learn robust unsupervised representations. Nonetheless, directly transplanting the Infomax technique from typical UGRL to robust UGRL may involve a biased assumption. In light of the limitation of Infomax, we propose a novel unbiased robust UGRL method called Robust Graph Information Bottleneck (RGIB), which is grounded in the Information Bottleneck (IB) principle. Our RGIB attempts to learn robust node representations against adversarial perturbations by preserving the original information in the benign graph while eliminating the adversarial information in the adversarial graph. There are mainly two challenges to optimize RGIB: 1) high complexity of adversarial attack to perturb node features and graph structure jointly in the training procedure; 2) mutual information estimation upon adversarially attacked graphs. To tackle these problems, we further propose an efficient adversarial training strategy with only feature perturbations and an effective mutual information estimator with subgraph-level summary. Moreover, we theoretically establish a connection between our proposed RGIB and the robustness of downstream classifiers, revealing that RGIB can provide a lower bound on the adversarial risk of downstream classifiers. Extensive experiments over several benchmarks and downstream tasks demonstrate the effectiveness and superiority of our proposed method. | ['Qinghua Zheng', 'Jun Zhou', 'Ziqi Liu', 'Jundong Li', 'Minnan Luo', 'Jihong Wang'] | 2022-01-21 | null | null | null | null | ['mutual-information-estimation'] | ['methodology'] | [ 4.71166402e-01 5.02012491e-01 -3.16289306e-01 5.52653382e-03
-7.08593488e-01 -9.43019032e-01 4.75345314e-01 4.43318561e-02
5.85677028e-02 6.08065605e-01 7.77203143e-02 -5.24649978e-01
-3.16437125e-01 -1.04423177e+00 -9.47342038e-01 -9.42654431e-01
-1.93227410e-01 -3.02513450e-01 -2.81859022e-02 -3.03586364e-01
-1.50150293e-02 8.05707991e-01 -7.24953711e-01 -2.78494447e-01
8.49268675e-01 8.16223741e-01 -5.02503037e-01 6.56235993e-01
2.65932471e-01 1.02595615e+00 -5.87276638e-01 -5.15159428e-01
6.35717869e-01 -4.43711370e-01 -8.88646781e-01 -1.79638676e-02
8.33924487e-02 3.86074930e-02 -1.11248910e+00 1.36269391e+00
5.32984436e-01 3.58904898e-01 8.47468555e-01 -1.47815835e+00
-6.18096590e-01 1.07186687e+00 -5.72774529e-01 1.21317767e-01
1.12913795e-01 2.59771973e-01 9.70061719e-01 -1.61723226e-01
4.47675377e-01 1.21690500e+00 5.70519626e-01 5.89188695e-01
-1.15926111e+00 -6.99203491e-01 6.57368898e-01 -1.10109098e-01
-1.42752671e+00 -3.31668764e-01 1.15226877e+00 -1.28521696e-01
5.90496361e-01 4.89951611e-01 2.00425014e-02 1.33842111e+00
1.64944038e-01 6.71967149e-01 8.68015110e-01 -1.93630144e-01
1.32256314e-01 -1.00818183e-02 3.51458788e-02 9.07682478e-01
3.60808194e-01 4.34020340e-01 -1.21769287e-01 -1.99694693e-01
6.37076616e-01 1.32374130e-02 -4.49502349e-01 -3.62978131e-01
-8.20944726e-01 8.79438758e-01 1.02002299e+00 7.19196796e-02
1.04108997e-01 4.34343308e-01 4.54708964e-01 6.74603045e-01
4.15945113e-01 3.84894520e-01 -2.88268536e-01 7.23836780e-01
-3.22351575e-01 -3.22542816e-01 8.87044907e-01 9.93188739e-01
7.20126569e-01 3.75040025e-01 -1.84164479e-01 4.06600684e-01
5.50702810e-01 3.90407950e-01 3.17532510e-01 -5.00416875e-01
6.03996217e-01 6.20580971e-01 -5.69690406e-01 -1.28758252e+00
-4.14434999e-01 -7.64641106e-01 -1.22161245e+00 1.23849817e-01
4.06608880e-01 -3.76360267e-01 -9.12096739e-01 2.09367752e+00
2.70314485e-01 3.47016633e-01 2.37934038e-01 5.44688106e-01
8.66400778e-01 3.40872258e-01 9.03759450e-02 -2.75408179e-01
6.55502379e-01 -6.41688466e-01 -4.91243124e-01 -3.29913139e-01
7.32582331e-01 -2.07548827e-01 5.80199242e-01 -5.17612323e-02
-6.64101243e-01 -2.00004026e-01 -1.34637153e+00 2.76427388e-01
-5.51069379e-01 -4.84811485e-01 8.58509302e-01 1.17883921e+00
-8.97142768e-01 7.98382103e-01 -6.94309890e-01 -4.76762047e-03
3.65659714e-01 6.08560145e-01 -7.08925426e-01 -9.14631635e-02
-1.47533262e+00 5.28314948e-01 4.39052880e-01 2.42848814e-01
-1.15108573e+00 -5.11423409e-01 -1.22676969e+00 1.63783878e-02
7.20108807e-01 -5.19711435e-01 5.73589087e-01 -9.58980203e-01
-1.60420012e+00 4.63219672e-01 3.77234936e-01 -4.62924719e-01
5.61444879e-01 2.45583564e-01 -3.29934061e-01 4.75034341e-02
-3.03737432e-01 8.84964764e-02 1.07717288e+00 -1.24606073e+00
2.18627572e-01 -4.53996420e-01 3.19991410e-01 1.30132660e-01
-4.07285094e-01 -1.88684285e-01 -3.11312765e-01 -1.00733137e+00
1.88838884e-01 -8.96286190e-01 -6.72194064e-01 -3.29975843e-01
-9.44591284e-01 3.37321945e-02 4.77948308e-01 -4.96526539e-01
1.17362070e+00 -2.03956604e+00 2.32050896e-01 7.58680403e-01
3.59093159e-01 1.89666599e-01 -3.41049820e-01 5.66939712e-01
-4.81319457e-01 6.31418765e-01 -3.73344004e-01 6.80908039e-02
3.44804078e-02 3.46225239e-02 -3.98845583e-01 9.57572579e-01
1.71951115e-01 1.10264659e+00 -9.73504543e-01 -3.36529702e-01
3.60101387e-02 3.53350103e-01 -3.59793037e-01 2.65499711e-01
3.28581668e-02 5.22645473e-01 -6.76322758e-01 7.33164728e-01
8.08824897e-01 -1.11258969e-01 3.63233864e-01 -1.98746681e-01
6.36857092e-01 -7.68566877e-03 -1.16772568e+00 1.34075177e+00
-2.61794835e-01 1.94316298e-01 -3.30349128e-03 -1.26337373e+00
8.48928452e-01 2.01532409e-01 3.56270045e-01 -2.17773855e-01
4.37230796e-01 -2.08511904e-01 -1.62109897e-01 -8.06145743e-02
-1.93120167e-01 -2.14722715e-02 -5.27256846e-01 3.69588882e-01
1.86523542e-01 9.22579169e-02 -1.52474180e-01 5.82327604e-01
1.48830497e+00 -2.03336552e-01 3.78959537e-01 -8.12245682e-02
6.48751497e-01 -6.91731691e-01 6.74962461e-01 9.75739896e-01
-2.67869443e-01 4.43297088e-01 8.71512711e-01 -4.36172932e-02
-2.15379462e-01 -1.29132080e+00 3.11038077e-01 9.12746072e-01
1.12451568e-01 -3.73967856e-01 -7.34420300e-01 -1.60374737e+00
-2.05239886e-03 3.76537651e-01 -7.21747518e-01 -8.48352313e-01
-3.40670645e-01 -9.13969934e-01 9.35722828e-01 4.06091094e-01
4.05127108e-01 -6.65654957e-01 6.10815227e-01 -1.60815418e-01
1.37497857e-01 -9.83916819e-01 -5.29793680e-01 3.06434959e-01
-7.60232329e-01 -1.18529153e+00 -1.99647948e-01 -5.43299973e-01
1.13212407e+00 1.14551879e-01 7.63599694e-01 1.87566966e-01
-1.07503362e-01 3.19111735e-01 -3.36780518e-01 -1.90975785e-01
-6.49115980e-01 2.19846919e-01 1.78458765e-01 1.99916214e-01
-3.88379812e-01 -7.82217622e-01 -3.42300773e-01 2.87994117e-01
-1.15355527e+00 -4.33819145e-01 7.60137856e-01 6.04857564e-01
4.74987119e-01 4.45973754e-01 8.76575470e-01 -1.36405373e+00
5.92815936e-01 -8.09421360e-01 -6.51342869e-01 3.97562176e-01
-7.66286314e-01 3.17601383e-01 9.36385095e-01 -3.63295674e-01
-6.72438979e-01 1.96288541e-01 -1.14759728e-01 -6.17709398e-01
1.26764253e-01 5.88595867e-01 -8.66192877e-01 -6.14602864e-01
7.57975698e-01 1.55144213e-02 -2.59527899e-02 -1.75684765e-01
7.31947839e-01 2.38364026e-01 4.66986984e-01 -3.92513365e-01
1.49772441e+00 3.53165686e-01 4.66571927e-01 -4.53578234e-01
-7.59027600e-01 -2.13956296e-01 -4.86882746e-01 -2.00651899e-01
4.16430265e-01 -7.30970621e-01 -1.00597811e+00 4.39329356e-01
-5.41448355e-01 -2.32562393e-01 -1.98486507e-01 2.48718411e-01
-3.96689385e-01 7.57816434e-01 -6.38649464e-01 -7.05416322e-01
-3.95640582e-01 -1.08144474e+00 5.15879333e-01 1.87051356e-01
3.37169230e-01 -1.27153814e+00 -8.01732987e-02 1.95790172e-01
1.91545784e-01 8.20989132e-01 8.90913129e-01 -9.52942908e-01
-4.42981750e-01 -7.00140357e-01 -1.67119622e-01 5.04927278e-01
3.30867976e-01 -1.78800702e-01 -9.69521165e-01 -6.95514381e-01
-8.03755149e-02 -3.16864997e-01 1.08520567e+00 1.66351840e-01
1.27567995e+00 -6.65016949e-01 -3.94971579e-01 9.32861507e-01
1.53342712e+00 -4.05276507e-01 5.99442601e-01 -2.02965736e-01
1.19517195e+00 4.39977050e-01 1.17155515e-01 1.43466592e-01
2.39964761e-02 2.26437464e-01 8.22258770e-01 -9.28204358e-02
-7.02145100e-02 -6.23366773e-01 7.49440908e-01 8.30955267e-01
1.44189605e-02 -6.50273383e-01 -5.29908717e-01 8.25102702e-02
-1.70311093e+00 -7.17849553e-01 2.08671734e-01 2.32485557e+00
7.00950325e-01 2.82458603e-01 -1.00211307e-01 2.26055592e-01
7.90972114e-01 4.04810190e-01 -6.68725669e-01 -3.34785193e-01
-1.35685116e-01 6.50295764e-02 1.02976346e+00 4.75367486e-01
-1.39009535e+00 1.01322579e+00 5.89087725e+00 9.42592263e-01
-8.15444350e-01 -2.02837959e-01 6.76109612e-01 1.97145894e-01
-4.20864224e-01 1.22479118e-01 -5.20994365e-01 1.93289548e-01
1.03518724e+00 -3.24556977e-01 6.72889948e-01 7.29692340e-01
-6.68740049e-02 6.12713754e-01 -1.03276896e+00 5.84082544e-01
5.15155606e-02 -1.02626741e+00 2.00022176e-01 1.56735733e-01
8.58458042e-01 -1.03152357e-01 1.55986473e-01 4.85628873e-01
8.49851370e-01 -1.19480455e+00 1.12750620e-01 5.97193420e-01
6.76949203e-01 -1.18835974e+00 5.99259436e-01 1.88448116e-01
-1.29461491e+00 -6.30821809e-02 -4.36226159e-01 3.30890715e-01
-3.22762311e-01 5.96202910e-01 -8.30964983e-01 1.12384748e+00
1.50159746e-01 7.39342928e-01 -7.61355162e-01 6.15985870e-01
-7.29449689e-01 7.37668216e-01 -2.24390984e-01 3.13692689e-01
1.04859926e-01 -2.92541534e-02 8.54331553e-01 9.79741812e-01
-1.89370662e-01 -2.93908834e-01 3.47118676e-01 7.09298432e-01
-6.21909618e-01 1.06723703e-01 -9.83899713e-01 -3.98929447e-01
4.22190428e-01 1.39032423e+00 -7.83329785e-01 2.42158875e-01
-1.91809028e-01 1.01248550e+00 4.47540015e-01 5.51868498e-01
-7.76320338e-01 -5.33390641e-01 5.62867880e-01 -2.50525713e-01
2.60004587e-02 7.92398155e-02 -1.49001635e-03 -1.12557042e+00
-1.75632134e-01 -1.08574545e+00 7.83344090e-01 -8.66793171e-02
-1.60523629e+00 4.34373260e-01 -1.78327054e-01 -1.24660385e+00
-2.55861461e-01 -4.60822910e-01 -8.09419572e-01 6.31319642e-01
-1.43329942e+00 -1.38481152e+00 -9.02831405e-02 9.21578228e-01
2.83572860e-02 -6.67625517e-02 8.43169451e-01 -3.15262340e-02
-1.10721529e+00 1.31040430e+00 1.02152139e-01 5.34709632e-01
5.55921853e-01 -1.31749129e+00 4.60776925e-01 1.31872523e+00
1.04808345e-01 7.01814950e-01 3.89248639e-01 -8.01857293e-01
-1.67306125e+00 -1.49758053e+00 1.07951351e-01 -2.68805265e-01
8.49560797e-01 -4.87651020e-01 -7.93867767e-01 9.33625102e-01
-2.15995029e-01 3.80741030e-01 7.87901402e-01 -5.96471764e-02
-7.57445693e-01 -1.27623335e-01 -1.22844398e+00 8.48189294e-01
1.22887719e+00 -6.44621372e-01 -4.38120179e-02 5.82912385e-01
1.03798544e+00 -1.72658294e-01 -1.08518898e+00 7.38130510e-01
8.66213813e-02 -4.33438152e-01 1.10750949e+00 -9.13171291e-01
-2.43541040e-02 -2.71325529e-01 -7.06372336e-02 -1.38123477e+00
-2.81994194e-01 -1.24719238e+00 -2.80729324e-01 1.41423190e+00
5.82262814e-01 -8.36054385e-01 7.11634815e-01 4.75683719e-01
-2.56274338e-03 -5.91271877e-01 -7.89489925e-01 -8.60079944e-01
2.04395026e-01 -4.14635092e-01 5.17282486e-01 1.05892026e+00
1.09173700e-01 1.40596911e-01 -3.90468806e-01 7.41271794e-01
7.19556451e-01 -2.50514627e-01 6.02799892e-01 -1.06019616e+00
-4.81150717e-01 -4.50740486e-01 -7.58763969e-01 -5.15554130e-01
6.41180933e-01 -1.49469197e+00 -7.75567293e-02 -1.16953552e+00
3.49399783e-02 -2.95825660e-01 -7.61689603e-01 5.46151459e-01
-4.82101023e-01 1.17649324e-01 -5.17381867e-03 -7.76140764e-02
-5.02098799e-01 3.82112294e-01 9.69622552e-01 -4.19811457e-01
-6.04200996e-02 3.90158057e-01 -1.13050807e+00 6.77279711e-01
8.41466606e-01 -5.73946476e-01 -7.31866360e-01 7.07799941e-02
3.13380688e-01 -6.17249943e-02 2.07883492e-01 -8.21420491e-01
1.93881214e-01 -3.22254337e-02 3.23219895e-01 3.29001844e-02
-1.04887918e-01 -7.11781979e-01 -3.25118415e-02 4.25975859e-01
-4.60334063e-01 -3.86201829e-01 -3.04981340e-02 1.15318990e+00
4.90751863e-02 -1.58286899e-01 8.53118062e-01 -4.42790724e-02
-3.83414716e-01 7.31875420e-01 -2.64149785e-01 8.45355690e-02
9.72612441e-01 2.27697417e-01 -3.46991390e-01 -4.89511698e-01
-8.19671631e-01 2.41835892e-01 2.59282678e-01 3.37041587e-01
6.24421954e-01 -1.24033511e+00 -7.62821853e-01 3.95589709e-01
8.70515481e-02 -2.49993518e-01 2.41359428e-01 5.71767986e-01
-9.31629986e-02 -4.04513627e-02 1.65641323e-01 -1.65827218e-02
-9.50676858e-01 9.79180813e-01 4.51064348e-01 -7.27453113e-01
-3.62900376e-01 9.73254681e-01 2.62737662e-01 -6.18754923e-01
3.45247597e-01 1.36502296e-01 -2.93359250e-01 -1.79876894e-01
2.60254383e-01 3.47147286e-01 1.19201086e-01 -5.33811212e-01
-2.77501911e-01 2.43949935e-01 -2.82688051e-01 2.70645916e-01
9.25711513e-01 -1.48653492e-01 -6.35058135e-02 2.65736468e-02
1.46263802e+00 2.29976311e-01 -1.08824050e+00 -2.38413543e-01
-5.86290890e-03 -1.87341750e-01 1.99197903e-01 -4.78125125e-01
-1.52041900e+00 5.29687047e-01 3.65972042e-01 3.22249472e-01
1.30003130e+00 -6.71461746e-02 4.83999133e-01 4.40096378e-01
1.98430955e-01 -6.70202434e-01 -3.51023562e-02 2.77800113e-01
7.97129214e-01 -1.17083430e+00 1.05698794e-01 -8.26648295e-01
-3.77166688e-01 9.19656038e-01 4.48357522e-01 -3.30249250e-01
8.44383061e-01 6.53685629e-02 -1.28329456e-01 -1.20555669e-01
-4.31692064e-01 -2.88477093e-01 4.70692843e-01 7.38386631e-01
5.49158640e-02 1.30109668e-01 4.85027917e-02 7.40893722e-01
2.42077908e-03 -6.85268521e-01 3.69136214e-01 6.42426550e-01
-3.33898887e-02 -1.05793369e+00 -6.19211458e-02 3.30196381e-01
-7.81435072e-01 -6.12285882e-02 -8.36714745e-01 6.76654518e-01
-3.86126906e-01 1.28531754e+00 -6.73073232e-01 -7.73833334e-01
2.36486793e-01 -1.67394802e-01 3.19996774e-01 -5.20095885e-01
-6.16675735e-01 -2.05697969e-01 -6.32098466e-02 -6.78820968e-01
4.05686200e-02 -1.96011394e-01 -1.03805351e+00 -2.63027638e-01
-6.84957206e-01 5.99423870e-02 4.67364728e-01 9.44210649e-01
1.18407972e-01 8.11517715e-01 1.31426466e+00 -5.29021204e-01
-7.83968687e-01 -8.67001891e-01 -6.93006337e-01 3.58174622e-01
1.66977286e-01 -3.92215967e-01 -9.44705486e-01 -3.53204936e-01] | [6.1593146324157715, 7.290466785430908] |
a5b0c5e2-4883-4f8b-a191-4b88e32fc89d | how-to-train-good-word-embeddings-for | null | null | https://aclanthology.org/W16-2922 | https://aclanthology.org/W16-2922.pdf | How to Train good Word Embeddings for Biomedical NLP | null | ['Gamal Crichton', 'Billy Chiu', 'Anna Korhonen', 'Sampo Pyysalo'] | 2016-08-01 | null | null | null | ws-2016-8 | ['learning-word-embeddings'] | ['methodology'] | [-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] | [-7.43160343170166, 3.6355607509613037] |
a09e669e-44b7-40ec-b23b-54d7d34ae860 | feature-reduction-for-machine-learning-on | 2101.05546 | null | https://arxiv.org/abs/2101.05546v1 | https://arxiv.org/pdf/2101.05546v1.pdf | Feature reduction for machine learning on molecular features: The GeneScore | We present the GeneScore, a concept of feature reduction for Machine Learning analysis of biomedical data. Using expert knowledge, the GeneScore integrates different molecular data types into a single score. We show that the GeneScore is superior to a binary matrix in the classification of cancer entities from SNV, Indel, CNV, gene fusion and gene expression data. The GeneScore is a straightforward way to facilitate state-of-the-art analysis, while making use of the available scientific knowledge on the nature of molecular data features used. | ['Sylvia Nürnberg', 'Frank Ueckert', 'Theresa Grooss', 'Anastasia Steshina', 'Alexander Denker'] | 2021-01-14 | null | null | null | null | ['art-analysis'] | ['computer-vision'] | [ 2.93036312e-01 -8.12916011e-02 -3.78562212e-01 -4.06610221e-01
-6.12379253e-01 -4.99657184e-01 2.90396750e-01 9.57600534e-01
-3.50151628e-01 9.50452566e-01 1.17025621e-01 -5.41615844e-01
-7.08390355e-01 -8.35084856e-01 -1.22991078e-01 -8.18408012e-01
-1.20176151e-01 4.51145023e-01 -1.01075031e-01 -1.61361307e-01
1.92239478e-01 6.90612793e-01 -1.51345515e+00 3.32719237e-01
7.80550003e-01 1.02073550e+00 -6.09651543e-02 7.23777413e-01
-2.57436097e-01 3.08554739e-01 -4.90248144e-01 -3.74498874e-01
-1.93138137e-01 -3.96858782e-01 -5.29089153e-01 -6.63558602e-01
-1.26343265e-01 5.41238427e-01 -1.37392476e-01 1.07940960e+00
6.90977991e-01 -3.56568307e-01 9.20982122e-01 -1.18574655e+00
-1.78851053e-01 5.66523492e-01 -2.54973590e-01 1.58738509e-01
6.20855451e-01 -2.80182660e-02 1.14024222e+00 -8.16434324e-01
1.09403265e+00 1.21195114e+00 8.10044050e-01 5.65500893e-02
-1.26461792e+00 -3.49113613e-01 -5.01818597e-01 2.49059111e-01
-1.64584613e+00 -9.56382155e-02 1.88534632e-01 -7.39412665e-01
1.19462764e+00 8.95546675e-01 7.83565283e-01 3.66975993e-01
6.64019704e-01 2.03653306e-01 9.08533454e-01 -6.90013587e-01
1.51989087e-01 -7.37031177e-02 5.03434360e-01 9.93794382e-01
6.95583463e-01 6.91225827e-02 -4.84128624e-01 -7.80745864e-01
-1.45076858e-02 1.85857341e-01 -7.29237450e-03 -5.77921197e-02
-1.17943466e+00 7.28944540e-01 7.12647662e-02 6.79075599e-01
-2.57514983e-01 -4.78935689e-02 5.78739643e-01 6.34607911e-01
2.30590746e-01 5.57495236e-01 -7.18116283e-01 -1.02451809e-01
-6.63856268e-01 -1.07773647e-01 6.01270020e-01 4.60590541e-01
5.77627122e-01 -4.73928660e-01 -1.22329675e-01 8.23265195e-01
3.77155244e-01 4.27959621e-01 8.01600099e-01 -4.95121568e-01
-2.54576117e-01 1.03237140e+00 -5.09203374e-01 -7.54761577e-01
-9.55924928e-01 -2.66834646e-01 -7.98326015e-01 -7.35753700e-02
-1.42372679e-02 -2.52433121e-02 -9.14747417e-01 1.40575540e+00
2.13328958e-01 -1.13202758e-01 6.98034232e-03 1.43398255e-01
1.15531051e+00 7.92197362e-02 1.37009948e-01 -2.14545906e-01
1.50725687e+00 -1.00841038e-01 -7.97001302e-01 5.35697162e-01
1.28914797e+00 -5.83301961e-01 2.23199815e-01 5.22637367e-01
-5.77984810e-01 -9.31753069e-02 -8.62167656e-01 -3.42815928e-02
-9.85353291e-01 1.48378015e-01 1.14132369e+00 9.96997893e-01
-1.02084017e+00 7.16925383e-01 -7.57041216e-01 -6.66285276e-01
4.59611058e-01 7.08052039e-01 -1.07487667e+00 -5.35891801e-02
-1.12845719e+00 1.06793296e+00 5.16118884e-01 -2.37718314e-01
-3.21657121e-01 -8.91996026e-01 -7.92633474e-01 5.18738292e-02
9.18575525e-02 -8.49136353e-01 4.48036194e-01 -6.69005513e-02
-1.06005096e+00 1.16180849e+00 -2.25351512e-01 -1.72378719e-01
-2.01055408e-02 4.02081698e-01 -6.16589606e-01 -1.38290627e-02
6.05037808e-02 2.32413873e-01 8.20761770e-02 -2.40705371e-01
-5.68729222e-01 -6.06282294e-01 -4.47460562e-01 -3.55014056e-01
-1.30774900e-01 4.31704372e-02 8.86647329e-02 -3.57446074e-01
3.69189084e-02 -9.12127793e-01 -4.42343891e-01 -7.89505523e-03
-6.62276566e-01 -3.96980256e-01 3.45542073e-01 -3.76002371e-01
1.21037459e+00 -2.06331062e+00 3.99144262e-01 7.15864301e-01
4.51824963e-01 1.91871971e-01 -4.28613946e-02 5.16871035e-01
-6.20172679e-01 5.19372523e-01 1.39677629e-01 3.70449752e-01
-1.77657977e-01 8.40780325e-03 5.95748425e-01 6.65393233e-01
2.80147195e-01 1.02944458e+00 -7.35548139e-01 -4.16313767e-01
2.20651746e-01 2.28636339e-01 -4.89286363e-01 -1.11841761e-01
2.51690656e-01 -1.50117040e-01 -5.97625434e-01 1.07317269e+00
5.25627792e-01 -3.31796199e-01 3.31299663e-01 -2.07584769e-01
-4.93409904e-03 -8.81559625e-02 -9.21447992e-01 1.65707564e+00
2.22842351e-01 5.27406633e-01 -7.51814470e-02 -8.32075596e-01
8.57703686e-01 3.29971373e-01 6.97410643e-01 -7.85216540e-02
3.63331318e-01 2.33059809e-01 4.90657330e-01 -5.50167382e-01
-9.66108367e-02 -2.79417425e-01 -8.90117362e-02 2.83638220e-02
5.15597403e-01 1.96291804e-01 5.15070736e-01 2.35867873e-01
1.75096977e+00 -6.26491368e-01 1.09308314e+00 -3.93106610e-01
6.87630773e-01 -3.12145799e-02 6.37426436e-01 6.07851088e-01
-8.32427889e-02 2.53476053e-01 9.84716356e-01 -3.16334724e-01
-7.19128132e-01 -8.67470980e-01 -5.43870389e-01 8.08764040e-01
-5.80272615e-01 -7.94173062e-01 -1.78940818e-01 -4.91411090e-01
5.73012173e-01 1.84826151e-01 -9.77924228e-01 -2.28752241e-01
1.05369352e-01 -1.37740850e+00 8.84145856e-01 5.52244820e-02
-5.17679274e-01 -3.82131696e-01 -1.04564354e-01 1.11451022e-01
3.03730696e-01 -5.42142093e-01 1.11911863e-01 7.91935861e-01
-6.99073553e-01 -1.62792337e+00 -2.35354766e-01 -4.11020786e-01
7.55145907e-01 -2.70720363e-01 8.05016518e-01 2.49111637e-01
-1.10706365e+00 3.57780159e-02 -4.37103838e-01 -5.89897513e-01
-5.71117342e-01 -4.84288447e-02 -2.65867598e-02 -4.09853578e-01
5.74949563e-01 -3.09838474e-01 -2.86544651e-01 1.21544331e-01
-8.62478197e-01 -3.49927813e-01 7.80033231e-01 1.15967512e+00
6.43385112e-01 9.39679369e-02 7.10129499e-01 -1.18335378e+00
5.90072930e-01 -7.19506800e-01 -3.11124772e-01 3.65198910e-01
-9.08427179e-01 1.55579641e-01 1.56608552e-01 1.82734951e-01
-4.93826479e-01 1.97976619e-01 -5.09456336e-01 1.91170305e-01
-3.19869787e-01 1.15319622e+00 -2.39046112e-01 -5.14914870e-01
5.91388583e-01 -6.47963509e-02 3.06263000e-01 -4.41766441e-01
2.85369903e-01 5.78289986e-01 2.51013517e-01 -1.12770416e-01
5.70197031e-02 4.80933934e-01 7.87663400e-01 -6.61566317e-01
-1.55119359e-01 -7.52344847e-01 -8.52470338e-01 1.73366785e-01
5.27104020e-01 -6.52001441e-01 -8.98768961e-01 2.23910555e-01
-7.88186550e-01 5.47252834e-01 -2.05960318e-01 6.06067538e-01
-2.24309668e-01 2.62370348e-01 -5.28912544e-01 -4.33609337e-01
-3.35130841e-01 -9.96656537e-01 8.95262241e-01 -6.49193004e-02
-4.96379614e-01 -9.25691247e-01 4.47014987e-01 1.96869392e-02
1.66426241e-01 4.86765921e-01 1.41090405e+00 -1.16284096e+00
1.80466667e-01 -7.76443899e-01 -3.69599350e-02 -1.26687080e-01
4.84223336e-01 3.76063704e-01 -7.85785377e-01 -8.66524503e-02
-4.67288375e-01 9.12534595e-02 1.09338939e+00 2.65180618e-01
1.13735259e+00 5.22920229e-02 -1.01788521e+00 5.92400730e-01
1.52501070e+00 4.25699174e-01 6.28326535e-01 6.50431886e-02
3.68665606e-01 6.87040687e-01 6.07326925e-01 4.16883588e-01
1.31138386e-02 5.19143045e-01 2.73823142e-01 -1.77775711e-01
1.62011027e-01 1.65301070e-01 -2.15176016e-01 4.17183876e-01
-3.52364071e-02 -2.13223681e-01 -1.09796989e+00 4.24525321e-01
-1.52842641e+00 -7.00262070e-01 -3.62790257e-01 1.94564652e+00
7.33110607e-01 -9.36341956e-02 -1.54504389e-01 3.51305872e-01
6.98547781e-01 -4.75832462e-01 -2.30593204e-01 -4.91917491e-01
-4.88510728e-01 1.84243858e-01 5.65718830e-01 1.55127030e-02
-7.90296614e-01 4.52314079e-01 8.24070740e+00 9.86473680e-01
-1.11936057e+00 -7.03205913e-02 4.93202358e-01 -8.69786441e-02
-3.05897713e-01 -1.72704637e-01 -6.77671134e-01 5.08642972e-01
1.24959612e+00 -5.19359648e-01 -2.34339520e-01 5.94435453e-01
1.48362353e-01 -2.00419530e-01 -1.33157194e+00 9.34089601e-01
3.20345443e-03 -1.60102046e+00 -1.81977302e-01 2.41896480e-01
4.05614108e-01 2.47717991e-01 -1.17712328e-02 -1.52482435e-01
1.04229555e-01 -1.13940549e+00 -8.05531293e-02 6.89653337e-01
1.02176905e+00 -5.48553884e-01 1.26967013e+00 -2.81095743e-01
-6.52317643e-01 2.31624227e-02 -2.74424672e-01 8.67161080e-02
7.09897839e-03 1.27268398e+00 -1.27065921e+00 7.17957914e-01
4.12480056e-01 7.66548276e-01 -8.40804935e-01 1.30501866e+00
1.79606825e-01 3.67412537e-01 -1.49492204e-01 -2.08093738e-03
-2.71515846e-01 1.97145447e-01 4.35683101e-01 1.35169184e+00
5.56443214e-01 1.60571814e-01 -1.92445472e-01 4.86953169e-01
2.06952929e-01 5.07417023e-01 -6.03687406e-01 -4.54726189e-01
3.60916436e-01 1.46062100e+00 -7.96454608e-01 -2.85990626e-01
-4.57489222e-01 3.67106646e-01 9.25917253e-02 -3.07498515e-01
-1.06338292e-01 -7.09718525e-01 8.97247195e-01 -1.51089013e-01
1.63099483e-01 5.26395738e-01 -2.13600308e-01 -8.14332604e-01
-5.24151504e-01 -6.63587928e-01 7.63998866e-01 -5.47139108e-01
-1.37402153e+00 2.71788388e-01 -2.59032428e-01 -1.01406670e+00
-5.45375586e-01 -1.14151502e+00 -3.57093424e-01 7.23223150e-01
-1.26186454e+00 -7.10439384e-01 1.31457234e-02 1.47591561e-01
-4.24104810e-01 -4.82509792e-01 1.44045222e+00 2.62454569e-01
-6.76827550e-01 3.90861362e-01 5.39095700e-01 -1.09009005e-01
8.77697170e-01 -1.25836861e+00 6.85419366e-02 -6.96609449e-03
-3.24849784e-01 8.12635362e-01 6.45038068e-01 -8.25490952e-01
-1.57699347e+00 -8.66218925e-01 8.83186400e-01 -3.17868531e-01
8.22675169e-01 -5.33996802e-03 -5.32973170e-01 4.06370968e-01
-1.99693665e-01 2.34073728e-01 1.80226016e+00 5.79093158e-01
-2.82451451e-01 -9.11210999e-02 -1.52890444e+00 7.59414583e-02
7.20658422e-01 -3.99676114e-01 -6.02500260e-01 3.27818602e-01
3.40147316e-01 -6.82619661e-02 -1.52477312e+00 4.79169399e-01
8.92689645e-01 -6.01561785e-01 9.38575685e-01 -9.68020856e-01
5.34820184e-02 -2.59792417e-01 -1.05589949e-01 -1.58004236e+00
-6.00924253e-01 -3.52319300e-01 3.70468378e-01 8.73210430e-01
8.69623482e-01 -8.08165550e-01 6.84182882e-01 4.44113821e-01
-1.61098912e-01 -7.28844881e-01 -1.19558561e+00 -5.74823916e-01
-7.41222426e-02 -1.12339899e-01 7.38434553e-01 1.01956999e+00
8.41240168e-01 -3.16615589e-02 2.45583989e-02 -3.80413353e-01
2.11011082e-01 -9.13581029e-02 2.95343429e-01 -1.92252159e+00
-5.88665232e-02 -6.69682145e-01 -1.59629703e+00 3.74223560e-01
1.64453425e-02 -1.27360547e+00 -4.92527127e-01 -1.29751301e+00
4.97394681e-01 -2.96615958e-01 -8.71442735e-01 5.97526133e-01
-9.55704674e-02 1.65100202e-01 -1.72546774e-01 -3.83883744e-01
-4.69697386e-01 -1.05347574e-01 7.36809850e-01 -2.72045076e-01
1.97905436e-01 -2.79073834e-01 -1.09801280e+00 7.07021117e-01
5.47482908e-01 -7.91132867e-01 1.10958725e-01 4.39288527e-01
6.48538888e-01 7.30755404e-02 -1.58670153e-02 -5.78579903e-01
3.57646912e-01 -1.61293596e-01 5.91558278e-01 -7.72012115e-01
2.37121791e-01 -6.11487627e-01 7.25226045e-01 9.64918137e-01
-3.77245545e-01 7.43952319e-02 2.50405431e-01 5.77561200e-01
-3.67325127e-01 -3.19541216e-01 3.03647190e-01 -1.01802506e-01
-4.73238945e-01 -4.30875756e-02 -4.86685634e-01 -4.81920272e-01
1.13293004e+00 -2.14390546e-01 -6.87642753e-01 1.34181708e-01
-1.04892468e+00 8.06667656e-02 3.29344928e-01 1.80179134e-01
4.57865924e-01 -1.27566338e+00 -9.07504797e-01 2.55443782e-01
5.14893889e-01 -7.86458254e-01 3.45131904e-01 1.07947934e+00
-5.14701545e-01 1.04234993e+00 -5.20736814e-01 -5.32987058e-01
-1.91996634e+00 4.99328375e-01 7.89796636e-02 -2.70115554e-01
5.25933988e-02 8.15999329e-01 -6.49205595e-02 -3.50322485e-01
-1.40467957e-01 -1.79392189e-01 -5.06040990e-01 5.20282865e-01
5.60009003e-01 6.55773580e-01 5.86391687e-01 -4.01348174e-01
-8.83631647e-01 4.34324503e-01 -8.03334340e-02 2.74464041e-01
1.40715313e+00 1.35316387e-01 -8.41081917e-01 4.53496516e-01
1.37204278e+00 1.14281252e-01 1.74745232e-01 1.65788323e-01
1.62732229e-01 -3.98597151e-01 -1.62510145e-02 -1.03805101e+00
-7.31235445e-01 4.52003270e-01 7.64912963e-01 9.05138776e-02
9.05848980e-01 5.31675406e-02 -1.19499236e-01 4.77106243e-01
2.46592328e-01 -9.57451820e-01 -7.07578599e-01 2.86748976e-01
4.65142995e-01 -1.15274858e+00 3.63324463e-01 -8.74782622e-01
-6.00933805e-02 1.17390037e+00 -5.14571108e-02 3.35836887e-01
1.04286218e+00 5.91852784e-01 -2.28495881e-01 -4.26081568e-01
-1.22504771e+00 -2.42518902e-01 4.20878023e-01 7.63655305e-01
8.36421371e-01 4.44486648e-01 -1.04253364e+00 8.65763843e-01
-2.17222288e-01 1.43068716e-01 4.31049198e-01 8.10444176e-01
-4.74697798e-01 -1.55886400e+00 -2.84157455e-01 1.23968017e+00
-8.99829805e-01 -2.24649772e-01 -8.02948654e-01 6.34181678e-01
3.75715673e-01 6.71699464e-01 -1.54476270e-01 -4.66300458e-01
2.11957842e-01 1.64147526e-01 3.96373868e-01 -7.42784381e-01
-6.39339149e-01 1.70289557e-02 2.53778964e-01 -4.78072971e-01
-2.54625887e-01 -9.47250724e-01 -1.05163515e+00 -1.57274231e-01
-6.10969901e-01 3.11844736e-01 9.49374616e-01 8.22794139e-01
7.04438925e-01 7.80923665e-01 3.61525148e-01 -1.25185356e-01
-2.03538015e-01 -8.16954494e-01 -9.02718961e-01 1.74457014e-01
1.79821119e-01 -7.61069298e-01 -3.43627632e-01 -2.70680577e-01] | [6.228966236114502, 5.629906177520752] |
2c43d9ca-282e-4305-9d93-3ac27b8115b7 | top-down-beats-bottom-up-in-3d-instance | 2302.02871 | null | https://arxiv.org/abs/2302.02871v3 | https://arxiv.org/pdf/2302.02871v3.pdf | Top-Down Beats Bottom-Up in 3D Instance Segmentation | Most 3D instance segmentation methods exploit a bottom-up strategy, typically including resource-exhaustive post-processing. For point grouping, bottom-up methods rely on prior assumptions about the objects in the form of hyperparameters, which are domain-specific and need to be carefully tuned. On the contrary, we address 3D instance segmentation with a TD3D: top-down, fully data-driven, simple approach trained in an end-to-end manner. With its straightforward fully-convolutional pipeline, it performs surprisingly well on the standard benchmarks: ScanNet v2, its extension ScanNet200, and S3DIS. Besides, our method is much faster on inference than the current state-of-the-art grouping-based approaches. Code is available at https://github.com/SamsungLabs/td3d . | ['Anton Konushin', 'Anna Vorontsova', 'Danila Rukhovich', 'Maksim Kolodiazhnyi'] | 2023-02-06 | null | null | null | null | ['3d-instance-segmentation-1'] | ['computer-vision'] | [ 0.03941243 0.20798756 -0.1640212 -0.5384459 -0.8457157 -0.7651859
0.6854952 0.02019189 -0.37431064 0.05815756 -0.19156608 -0.6183239
0.22254954 -0.7372163 -0.9383805 -0.2921008 0.05501155 0.9402181
0.848351 -0.1790176 0.38968366 0.43219736 -1.4942499 -0.07035264
0.70893484 1.2157217 0.0530939 0.64775205 -0.27950788 0.06901265
-0.10481584 -0.26920754 0.5611508 -0.19183454 -0.8833693 0.31013736
0.56093 -0.37145177 -0.07600506 0.86651194 0.53946275 -0.04668349
0.6112417 -0.8169908 -0.25112414 0.4217532 -0.8963256 -0.04008975
0.06939984 0.49190283 0.9537446 -1.1486022 0.5290531 1.1639485
0.9281609 0.23382813 -1.3765881 -0.38740453 0.41996813 -0.06744403
-1.2625177 -0.13216166 0.72926366 -0.5060586 1.011456 -0.05520958
0.68842 0.79974174 -0.41836208 1.0731022 0.9490659 -0.30566534
0.40062273 -0.43004027 0.23326023 0.7470456 0.33146307 -0.13633654
-0.30084273 0.04670507 1.0175987 -0.19368373 0.09803493 -0.8351397
-1.2987864 0.6410691 0.54789126 0.0754088 -0.42602777 0.24329466
0.45269543 -0.0057015 0.77245075 0.2776669 -0.89176965 -0.08919127
-1.2404722 0.40811402 0.7445129 1.2084857 1.0036417 -0.4376755
-0.06663396 0.63877195 0.34476173 0.2814254 -0.03096562 -0.94552946
0.38872716 0.7166661 0.03558894 -0.46319807 -0.5840567 -0.4363183
-0.36569571 0.35839194 0.70051324 -0.03726864 -1.5316046 1.2082454
0.7001629 0.32285684 -0.40034318 0.9430127 0.8107767 0.17842518
-0.06298679 0.55229163 1.3854752 -1.2390387 -0.07937149 -0.3094747
0.60241467 -0.6947472 1.3509957 0.39468375 -1.3330076 -0.50317943
-0.92529523 -0.5156073 -0.42838144 0.05842188 0.80592304 0.517299
-0.8901522 0.48248142 -1.1686417 -0.3585336 0.8803214 0.27450448
-0.19929697 -0.10440096 -0.6161129 0.5046542 0.44977343 0.11026239
-0.96074426 -1.1168052 -0.92940706 -0.25453454 0.8581313 -0.9711954
1.6823176 -0.43934605 -1.5332763 1.2732863 -0.03044644 -0.39133912
0.8724137 -0.5379713 0.4633046 0.29612195 0.1286784 0.89102453
0.68792135 -1.3221722 -0.5468771 -0.39046675 0.24899499 0.05719183
0.23359472 -0.25607517 -1.1787233 -0.5040626 0.3473204 -0.95131826
-0.5181252 0.18761636 -0.7438127 -0.32875293 0.7526296 -0.2616805
0.8076798 -2.1570916 0.02051361 0.16285978 0.15991019 0.21836326
0.01387083 0.31918496 0.17833808 0.26274505 -0.7023972 -0.72707236
0.41646108 0.05770389 0.15185459 0.5140379 0.42943975 1.095764
-0.8163589 -0.50552094 0.67521626 0.44887695 -0.76736677 -0.04768887
-0.7340376 0.4858549 -0.56240904 0.9265118 0.7983259 -0.44725403
-0.2755931 -0.28330937 -0.3288328 0.57600605 -1.0531011 2.4144866
-0.4186798 0.38806263 0.04260078 -0.80431765 0.5420734 -0.06833562
0.41450763 -0.42092574 0.23465568 0.44735274 -0.21454747 -0.2033582
0.4641 0.26461592 -0.10411253 0.27840218 0.12510614 -0.5495994
0.26212487 0.24215664 1.0519775 0.7512648 0.17576483 -0.2748794
0.15661293 0.4446995 0.37264928 0.90099865 0.14284073 1.1520176
0.5072816 -0.24058802 -1.0696024 -0.9621036 -0.40235117 0.8077451
0.33374983 -0.3368685 -0.96901596 -0.8108217 0.17769466 0.65085447
-0.73933977 0.2488896 -0.5333459 -0.6262829 0.49690896 0.7382357
0.67668617 -0.76850814 -0.64075154 0.34196964 0.41147408 -1.2682987
-0.39954364 0.37588096 -1.0643065 -1.0724069 -0.79877007 -0.41405264
0.7099042 0.15964127 1.4674822 0.10910802 -0.3554529 0.26259425
-0.45270678 -0.5384107 0.2165888 0.57589644 -0.58292836 -0.39780274
0.4316538 -0.616762 -1.0084493 0.32559523 -0.638297 0.15327325
0.8213714 0.33675572 1.0334451 -0.2249543 0.11696877 -1.1268405
0.02636047 -0.37859035 -0.67692304 -0.10920218 -0.2573052 -0.07883511
0.29684722 -0.11068608 -0.7463298 0.33717015 -0.44348636 -0.508181
-0.6995738 0.22225307 -0.25884876 -0.02283249 0.46270517 0.03029195
-0.1964386 -0.895256 0.6031374 0.26836875 0.57417715 -0.62324643
1.0064052 0.6719728 -0.01923479 -0.77135813 -1.0433685 -0.661034
-1.151231 0.0519496 1.0962291 -0.86709857 -0.34330457 0.7321659
-1.0198027 -0.9851468 -0.2864426 0.24069656 -0.68399024 0.14704016
-0.51523477 -0.34605935 -0.17415453 -1.1423774 1.6241015 0.06013419
-0.00619678 -0.9879585 -0.17794211 0.41589203 0.14072534 0.46149066
0.77683055 -0.76948607 -0.86812264 -0.19926457 -0.34364206 0.2620857
-0.04367306 0.09303216 -0.9649168 0.17375927 -0.44500184 -0.1389083
1.1086339 0.55832696 1.5216956 0.18405369 -0.4192078 0.93236494
1.3782291 -0.29490164 0.43558717 0.47174412 0.9957765 0.39371964
0.6906476 0.27524534 0.8108969 0.60816723 0.7639891 -0.463476
-0.22337694 -0.20867732 -0.2924179 0.21896349 -0.15500127 -0.21485412
-1.0683271 0.6030586 -1.871969 -0.6018692 -0.4897025 2.0301852
0.7638388 0.5703324 0.3740014 0.1564756 0.39547384 0.2791797
-0.90115863 -0.08945101 0.1812163 0.2353073 0.7321482 0.39714557
-1.3345253 1.2422392 5.613786 0.82324916 -0.97976005 -0.03832458
0.6690822 -0.19768332 -0.30082884 0.11499747 -0.8009396 0.30283093
0.50608164 0.33977196 0.11013907 0.86004186 0.32802847 -0.3208256
-1.1760145 0.7848867 -0.35387167 -1.3412731 -0.25515845 0.2350708
0.64550614 0.45515084 -0.06173616 0.18240362 0.40059963 -0.6433071
1.0267427 0.2803192 0.62538695 -0.7231869 0.49555895 0.2709723
-1.1541309 0.40834448 -0.2753298 0.17132407 0.41817802 1.0019
-0.59965837 0.5731647 0.8499168 0.77319765 -0.7070882 1.0422534
-0.38122398 0.8566978 -0.6276071 0.4027082 0.7052335 -0.03256581
0.38484925 1.4673295 0.08760796 0.14728002 0.36647224 0.87335885
-0.12792715 -0.22287777 -0.38797715 0.18672678 0.43267334 1.3629602
-1.1407663 -0.40631914 -0.59747875 0.77320915 0.13057421 0.13582456
-0.84725183 -0.33991927 0.5410328 0.4393906 0.89935684 -0.5598517
-0.6457226 -0.980824 0.0365885 -0.54107744 0.16569975 -0.50517166
-1.3531795 0.34085655 0.21425252 -1.0897894 -0.0802551 -0.79196614
-0.71259695 0.5489444 -1.7908789 -1.2587333 -0.38771474 0.5391283
0.8056114 0.3085882 0.27993664 0.29231894 -0.55144864 0.36704755
-0.37079948 0.20533426 0.42821917 -1.6266491 1.003152 0.83076096
-0.01781762 0.4755101 0.52638996 -0.5196722 -1.2779528 -1.1544377
0.47084495 -0.5010963 0.6132818 -0.6063154 -0.8455678 0.6376142
0.08911071 0.27080947 0.43487823 0.27146804 -0.29561782 0.2910858
-0.96491796 0.53906304 1.4931362 -0.23192531 -0.4374581 0.38847405
0.8514753 -0.85503834 -0.86004376 0.51448476 0.40192086 -1.1510508
1.148678 -0.20457327 0.5121177 -0.5159178 -0.02989566 -0.90099305
-0.17026372 -0.52926403 -0.07886066 1.0200988 0.6234801 -0.4808976
0.96026725 0.49614006 -0.69499415 -1.0796697 -0.60963434 -0.8300958
0.16998431 -0.77312917 0.8009591 0.6248141 -0.71747994 0.1587572
0.19411142 0.333552 0.8244343 0.1510519 1.0561078 -1.3049163
-0.18196157 -0.7418848 -0.23466931 -1.4967432 -0.07780562 -0.71784747
0.14326264 -1.8130121 -0.2240705 -0.6441973 -0.05542175 0.6285573
-0.23251231 0.5158582 -0.05090318 0.12808186 -0.7696573 0.35475576
1.294051 0.12518966 -0.41012618 0.10234677 -0.65029645 1.0672104
0.9741733 -0.43037346 -0.34105334 -0.7248607 0.03258783 -0.30211717
0.7168901 -0.9121158 0.2053329 0.07050055 0.24479507 -1.2910343
0.36926296 -0.7135969 -0.25913528 0.1912602 -0.15110877 -0.09643294
0.29584962 0.38938615 0.0210209 -0.20370871 0.72846365 -0.3339159
-0.8023015 0.76920605 0.09801723 0.21906993 0.9829625 -0.5782603
-0.12101009 0.13796972 -0.6333041 0.318122 0.680597 0.28162915
0.27150518 -0.8681523 -0.6412445 0.05794982 0.10133692 1.0163008
0.19073167 1.0186211 -0.7544512 0.2201969 0.22351654 -1.0633904
-0.81309813 0.32399383 0.30922613 -0.04057369 -1.0333134 0.8839036
0.26677817 -0.53976005 0.1529333 -0.5760075 0.29053319 -0.10245085
0.20471732 0.18231203 0.24480319 -0.24661067 -0.59095836 0.77345276
-0.16385376 -0.12261254 1.5486536 -0.03446532 0.17481822 0.3820616
0.9919376 -0.18307911 -1.7252917 -0.18670528 0.11770137 -0.43807653
0.23176518 -0.76281756 -1.141135 0.93741673 0.0562888 0.21075481
0.9303203 0.38605195 0.8523592 0.27990893 0.3518386 -0.9391075
-0.10768567 0.45784825 0.6658522 -1.3701504 0.08983049 -0.53509516
-0.34024447 0.95947486 0.5860988 -0.43731463 0.7834309 0.22325063
-0.03945557 -0.43330836 -0.59643275 -0.4927288 0.42248052 0.43937212
0.3336981 -0.22716343 0.01309502 0.27296135 -0.2480765 -0.05450966
0.17817889 0.9999427 -0.43188083 -1.058474 -0.21301264 0.36232108
-0.2827281 -0.07532904 -0.3606489 1.2814618 0.13186924 0.6056258
0.26950294 -0.06312688 0.63321906 -0.17516968 0.5292786 -0.8634815
-0.5247111 0.33217755 0.04845265 -0.8971773 -0.5634661 -1.0313765
-1.468246 -0.17522258 -0.1232336 -0.43856052 0.8167698 0.91303194
0.5654246 0.479764 0.3010596 -1.5239947 -0.19738519 -0.8096747
-0.2563061 0.25854856 0.20958818 -0.72934324 -0.2848659 -0.10802234] | [8.093801498413086, -3.096825122833252] |
7b4ca608-345d-4a31-a086-91de41744b88 | stochastic-talking-face-generation-using | 2011.10727 | null | https://arxiv.org/abs/2011.10727v1 | https://arxiv.org/pdf/2011.10727v1.pdf | Stochastic Talking Face Generation Using Latent Distribution Matching | The ability to envisage the visual of a talking face based just on hearing a voice is a unique human capability. There have been a number of works that have solved for this ability recently. We differ from these approaches by enabling a variety of talking face generations based on single audio input. Indeed, just having the ability to generate a single talking face would make a system almost robotic in nature. In contrast, our unsupervised stochastic audio-to-video generation model allows for diverse generations from a single audio input. Particularly, we present an unsupervised stochastic audio-to-video generation model that can capture multiple modes of the video distribution. We ensure that all the diverse generations are plausible. We do so through a principled multi-modal variational autoencoder framework. We demonstrate its efficacy on the challenging LRW and GRID datasets and demonstrate performance better than the baseline, while having the ability to generate multiple diverse lip synchronized videos. | ['Rajesh M Hegde', 'Vinay P Namboodiri', 'Ashish Sardana', 'Ravindra Yadav'] | 2020-11-21 | null | null | null | null | ['talking-face-generation'] | ['computer-vision'] | [ 3.13216448e-01 3.78890246e-01 2.08378151e-01 -3.83218378e-01
-1.25635552e+00 -6.38145268e-01 9.10342336e-01 -9.19690430e-01
1.60136744e-01 5.97600341e-01 4.33894873e-01 4.16908115e-02
2.02649027e-01 -3.93971086e-01 -8.49693000e-01 -7.74711967e-01
2.25345328e-01 5.25215507e-01 -1.34459749e-01 -9.34644639e-02
-1.66624829e-01 1.23804577e-01 -2.21613812e+00 6.14074290e-01
4.43570524e-01 8.39535117e-01 4.45658922e-01 1.20569623e+00
8.47447813e-02 7.14469850e-01 -6.72808290e-01 -4.68248367e-01
1.63425341e-01 -6.41300857e-01 -4.74593252e-01 3.87397468e-01
6.12109184e-01 -5.98533332e-01 -1.99025795e-01 6.80079460e-01
8.53453159e-01 7.28639364e-02 8.71536911e-01 -1.60699391e+00
-6.60794556e-01 5.99250317e-01 -2.13256508e-01 -2.96128303e-01
8.38254869e-01 4.52578902e-01 1.08003008e+00 -8.79366398e-01
7.55907059e-01 1.53299582e+00 5.59322894e-01 1.06075525e+00
-1.32178080e+00 -7.14902401e-01 -2.12829918e-01 -2.04320759e-01
-1.55096364e+00 -1.03438735e+00 6.99973464e-01 -5.61778307e-01
6.46660447e-01 9.83134285e-02 6.56616688e-01 1.63932490e+00
9.10445675e-02 6.73328638e-01 7.18633056e-01 -3.53720218e-01
8.90737548e-02 1.33203551e-01 -8.32113385e-01 4.74894494e-01
-4.05065656e-01 7.83632323e-02 -9.58267629e-01 -2.64435917e-01
6.61409438e-01 -4.78143156e-01 -3.15001070e-01 -2.04294190e-01
-1.04275370e+00 8.89409423e-01 -3.00146997e-01 8.67984071e-03
-2.68409282e-01 2.73083508e-01 -6.19352683e-02 2.18339972e-02
2.10174441e-01 2.15999886e-01 -2.06546560e-01 -3.86151135e-01
-1.35918653e+00 4.54715639e-01 8.75532627e-01 1.01401520e+00
4.27490890e-01 3.97134572e-01 -3.33286524e-01 7.13070452e-01
4.82058913e-01 6.21218443e-01 6.50247633e-01 -1.74985993e+00
-6.03491999e-03 -2.01914147e-01 1.51582226e-01 -6.18157983e-01
7.44916946e-02 1.57766625e-01 -5.17994702e-01 3.33183289e-01
4.16128486e-01 -4.45896208e-01 -9.04223859e-01 2.10152030e+00
2.82591552e-01 4.48981136e-01 1.50530517e-01 6.07186615e-01
8.04531455e-01 8.81199777e-01 -3.31447631e-01 -5.16014457e-01
9.63983297e-01 -6.80303812e-01 -7.40121722e-01 2.16768011e-01
-1.71449974e-01 -1.01905346e+00 1.12876475e+00 6.45088255e-01
-1.38443172e+00 -7.23464549e-01 -7.70864129e-01 -3.40834893e-02
1.15128711e-01 -8.62762257e-02 5.44280112e-01 8.72770250e-01
-1.51634359e+00 3.04720134e-01 -6.78981364e-01 -1.60334185e-01
1.94370925e-01 2.15654656e-01 -4.15083110e-01 1.57585710e-01
-9.85304296e-01 4.23791677e-01 6.34454116e-02 -1.38422832e-01
-1.19015920e+00 -8.31735969e-01 -8.54532123e-01 2.26236284e-02
3.37026380e-02 -8.55191708e-01 1.66364217e+00 -1.06738710e+00
-1.90777457e+00 5.36080778e-01 -5.56749105e-01 -3.12040210e-01
6.67680085e-01 6.78041130e-02 -1.11390665e-01 4.38189954e-01
-1.33023068e-01 1.34215999e+00 1.30653942e+00 -1.48423541e+00
-3.34990174e-01 1.86345547e-01 -1.30260423e-01 1.72097161e-01
-3.44653159e-01 -5.15060388e-02 -4.33403879e-01 -5.37887692e-01
-4.23087955e-01 -1.05476737e+00 3.32729578e-01 7.62816742e-02
-2.79118299e-01 -2.98420757e-01 8.59866858e-01 -4.81512159e-01
8.40392232e-01 -2.06892180e+00 2.96972364e-01 -3.87832731e-01
2.92481966e-02 -1.65683433e-01 -2.81123757e-01 4.76332664e-01
9.60706994e-02 3.73231083e-01 -7.71027878e-02 -8.63754213e-01
2.28528202e-01 1.27643868e-01 -5.71322799e-01 6.71487674e-02
3.08489025e-01 6.76060438e-01 -6.72858238e-01 -7.51392007e-01
1.63247231e-02 1.07404459e+00 -1.01539934e+00 3.37376028e-01
-4.38043207e-01 5.02888262e-01 3.36174890e-02 5.49587429e-01
4.41345870e-01 2.56777741e-02 2.81611025e-01 -1.08155096e-02
2.07916781e-01 -1.15415163e-01 -1.15059257e+00 1.70090795e+00
-5.43144405e-01 8.21019351e-01 2.78456450e-01 -4.61384952e-01
5.00884533e-01 9.00674045e-01 5.90950489e-01 5.28214015e-02
-4.68537509e-02 4.53530550e-02 -5.60305528e-02 -4.10000831e-01
4.12844598e-01 -5.91620803e-01 9.11048278e-02 6.06409431e-01
5.40991127e-01 -7.82560766e-01 1.33283332e-01 3.24772060e-01
7.83364773e-01 3.83637726e-01 -2.84654021e-01 9.83638763e-02
1.59414753e-01 -5.25384009e-01 4.92557913e-01 7.03627467e-01
-1.78833932e-01 1.21758389e+00 4.55385953e-01 3.48809063e-02
-1.22167230e+00 -1.56796229e+00 1.07423730e-01 8.28717947e-01
-2.12264970e-01 -4.17906135e-01 -9.61418867e-01 -1.95323959e-01
-1.82644606e-01 8.02878141e-01 -3.74978960e-01 1.34396225e-01
-1.28091544e-01 -3.56661826e-01 7.43953049e-01 2.40480065e-01
-2.76331082e-02 -1.20912623e+00 -3.22289109e-01 2.54342593e-02
-6.30476892e-01 -1.22928584e+00 -6.54772222e-01 -4.11869824e-01
-3.35777313e-01 -6.71432078e-01 -8.46629858e-01 -7.49133110e-01
2.42516100e-01 -1.12906210e-01 1.20625198e+00 -1.83230832e-01
-4.97690231e-01 8.64628911e-01 -6.99684694e-02 -4.53106970e-01
-7.15946019e-01 -2.34188288e-01 5.31288266e-01 1.50369823e-01
-6.34754151e-02 -7.56258667e-01 -3.65040213e-01 4.25937325e-02
-9.82079923e-01 1.20832354e-01 1.62530586e-01 7.87175417e-01
5.66296816e-01 1.81318745e-01 7.45862782e-01 -3.79918844e-01
7.62227297e-01 -5.40547609e-01 -2.97723562e-01 1.49797127e-02
-1.50247496e-02 -1.31586511e-02 6.03241146e-01 -7.63063490e-01
-1.41547418e+00 2.45901689e-01 -2.64668822e-01 -8.41631889e-01
-3.13659728e-01 -3.78416255e-02 -2.96034962e-01 3.85326535e-01
5.09808183e-01 4.09198850e-01 2.31617987e-01 -6.52182847e-02
6.74693823e-01 7.79661953e-01 7.28278160e-01 -7.09510267e-01
9.22450244e-01 5.32296419e-01 -3.02872539e-01 -1.20211995e+00
-4.05819654e-01 3.05012357e-03 -4.27906632e-01 -5.69743454e-01
9.27218735e-01 -9.89840925e-01 -9.62172389e-01 5.82882345e-01
-1.19015956e+00 -6.26693785e-01 -3.07142228e-01 3.67864251e-01
-1.15837419e+00 2.28900641e-01 -5.27722716e-01 -1.22195172e+00
3.95396575e-02 -1.28079736e+00 1.36389792e+00 2.83557206e-01
-4.99529898e-01 -7.31707990e-01 5.53111322e-02 4.35728937e-01
3.29169154e-01 2.53131390e-01 3.79167229e-01 -9.34026837e-02
-5.99572062e-01 1.29425138e-01 4.55765277e-01 3.33597124e-01
3.11304212e-01 6.29876196e-01 -1.28808284e+00 -3.54641229e-01
5.72645552e-02 -7.44275212e-01 7.12354481e-01 5.64006746e-01
9.56566751e-01 -4.15417641e-01 2.44837478e-02 5.77770352e-01
1.07340777e+00 5.88213131e-02 5.35818577e-01 -5.02307177e-01
4.04703289e-01 7.50261784e-01 1.23358689e-01 5.29231071e-01
4.66078639e-01 6.15019202e-01 2.52608687e-01 3.20025951e-01
-3.88136178e-01 -5.30330658e-01 7.35999882e-01 7.05419362e-01
-4.02470455e-02 -5.71402848e-01 -5.98885059e-01 6.91290438e-01
-1.36710489e+00 -1.53502691e+00 4.19729322e-01 2.05558872e+00
9.24011290e-01 -1.80038705e-01 3.43346655e-01 3.77609394e-02
7.25438297e-01 6.15615547e-02 -4.68804359e-01 -3.87550712e-01
-2.43120342e-01 2.13979200e-01 -2.82749444e-01 7.81072974e-01
-8.11388016e-01 9.00658667e-01 7.47525597e+00 8.00077677e-01
-1.08420610e+00 -1.01359434e-01 6.47869229e-01 -6.30167425e-01
-7.13486314e-01 -2.63324827e-01 -8.06039870e-01 6.20631576e-01
1.11207008e+00 -2.30417907e-01 7.39713252e-01 5.23398757e-01
4.04974252e-01 1.11299334e-03 -1.21051097e+00 1.09435058e+00
4.22821611e-01 -1.19872332e+00 3.68157804e-01 1.02473758e-01
8.18047941e-01 -2.77562082e-01 6.07980907e-01 8.42435136e-02
4.75959569e-01 -1.40740466e+00 1.05529165e+00 6.89186990e-01
1.08741295e+00 -7.51014352e-01 8.99229944e-02 5.17704964e-01
-1.08132470e+00 -1.54004321e-01 -1.10015891e-01 1.64401099e-01
4.60791707e-01 2.88833052e-01 -8.83731186e-01 1.72536403e-01
7.30341256e-01 1.33534238e-01 -2.48093233e-01 5.37436306e-01
-3.18483934e-02 6.20922089e-01 -4.77201641e-01 3.34314406e-01
-2.66684622e-01 7.93571621e-02 6.66273236e-01 8.98074567e-01
8.54482114e-01 -2.57087778e-02 3.42270471e-02 9.61129904e-01
-2.13730082e-01 -2.09856808e-01 -7.83086479e-01 -2.42447689e-01
5.70587635e-01 1.07975495e+00 -2.49494359e-01 -1.17345199e-01
-2.85919070e-01 1.15760827e+00 -9.17538628e-02 3.83058786e-01
-9.14920449e-01 -5.20451255e-02 7.76806653e-01 5.31137735e-02
5.16470909e-01 -2.55053759e-01 2.07694829e-01 -1.19374490e+00
-2.96792854e-03 -1.05424488e+00 -3.24867629e-02 -1.29567206e+00
-1.17750514e+00 7.16471136e-01 -4.51627001e-02 -9.03896987e-01
-1.05118370e+00 -2.68429101e-01 -5.97071528e-01 8.23166490e-01
-1.18077314e+00 -1.38921475e+00 -1.35200113e-01 7.66632915e-01
9.45276201e-01 -3.32921952e-01 8.99503291e-01 2.19760323e-03
-8.36317614e-02 7.16077745e-01 -2.87335753e-01 -2.40631059e-01
9.74160671e-01 -1.07305944e+00 1.91500202e-01 6.29937232e-01
3.91024947e-01 6.58711314e-01 9.56423938e-01 -3.95038664e-01
-1.52128804e+00 -7.25342929e-01 8.61472070e-01 -8.37841153e-01
3.70755017e-01 -5.55279970e-01 -5.21524131e-01 7.15441048e-01
5.43782890e-01 -2.78502524e-01 7.98463583e-01 -2.51893610e-01
-2.41523251e-01 1.02249540e-01 -1.20213258e+00 6.62415206e-01
9.15229321e-01 -7.36495912e-01 -4.33170229e-01 9.53304842e-02
7.49373078e-01 -1.90323412e-01 -8.13495040e-01 1.41550347e-01
9.40419137e-01 -1.23759389e+00 1.01663601e+00 -2.47102439e-01
6.88529253e-01 -3.45340908e-01 -3.32482219e-01 -1.33210409e+00
7.82910287e-02 -1.31287742e+00 -1.71805620e-01 1.67716527e+00
4.01933908e-01 -3.52446675e-01 6.61237836e-01 5.09214342e-01
6.76622540e-02 -4.37078148e-01 -8.74598503e-01 -6.19926810e-01
2.84563422e-01 -6.20599866e-01 7.02808201e-01 4.66657728e-01
-1.69346333e-01 1.39514506e-01 -6.97256565e-01 1.56460971e-01
6.98942780e-01 -7.54230714e-04 9.46915746e-01 -1.16723716e+00
-6.56592190e-01 -3.28087687e-01 -1.62861615e-01 -9.99133050e-01
4.84183818e-01 -6.43072486e-01 4.16439861e-01 -1.24292433e+00
3.26412797e-01 2.82899700e-02 1.91393048e-01 1.89709231e-01
-2.52967365e-02 4.62210983e-01 3.39837134e-01 1.59051046e-01
-1.97990462e-01 6.05441809e-01 1.13898373e+00 -5.22942729e-02
-1.76293924e-01 -1.54222637e-01 -8.15880358e-01 7.96440065e-01
6.52584016e-01 -1.68348387e-01 -7.54957199e-01 -4.27133679e-01
7.96211734e-02 4.29926932e-01 3.45315427e-01 -1.06615877e+00
2.38493741e-01 -9.87645760e-02 4.65300620e-01 -2.92366028e-01
9.51984644e-01 -3.07346523e-01 4.54312295e-01 -2.72478648e-02
-3.52459967e-01 -2.28743523e-01 1.10839903e-01 6.04275584e-01
-1.95454761e-01 1.75508991e-01 8.12323511e-01 -1.48261502e-01
-2.86038965e-01 3.55468303e-01 -6.79085135e-01 1.49830505e-01
8.72522652e-01 -3.41426700e-01 1.59769468e-02 -1.16884339e+00
-8.83845150e-01 -2.23236089e-03 6.17511749e-01 5.50803542e-01
6.20772958e-01 -1.31176364e+00 -8.15229118e-01 4.04857337e-01
-2.18613163e-01 -1.93735689e-01 3.98116887e-01 2.78466135e-01
-2.48511285e-01 1.69855103e-01 -1.52748987e-01 -7.80217171e-01
-1.29237974e+00 4.44500893e-01 2.64333427e-01 2.33335793e-01
-2.46321484e-01 1.02268517e+00 3.03237915e-01 -3.32415432e-01
1.01323150e-01 2.47992441e-01 -5.26330294e-03 1.66669130e-01
5.96489966e-01 5.02726287e-02 -4.77241725e-01 -1.00032389e+00
-8.04169476e-02 6.39442146e-01 3.78891736e-01 -8.75616133e-01
1.20907116e+00 -2.23196045e-01 3.56761783e-01 6.68693423e-01
1.02837420e+00 4.21777666e-01 -1.72132635e+00 4.67804283e-01
-8.67822766e-01 -2.58143604e-01 -1.36431068e-01 -4.26316470e-01
-9.79577303e-01 1.03595603e+00 3.31794292e-01 3.00969213e-01
1.09423840e+00 9.52258147e-03 8.23004961e-01 7.25922734e-02
3.44323874e-01 -9.27636385e-01 3.37845892e-01 2.54181355e-01
8.69756997e-01 -1.12314332e+00 -3.62378597e-01 -2.90062726e-01
-9.46778595e-01 1.15154719e+00 4.37799394e-01 1.30390584e-01
6.07823789e-01 5.46080053e-01 2.36623004e-01 2.33139783e-01
-1.09293926e+00 4.10427935e-02 2.08103701e-01 9.03348505e-01
5.10755360e-01 -8.79205689e-02 2.43042618e-01 4.73899424e-01
-7.30327129e-01 1.01226047e-01 6.97120130e-01 5.05052567e-01
-2.31309548e-01 -1.08625138e+00 -4.33463752e-01 1.67283505e-01
-6.08443320e-01 -1.09730929e-01 -4.32627916e-01 4.71827865e-01
1.34443372e-01 1.45222294e+00 1.84658602e-01 -3.26564521e-01
-2.07912877e-01 5.80353439e-01 7.03158319e-01 -5.83181620e-01
-1.02476224e-01 3.23648006e-01 -1.80977628e-01 -5.03762722e-01
-3.75972778e-01 -8.79520297e-01 -8.63530159e-01 -3.14479262e-01
-2.15755522e-01 -7.12914346e-03 4.91292596e-01 6.64664745e-01
3.90917808e-01 3.42498362e-01 7.53564715e-01 -1.42165101e+00
-3.34182024e-01 -8.10563624e-01 -5.77062488e-01 4.10829604e-01
5.02327681e-01 -5.94239295e-01 -8.09183836e-01 6.98699594e-01] | [13.154438972473145, -0.38568711280822754] |
af169a1e-8880-449f-a257-aa0ae08f9b19 | on-the-susceptibility-and-robustness-of-time | 2301.03703 | null | https://arxiv.org/abs/2301.03703v1 | https://arxiv.org/pdf/2301.03703v1.pdf | On the Susceptibility and Robustness of Time Series Models through Adversarial Attack and Defense | Under adversarial attacks, time series regression and classification are vulnerable. Adversarial defense, on the other hand, can make the models more resilient. It is important to evaluate how vulnerable different time series models are to attacks and how well they recover using defense. The sensitivity to various attacks and the robustness using the defense of several time series models are investigated in this study. Experiments are run on seven-time series models with three adversarial attacks and one adversarial defense. According to the findings, all models, particularly GRU and RNN, appear to be vulnerable. LSTM and GRU also have better defense recovery. FGSM exceeds the competitors in terms of attacks. PGD attacks are more difficult to recover from than other sorts of attacks. | ['Bidhan Bashyal', 'Asadullah Hill Galib'] | 2023-01-09 | null | null | null | null | ['adversarial-defense', 'time-series-regression'] | ['adversarial', 'time-series'] | [-1.65027767e-01 -4.62637961e-01 2.27279946e-01 -1.29417581e-02
-6.44625366e-01 -1.29441512e+00 7.07265615e-01 -3.22355270e-01
-4.37436886e-02 5.20685315e-01 8.99235457e-02 -7.00555027e-01
-1.67867959e-01 -8.76006901e-01 -5.18667579e-01 -7.27260113e-01
-6.32867754e-01 -1.13203198e-01 3.14861685e-02 -6.84084654e-01
2.95869619e-01 9.80589926e-01 -9.50175524e-01 1.59365699e-01
6.07256830e-01 9.01328444e-01 -7.72575557e-01 7.03367889e-01
3.17132846e-02 9.56593275e-01 -1.34525037e+00 -3.67876738e-01
7.45551646e-01 -1.34405166e-01 -3.66343945e-01 -7.20502913e-01
2.01849416e-01 -5.50808728e-01 -8.17108989e-01 1.14987481e+00
5.85269451e-01 4.87336725e-01 6.02228463e-01 -1.21974456e+00
-7.98144817e-01 7.28610218e-01 -3.70844334e-01 8.09235334e-01
4.10648406e-01 2.84630179e-01 3.66751909e-01 -1.14477456e-01
1.81980327e-01 1.68140411e+00 7.74210870e-01 8.65336180e-01
-1.04753852e+00 -1.18241620e+00 3.27989787e-01 -1.50566190e-01
-9.01521385e-01 -1.35503009e-01 8.43170345e-01 -4.02269289e-02
1.15875340e+00 6.94575310e-01 -4.66775820e-02 1.85067666e+00
9.21012878e-01 3.07557613e-01 1.37722147e+00 -1.96296182e-02
1.44784406e-01 -2.39754081e-01 3.00402999e-01 -4.26272936e-02
-2.21269168e-02 8.37958157e-01 1.44612834e-01 -6.82595491e-01
4.92443085e-01 4.06821907e-01 -1.22839376e-01 6.82966650e-01
-4.76450831e-01 9.23648357e-01 5.52300394e-01 5.66974223e-01
-4.55427766e-01 1.70254186e-01 8.58151078e-01 1.07331109e+00
5.98103881e-01 7.44119644e-01 -6.82379723e-01 2.11595278e-03
-4.52239603e-01 -1.38217462e-02 9.67446387e-01 6.92007244e-01
-1.81206968e-02 1.27119505e+00 5.48016727e-02 5.23620129e-01
-3.34610254e-01 8.05831134e-01 7.84182727e-01 -5.14740050e-01
6.03863418e-01 5.38876131e-02 -3.58110279e-01 -1.39058566e+00
-2.79404074e-01 -4.79569286e-01 -9.60474193e-01 2.17363238e-01
4.47049439e-01 -4.99570280e-01 -1.17389011e+00 1.80736947e+00
-1.80510804e-01 5.89068592e-01 3.61482710e-01 2.31266066e-01
3.13519925e-01 9.38917041e-01 1.98382899e-01 -3.71384472e-01
6.41868412e-01 -3.08123171e-01 -6.99201345e-01 -1.26026794e-01
3.56228262e-01 -6.61582828e-01 5.33364773e-01 4.95435804e-01
-7.97269225e-01 -4.48255032e-01 -9.14921999e-01 6.92774117e-01
-8.61237347e-01 -8.31726193e-01 5.99305332e-01 1.05433881e+00
-6.77836180e-01 1.04921305e+00 -9.25160170e-01 -2.07031950e-01
-1.06495053e-01 4.58363324e-01 -3.28549743e-01 3.44154298e-01
-1.83596838e+00 1.28017330e+00 3.07731718e-01 -1.12279631e-01
-1.27253640e+00 -5.23032129e-01 -6.28920615e-01 -8.22103918e-02
-1.41440943e-01 1.87222455e-02 9.35306072e-01 -8.42668116e-01
-1.14905608e+00 2.62002677e-01 5.06449521e-01 -1.08809578e+00
5.44017911e-01 -1.35213330e-01 -1.19015658e+00 2.07584575e-01
-3.07503879e-01 -1.35825783e-01 1.24879479e+00 -7.26637185e-01
-1.65635392e-01 -4.23928559e-01 1.54253080e-01 -4.83263582e-01
-5.85346043e-01 6.82254314e-01 5.93120873e-01 -1.07641888e+00
-2.53502280e-02 -1.16283488e+00 -3.81721795e-01 -8.26692164e-01
-4.69213873e-01 6.18344545e-02 1.27606404e+00 -9.47762311e-01
1.41967452e+00 -2.16929722e+00 -2.11398318e-01 3.96385700e-01
-9.73876938e-02 5.64087689e-01 -1.89038903e-01 8.09293449e-01
-8.31624687e-01 5.77173650e-01 2.56498903e-01 1.28972858e-01
-1.25651676e-02 3.59971762e-01 -1.37551093e+00 6.49103582e-01
4.63575916e-03 6.45431519e-01 -5.46568394e-01 3.60045195e-01
2.11864918e-01 4.14406151e-01 -1.38582960e-01 6.47995844e-02
1.40713617e-01 4.77704138e-01 -7.44408906e-01 8.55214596e-01
5.30451834e-01 4.65924323e-01 -2.91686505e-01 2.20588803e-01
1.67902753e-01 -3.21238339e-02 -5.59074879e-01 5.28921723e-01
-2.00067043e-01 6.30972505e-01 -4.18969542e-01 -8.56099963e-01
1.16423714e+00 6.74221754e-01 1.83542222e-01 -7.23698258e-01
4.01143938e-01 3.04676294e-01 4.16452259e-01 -1.68341145e-01
1.95578292e-01 -1.36998206e-01 -5.60585201e-01 6.97722256e-01
-2.23373920e-01 -7.09609464e-02 -3.13988298e-01 6.55843616e-02
1.44243908e+00 -2.69824415e-01 1.12894997e-01 -7.15424046e-02
2.90864497e-01 -3.96386087e-01 6.99112833e-01 9.22495365e-01
-2.62461394e-01 5.57600195e-03 3.47092450e-01 -7.11858690e-01
-8.11209917e-01 -1.19127858e+00 1.74417675e-01 1.04418600e+00
-4.02303934e-01 -7.18170479e-02 -6.08454227e-01 -7.85511196e-01
1.40650436e-01 1.11517239e+00 -7.52231061e-01 -7.87016094e-01
-7.09482908e-01 -7.00979829e-01 1.35408342e+00 7.90010571e-01
3.36118132e-01 -1.05216956e+00 -2.72987604e-01 3.51482779e-01
1.36768520e-02 -9.21603560e-01 -3.57758641e-01 3.95063609e-01
-1.14527833e+00 -7.34887362e-01 -2.70224482e-01 -6.44635633e-02
3.88686031e-01 4.22651887e-01 7.95044780e-01 2.73748320e-02
-1.10493600e-01 4.51207459e-01 -5.23697913e-01 -1.00782967e+00
-7.83662319e-01 -2.30023012e-01 6.22643292e-01 -3.42972457e-01
3.66335928e-01 -9.91710961e-01 -6.92175180e-02 4.82757807e-01
-1.01691902e+00 -1.24659097e+00 1.72656588e-02 6.11623824e-01
1.16485432e-01 6.79148257e-01 7.46467531e-01 -7.63193548e-01
9.85274673e-01 -7.95013130e-01 -3.62705320e-01 7.80795664e-02
-4.31730181e-01 -3.46021622e-01 1.37569666e+00 -1.20413482e+00
-6.85103238e-01 -5.75477898e-01 -2.08326161e-01 -1.05805302e+00
-1.23420767e-01 4.16653872e-01 2.57235765e-02 -4.96616274e-01
1.16964626e+00 1.15928598e-01 -2.06113055e-01 -5.13198137e-01
1.03411786e-01 3.16110313e-01 3.91695291e-01 -7.01551914e-01
1.43217027e+00 2.93558717e-01 -1.32025138e-01 -9.61166978e-01
-5.32472908e-01 2.04251513e-01 -3.68171781e-01 -1.91229120e-01
4.35388803e-01 -5.77660382e-01 -4.11547452e-01 7.45508313e-01
-1.01767027e+00 -2.44495422e-01 9.22325402e-02 3.27667654e-01
-1.69222310e-01 3.82553488e-01 -1.08383477e+00 -9.00094867e-01
-4.57531989e-01 -7.15397060e-01 1.98124409e-01 2.94959825e-02
-2.68936515e-01 -1.36447525e+00 -1.38487533e-01 -2.25323677e-01
8.18390310e-01 1.11428881e+00 9.55571413e-01 -1.63743269e+00
-3.68325450e-02 -9.64063525e-01 4.62846845e-01 7.52748370e-01
2.48426184e-01 4.40169901e-01 -1.07688618e+00 -7.29594231e-01
8.76723051e-01 -1.40837595e-01 5.46415508e-01 3.16710562e-01
9.53752220e-01 -7.51022696e-01 -8.91646817e-02 7.34629631e-01
1.15721095e+00 1.01322877e+00 9.88785565e-01 4.69303638e-01
5.69174111e-01 6.11574054e-01 5.94527721e-01 1.53146639e-01
-4.30496424e-01 1.43512443e-01 7.34305441e-01 1.62497431e-01
5.90582013e-01 -1.46161005e-01 1.08763802e+00 9.94978726e-01
-1.02389231e-01 -4.43781406e-01 -1.02437866e+00 7.62931863e-03
-1.20884550e+00 -1.58710182e+00 -1.07670307e-01 2.04702854e+00
3.13795954e-01 4.92974520e-01 1.29150197e-01 5.32099485e-01
7.31301904e-01 5.02878368e-01 -5.91580391e-01 -9.45552051e-01
-4.24108207e-01 3.13040286e-01 6.46145463e-01 2.09154725e-01
-1.23887491e+00 1.06457162e+00 8.00753784e+00 7.28396177e-01
-1.50859880e+00 -5.43698519e-02 7.11351991e-01 -1.75885469e-01
3.90369184e-02 -9.34662595e-02 -4.75447804e-01 6.30810738e-01
1.65078604e+00 -6.25920713e-01 5.09597242e-01 8.35517347e-01
1.25587761e-01 7.78305233e-01 -8.19307446e-01 4.09755498e-01
5.20838890e-03 -7.60072351e-01 1.39552400e-01 1.55250117e-01
6.09683096e-01 8.36751312e-02 4.77508485e-01 7.78604031e-01
7.84576535e-01 -1.24749517e+00 4.22835499e-01 3.34972531e-01
3.95867586e-01 -1.02998447e+00 6.52703583e-01 3.81223500e-01
-9.05192852e-01 -5.54022551e-01 -4.59931672e-01 -2.84113586e-01
-1.27469018e-01 -7.96490610e-02 -2.55700678e-01 6.57428324e-01
6.30901635e-01 3.89704019e-01 -6.39820516e-01 4.29196924e-01
-1.71966255e-01 9.56482708e-01 -2.88913071e-01 1.72773719e-01
4.94604498e-01 8.04674849e-02 8.12897682e-01 1.01231909e+00
4.53750014e-01 2.34548569e-01 2.90697753e-01 2.98091829e-01
4.61706758e-01 -3.53194833e-01 -1.25218678e+00 -4.09765691e-01
8.00139964e-01 7.53409564e-01 -5.14216721e-01 -2.32933406e-02
-1.06610075e-01 6.28768384e-01 -2.15300664e-01 5.18840730e-01
-1.02799749e+00 -4.59081650e-01 8.33596766e-01 -3.58635515e-01
-2.65891671e-01 -2.32416913e-01 1.99116260e-01 -9.90477502e-01
-2.67186314e-01 -1.45790339e+00 8.35807800e-01 -5.63173473e-01
-1.77578580e+00 1.07148862e+00 4.14807871e-02 -1.60419369e+00
-7.24897861e-01 -5.70930064e-01 -1.11961806e+00 8.99908662e-01
-8.58093143e-01 -1.07796097e+00 2.86089331e-01 1.01041543e+00
3.39870453e-01 -7.27441013e-01 1.10739744e+00 1.73566774e-01
-8.59682977e-01 8.85664880e-01 1.45652983e-02 3.83612812e-01
5.71134806e-01 -9.07784939e-01 1.01782453e+00 1.39727986e+00
-2.19909489e-01 1.14581800e+00 9.69099760e-01 -8.31353664e-01
-1.35050464e+00 -1.27654302e+00 1.96061268e-01 -5.55979550e-01
1.21513271e+00 -1.46755073e-02 -1.26537430e+00 1.11672878e+00
2.35724486e-02 -2.48098120e-01 8.58626127e-01 -9.78160799e-02
-8.83508563e-01 1.18884541e-01 -1.47713339e+00 8.21162105e-01
6.65310085e-01 -1.05160487e+00 -8.12198758e-01 3.56413603e-01
9.44519639e-01 -2.45076731e-01 -1.17015982e+00 2.49681428e-01
2.81127155e-01 -7.45250583e-01 1.33449519e+00 -1.20770741e+00
1.15912348e-01 1.20705657e-01 -4.97672468e-01 -1.40438664e+00
-5.57240546e-01 -1.10353911e+00 -2.72611707e-01 1.40439987e+00
1.77637622e-01 -1.34078693e+00 1.93599224e-01 7.51929164e-01
1.60799678e-02 -1.63957924e-01 -9.14263248e-01 -1.55886996e+00
6.69240475e-01 -5.33340931e-01 8.82368684e-01 1.44439757e+00
-3.64311606e-01 -9.60868299e-02 -7.31000602e-01 4.46376294e-01
4.36297864e-01 -3.47658068e-01 6.88060880e-01 -1.12187910e+00
-1.78457990e-01 -4.10503775e-01 -5.76962233e-01 -3.33712134e-03
4.83568341e-01 -6.83540881e-01 -5.62141597e-01 -6.65030003e-01
-7.11834133e-01 -1.17715470e-01 -7.80765116e-01 6.30075932e-01
-1.39543459e-01 1.82510033e-01 3.99452597e-01 3.50841165e-01
1.85065344e-01 1.29637614e-01 6.36049211e-01 -2.24898741e-01
-7.26884678e-02 3.94036531e-01 -8.22462976e-01 9.09354687e-01
1.25923181e+00 -6.97587788e-01 -4.59870279e-01 -3.26477259e-01
-3.41478318e-01 4.94433045e-01 1.54846773e-01 -9.19729531e-01
1.01810768e-01 -3.09258729e-01 5.16693294e-01 -4.52219367e-01
4.19288576e-01 -7.90148914e-01 4.63348776e-01 7.96573162e-01
-1.96959063e-01 8.43187451e-01 7.33318746e-01 4.95218664e-01
-6.17027991e-02 -4.13062833e-02 8.76602471e-01 -1.65452540e-01
-3.86825562e-01 4.15214181e-01 -6.02039635e-01 -2.97845230e-02
1.08166230e+00 -3.51370573e-01 -6.67954922e-01 -7.03278780e-01
-8.88291895e-01 -4.87062037e-02 2.01156184e-01 6.77361190e-01
5.72836459e-01 -1.41475832e+00 -6.58574998e-01 1.18793286e-01
-5.29093087e-01 -1.07838607e+00 3.96049619e-01 4.87888694e-01
-2.53367126e-01 2.99638063e-01 -7.18119144e-01 1.21471353e-01
-1.32843280e+00 1.08577538e+00 5.22937179e-01 -2.98630953e-01
-5.43668807e-01 6.79157257e-01 1.82417445e-02 -2.49499142e-01
3.54910433e-01 -9.34577547e-03 -1.96106315e-01 1.03982635e-01
8.13790083e-01 6.72378302e-01 -2.15550527e-01 -6.20814264e-01
-3.84934902e-01 2.83897012e-01 -4.66512889e-01 1.60482243e-01
1.27858353e+00 4.47012514e-01 -2.32269034e-01 4.12200898e-01
1.21161008e+00 6.36734292e-02 -8.07842731e-01 4.82800044e-02
1.67163953e-01 -6.67359531e-01 -3.77549917e-01 -7.69557655e-01
-1.08449006e+00 8.54604542e-01 4.09091860e-01 1.08030999e+00
1.32977235e+00 -9.33028400e-01 8.01559627e-01 1.12788878e-01
5.44761479e-01 -7.47081399e-01 6.80806860e-02 9.37300026e-01
9.78276312e-01 -5.09522676e-01 -1.68937877e-01 -5.05726337e-02
-4.99165386e-01 1.09580553e+00 4.74908084e-01 -7.13497102e-01
5.57346642e-01 4.03977543e-01 2.25631416e-01 1.96819544e-01
-9.76285219e-01 4.64788288e-01 1.07224189e-01 8.36586356e-01
9.98838171e-02 6.09931201e-02 -1.20308753e-02 4.41375732e-01
-4.40132201e-01 -9.32464421e-01 6.63083851e-01 7.12055087e-01
-3.29305828e-02 -9.60110724e-01 -8.85432184e-01 4.63476777e-01
-1.19265544e+00 9.23810527e-02 -8.14614236e-01 9.45050001e-01
-5.69180965e-01 1.46354866e+00 -3.64552021e-01 -1.04601789e+00
4.96710032e-01 -3.53869004e-03 -3.88107561e-02 -3.46820772e-01
-1.52748942e+00 3.25448774e-02 8.08144510e-02 -7.07551122e-01
3.48916978e-01 -6.06951892e-01 -8.80864620e-01 -1.07014894e+00
-3.62031192e-01 5.43109737e-02 3.52287203e-01 6.26897991e-01
5.42462990e-02 7.42415905e-01 1.21043491e+00 -6.97254360e-01
-1.29961455e+00 -8.68749738e-01 -6.15604103e-01 4.29610044e-01
2.62962490e-01 -3.87870252e-01 -9.56941187e-01 -5.75810611e-01] | [5.652348041534424, 7.742506504058838] |
a48cfd94-e1e8-4271-afdd-c7ac7803721a | alpcah-sample-wise-heteroscedastic-pca-with | 2307.02745 | null | https://arxiv.org/abs/2307.02745v1 | https://arxiv.org/pdf/2307.02745v1.pdf | ALPCAH: Sample-wise Heteroscedastic PCA with Tail Singular Value Regularization | Principal component analysis (PCA) is a key tool in the field of data dimensionality reduction that is useful for various data science problems. However, many applications involve heterogeneous data that varies in quality due to noise characteristics associated with different sources of the data. Methods that deal with this mixed dataset are known as heteroscedastic methods. Current methods like HePPCAT make Gaussian assumptions of the basis coefficients that may not hold in practice. Other methods such as Weighted PCA (WPCA) assume the noise variances are known, which may be difficult to know in practice. This paper develops a PCA method that can estimate the sample-wise noise variances and use this information in the model to improve the estimate of the subspace basis associated with the low-rank structure of the data. This is done without distributional assumptions of the low-rank component and without assuming the noise variances are known. Simulations show the effectiveness of accounting for such heteroscedasticity in the data, the benefits of using such a method with all of the data versus retaining only good data, and comparisons are made against other PCA methods established in the literature like PCA, Robust PCA (RPCA), and HePPCAT. Code available at https://github.com/javiersc1/ALPCAH | ['Laura Balzano', 'Jeffrey A. Fessler', 'Javier Salazar Cavazos'] | 2023-07-06 | null | null | null | null | ['dimensionality-reduction'] | ['methodology'] | [-3.06177497e-01 -5.58283925e-01 2.28934169e-01 -1.19786881e-01
-7.01003671e-01 -5.94921589e-01 4.22745287e-01 -4.13661689e-01
-9.19482186e-02 5.27385533e-01 6.31976426e-01 6.84191585e-02
-7.21337080e-01 -4.99250025e-01 -5.52962005e-01 -1.29728353e+00
2.75722314e-02 4.39500719e-01 -2.56700516e-01 1.82731107e-01
1.75710484e-01 5.36842823e-01 -1.57543087e+00 -1.31556779e-01
6.11548364e-01 6.64846718e-01 9.10660177e-02 5.55966198e-01
-6.11529946e-02 3.87079746e-01 -3.21854919e-01 -1.66720778e-01
7.31752753e-01 -1.33598611e-01 -1.84586212e-01 3.01505744e-01
-9.37840436e-03 -5.82712591e-02 -4.71833438e-01 1.19569552e+00
4.62306887e-01 2.81977504e-01 7.06018686e-01 -1.34544241e+00
-3.67801994e-01 2.82082349e-01 -9.31423903e-01 1.07742451e-01
-3.80654441e-04 -1.60297602e-01 5.67400873e-01 -1.22039545e+00
3.28393549e-01 1.29720807e+00 6.73324883e-01 1.75919205e-01
-1.31271911e+00 -7.26493478e-01 -2.76390612e-01 2.84875482e-01
-1.70571005e+00 -5.12302279e-01 1.15965366e+00 -6.06972754e-01
3.19284528e-01 3.55105758e-01 3.31462473e-01 1.03089535e+00
6.37693629e-02 5.27610123e-01 1.46815455e+00 -3.58002841e-01
4.07118678e-01 2.75905505e-02 6.67675197e-01 -5.21424003e-02
6.76349819e-01 1.93768233e-01 -6.24937952e-01 -5.71259677e-01
4.46341187e-01 3.73722687e-02 -2.48386249e-01 -8.67446303e-01
-8.73377919e-01 8.79329443e-01 -2.62789607e-01 2.39648402e-01
-7.79869735e-01 -1.62713096e-01 2.50402540e-01 5.49420379e-02
2.61279941e-01 1.39286295e-01 -4.48942721e-01 -3.84767115e-01
-9.09493983e-01 1.39092177e-01 7.77111769e-01 1.08561742e+00
6.83115780e-01 2.40645945e-01 4.43012901e-02 9.44518685e-01
2.64035344e-01 8.62256289e-01 7.04279006e-01 -1.13952374e+00
3.62526566e-01 3.08807313e-01 1.00206368e-01 -1.12230980e+00
-4.39517945e-01 -2.32348099e-01 -1.03269684e+00 7.01064616e-02
5.80299437e-01 -3.72594953e-01 -8.74708772e-01 1.57777679e+00
2.97661334e-01 2.98657000e-01 2.94773787e-01 7.28475392e-01
2.79950500e-01 6.19457304e-01 -3.07729155e-01 -4.09202665e-01
1.12179029e+00 -3.85375738e-01 -1.00773132e+00 -8.69349167e-02
1.37800545e-01 -9.73627210e-01 8.66558254e-01 6.82692468e-01
-6.81056499e-01 -4.20194745e-01 -7.63016880e-01 2.53482968e-01
-3.42047751e-01 2.76858658e-01 5.02767563e-01 1.03483200e+00
-8.53562713e-01 3.18066955e-01 -1.11088479e+00 -3.44360173e-01
2.55622953e-01 3.56743783e-01 -5.66736817e-01 -3.20437074e-01
-7.10671484e-01 6.73469186e-01 1.31791621e-01 4.35311347e-01
-6.44311666e-01 -7.13377357e-01 -5.28501272e-01 9.63409096e-02
3.26221794e-01 -2.49793023e-01 6.65249586e-01 -6.69090986e-01
-1.08819079e+00 3.93881239e-02 -4.43225652e-01 -2.20443066e-02
4.08986032e-01 -3.71189922e-01 -3.63247037e-01 -1.52852563e-02
-5.74579127e-02 1.56849846e-02 9.97551918e-01 -1.24703693e+00
-2.39692762e-01 -6.08034790e-01 -6.91188931e-01 1.59946561e-01
-2.14778557e-01 -1.31396586e-02 -6.19302690e-01 -5.96553326e-01
4.96600688e-01 -1.40475702e+00 -1.44493401e-01 -3.33853930e-01
-4.16857600e-01 2.34953374e-01 1.03794348e+00 -1.08755183e+00
9.66208398e-01 -2.68714714e+00 9.22543406e-02 7.96442389e-01
-2.01813310e-01 -7.83826709e-02 -1.28714681e-01 7.73096740e-01
-2.93731898e-01 -1.72863469e-01 -2.77846664e-01 -1.78631052e-01
-1.57519445e-01 3.38933706e-01 -1.84810132e-01 7.27760673e-01
-1.43375680e-01 1.22599944e-01 -4.82815117e-01 1.41685493e-02
2.88302362e-01 6.63823485e-01 -3.48193586e-01 1.67678654e-01
5.26241958e-01 3.70205224e-01 -1.85595617e-01 4.35795873e-01
1.10494065e+00 5.03517501e-02 3.16556990e-01 -5.45465827e-01
-1.29388928e-01 -3.62706810e-01 -2.01558161e+00 1.08810961e+00
3.45102288e-02 5.19173801e-01 2.29727298e-01 -8.91941786e-01
7.95345604e-01 3.09708238e-01 8.65528822e-01 -7.70573467e-02
7.69711658e-02 1.08314231e-01 3.83164555e-01 -4.49131250e-01
2.35578522e-01 -8.84783417e-02 3.04727465e-01 1.41696408e-01
8.35423172e-02 -1.15046568e-01 3.57772112e-01 5.28074242e-02
9.76841509e-01 -6.23382144e-02 1.96040392e-01 -5.48547268e-01
3.58808637e-01 -1.29042283e-01 8.80082786e-01 6.27568245e-01
-2.36246940e-02 8.77090931e-01 4.90841150e-01 5.07071689e-02
-1.25074160e+00 -9.30301070e-01 -3.34566772e-01 3.72591853e-01
-2.65376151e-01 -4.38302547e-01 -4.38576013e-01 -8.49215165e-02
5.10630906e-02 5.64399302e-01 -5.92166543e-01 -9.68864039e-02
-1.89969271e-01 -1.25562608e+00 6.96922690e-02 5.18342018e-01
3.72750044e-01 -3.71645242e-01 -1.18858412e-01 8.13029800e-03
1.54244145e-02 -9.60781813e-01 -1.63923651e-01 2.31238827e-01
-9.42610204e-01 -1.27222431e+00 -6.14676416e-01 -2.24451348e-01
8.21271479e-01 5.52197993e-01 4.46366876e-01 -5.65924644e-01
-1.31780818e-01 6.90669477e-01 -3.45235676e-01 -4.58612174e-01
-4.36665602e-02 -5.48455954e-01 3.38054210e-01 2.06124321e-01
7.75520205e-01 -5.32672882e-01 -2.55840123e-01 3.03271711e-01
-1.03066719e+00 -3.54389668e-01 5.79906821e-01 1.03772247e+00
6.67242169e-01 9.08925951e-01 6.51261434e-02 -9.87664342e-01
7.95686364e-01 -6.39233768e-01 -6.65731430e-01 1.12667412e-01
-6.77671432e-01 3.68229076e-02 2.25103885e-01 -6.28160715e-01
-1.20109797e+00 3.02360833e-01 3.03854018e-01 -9.78586495e-01
-2.07860798e-01 7.23999023e-01 -4.83323991e-01 8.28694329e-02
4.77957428e-01 9.83437002e-02 3.00422788e-01 -8.46468151e-01
1.20058127e-01 6.00031316e-01 2.15082541e-01 -5.76263905e-01
9.91093874e-01 6.05594635e-01 2.98761874e-01 -1.13390791e+00
-3.98498625e-01 -8.60258818e-01 -6.72194183e-01 1.42497241e-01
5.17960846e-01 -1.03936970e+00 -2.84366935e-01 7.20943570e-01
-5.28404534e-01 1.48598522e-01 -2.23162368e-01 9.71314669e-01
-4.15864706e-01 4.31885004e-01 -4.47016627e-01 -9.52019334e-01
-3.91783156e-02 -1.16086733e+00 5.86647928e-01 -7.73046724e-03
-1.09605722e-01 -8.07519078e-01 1.75410718e-01 2.28652596e-01
4.94719297e-01 -2.84610149e-02 8.39094937e-01 -8.10190380e-01
-9.22587886e-02 -4.95734036e-01 5.22260368e-02 7.19008148e-01
4.49094862e-01 2.54158467e-01 -9.19908464e-01 -3.48717779e-01
4.25489902e-01 4.27777231e-01 3.32045406e-01 7.49382973e-01
1.06746757e+00 -3.41343880e-01 -4.89072166e-02 4.31147069e-01
1.35530400e+00 3.12660605e-01 6.39158487e-01 1.10714607e-01
8.09216559e-01 8.91382575e-01 4.93470371e-01 6.74703240e-01
1.27588570e-01 5.68406940e-01 5.80429612e-03 1.78774163e-01
1.12914883e-01 1.09677136e-01 3.09038728e-01 1.13175738e+00
-3.51882316e-02 2.03147143e-01 -1.04664624e+00 4.96683270e-01
-1.85010219e+00 -1.05770576e+00 -5.37696123e-01 2.47234511e+00
5.28194785e-01 -3.49822640e-01 1.33791327e-01 4.21315342e-01
7.84669220e-01 -2.02347159e-01 -3.43405932e-01 -1.15785383e-01
-4.73934084e-01 -5.39012328e-02 8.06977987e-01 3.47178191e-01
-1.14953029e+00 2.75613725e-01 6.42011404e+00 5.80400050e-01
-7.31149554e-01 -9.64874215e-03 2.72925049e-01 4.17197198e-02
-6.24026731e-02 2.54820168e-01 -6.34404838e-01 7.16710567e-01
1.08957314e+00 -3.78343880e-01 6.64240301e-01 7.93739200e-01
4.79670167e-01 -2.65858889e-01 -7.49533653e-01 1.15030253e+00
2.23125473e-01 -5.08566022e-01 -2.79548705e-01 4.64753807e-01
7.82591581e-01 -7.29239061e-02 5.99262714e-02 4.35738638e-02
3.67354602e-01 -5.25843263e-01 3.08443725e-01 8.52371395e-01
6.38611764e-02 -8.90707552e-01 9.99758542e-01 2.06394389e-01
-7.45838761e-01 -2.63432503e-01 -7.21698582e-01 1.72326356e-01
-2.55322885e-02 9.77955997e-01 -5.60597539e-01 5.65745473e-01
8.91503453e-01 6.65819585e-01 -6.47943676e-01 1.43671811e+00
1.30825564e-02 8.51807415e-01 -5.60379446e-01 5.40921926e-01
-1.90594867e-01 -8.85353982e-01 6.84709251e-01 8.09285820e-01
9.42715168e-01 8.87380447e-03 -5.46229631e-02 3.78645748e-01
3.28568310e-01 1.87061757e-01 -5.63960671e-01 -2.29001686e-01
5.63704669e-01 1.05757618e+00 -4.05311197e-01 -9.38847885e-02
-8.37385476e-01 5.74763954e-01 -2.71145910e-01 6.29464507e-01
-1.76006183e-01 2.14052927e-02 7.79856086e-01 9.43252724e-03
3.61132205e-01 -4.82559502e-01 -3.06475133e-01 -1.16441095e+00
5.71373887e-02 -1.26524127e+00 5.73881626e-01 -8.15218627e-01
-1.70021069e+00 1.36643469e-01 3.05130273e-01 -1.47728455e+00
-2.26037771e-01 -6.82956815e-01 -3.28799665e-01 1.07776952e+00
-9.21572506e-01 -9.68953431e-01 -2.09625661e-01 8.02612543e-01
2.33351365e-01 -3.76707703e-01 7.26429820e-01 3.01158071e-01
-7.64170110e-01 -4.24879678e-02 8.31905484e-01 4.47798707e-02
8.67611170e-01 -1.09399164e+00 -2.16600522e-01 1.18285131e+00
1.03461267e-02 1.17067063e+00 1.03341722e+00 -8.35577488e-01
-1.82992017e+00 -7.08107054e-01 2.53348827e-01 -2.87917227e-01
9.16450262e-01 -2.18821004e-01 -1.08859885e+00 6.00406110e-01
4.95700054e-02 1.80739593e-02 1.07534790e+00 3.44875067e-01
-2.71169960e-01 -2.67703652e-01 -1.05465019e+00 3.44777137e-01
2.60382116e-01 -3.57543200e-01 -3.66606057e-01 1.28140375e-01
-2.33719870e-02 -1.39067814e-01 -9.11552012e-01 3.07089120e-01
5.47790647e-01 -6.89665735e-01 8.82248700e-01 -5.41363537e-01
-1.59549490e-01 -5.64910293e-01 -4.92060095e-01 -1.41071594e+00
-5.08636951e-01 -2.06615865e-01 -1.40593633e-01 1.49646199e+00
1.58564314e-01 -5.62471509e-01 7.23640084e-01 1.32068324e+00
1.74543172e-01 1.21016055e-01 -1.04597926e+00 -1.13507652e+00
2.11296938e-02 -4.50024307e-01 3.91537011e-01 1.03330290e+00
-2.52460390e-01 -3.66805345e-02 -8.26436698e-01 5.19301414e-01
9.26265597e-01 -2.17860654e-01 9.07417178e-01 -1.39259410e+00
-1.54784948e-01 -9.69406292e-02 -5.37431002e-01 -1.84227526e-01
1.17754065e-01 -3.68397951e-01 -6.25132537e-03 -1.20447147e+00
3.37514192e-01 -2.57971495e-01 -2.01430991e-01 3.14253330e-01
-2.11641490e-01 -1.66882828e-01 2.13371500e-01 5.31737685e-01
2.33234286e-01 4.71398324e-01 5.92311263e-01 4.71786000e-02
-3.10496897e-01 1.74774066e-01 -7.02505827e-01 6.63792968e-01
7.97858238e-01 -6.45513475e-01 -5.92054427e-01 -1.47779316e-01
-3.55842821e-02 -1.43655822e-01 2.70112213e-02 -1.11733687e+00
3.25356215e-01 -2.12748751e-01 5.67262888e-01 -6.12353384e-01
6.00209296e-01 -1.46152520e+00 1.01159549e+00 2.63734628e-02
8.34441110e-02 2.32565075e-01 1.70322254e-01 7.04953611e-01
-3.47595513e-01 -3.89789730e-01 5.03367424e-01 9.28961784e-02
-4.48011696e-01 2.95244623e-02 -4.20964569e-01 -3.45560700e-01
9.86687779e-01 -8.05244818e-02 -3.11537206e-01 -5.81933141e-01
-7.93064892e-01 -4.31794487e-02 4.42281693e-01 3.09018910e-01
4.29020762e-01 -1.26234579e+00 -6.00214124e-01 3.57794493e-01
1.08789280e-02 -3.93658042e-01 5.38880408e-01 1.03378606e+00
-2.01638579e-01 2.75755167e-01 -2.24805638e-01 -3.65235209e-01
-1.33060420e+00 9.67408419e-01 -6.20357245e-02 2.19961688e-01
-3.75551105e-01 2.57137895e-01 6.87417388e-02 -2.40931571e-01
9.37309396e-03 2.97549278e-01 -1.59080684e-01 3.02232742e-01
5.32321990e-01 8.00914764e-01 1.20797314e-01 -9.48383451e-01
-3.09419930e-01 5.69687665e-01 1.69694632e-01 -2.58642048e-01
1.47991323e+00 -3.50235820e-01 -3.14488620e-01 6.96493268e-01
1.12624943e+00 5.02853155e-01 -1.17401791e+00 -2.59429693e-01
1.94525748e-01 -8.68153751e-01 2.24763572e-01 -4.09313291e-01
-1.13586187e+00 7.66741931e-01 8.80791664e-01 8.21345598e-02
1.22048044e+00 -4.22736555e-01 -1.95813268e-01 2.26089984e-01
2.88592994e-01 -1.18279433e+00 -3.40416759e-01 2.59612501e-01
9.08259213e-01 -1.15185165e+00 3.05393428e-01 -4.45683420e-01
-9.17934716e-01 1.08101428e+00 3.54886353e-01 -1.46976337e-01
1.19316506e+00 3.37364644e-01 2.06044778e-01 -6.92162588e-02
-5.18800616e-01 -1.62511587e-01 3.78700942e-01 6.30042613e-01
1.87974066e-01 1.09055541e-01 -5.46745002e-01 5.54036617e-01
-1.55219063e-02 -3.07848811e-01 7.85896719e-01 1.02110481e+00
1.02237888e-01 -1.36481285e+00 -9.37896311e-01 6.80390000e-01
-5.53479791e-01 -3.85400243e-02 -2.25736499e-01 6.50332928e-01
-3.44147593e-01 1.13901496e+00 -1.03445150e-01 -2.10996583e-01
3.62955332e-01 2.03943625e-01 -1.03011802e-01 -3.58879119e-01
-1.69853326e-02 8.53670955e-01 -1.09131038e-02 -3.72912169e-01
-6.04030430e-01 -1.25793898e+00 -6.55563772e-01 -3.79175961e-01
-3.06473315e-01 4.01202679e-01 9.69320416e-01 5.65847814e-01
5.30349314e-01 1.37766138e-01 6.39753640e-01 -6.91319346e-01
-6.59497619e-01 -1.03068149e+00 -1.03856063e+00 5.78867316e-01
6.66825399e-02 -8.60981047e-01 -6.87555313e-01 2.14229450e-01] | [7.701168060302734, 4.22667121887207] |
da517ede-0b10-41e4-a868-28026e7b25f4 | food-recognition-and-recipe-analysis | 1801.07239 | null | http://arxiv.org/abs/1801.07239v1 | http://arxiv.org/pdf/1801.07239v1.pdf | Food recognition and recipe analysis: integrating visual content, context and external knowledge | The central role of food in our individual and social life, combined with
recent technological advances, has motivated a growing interest in applications
that help to better monitor dietary habits as well as the exploration and
retrieval of food-related information. We review how visual content, context
and external knowledge can be integrated effectively into food-oriented
applications, with special focus on recipe analysis and retrieval, food
recommendation, and the restaurant context as emerging directions. | ['Weiqing Min', 'Luis Herranz', 'Shuqiang Jiang'] | 2018-01-22 | null | null | null | null | ['food-recognition', 'food-recommendation'] | ['computer-vision', 'miscellaneous'] | [-3.24024081e-01 -4.78229791e-01 -6.73378229e-01 -3.55380654e-01
2.09339499e-01 -6.23374760e-01 3.82005535e-02 1.52609110e+00
-1.62577525e-01 9.96742919e-02 9.00410831e-01 -6.07423894e-02
1.37952104e-01 -1.16396904e+00 -7.88111389e-02 -3.65349233e-01
-2.54488409e-01 -3.87884647e-01 1.05783559e-01 -4.91399825e-01
1.78895727e-01 1.31097697e-02 -1.70847189e+00 3.43495756e-01
8.25473070e-01 5.64810514e-01 5.51886797e-01 5.00128388e-01
-6.18836641e-01 4.18599159e-01 -1.28280848e-01 -1.58278011e-02
-3.14450353e-01 -4.82104063e-01 -9.91370715e-03 -7.34248832e-02
-5.13258092e-02 -1.46404088e-01 2.33981058e-01 1.27399004e+00
4.00890559e-01 4.92615134e-01 6.66368753e-02 -3.93201590e-01
-1.40319407e+00 6.45939350e-01 -3.57942611e-01 2.75083899e-01
9.51072931e-01 5.36128581e-02 3.69829595e-01 -5.77566087e-01
5.08278608e-01 1.20337570e+00 4.87610132e-01 1.02072306e-01
-1.12513494e+00 -1.01091966e-01 5.95386446e-01 3.01756859e-01
-1.12133241e+00 -6.34262189e-02 9.13601816e-01 -3.95665556e-01
8.92769217e-01 4.58268940e-01 1.49239922e+00 5.65484285e-01
1.07350886e-01 8.11791599e-01 8.96851301e-01 -8.26390147e-01
3.13683838e-01 4.85673100e-01 5.93058527e-01 6.68288648e-01
5.66956401e-01 1.42849654e-01 -4.83948141e-01 8.74030590e-02
4.95957375e-01 5.25714993e-01 7.03347921e-02 -5.09953558e-01
-1.23590016e+00 9.70833421e-01 5.81556380e-01 5.04553616e-01
-6.23034120e-01 -6.14841022e-02 4.71440077e-01 2.42363408e-01
7.89963365e-01 2.31336728e-01 1.35729817e-04 2.93352962e-01
-6.05743229e-01 1.83201358e-01 7.24969745e-01 6.54239953e-01
6.78877473e-01 -1.21086411e-01 6.48944005e-02 9.18610573e-01
7.95424581e-01 1.02031887e+00 3.67012441e-01 -4.57889974e-01
-1.22759849e-01 9.23906624e-01 3.52171719e-01 -1.51394844e+00
-7.49555290e-01 -4.86970060e-02 -3.36501032e-01 3.29683386e-02
2.50314027e-01 1.68296114e-01 -5.63529551e-01 1.26034534e+00
9.56567943e-01 -5.55863023e-01 -3.28306377e-01 7.99946964e-01
1.33668363e+00 5.24161339e-01 1.09426248e+00 -1.21110767e-01
1.94705737e+00 -6.54949844e-01 -1.16108561e+00 -1.72223911e-01
6.00167155e-01 -1.06617749e+00 9.39796627e-01 2.18581617e-01
-1.02370226e+00 -6.64264381e-01 -1.08237529e+00 -3.49114418e-01
-1.33991170e+00 -1.38181942e-02 7.26340055e-01 8.41650546e-01
-8.34511161e-01 4.97722894e-01 -7.41469622e-01 -1.04847574e+00
1.25154212e-01 -1.45779654e-01 -7.37677887e-02 -2.21619517e-01
-1.10028911e+00 8.53380203e-01 5.42290747e-01 -6.65197968e-02
1.91959124e-02 -8.76916409e-01 -1.25541329e+00 -2.59747624e-01
4.16672617e-01 -6.32354379e-01 8.71598899e-01 -4.79125381e-01
-1.36894178e+00 9.20435011e-01 2.33625352e-01 -3.61314253e-03
-2.36979574e-01 -5.02791047e-01 -1.21614182e+00 1.24879658e-01
1.82575315e-01 6.10709012e-01 2.03075796e-01 -6.15088403e-01
-6.95532620e-01 -8.03981662e-01 -8.16753209e-02 4.36521471e-01
-2.89764285e-01 3.89036804e-01 7.39168674e-02 -7.64302433e-01
4.81610931e-02 -2.87050962e-01 -3.75992179e-01 5.07163167e-01
-3.81281413e-02 -3.06435019e-01 6.55174136e-01 -8.50528955e-01
1.57116866e+00 -2.51518846e+00 -3.09860528e-01 2.48649061e-01
-5.20492941e-02 2.86614180e-01 5.98686747e-02 6.52762949e-01
8.66041109e-02 1.37205824e-01 5.50450921e-01 4.53293443e-01
1.79197490e-01 7.48852789e-02 4.23595220e-01 5.04515767e-01
-3.66337419e-01 9.37195659e-01 -1.64422953e+00 -3.87157261e-01
8.68221223e-01 7.67896652e-01 -3.31068337e-01 -1.85473323e-01
-5.89533091e-01 2.41772428e-01 -6.60314262e-01 6.58145368e-01
4.31543857e-01 -4.50492680e-01 4.40262437e-01 -5.08719385e-01
-7.93648303e-01 3.23979139e-01 -1.18412495e+00 1.83820438e+00
-7.37093613e-02 3.78221385e-02 2.82467872e-01 -5.38091719e-01
8.14716637e-01 1.66893750e-01 3.99729282e-01 -1.16153979e+00
-4.48428877e-02 -1.18712328e-01 -3.43946248e-01 -1.14095414e+00
5.37604392e-01 1.52924135e-01 2.27986246e-01 5.72132707e-01
-5.15888333e-01 5.87283850e-01 6.63965464e-01 -6.54078722e-02
1.64025918e-01 6.78577900e-01 1.11967611e+00 -6.93373442e-01
3.34066153e-01 3.98622274e-01 -1.22471608e-01 6.57663941e-02
-4.10347253e-01 -2.27850199e-01 -6.67506337e-01 -7.11578190e-01
-6.98664129e-01 -8.89449358e-01 -3.01133916e-02 1.54911029e+00
3.10161889e-01 -6.89461350e-01 -2.75226325e-01 -5.09359360e-01
2.13562191e-01 6.23264432e-01 -7.24247277e-01 -7.05367774e-02
-7.02634096e-01 -6.70154214e-01 -6.66816652e-01 5.15990496e-01
5.47745168e-01 -1.21492076e+00 -1.07465231e+00 5.89527190e-01
-2.79801816e-01 -7.72190019e-02 -7.45963037e-01 -2.42322721e-02
-9.50442374e-01 -1.23177290e+00 -7.89405763e-01 -8.68896842e-01
4.37241644e-01 8.78139973e-01 1.50170588e+00 1.94307715e-01
-6.33852184e-01 8.76329720e-01 -5.58511198e-01 -5.06169975e-01
-3.91995162e-01 -2.77245253e-01 -3.71262699e-01 -4.37079012e-01
1.27477360e+00 -4.90591854e-01 -1.50602651e+00 1.37944862e-01
-8.38755548e-01 1.56170383e-01 -1.63520679e-01 -7.87118822e-03
7.35819697e-01 -2.72318691e-01 4.81919438e-01 -6.93890333e-01
3.91226947e-01 -1.04957139e+00 -2.42757738e-01 4.07359660e-01
-5.33341408e-01 -1.58160433e-01 -3.80316004e-02 -5.51226258e-01
-8.39021146e-01 -1.07709011e-02 -1.81694210e-01 6.74757898e-01
-5.58363914e-01 7.76587486e-01 -5.67557290e-02 2.35375866e-01
8.53406608e-01 -2.20450118e-01 6.16084337e-02 -9.87919688e-01
9.09395635e-01 3.17503333e-01 2.56801575e-01 -7.37402737e-02
4.23909947e-02 3.40075552e-01 -3.37977171e-01 -1.11027384e+00
-7.22860038e-01 -8.17902446e-01 -3.08622986e-01 -4.59159642e-01
1.03818679e+00 -8.40995073e-01 -9.78811026e-01 -1.39216900e-01
-3.50196242e-01 -2.47308940e-01 -4.92806435e-01 6.75534070e-01
-1.56434420e-02 3.87256145e-01 -3.66636992e-01 -7.67800808e-01
-4.72547263e-01 -4.48687255e-01 4.92852628e-01 4.20865834e-01
-5.53092897e-01 -1.34017599e+00 2.65291631e-01 4.06467728e-02
7.78203547e-01 6.59867942e-01 1.04859889e+00 -2.27154121e-01
-3.00079435e-01 6.24489412e-02 -7.80465007e-02 -4.82032776e-01
6.41689122e-01 -3.60181481e-01 -6.31302118e-01 3.89861614e-02
-1.17287591e-01 9.70155522e-02 5.57580292e-01 6.74912333e-01
2.54560739e-01 -3.77509743e-02 -7.00235605e-01 -6.84892982e-02
1.49362803e+00 4.74219501e-01 2.18283594e-01 2.94144422e-01
3.42890501e-01 8.05762529e-01 6.46760941e-01 5.69648445e-01
5.49497068e-01 7.11900175e-01 2.25452825e-01 -3.31594586e-01
-6.80082381e-01 -4.97759521e-01 1.45061061e-01 7.46978760e-01
2.35381991e-01 1.44415081e-01 -2.20694616e-01 5.36783755e-01
-1.72021830e+00 -9.54767466e-01 -1.13797657e-01 2.15792847e+00
6.53110504e-01 -7.33259022e-01 7.43999839e-01 -7.37546906e-02
6.14992797e-01 6.76231310e-02 -6.77858949e-01 -1.33402467e-01
2.00028747e-01 -1.08167335e-01 1.82082012e-01 1.28326654e-01
-1.25098956e+00 3.65630507e-01 7.98838139e+00 1.91345394e-01
-9.83937144e-01 1.50735304e-01 5.06061852e-01 1.26578677e-02
-3.76420945e-01 -7.97707379e-01 -6.02726161e-01 4.10941690e-01
9.75911498e-01 -3.71158756e-02 4.26243037e-01 8.16523492e-01
4.39002424e-01 -4.67773139e-01 -1.01770401e+00 6.56191945e-01
-1.65715758e-02 -1.26401699e+00 -5.18987834e-01 -1.39766023e-01
3.02077383e-01 -1.35933787e-01 1.22499898e-01 -4.57549468e-02
3.56602132e-01 -3.55793118e-01 7.81163692e-01 3.10696572e-01
4.45576668e-01 -5.12642860e-01 -3.67509387e-02 -9.69917178e-02
-1.75386751e+00 3.44877206e-02 -2.47967392e-01 -1.14988662e-01
3.28913242e-01 6.61498487e-01 -2.35512465e-01 4.01933610e-01
1.10296953e+00 1.12716913e+00 -4.20257658e-01 1.38972294e+00
2.51571793e-04 1.25422493e-01 -5.89255273e-01 -3.75633717e-01
-2.81385154e-01 -2.56289184e-01 3.11614454e-01 1.27553415e+00
4.49603319e-01 3.28091025e-01 5.71856380e-01 7.83182800e-01
5.97469628e-01 1.12711203e+00 -5.21847904e-01 -2.92684376e-01
3.36728729e-02 1.15894890e+00 -1.32344675e+00 -2.55003244e-01
-7.61206150e-01 6.08049572e-01 7.44462162e-02 2.23938078e-01
-2.54866958e-01 1.97061062e-01 6.66858673e-01 4.15905088e-01
2.43523389e-01 -2.43966505e-01 1.33342981e-01 -8.63297462e-01
-4.06045645e-01 -7.18167603e-01 8.70096385e-01 -4.36720699e-01
-1.25044358e+00 -9.96121615e-02 3.34778041e-01 -8.99546921e-01
-6.06155507e-02 -5.03942192e-01 -1.85859978e-01 5.43205023e-01
-1.47615755e+00 -1.00542223e+00 -1.18955508e-01 3.74137014e-01
3.29587191e-01 4.45516914e-01 1.34938109e+00 6.26611173e-01
-3.58527869e-01 -4.25853580e-02 2.16949284e-01 -3.32086891e-01
5.98837376e-01 -1.19765592e+00 2.08351955e-01 2.73874998e-01
1.04702771e-01 6.47142112e-01 6.87537611e-01 -1.12381268e+00
-1.48688221e+00 -9.56209362e-01 7.64368296e-01 -2.64602631e-01
3.61206710e-01 -1.71454325e-01 -6.61680043e-01 2.40341365e-01
3.36903274e-01 -3.32811296e-01 1.42257285e+00 2.80455738e-01
-2.91682094e-01 -7.37282932e-02 -1.45146036e+00 7.82836199e-01
9.19264019e-01 1.93380080e-02 -2.41448805e-01 4.98047054e-01
6.09141409e-01 -2.17632040e-01 -1.04123187e+00 -2.21464053e-01
8.49209189e-01 -7.42472112e-01 1.54802370e+00 -1.90818116e-01
-1.36230320e-01 -4.34897393e-01 -8.05222392e-02 -1.14058936e+00
-7.96880901e-01 -9.90505293e-02 -1.92434415e-01 8.16412747e-01
5.25451936e-02 -3.45121682e-01 5.15123546e-01 8.14023733e-01
-3.54914777e-02 -5.90859279e-02 1.09684482e-01 -1.40550345e-01
-4.49370831e-01 1.65668696e-01 1.03009200e+00 8.63262057e-01
6.53008699e-01 -1.24326505e-01 -3.65628958e-01 -5.86531572e-02
4.00440931e-01 5.40180326e-01 1.64671838e-01 -1.31682801e+00
-8.91516358e-02 -4.42871988e-01 -3.09407953e-02 -9.18189764e-01
-8.92378211e-01 -8.96190882e-01 -1.95768833e-01 -1.89682662e+00
1.32218227e-01 -1.24233097e-01 -5.45674145e-01 3.63063514e-01
-1.50869668e-01 5.79200923e-01 1.45711586e-01 -3.52574378e-01
-7.24646151e-01 -1.38762787e-01 1.47915375e+00 -1.70735225e-01
-7.99687922e-01 -2.42874742e-01 -1.10275960e+00 6.00368083e-01
7.33546078e-01 -1.04455017e-01 -4.54556465e-01 -1.20128438e-01
1.04729176e+00 -4.31527078e-01 2.82531115e-03 -7.66726196e-01
-9.56647620e-02 -4.86627996e-01 7.45540500e-01 -8.63543391e-01
-1.57345787e-01 -9.55412626e-01 6.67988539e-01 7.48457670e-01
-4.03011292e-01 1.83265686e-01 2.91535050e-01 7.27730393e-01
2.03191027e-01 -2.28162661e-01 3.49548042e-01 -7.45920837e-01
-1.00284481e+00 -2.35672936e-01 -6.87300861e-01 -4.92242545e-01
8.69016647e-01 -2.85464048e-01 -4.27006871e-01 -8.69133137e-03
-1.42168081e+00 2.83054173e-01 3.71083975e-01 6.75432980e-01
3.50861877e-01 -1.65003026e+00 -4.21121150e-01 5.08248925e-01
4.22828645e-01 -8.70315731e-01 6.01577401e-01 4.83397484e-01
-5.17465651e-01 3.47877473e-01 -1.82980895e-01 -3.42014819e-01
-1.30553234e+00 1.23909402e+00 -6.94501325e-02 -1.16139263e-01
-1.10872817e+00 -4.06778511e-03 2.15560079e-01 7.48793110e-02
3.45494509e-01 -6.40756965e-01 -1.04947412e+00 6.08197749e-02
1.36871111e+00 7.05675840e-01 -2.62060106e-01 -4.02811021e-01
-2.62398273e-01 5.87380469e-01 2.68756956e-01 5.82109272e-01
1.29948080e+00 -8.60548377e-01 1.02513038e-01 7.37716496e-01
5.45715749e-01 -1.37767270e-01 -5.38814425e-01 -2.79068023e-01
4.13131081e-02 -4.79913533e-01 1.23067193e-01 -1.37301195e+00
-9.93850887e-01 6.21275127e-01 1.44638956e+00 8.93680334e-01
1.08407426e+00 2.65481025e-02 9.09223914e-01 -4.74296808e-02
3.13347965e-01 -1.10297847e+00 -6.67936802e-02 -3.36315572e-01
7.37578988e-01 -1.39142382e+00 2.22208008e-01 -6.83357954e-01
-5.33813536e-02 1.03353846e+00 -8.58432148e-03 9.74282250e-02
1.46090460e+00 -2.61987466e-02 3.64127427e-01 -4.35316265e-01
-1.03175581e-01 -4.16931063e-01 5.07627666e-01 1.00148606e+00
1.00465536e+00 3.81644726e-01 -6.04880571e-01 2.68133432e-01
2.64446855e-01 6.41294599e-01 -3.19864452e-01 1.04611158e+00
-9.39409673e-01 -1.54118395e+00 -4.61129785e-01 1.89850926e-01
-4.97393191e-01 -2.35327035e-02 3.60398710e-01 5.56398094e-01
3.05592000e-01 1.25758588e+00 -1.05942830e-01 3.36877257e-01
3.02501649e-01 2.15479776e-01 5.55293262e-01 -7.12083340e-01
-1.03673530e+00 6.83504283e-01 2.93678850e-01 -6.57972097e-01
-1.03228652e+00 -6.37586296e-01 -1.15885580e+00 -3.43452275e-01
2.64586061e-02 -5.53931817e-02 8.34910989e-01 6.23061419e-01
4.39585179e-01 2.81026036e-01 -4.63010930e-02 -4.64112788e-01
1.88223436e-01 -7.99370229e-01 -7.20658362e-01 6.78550541e-01
6.81359470e-02 -4.49354023e-01 3.01680744e-01 2.20454782e-01] | [11.546428680419922, 4.478516101837158] |
635be74a-6998-43ef-840c-03013f3c2ed1 | low-resource-white-box-semantic-segmentation | 2306.07809 | null | https://arxiv.org/abs/2306.07809v1 | https://arxiv.org/pdf/2306.07809v1.pdf | Low-Resource White-Box Semantic Segmentation of Supporting Towers on 3D Point Clouds via Signature Shape Identification | Research in 3D semantic segmentation has been increasing performance metrics, like the IoU, by scaling model complexity and computational resources, leaving behind researchers and practitioners that (1) cannot access the necessary resources and (2) do need transparency on the model decision mechanisms. In this paper, we propose SCENE-Net, a low-resource white-box model for 3D point cloud semantic segmentation. SCENE-Net identifies signature shapes on the point cloud via group equivariant non-expansive operators (GENEOs), providing intrinsic geometric interpretability. Our training time on a laptop is 85~min, and our inference time is 20~ms. SCENE-Net has 11 trainable geometrical parameters and requires fewer data than black-box models. SCENE--Net offers robustness to noisy labeling and data imbalance and has comparable IoU to state-of-the-art methods. With this paper, we release a 40~000 Km labeled dataset of rural terrain point clouds and our code implementation. | ['Patrizio Frosini', 'Manuel Silva', 'Alex Coronati', 'Giovanni Bocchi', 'Alessandra Micheletti', 'Cláudia Soares', 'Diogo Lavado'] | 2023-06-13 | null | null | null | null | ['3d-semantic-segmentation'] | ['computer-vision'] | [ 1.06215410e-01 2.83301950e-01 -1.31002754e-01 -5.32802165e-01
-6.89872682e-01 -9.72517848e-01 1.77704412e-02 1.42258182e-01
-2.18978226e-01 1.60183564e-01 -4.98877794e-01 -9.73043263e-01
6.72165230e-02 -1.03144467e+00 -8.44499648e-01 -1.99577555e-01
-1.76081091e-01 1.12417042e+00 5.64104855e-01 1.00428285e-02
3.76951814e-01 8.47664595e-01 -1.43393981e+00 -1.46116406e-01
1.24281204e+00 1.03240478e+00 2.51072973e-01 8.21313560e-01
-5.50676525e-01 2.34705329e-01 -2.89935827e-01 -1.48787454e-01
7.07792580e-01 2.37609565e-01 -1.07168579e+00 1.44789666e-01
6.32489145e-01 -2.97443271e-01 -3.65435891e-02 1.14341116e+00
3.99779499e-01 -1.78521331e-02 5.36898017e-01 -1.17769527e+00
-1.06686518e-01 1.83359876e-01 -6.96561754e-01 -2.49311631e-03
-1.00581020e-01 1.22139513e-01 8.78256083e-01 -8.02424192e-01
6.92719579e-01 1.34447968e+00 9.04908419e-01 2.95146197e-01
-1.18981755e+00 -5.86679935e-01 2.63223559e-01 -3.56983066e-01
-1.60721517e+00 -1.76072344e-01 4.26638633e-01 -5.62921047e-01
1.16721082e+00 5.36384523e-01 7.26389468e-01 2.19411373e-01
-2.60634899e-01 7.74143696e-01 7.81313956e-01 -1.42631114e-01
4.59073633e-01 -2.20379442e-01 2.30187058e-01 8.30031872e-01
3.52364302e-01 -2.78938651e-01 -8.32338035e-02 -3.03550303e-01
1.05364537e+00 -1.37838885e-01 3.05323452e-01 -4.96479303e-01
-8.70732129e-01 6.59723759e-01 5.98465502e-01 -3.29383850e-01
1.11715965e-01 3.77135545e-01 1.18657798e-01 3.70638743e-02
8.38095248e-01 4.94004369e-01 -7.73369551e-01 -2.45352387e-01
-9.94789064e-01 4.17553782e-01 6.54205501e-01 1.59966898e+00
1.11578524e+00 -1.01088434e-01 6.07824624e-01 6.06176317e-01
3.50841612e-01 8.64293993e-01 -3.97463500e-01 -1.26444089e+00
6.31667614e-01 1.01367640e+00 -1.40009355e-02 -1.05652630e+00
-7.66362071e-01 -2.43289396e-01 -5.45862973e-01 2.83015251e-01
2.10072517e-01 -5.37561327e-02 -1.41229868e+00 1.12543154e+00
6.12627864e-01 1.51769027e-01 -1.44518659e-01 6.93571031e-01
8.66384268e-01 5.62046587e-01 2.17787996e-02 5.67594707e-01
1.18557012e+00 -8.68632674e-01 1.60713717e-02 -5.98664522e-01
1.01815033e+00 -7.20504582e-01 1.27686715e+00 5.04087172e-02
-1.04947317e+00 -3.62466276e-01 -8.96816730e-01 -4.77309048e-01
-4.64750141e-01 -6.04300126e-02 1.12715280e+00 7.93007672e-01
-1.05466545e+00 6.39335394e-01 -1.23862493e+00 -4.71857309e-01
8.61327529e-01 5.79016268e-01 1.10305082e-02 7.49203265e-02
-4.26350862e-01 6.48432255e-01 4.33144599e-01 -3.74752991e-02
-5.20958662e-01 -9.05315340e-01 -8.94581616e-01 -9.43943411e-02
3.65067363e-01 -7.99153984e-01 1.32447696e+00 -4.90794212e-01
-1.11665952e+00 1.17461360e+00 -2.02299461e-01 -2.82221019e-01
6.17067277e-01 -2.02055663e-01 2.78192163e-02 2.14260250e-01
4.21590269e-01 1.04158354e+00 3.34796250e-01 -1.54821837e+00
-7.43176699e-01 -5.20072997e-01 1.39955372e-01 4.88426596e-01
1.85022637e-01 -1.76116183e-01 -8.34742129e-01 -3.18147510e-01
9.57380831e-01 -1.13828719e+00 -7.06654072e-01 2.81922340e-01
-5.91024399e-01 4.44517620e-02 1.14680135e+00 -4.01983947e-01
9.66945171e-01 -2.22690868e+00 -3.90020192e-01 5.68209469e-01
2.79717624e-01 1.41703542e-02 1.33643940e-01 6.60303757e-02
2.49940246e-01 7.70278752e-01 -6.89945936e-01 -2.63262212e-01
4.95198257e-02 5.42421877e-01 -3.89512926e-01 3.50192904e-01
3.73799764e-02 7.55187452e-01 -7.54360020e-01 -6.69153452e-01
4.41687137e-01 -7.95737952e-02 -8.90753508e-01 -1.67727008e-01
-6.21925116e-01 1.83057562e-01 -7.80450761e-01 1.10852718e+00
1.33868515e+00 -4.99056846e-01 -3.16228718e-02 1.37171730e-01
-2.26767495e-01 2.99728692e-01 -1.19601703e+00 1.89934540e+00
-2.04523206e-01 5.44068873e-01 5.59894741e-02 -4.64842081e-01
9.91827369e-01 -2.05300868e-01 4.84282851e-01 -2.90592462e-01
1.00162581e-01 4.98495936e-01 -3.82684201e-01 -2.86344230e-01
7.71385193e-01 3.79321694e-01 -2.24156231e-01 2.51845568e-01
-3.77342820e-01 -8.65997791e-01 -6.87071979e-02 1.95824340e-01
1.00498700e+00 3.94769907e-01 -7.97497761e-03 -6.08546615e-01
-1.79432452e-01 7.04345107e-01 5.54326773e-01 8.67043436e-01
1.20388769e-01 5.79650044e-01 4.65053588e-01 -4.75366175e-01
-1.17937374e+00 -1.09171951e+00 -3.97232413e-01 7.90057957e-01
6.86072409e-01 -4.06496853e-01 -9.29906845e-01 -3.89641345e-01
2.16161117e-01 6.22518063e-01 -1.29128516e-01 5.80089509e-01
-5.70576727e-01 -7.68618226e-01 7.18331993e-01 7.32532203e-01
8.04709494e-01 -4.68541592e-01 -6.57792687e-01 8.47675726e-02
-2.73443181e-02 -1.20461011e+00 -1.10235378e-01 2.31419906e-01
-1.56230748e+00 -8.73953581e-01 -7.66049027e-02 -7.30498970e-01
1.12630081e+00 5.04318833e-01 1.31174314e+00 2.04945445e-01
-2.40511507e-01 -3.57298963e-02 -2.76906848e-01 -7.67155886e-01
8.35545436e-02 5.71091175e-01 -1.26064494e-01 -8.09790969e-01
5.00318587e-01 -6.11323118e-01 -4.28245306e-01 5.56155324e-01
-6.63860917e-01 4.53394383e-01 2.03178212e-01 3.19108069e-01
1.07047129e+00 1.05926722e-01 -2.36447290e-01 -1.09192526e+00
-1.64622098e-01 -2.88790137e-01 -1.01638687e+00 -1.83043126e-02
-4.57375169e-01 -1.25589073e-01 1.18639223e-01 1.82579607e-01
-8.84974957e-01 4.06420052e-01 -1.90384984e-01 -3.32106292e-01
-3.12065929e-01 4.08702433e-01 -3.19518805e-01 -4.71015334e-01
6.78515613e-01 -4.06361073e-01 -4.18679118e-01 -7.44895339e-01
4.67278600e-01 7.24609196e-01 5.53113699e-01 -8.16722810e-01
1.06109750e+00 9.53516185e-01 1.79786831e-01 -1.10840178e+00
-8.64070058e-01 -6.67621613e-01 -1.01185060e+00 1.06945492e-01
9.94128585e-01 -1.07324064e+00 -4.88263160e-01 5.44864833e-01
-1.21572316e+00 -6.61891103e-01 -1.87533870e-01 9.80916023e-02
-6.09971702e-01 1.11087874e-01 -5.12343228e-01 -5.81821680e-01
-1.92065164e-01 -1.08408570e+00 1.44468904e+00 1.90887332e-01
-2.42368262e-02 -9.05289292e-01 -3.96605134e-01 4.74833608e-01
-7.44026825e-02 3.78934860e-01 9.75182772e-01 -2.64625698e-01
-1.30506635e+00 -7.49156252e-02 -5.73814631e-01 -1.47933662e-01
-1.35944992e-01 2.14029849e-01 -1.14276326e+00 7.44779631e-02
-3.16186219e-01 1.22244395e-01 5.29477835e-01 6.00297868e-01
1.71064413e+00 -6.41405359e-02 -5.40695429e-01 1.27900147e+00
1.43588185e+00 1.22731172e-01 6.88759744e-01 4.00680721e-01
1.06470895e+00 4.71872568e-01 8.23262930e-01 1.84990391e-01
6.32177889e-01 3.89385372e-01 5.33916235e-01 -6.95467234e-01
2.89240003e-01 -4.27836955e-01 -3.60611469e-01 7.45179594e-01
-1.66042387e-01 -3.04864705e-01 -1.50746679e+00 4.95615542e-01
-1.94523227e+00 -5.26509106e-01 -6.74719930e-01 1.74014592e+00
3.30636621e-01 3.73226345e-01 -1.84272304e-01 -1.44371718e-01
3.87163848e-01 6.58752546e-02 -8.27170134e-01 -3.68984818e-01
-1.66174471e-01 2.12422326e-01 1.39735460e+00 6.20230198e-01
-1.36984980e+00 1.58409119e+00 6.56711197e+00 9.01284158e-01
-8.39703977e-01 -7.52789080e-02 8.15365016e-01 1.66326165e-01
-4.05477285e-01 2.95583487e-01 -8.55787992e-01 1.07732736e-01
5.66291094e-01 3.73653807e-02 2.67726392e-01 1.35975504e+00
2.62835264e-01 -3.05013627e-01 -7.91623294e-01 1.01711440e+00
-3.33371878e-01 -1.52437520e+00 -7.73515627e-02 3.07439983e-01
8.14029217e-01 5.55701971e-01 -1.30392939e-01 -2.24738419e-01
6.07791901e-01 -1.06132221e+00 8.01235795e-01 1.95650309e-01
1.12601125e+00 -6.95229113e-01 5.32374322e-01 2.63261527e-01
-1.43612051e+00 2.82479346e-01 -6.06211901e-01 -1.21744186e-01
1.84339717e-01 4.71874386e-01 -1.02386045e+00 5.20039678e-01
1.07527316e+00 5.91987789e-01 -6.78342760e-01 9.75024462e-01
-6.55038282e-02 7.00895250e-01 -8.91294241e-01 1.62511304e-01
4.62127537e-01 -4.93209749e-01 5.05135417e-01 1.00833189e+00
2.83470660e-01 4.91946131e-01 4.09329653e-01 8.64273548e-01
-3.59659665e-03 -1.20061062e-01 -6.26604259e-01 3.37385498e-02
7.20520794e-01 1.04895353e+00 -1.42958128e+00 -3.48394215e-01
-7.50708114e-03 7.65691698e-01 -1.84201464e-01 2.47230038e-01
-6.65733516e-01 -4.38335568e-01 1.02144623e+00 1.81918144e-01
1.83284491e-01 -6.61992311e-01 -1.27902448e+00 -8.70307744e-01
-8.54537040e-02 -4.20222640e-01 3.78420986e-02 -9.20378029e-01
-9.62134421e-01 2.59655356e-01 7.69457519e-02 -1.25705791e+00
3.18204761e-01 -8.66914392e-01 -2.10965499e-01 7.40489185e-01
-1.36963320e+00 -1.22119260e+00 -4.92850125e-01 1.93567708e-01
5.06254017e-01 3.78894120e-01 6.68771505e-01 3.42251509e-01
-2.35629275e-01 1.79598346e-01 5.92643046e-04 1.28014863e-01
5.76330870e-02 -1.38863635e+00 1.37679291e+00 7.64691412e-01
-1.37483045e-01 5.89822531e-01 4.80675519e-01 -8.70945275e-01
-1.27270329e+00 -1.33101988e+00 7.79903948e-01 -8.61784816e-01
4.38699424e-01 -5.88478029e-01 -7.82291651e-01 7.79746473e-01
-6.75962210e-01 -3.66979390e-02 5.22917032e-01 1.37139127e-01
-8.11607167e-02 8.18791687e-02 -1.30148411e+00 7.45416284e-01
1.68964756e+00 -4.66820061e-01 -2.11611941e-01 5.30129790e-01
1.03799617e+00 -1.08270466e+00 -7.18726218e-01 4.85644996e-01
2.12842122e-01 -6.51987791e-01 1.06531155e+00 -4.40091938e-01
2.19649654e-02 -7.19958484e-01 -2.94813663e-01 -8.18549097e-01
-1.93475768e-01 -4.70582187e-01 3.62457544e-01 7.28160381e-01
6.27779603e-01 -5.85726678e-01 1.24398315e+00 7.88908839e-01
-6.45865023e-01 -5.38332880e-01 -7.87733614e-01 -7.17228651e-01
8.35344288e-03 -9.51875925e-01 1.09446859e+00 9.41549778e-01
-5.88352025e-01 -2.88836728e-03 1.81459829e-01 6.46460891e-01
6.63987100e-01 2.76290447e-01 1.19022655e+00 -1.51881266e+00
1.55894130e-01 -3.27672422e-01 -3.63274068e-01 -1.51791716e+00
-2.46226877e-01 -1.01644695e+00 -4.81755547e-02 -1.77698660e+00
-1.73334315e-01 -1.14990199e+00 5.58647215e-01 8.23283553e-01
1.13744728e-01 2.05797002e-01 3.30587551e-02 3.60294729e-01
-4.54275578e-01 1.40090197e-01 1.09653413e+00 -5.19710556e-02
-4.81687784e-01 -6.51464537e-02 -5.89111388e-01 1.20531726e+00
8.97401750e-01 -4.95537847e-01 -4.44294482e-01 -1.03902125e+00
4.70868915e-01 -2.43201852e-01 4.50872451e-01 -9.84928191e-01
-1.02631580e-02 -4.08452541e-01 2.26944506e-01 -1.20353591e+00
3.05775523e-01 -8.94702911e-01 3.28435898e-01 1.23438872e-01
2.62630343e-01 9.32947639e-03 5.03012359e-01 2.09365770e-01
2.47507215e-01 -2.45106518e-01 5.30596912e-01 -3.99658233e-01
-9.22098637e-01 6.38419986e-01 -6.89640939e-02 4.22608852e-02
8.97617519e-01 -8.11419606e-01 -3.85091871e-01 2.34669805e-01
-5.09813309e-01 5.62594771e-01 1.10050285e+00 1.77738488e-01
3.83083940e-01 -8.28092337e-01 -2.37399906e-01 1.67602330e-01
4.07555811e-02 1.02979434e+00 3.45481724e-01 2.40120634e-01
-1.54959595e+00 2.98126876e-01 1.36509225e-01 -1.16270554e+00
-1.17485905e+00 2.21176296e-02 2.95510590e-01 2.76666790e-01
-8.36701453e-01 1.10795891e+00 1.12994596e-01 -1.03531432e+00
-1.20004654e-01 -6.65756702e-01 4.30857092e-01 -1.76088750e-01
-7.27640837e-02 5.79398215e-01 1.99880242e-01 -4.31290060e-01
-4.65913475e-01 9.11237121e-01 3.68714362e-01 -9.63551328e-02
1.36211658e+00 5.59934750e-02 -2.28404596e-01 2.76008755e-01
8.64667535e-01 -2.14887619e-01 -1.36801529e+00 1.46138668e-01
1.79224357e-01 -6.91056192e-01 2.05228686e-01 -5.64532995e-01
-7.75072694e-01 9.16885555e-01 6.08139396e-01 1.99469492e-01
8.52039874e-01 1.33009240e-01 7.72141278e-01 6.32494748e-01
7.06447780e-01 -1.37018406e+00 -4.97757107e-01 7.07505465e-01
3.17490667e-01 -1.06409895e+00 2.91621119e-01 -1.11613405e+00
-4.05436635e-01 9.20297623e-01 6.08492672e-01 -7.34328777e-02
8.02531660e-01 4.11211163e-01 2.76447475e-01 -6.42814636e-01
-1.90843314e-01 -3.50937456e-01 1.23791229e-02 6.12416506e-01
-1.10122830e-01 3.76667798e-01 2.09974721e-01 7.62935132e-02
-7.81113863e-01 -2.06814766e-01 2.24556118e-01 9.61450815e-01
-6.98444366e-01 -8.32807899e-01 -4.22103286e-01 6.38879538e-01
-7.82616145e-04 -1.26438051e-01 -2.75234640e-01 8.50354612e-01
2.48001009e-01 8.11630487e-01 5.56832492e-01 -3.66209030e-01
4.67634290e-01 -3.08930665e-01 2.01023579e-01 -8.08766484e-01
-8.23336765e-02 5.71458340e-02 2.15563416e-01 -6.22255921e-01
-3.11323285e-01 -7.24770188e-01 -1.68694341e+00 -6.24830961e-01
-4.38269377e-01 -1.51269808e-01 1.02720797e+00 6.92885518e-01
6.95599496e-01 2.05384791e-01 5.23611367e-01 -9.69427586e-01
-3.84922326e-02 -4.94166225e-01 -5.50905168e-01 -1.50031209e-01
-7.64280111e-02 -5.14736652e-01 -2.15176672e-01 1.28153548e-01] | [7.964091777801514, -3.2657992839813232] |
f566bc21-b6bf-44e7-9ee8-8c442451ead9 | building-dynamic-knowledge-graphs-from-text-2 | 1910.09532 | null | https://arxiv.org/abs/1910.09532v3 | https://arxiv.org/pdf/1910.09532v3.pdf | Building Dynamic Knowledge Graphs from Text-based Games | We are interested in learning how to update Knowledge Graphs (KG) from text. In this preliminary work, we propose a novel Sequence-to-Sequence (Seq2Seq) architecture to generate elementary KG operations. Furthermore, we introduce a new dataset for KG extraction built upon text-based game transitions (over 300k data points). We conduct experiments and discuss the results. | ['Marc-Alexandre Côté', 'Mikuláš Zelinka', 'Xingdi Yuan', 'Romain Laroche', 'Adam Trischler'] | 2019-10-21 | null | null | null | null | ['text-based-games'] | ['playing-games'] | [ 1.00035384e-01 1.66375667e-01 -2.75135607e-01 -3.03422272e-01
-4.67661113e-01 -7.20545709e-01 5.06926715e-01 3.69636625e-01
-7.57125854e-01 1.10891140e+00 2.59693414e-01 -6.01818681e-01
-1.85859129e-01 -1.20099723e+00 -9.80085313e-01 5.82271582e-03
-5.53154409e-01 6.74578607e-01 6.27991736e-01 -5.06996572e-01
4.41724271e-01 1.76431030e-01 -1.37874281e+00 4.79078025e-01
8.50082219e-01 5.27445912e-01 -5.72108477e-02 1.14302206e+00
-3.51322472e-01 1.32057965e+00 -7.29394019e-01 -8.38891089e-01
1.98201418e-01 -4.83793765e-01 -1.27671111e+00 -5.15562892e-01
1.99844256e-01 -2.42682174e-01 -5.19056916e-01 8.81851375e-01
5.94510972e-01 5.58190167e-01 6.54333532e-01 -1.14506888e+00
-1.95522651e-01 1.43829429e+00 -7.67424852e-02 3.48871768e-01
7.23189712e-01 2.00712919e-01 8.76789927e-01 -5.14198303e-01
9.57286716e-01 9.45967913e-01 8.33625495e-01 4.93989438e-01
-8.21107924e-01 -4.50817198e-01 1.78718179e-01 7.96232641e-01
-1.42286563e+00 -3.98656428e-01 4.98013526e-01 -3.05496156e-01
1.68459129e+00 1.09936148e-01 1.12216222e+00 1.00738287e+00
4.18171585e-01 1.15442157e+00 6.83504999e-01 -6.00034416e-01
1.99561059e-01 -5.86310983e-01 1.50919586e-01 1.16637337e+00
4.70094353e-01 2.07836345e-01 -9.68607903e-01 1.09440416e-01
6.12885237e-01 -6.56150401e-01 1.12609543e-01 -4.16645437e-01
-9.44761395e-01 7.45663822e-01 2.04918105e-02 -1.18805699e-01
-7.38386512e-02 6.20879710e-01 7.60527551e-01 5.38534880e-01
9.29950625e-02 5.88801205e-01 -6.85704052e-01 -9.53955114e-01
-7.26032972e-01 4.90685105e-01 1.23334706e+00 1.27144384e+00
7.03999877e-01 1.45925745e-01 -2.42445916e-01 4.13802713e-01
-8.42469111e-02 1.41390055e-01 5.35411000e-01 -5.20159423e-01
6.92472816e-01 3.45407963e-01 -3.59842986e-01 -6.38036966e-01
-5.89557171e-01 -1.55922949e-01 -5.10742366e-01 -2.84407258e-01
1.92424297e-01 -3.52634609e-01 -1.02261496e+00 1.31355786e+00
1.77784115e-01 2.52501100e-01 2.18138948e-01 2.43425325e-01
1.10271955e+00 3.27544957e-01 1.20292701e-01 -1.81040451e-01
8.84131074e-01 -1.01073313e+00 -7.12595820e-01 -1.71540156e-01
1.10076785e+00 -1.10644244e-01 9.08599436e-01 7.69061327e-01
-1.15143681e+00 -2.76608974e-01 -9.12251651e-01 1.80674866e-01
-7.33290970e-01 -3.59496623e-01 9.97145593e-01 7.88959980e-01
-1.09510052e+00 7.79504776e-01 -7.53484666e-01 -2.92316854e-01
2.56717026e-01 4.46658045e-01 -2.85121530e-01 3.72785442e-02
-1.63403857e+00 1.01157665e+00 1.52998364e+00 1.13177486e-02
-8.44956815e-01 -5.55587828e-01 -1.20988965e+00 -2.22761080e-01
1.03259826e+00 -8.10540140e-01 1.45878124e+00 -8.71571153e-02
-1.89236009e+00 3.38730305e-01 4.23382252e-01 -6.05766416e-01
2.89316118e-01 -2.49136150e-01 -4.77528423e-01 -1.24585390e-01
-3.44793051e-01 4.42777365e-01 7.11345792e-01 -7.81585693e-01
-9.08149123e-01 -1.05494596e-01 2.96495736e-01 2.17085257e-01
-1.41051844e-01 -1.31026596e-01 -7.87443399e-01 -8.03387582e-01
-3.22263539e-01 -9.23150003e-01 -3.84900033e-01 -1.13977599e+00
-5.69099009e-01 -1.34808525e-01 4.45550270e-02 -8.82985115e-01
1.75453877e+00 -1.60198212e+00 3.27061683e-01 4.40395892e-01
2.77952403e-01 2.34441429e-01 -6.03006147e-02 9.61145759e-01
4.52850126e-02 1.19191281e-01 -1.36712930e-04 1.61552168e-02
2.58871913e-01 3.08107644e-01 -2.58505289e-02 -1.39354542e-01
-3.13319974e-02 1.59057236e+00 -1.32417512e+00 -7.99491584e-01
5.56223355e-02 -4.85612929e-01 -7.69046664e-01 -1.47954300e-01
-4.16218549e-01 -1.76560029e-01 -3.15755159e-01 6.68740332e-01
2.64631301e-01 1.89147055e-01 6.06121659e-01 5.16930260e-02
1.42903507e-01 3.83463740e-01 -1.18537903e+00 2.17188978e+00
-2.12893948e-01 3.78105372e-01 -6.33217216e-01 -9.10987556e-01
6.81872189e-01 -2.08483070e-01 2.02449128e-01 -6.59985244e-01
5.30242063e-02 -1.87582821e-01 2.25490313e-02 -3.87423575e-01
1.16351604e+00 5.87366261e-02 -7.55025923e-01 2.65565664e-01
5.30352831e-01 -4.32796121e-01 8.84626031e-01 5.61934710e-01
1.53810263e+00 2.01864138e-01 7.52882242e-01 -2.30880128e-03
2.44350284e-01 3.80155206e-01 2.96677291e-01 1.06520283e+00
6.15124330e-02 -1.74951151e-01 7.29289949e-01 -4.90960956e-01
-6.15971625e-01 -1.08301222e+00 4.86182481e-01 1.30396461e+00
-1.34856313e-01 -1.32783949e+00 -5.18699646e-01 -1.10177565e+00
7.06999153e-02 7.23156512e-01 -5.16751766e-01 -3.96029592e-01
-6.10100329e-01 -5.53613424e-01 1.19885683e+00 8.32009256e-01
4.75459009e-01 -1.34498799e+00 -1.65740371e-01 5.51584244e-01
-1.01345077e-01 -1.02358830e+00 -4.91474956e-01 5.06518960e-01
-7.87025869e-01 -1.24022758e+00 9.37979296e-02 -6.82584047e-01
3.14323336e-01 -5.02714992e-01 1.28508019e+00 -2.19152570e-01
-2.93860644e-01 4.01173472e-01 -8.16080868e-01 -5.45617998e-01
-5.35941720e-01 5.99344909e-01 1.69355690e-01 -8.97361457e-01
2.48664141e-01 -5.00730515e-01 4.89166304e-02 -2.38839090e-01
-8.55436265e-01 2.18409553e-01 7.40503609e-01 6.55404687e-01
3.61063123e-01 6.44712746e-01 4.41315681e-01 -1.27295995e+00
1.14294767e+00 -1.42918751e-01 -5.74817598e-01 2.57363021e-01
-6.54519498e-01 5.83103061e-01 6.85639322e-01 -3.59212518e-01
-8.71596575e-01 3.34979504e-01 -3.71183455e-01 -1.41875982e-01
3.12816463e-02 1.21398628e+00 -6.02936000e-02 -1.96438804e-01
7.33099699e-01 3.32073897e-01 -3.62254441e-01 -1.82913527e-01
8.83647859e-01 2.96441257e-01 7.69026101e-01 -1.07838631e+00
7.27856219e-01 -2.26706475e-01 -3.78225297e-02 -6.03481710e-01
-3.82691324e-01 -3.59390110e-01 -8.72542739e-01 -4.53170180e-01
3.36323291e-01 -5.61686456e-01 -8.58744025e-01 6.83724403e-01
-7.23504007e-01 -9.62435246e-01 -5.47190368e-01 2.86583036e-01
-8.38991821e-01 5.61090112e-01 -8.35489631e-01 -5.05829930e-01
-6.12017691e-01 -5.98733187e-01 5.22879243e-01 8.23738277e-02
-2.64148325e-01 -1.04630828e+00 5.01833558e-01 1.31581843e-01
-3.24069262e-02 9.64405462e-02 8.53780985e-01 -8.24184656e-01
-4.58620340e-01 -7.64762312e-02 3.64005089e-01 1.20336406e-01
9.70532075e-02 -9.13591236e-02 -1.94240302e-01 -2.56039768e-01
-9.16288316e-01 -5.98431587e-01 9.70332980e-01 1.48807928e-01
1.40790725e+00 -2.03348786e-01 -3.43129158e-01 5.89915514e-01
1.29403222e+00 1.96921557e-01 9.58999336e-01 4.08617079e-01
9.35281038e-01 1.14459351e-01 7.43037879e-01 5.99899709e-01
5.53818524e-01 5.78247070e-01 8.49399194e-02 4.95685130e-01
-4.40063924e-02 -7.93370962e-01 6.92505836e-01 1.19732976e+00
-3.21899265e-01 -3.48141283e-01 -1.25185835e+00 6.61785424e-01
-1.89979887e+00 -9.76968765e-01 -1.88342273e-01 1.87277818e+00
1.31208885e+00 7.12056339e-01 2.83176273e-01 2.09606916e-01
3.32656652e-01 -7.86925480e-02 -4.21976209e-01 -7.06514239e-01
1.13477223e-01 9.56750035e-01 8.49321425e-01 5.31129539e-01
-1.04427826e+00 1.71755767e+00 7.65917969e+00 1.38004780e+00
-5.81446648e-01 -5.09091139e-01 -1.63207814e-01 -5.34666963e-02
-1.77595407e-01 -6.22971635e-03 -9.33994412e-01 1.81487784e-01
1.07596540e+00 -5.51728308e-01 6.34826243e-01 6.83186233e-01
-4.08793747e-01 -2.98645914e-01 -8.73931885e-01 8.45284879e-01
5.12806997e-02 -1.42621136e+00 3.74096811e-01 -2.38965467e-01
6.89522862e-01 -1.93256035e-01 -4.01042402e-01 9.38810587e-01
1.38819909e+00 -1.02698958e+00 6.00353837e-01 6.60046279e-01
1.05774319e+00 -1.16958821e+00 5.63305736e-01 4.29947942e-01
-1.72204149e+00 7.36517981e-02 -1.98978022e-01 -3.04122835e-01
7.58160949e-02 4.34465080e-01 -1.49951625e+00 1.28255153e+00
5.97128093e-01 9.42028701e-01 -9.16362226e-01 8.19290757e-01
-7.54980803e-01 5.39976597e-01 -1.39505178e-01 -3.41316819e-01
3.65838297e-02 7.72889778e-02 3.15240353e-01 1.47326374e+00
7.33668404e-03 4.46251124e-01 1.41851768e-01 3.89497101e-01
-1.55194536e-01 9.99538898e-02 -7.71822870e-01 -4.90827888e-01
2.78580368e-01 1.03754163e+00 -8.60615730e-01 -5.69199324e-01
-1.36279687e-01 1.07938671e+00 6.82612956e-01 8.91526118e-02
-7.03755736e-01 -8.77828121e-01 5.69951594e-01 -2.15196058e-01
5.15907764e-01 -4.44649428e-01 2.51805395e-01 -1.24876547e+00
-2.21004710e-01 -9.87154067e-01 8.73298764e-01 -6.49116397e-01
-1.18071067e+00 1.56621605e-01 4.40636039e-01 -6.21985376e-01
-6.32795453e-01 -5.43126464e-01 -2.70837873e-01 4.27594781e-01
-1.07947135e+00 -8.29145551e-01 -1.34748682e-01 7.47616947e-01
3.62668633e-01 -4.92460579e-02 6.87065721e-01 9.71239135e-02
-3.46143544e-01 8.58904183e-01 8.50570351e-02 5.47487855e-01
4.50909108e-01 -1.60009575e+00 1.15993237e+00 7.36932755e-01
3.74857038e-01 5.35353720e-01 4.39491361e-01 -1.20025611e+00
-1.62219977e+00 -1.27061903e+00 5.47958195e-01 -5.85448682e-01
8.29501450e-01 -4.41900909e-01 -5.19314170e-01 1.07380044e+00
-7.73243383e-02 -4.22510624e-01 7.07082093e-01 4.29556221e-01
3.06449998e-02 1.74616620e-01 -7.25738406e-01 7.69753814e-01
1.78696156e+00 -4.49571550e-01 -8.75467956e-01 -2.49210283e-01
7.53757596e-01 -1.03289318e+00 -1.12442434e+00 7.00551689e-01
5.38452566e-01 -3.56375962e-01 9.26001430e-01 -1.24269664e+00
2.62927562e-01 -1.65079430e-01 2.86325097e-01 -1.88845205e+00
-3.53314072e-01 -7.13408828e-01 -6.67296648e-01 8.60629618e-01
4.98953730e-01 -3.37777168e-01 1.40648448e+00 -2.72447104e-03
-4.34770167e-01 -7.09506631e-01 -5.61690986e-01 -1.13038969e+00
-6.31781891e-02 -9.12459910e-01 7.42759228e-01 8.91521513e-01
7.49094784e-01 5.07664382e-01 -4.83316213e-01 -1.76170602e-01
3.12702537e-01 1.42499115e-02 9.33838904e-01 -9.46929455e-01
-3.99837375e-01 -2.94260144e-01 -7.04609275e-01 -1.15779889e+00
2.19766065e-01 -1.18386364e+00 1.95954099e-01 -1.68976581e+00
2.18497187e-01 4.36289515e-03 -2.43292838e-01 8.16708148e-01
-3.56752157e-01 -1.99611992e-01 -6.36728331e-02 -4.91268218e-01
-1.16398227e+00 6.38273835e-01 1.13183606e+00 -2.51967758e-01
-3.62772703e-01 -2.63931364e-01 -5.62888026e-01 3.57230157e-01
1.01704466e+00 -4.44570273e-01 -7.07100034e-01 -1.22890495e-01
8.74461114e-01 6.70489520e-02 -2.20742762e-01 -1.02992666e+00
6.14443898e-01 -3.70908946e-01 1.26920983e-01 -9.96174574e-01
6.49665250e-03 -1.42706186e-01 1.98306397e-01 5.98270655e-01
-1.16914377e-01 1.63775608e-01 6.90161169e-01 6.43455625e-01
-1.16919182e-01 -4.54048812e-01 4.38493714e-02 -2.84581810e-01
-1.44703853e+00 3.18671674e-01 -7.47128904e-01 4.05280322e-01
1.00703633e+00 -2.29310691e-01 -4.40044194e-01 -4.85439539e-01
-1.01055527e+00 4.49223101e-01 2.74520904e-01 3.84709746e-01
8.24515402e-01 -1.27959204e+00 -5.17033875e-01 -2.57196613e-02
4.15532798e-01 -8.94828737e-02 6.66681975e-02 1.79863960e-01
-9.09111321e-01 6.14080369e-01 -3.22706401e-01 1.40154436e-01
-1.34247816e+00 6.51434064e-01 1.41230881e-01 -7.42765963e-01
-3.56100827e-01 9.67426419e-01 -6.25401437e-01 -8.42476368e-01
7.13487491e-02 -6.41704440e-01 -3.51932734e-01 -5.01490831e-02
2.78269023e-01 4.15654182e-01 4.72711682e-01 2.07324684e-01
-3.18168461e-01 -5.07759005e-02 -2.51880795e-01 -1.15995727e-01
1.24918962e+00 5.03166437e-01 2.98324842e-02 3.62512261e-01
5.56447089e-01 6.44243881e-02 -6.42950892e-01 -1.47359639e-01
4.60391551e-01 -2.60386884e-01 -3.76051366e-01 -7.72352338e-01
-7.50856161e-01 1.95351243e-01 -1.06072985e-01 -4.71849442e-02
1.04377103e+00 -6.81181028e-02 6.41616106e-01 1.03496385e+00
8.31564784e-01 -1.49275112e+00 2.55862385e-01 1.21993303e+00
3.78848553e-01 -7.61951864e-01 1.94050625e-01 -4.96499956e-01
-5.63059926e-01 9.95165348e-01 7.66549528e-01 3.22856545e-01
4.93618399e-01 3.90206933e-01 -5.34108758e-01 -4.56649631e-01
-1.08683157e+00 -6.66115224e-01 1.18521161e-01 9.13384974e-01
1.06312996e-02 1.20697297e-01 -2.83466190e-01 8.94324958e-01
-8.07415545e-01 3.93558085e-01 7.22629011e-01 1.42400503e+00
-3.85828227e-01 -1.54712355e+00 3.43619240e-03 9.70835626e-01
-1.05180889e-01 -6.00652337e-01 -7.30132341e-01 1.00172496e+00
-7.27751255e-02 6.11339808e-01 -1.92408308e-01 -9.13432062e-01
6.44944727e-01 3.35460246e-01 1.06622100e+00 -9.58944917e-01
-4.77719247e-01 -7.09186256e-01 8.51627469e-01 -5.76710463e-01
6.62248507e-02 -6.23508334e-01 -1.39970672e+00 -7.97609687e-01
-3.06907028e-01 3.15916389e-01 2.76033282e-01 6.83507144e-01
6.60115033e-02 8.25323999e-01 8.44827853e-03 -1.40464455e-01
-1.17570803e-01 -1.04024136e+00 -8.94878447e-01 1.17895596e-01
-4.38254952e-01 -4.27584648e-01 1.70047745e-01 2.75971413e-01] | [9.254733085632324, 8.096877098083496] |
48aa36a5-7378-47df-9c29-301b6bebb945 | dcmt-a-direct-entire-space-causal-multi-task | 2302.06141 | null | https://arxiv.org/abs/2302.06141v1 | https://arxiv.org/pdf/2302.06141v1.pdf | DCMT: A Direct Entire-Space Causal Multi-Task Framework for Post-Click Conversion Estimation | In recommendation scenarios, there are two long-standing challenges, i.e., selection bias and data sparsity, which lead to a significant drop in prediction accuracy for both Click-Through Rate (CTR) and post-click Conversion Rate (CVR) tasks. To cope with these issues, existing works emphasize on leveraging Multi-Task Learning (MTL) frameworks (Category 1) or causal debiasing frameworks (Category 2) to incorporate more auxiliary data in the entire exposure/inference space D or debias the selection bias in the click/training space O. However, these two kinds of solutions cannot effectively address the not-missing-at-random problem and debias the selection bias in O to fit the inference in D. To fill the research gaps, we propose a Direct entire-space Causal Multi-Task framework, namely DCMT, for post-click conversion prediction in this paper. Specifically, inspired by users' decision process of conversion, we propose a new counterfactual mechanism to debias the selection bias in D, which can predict the factual CVR and the counterfactual CVR under the soft constraint of a counterfactual prior knowledge. Extensive experiments demonstrate that our DCMT can improve the state-of-the-art methods by an average of 1.07% in terms of CVR AUC on the five offline datasets and 0.75% in terms of PV-CVR on the online A/B test (the Alipay Search). Such improvements can increase millions of conversions per week in real industrial applications, e.g., the Alipay Search. | ['Yan Wang', 'Guanfeng Liu', 'Fei Wu', 'Chaochao Chen', 'Jun Zhou', 'Tiehua Zhang', 'Lu Yu', 'Longfei Li', 'Xinxing Yang', 'Mingjie Zhong', 'Feng Zhu'] | 2023-02-13 | null | null | null | null | ['selection-bias'] | ['natural-language-processing'] | [ 1.43476740e-01 -4.22721714e-01 -8.78804982e-01 -3.44536304e-01
-9.41505253e-01 -3.31084073e-01 4.27697510e-01 -1.88878357e-01
-1.19640924e-01 1.02086759e+00 1.74110994e-01 -7.69320369e-01
-4.13920611e-01 -7.64715075e-01 -9.04344857e-01 -4.77646351e-01
2.84468025e-01 6.78375661e-02 -4.12305705e-02 5.24429344e-02
4.26201582e-01 -3.38704944e-01 -1.18164945e+00 2.94152528e-01
1.34346235e+00 1.23198116e+00 9.09383893e-02 5.21012768e-03
2.40464117e-02 4.70655352e-01 -4.00824040e-01 -7.88730264e-01
2.53744841e-01 -1.47284195e-01 -2.93984413e-01 -5.10693669e-01
3.46248507e-01 -4.76655036e-01 -3.43974352e-01 9.98721778e-01
3.99606675e-01 1.96658280e-02 5.84738374e-01 -1.35559309e+00
-9.32623506e-01 7.94454873e-01 -9.72834527e-01 7.23956674e-02
9.32514146e-02 1.20102793e-01 1.23544908e+00 -7.97780097e-01
8.04786906e-02 1.30291581e+00 7.39689052e-01 1.06593698e-01
-1.25040090e+00 -1.21980608e+00 4.06829178e-01 2.32672974e-01
-1.00997734e+00 -8.81073847e-02 7.71972358e-01 -5.32804072e-01
2.70567626e-01 4.66981560e-01 2.55903602e-01 1.54828179e+00
4.59955633e-01 9.07181680e-01 1.47842777e+00 -6.62846118e-02
1.82787254e-01 2.22072423e-01 1.65690973e-01 1.11247376e-01
6.04624689e-01 5.04387796e-01 -6.48969173e-01 -4.32905763e-01
9.40900266e-01 4.35638994e-01 1.22182094e-01 -1.65856883e-01
-1.04924226e+00 1.04330397e+00 2.38721102e-01 -2.61945993e-01
-2.60796964e-01 1.58074364e-01 4.18030918e-01 3.96793365e-01
7.36450493e-01 3.08725715e-01 -8.94290268e-01 5.65288514e-02
-9.04777288e-01 5.69764376e-01 8.40213478e-01 9.99259651e-01
3.37998867e-01 -1.45378470e-01 -8.04968715e-01 7.68396676e-01
3.25955391e-01 6.54756486e-01 4.45853233e-01 -6.17535472e-01
9.47721779e-01 3.05506676e-01 4.80763197e-01 -8.28611612e-01
-1.82588138e-02 -7.60282576e-01 -8.74117315e-01 -3.40292841e-01
4.74738628e-01 -2.85657883e-01 -6.11140788e-01 1.85618901e+00
3.16860676e-01 1.55943185e-01 -5.15609026e-01 8.96397471e-01
1.19459145e-01 4.54649299e-01 2.23685279e-01 -6.22824311e-01
1.29302156e+00 -8.63566935e-01 -8.77901316e-01 -2.24089131e-01
3.78618091e-01 -9.67028439e-01 1.58230114e+00 6.11432672e-01
-8.71322930e-01 -5.37700772e-01 -9.09770906e-01 2.29368910e-01
-9.56338719e-02 3.01488936e-01 9.75238740e-01 7.70377874e-01
1.21890545e-01 6.11543417e-01 -4.65290189e-01 3.88031930e-01
6.25373721e-01 -7.90039636e-03 5.76450944e-01 -2.30178580e-01
-1.54220474e+00 5.32161891e-01 -1.15981400e-01 -8.83266404e-02
-1.16348505e+00 -1.41683722e+00 -5.33162914e-02 1.09157108e-01
9.87953305e-01 -7.07699537e-01 1.39251769e+00 -6.02326810e-01
-1.25998247e+00 1.79827601e-01 2.88056862e-02 -4.85115916e-01
1.14409852e+00 -6.81307435e-01 -5.58106780e-01 -8.18291605e-01
3.22292209e-01 -2.06309661e-01 8.57968032e-01 -1.08758414e+00
-8.69777501e-01 -5.85827827e-01 -1.06540889e-01 -3.99783179e-02
-3.70267808e-01 -3.75907511e-01 -3.10079455e-01 -1.23084724e+00
-3.90339792e-01 -8.56462002e-01 -2.59800076e-01 -2.80124784e-01
-6.71961725e-01 -5.12660384e-01 6.89411879e-01 -7.54139841e-01
1.79515874e+00 -1.85985398e+00 -4.19672221e-01 -2.25804262e-02
9.68030244e-02 1.99661508e-01 -6.83016852e-02 1.80667803e-01
-2.21141148e-02 4.94100392e-01 2.21707404e-01 -3.57321054e-02
2.06377923e-01 -1.74985573e-01 -7.89768219e-01 2.06889391e-01
-3.32654327e-01 7.78565705e-01 -8.66769850e-01 -1.65361032e-01
4.79974076e-02 5.00629432e-02 -6.49648011e-01 2.04459280e-01
-4.01238859e-01 2.72631735e-01 -8.04123521e-01 5.71994066e-01
9.49240208e-01 -3.23988736e-01 2.31765777e-01 -2.50775307e-01
-3.12211663e-01 6.88718677e-01 -1.17637408e+00 1.51725137e+00
-9.12464738e-01 8.04548524e-03 -2.77768582e-01 -1.02634335e+00
7.62473881e-01 1.05997205e-01 4.18284416e-01 -9.66662586e-01
-1.51800856e-01 4.36258256e-01 -1.23723604e-01 -3.56371582e-01
1.52994037e-01 -1.88453987e-01 -3.51327270e-01 5.42420089e-01
-3.10038835e-01 4.23717827e-01 -1.76343977e-01 -1.12691335e-01
8.46666038e-01 1.42103925e-01 2.67089903e-01 -3.05869043e-01
9.30947736e-02 -2.39970490e-01 8.53173673e-01 1.10112858e+00
-6.09523663e-03 4.11654115e-02 5.59896231e-01 -3.30397636e-01
-9.97549057e-01 -1.14490211e+00 -2.64176369e-01 1.28735936e+00
2.01538324e-01 -1.17759727e-01 -2.76083201e-01 -8.91098917e-01
3.94442558e-01 1.26237714e+00 -4.72292185e-01 -3.72256726e-01
-3.65255356e-01 -1.05738139e+00 2.08283961e-01 2.34794572e-01
7.61264682e-01 -3.56636465e-01 -3.20070684e-02 2.64682561e-01
-4.14373189e-01 -6.73609495e-01 -7.74483681e-01 -5.10274917e-02
-8.26729059e-01 -9.61634159e-01 -6.47684336e-01 -4.79672998e-02
1.23192601e-01 3.25957745e-01 9.11195874e-01 -4.97612119e-01
-4.87449840e-02 -1.57818615e-01 -2.51303703e-01 -6.34242892e-01
3.72444540e-02 7.95639604e-02 8.80912617e-02 2.75452398e-02
3.72394294e-01 -6.08030617e-01 -9.87162650e-01 8.67432594e-01
-5.36901116e-01 5.88158667e-02 7.68136501e-01 9.70553219e-01
6.05272651e-01 3.51646952e-02 1.06891859e+00 -1.15700316e+00
8.21464539e-01 -9.30971742e-01 -6.93944037e-01 4.98174042e-01
-1.43773830e+00 3.50473784e-02 7.11657584e-01 -9.42581058e-01
-1.42124593e+00 -4.35061395e-01 2.99505860e-01 -4.14797992e-01
4.30090666e-01 6.44874036e-01 -1.43414259e-01 4.07920718e-01
5.02667189e-01 -6.04318501e-03 -4.30477083e-01 -8.52370799e-01
4.52869982e-01 7.58231044e-01 2.08144441e-01 -5.79301119e-01
7.75938213e-01 2.03837261e-01 -1.10179624e-02 1.16754711e-01
-1.54507411e+00 -2.46140465e-01 -8.50856453e-02 8.17818567e-02
4.02440608e-01 -1.16137755e+00 -9.34598565e-01 1.67481780e-01
-9.83564258e-01 -1.82963714e-01 -6.47699982e-02 6.29391432e-01
-4.41366732e-01 1.42226502e-01 -2.53337651e-01 -9.53428149e-01
-3.62331301e-01 -8.43217492e-01 7.57250905e-01 1.80589557e-02
1.17540643e-01 -8.74986291e-01 -3.97498757e-02 7.12364912e-01
4.54031020e-01 2.53071561e-02 1.12978590e+00 -7.41366744e-01
-7.19557941e-01 -3.81239206e-01 -4.46325928e-01 3.07577372e-01
1.65234759e-01 -4.74854261e-01 -7.15157151e-01 -2.40348533e-01
1.58058017e-01 -2.25696266e-01 9.27455723e-01 6.68431759e-01
1.73596287e+00 -7.51065552e-01 -5.12076557e-01 2.33483180e-01
1.32619309e+00 1.33263677e-01 5.11241555e-01 7.07494169e-02
4.78608668e-01 2.90536642e-01 1.18276024e+00 8.95939231e-01
5.00627637e-01 9.50635135e-01 3.99319559e-01 1.74191684e-01
-3.26534174e-02 -9.51621175e-01 2.62698680e-01 4.29506689e-01
-1.31868934e-02 -2.57864982e-01 -3.84056479e-01 3.86310697e-01
-2.00881410e+00 -8.31603885e-01 -3.74040574e-01 2.70194912e+00
1.11691475e+00 2.03173742e-01 8.57375637e-02 -2.69153386e-01
9.64496374e-01 1.27881810e-01 -1.01725972e+00 -1.40866697e-01
2.43881896e-01 -2.70946950e-01 9.16002214e-01 1.05459765e-01
-1.02866304e+00 5.14746308e-01 5.45454979e+00 1.43020725e+00
-7.97034025e-01 6.06677294e-01 8.90983760e-01 -3.31019074e-01
-4.67640072e-01 2.23122507e-01 -1.12374926e+00 1.09622109e+00
8.92355084e-01 -5.71692348e-01 6.50108397e-01 9.10164714e-01
6.64114773e-01 2.24350505e-02 -1.15996265e+00 9.44603384e-01
-2.50628591e-01 -1.36926544e+00 1.64691359e-01 3.56512904e-01
9.73993957e-01 -2.12784097e-01 4.77063477e-01 7.73060620e-01
5.27199447e-01 -7.08548367e-01 7.13345766e-01 6.22116685e-01
9.89260614e-01 -5.05425155e-01 5.69760740e-01 5.12140512e-01
-8.73013020e-01 -4.55978334e-01 -3.83869082e-01 -1.63663551e-01
2.00254768e-01 1.28056479e+00 -4.04023528e-01 6.56022727e-01
6.34994209e-01 5.64496815e-01 -1.27755851e-01 1.10599720e+00
-1.83354422e-01 9.73661005e-01 -8.29730481e-02 -2.41688445e-01
-1.12396441e-01 -2.84848452e-01 4.72520173e-01 7.10389495e-01
4.81384009e-01 -1.38117105e-01 1.54022589e-01 1.21672952e+00
-3.94886345e-01 1.11127114e-02 -2.08122417e-01 2.08208397e-01
7.56723285e-01 9.36114311e-01 1.65797696e-01 -2.29511783e-01
-4.53818679e-01 6.81029439e-01 -1.32129192e-01 4.86752957e-01
-1.35032737e+00 -2.86960155e-01 5.89160681e-01 2.34710827e-01
2.92574078e-01 1.11758016e-01 -7.22042084e-01 -1.29796660e+00
2.46496618e-01 -7.72046268e-01 4.12527472e-01 -3.88691813e-01
-1.74981391e+00 -3.10756683e-01 7.62234777e-02 -1.28584409e+00
2.20985383e-01 -2.54281908e-01 -7.42301702e-01 8.67492318e-01
-1.73759592e+00 -1.11401451e+00 -4.34362963e-02 4.27629381e-01
8.02697778e-01 -3.30684073e-02 3.25119019e-01 6.38321877e-01
-7.27007091e-01 7.76907682e-01 4.59500104e-01 -3.37298393e-01
1.09607589e+00 -9.98372972e-01 1.84441119e-01 3.91877562e-01
-2.01163962e-01 8.29866648e-01 6.36481047e-01 -8.96487892e-01
-1.31606615e+00 -1.29120207e+00 7.78360844e-01 -4.87834692e-01
9.57630038e-01 -4.19581503e-01 -5.88155270e-01 5.39677441e-01
-3.96099746e-01 -2.07201779e-01 7.51797199e-01 7.48847365e-01
-5.89969933e-01 -4.08431202e-01 -9.20219481e-01 6.05107188e-01
1.09894657e+00 -2.95555383e-01 -5.09951830e-01 6.16862714e-01
1.03827846e+00 -1.04173243e-01 -1.01658773e+00 4.14058894e-01
7.83266604e-01 -6.74732089e-01 1.11351609e+00 -8.65812838e-01
7.14096427e-01 8.39158148e-02 -1.99709058e-01 -1.28457224e+00
-2.35903487e-01 -7.62775362e-01 -2.73692995e-01 1.47918200e+00
5.68474174e-01 -6.47608280e-01 5.71509480e-01 6.64696157e-01
5.66695854e-02 -7.17844844e-01 -1.13952267e+00 -9.49142039e-01
1.60453975e-01 -6.00575507e-01 6.73342228e-01 9.25270081e-01
-3.42204928e-01 5.27531683e-01 -9.54168558e-01 -1.83549330e-01
8.82179737e-01 5.74255347e-01 7.17465937e-01 -1.32978678e+00
-3.46294105e-01 -2.12926507e-01 6.41215742e-01 -1.40651846e+00
-1.06058381e-01 -7.05485225e-01 -3.98301244e-01 -1.07673717e+00
5.23221433e-01 -8.65464628e-01 -6.27648294e-01 3.58307481e-01
-3.12381476e-01 -4.00307834e-01 3.67550738e-02 4.65056807e-01
-5.97816765e-01 7.70273745e-01 1.28956854e+00 -3.61900143e-02
-2.86445349e-01 6.90739930e-01 -9.94462907e-01 6.38074994e-01
5.95861435e-01 -7.23359168e-01 -3.45006913e-01 -3.05514216e-01
4.54503059e-01 3.21436614e-01 5.62495768e-01 -3.06940287e-01
2.68871821e-02 -6.55043125e-01 2.39327848e-01 -6.20216310e-01
-8.28409344e-02 -6.20259404e-01 4.09750529e-02 4.15801585e-01
-6.80288255e-01 -3.09314400e-01 -1.22833185e-01 1.25793111e+00
9.76268500e-02 1.42913416e-01 4.04647142e-01 -5.04397936e-02
-1.45002052e-01 4.13086027e-01 9.52331275e-02 6.20999001e-02
7.93818951e-01 2.91882575e-01 -6.95782840e-01 -1.70623913e-01
-3.09471518e-01 5.21378219e-01 -2.59513259e-01 7.04211891e-01
1.58886030e-01 -1.57212114e+00 -5.76184452e-01 -1.38339356e-01
3.81939225e-02 -4.47879612e-01 5.62023699e-01 1.09640372e+00
5.63009918e-01 8.34468544e-01 1.67461321e-01 -1.16219550e-01
-7.65968323e-01 6.24089003e-01 6.91402238e-04 -7.47939050e-01
3.81034985e-02 5.27717829e-01 3.53372008e-01 -2.96342492e-01
1.52866170e-01 -3.09243262e-01 1.54598847e-01 1.50670737e-01
2.58516788e-01 8.16029191e-01 -1.18078329e-01 3.58517438e-01
-1.22353688e-01 1.93316370e-01 -2.41680205e-01 2.20126882e-01
1.24365735e+00 -3.40011746e-01 1.64667651e-01 5.46596885e-01
1.07271922e+00 1.53528184e-01 -1.40129304e+00 -4.02657360e-01
-4.92540933e-03 -9.02888119e-01 3.14997554e-01 -1.43586004e+00
-7.61014283e-01 8.02239358e-01 5.61108112e-01 3.21866035e-01
9.13331389e-01 -1.22085154e-01 9.29065108e-01 1.53925434e-01
5.19623160e-01 -1.26298571e+00 2.06815809e-01 -1.08443558e-01
9.37183797e-01 -1.33204937e+00 2.83786476e-01 -4.14101571e-01
-7.26647973e-01 6.04528725e-01 6.00213110e-01 1.25638116e-02
7.43794918e-01 -2.59598345e-01 -6.28718376e-01 1.33177698e-01
-1.00626552e+00 3.63416493e-01 2.73834825e-01 2.05651984e-01
4.94681537e-01 4.25243884e-01 -8.31636190e-01 1.38303387e+00
1.63679942e-01 2.95181423e-01 2.27598652e-01 2.26850316e-01
-1.23645127e-01 -1.04840839e+00 -1.26381621e-01 9.95508015e-01
-6.80184662e-01 -2.98370302e-01 -3.07592750e-02 7.07354248e-01
8.29287767e-02 1.03637922e+00 -2.42031410e-01 -4.89834338e-01
5.93599319e-01 -1.54488117e-01 -2.64745932e-02 -3.91984195e-01
-2.56245404e-01 3.02146643e-01 1.44128010e-01 -6.48502350e-01
-9.15718898e-02 -8.47915411e-01 -5.18889666e-01 -4.31444556e-01
-8.24561656e-01 3.55050191e-02 5.90415299e-01 8.68131578e-01
3.13260198e-01 5.42313397e-01 1.01555514e+00 -1.82602137e-01
-1.56525362e+00 -1.05922818e+00 -6.55760884e-01 2.72111744e-01
1.37146994e-01 -9.64501441e-01 -5.91255605e-01 -2.84410566e-01] | [9.727709770202637, 5.396203517913818] |
439057d1-df74-4000-b0aa-98ebbd5a2fd7 | mitosis-domain-generalization-in | 2204.03742 | null | https://arxiv.org/abs/2204.03742v1 | https://arxiv.org/pdf/2204.03742v1.pdf | Mitosis domain generalization in histopathology images -- The MIDOG challenge | The density of mitotic figures within tumor tissue is known to be highly correlated with tumor proliferation and thus is an important marker in tumor grading. Recognition of mitotic figures by pathologists is known to be subject to a strong inter-rater bias, which limits the prognostic value. State-of-the-art deep learning methods can support the expert in this assessment but are known to strongly deteriorate when applied in a different clinical environment than was used for training. One decisive component in the underlying domain shift has been identified as the variability caused by using different whole slide scanners. The goal of the MICCAI MIDOG 2021 challenge has been to propose and evaluate methods that counter this domain shift and derive scanner-agnostic mitosis detection algorithms. The challenge used a training set of 200 cases, split across four scanning systems. As a test set, an additional 100 cases split across four scanning systems, including two previously unseen scanners, were given. The best approaches performed on an expert level, with the winning algorithm yielding an F_1 score of 0.748 (CI95: 0.704-0.781). In this paper, we evaluate and compare the approaches that were submitted to the challenge and identify methodological factors contributing to better performance. | ['Katharina Breininger', 'Mitko Veta', 'Xiyue Wang', 'Sen yang', 'April Khademi', 'Salar Razavi', 'Jingxin Liu', 'Xi Long', 'YuBo Wang', 'Jingtang Liang', 'Viktor H. Koelzer', 'Maxime W. Lafarge', 'Satoshi Kondo', 'Thomas Wittenberg', 'Jakob Dexl', 'Nasir Rajpoot', 'Mostafa Jahanifar', 'Saima Ben Hadj', 'Rutger H. J. Fick', 'Fattaneh Pourakpour', 'Ramin Nateghi', 'Jinah Park', 'Youjin Chung', 'Nishant Ravikumar', 'Jack Breen', 'Andreas Maier', 'Taryn A. Donovan', 'Christian Marzahl', 'Frauke Wilm', 'Francesco Ciompi', 'Natalie ter Hoeve', 'Robert Klopleisch', 'Christof A. Bertram', 'Nikolas Stathonikos', 'Marc Aubreville'] | 2022-04-06 | null | null | null | null | ['mitosis-detection'] | ['medical'] | [ 2.07171842e-01 2.30079889e-02 -2.49736801e-01 -1.55705109e-01
-1.23438048e+00 -6.19666457e-01 5.83008230e-01 7.73534894e-01
-8.12096596e-01 8.66358280e-01 2.95934007e-02 -3.62485409e-01
-1.84100389e-01 -6.67361498e-01 -4.80581433e-01 -1.22589922e+00
2.14291230e-01 1.02057815e+00 3.43964577e-01 1.17207490e-01
3.04419845e-01 6.35000467e-01 -1.18550980e+00 5.77012300e-01
6.03289068e-01 6.81297481e-01 1.84099540e-01 9.25117552e-01
-2.42297962e-01 4.17102516e-01 -8.59640896e-01 -1.77530900e-01
-1.30253077e-01 -2.63004065e-01 -8.51548254e-01 -1.13936916e-01
4.72856462e-01 1.04865514e-01 2.98417872e-03 8.65537524e-01
8.53716075e-01 -5.38158238e-01 1.00371599e+00 -7.74586380e-01
5.81987798e-02 6.38557971e-01 -5.08338332e-01 5.30606329e-01
-1.63589925e-01 3.98310632e-01 6.54437721e-01 -4.74123299e-01
9.29889381e-01 3.85734826e-01 8.16551566e-01 5.70479691e-01
-1.30382895e+00 -5.49934685e-01 -4.70665544e-01 3.03217560e-01
-1.45622289e+00 -2.60470927e-01 2.92448312e-01 -7.16220438e-01
7.16805577e-01 2.41041929e-01 7.61335671e-01 9.81971145e-01
6.20449185e-01 4.97575432e-01 1.37855613e+00 -3.94838035e-01
4.37857151e-01 4.17920053e-01 3.01389564e-02 3.83392602e-01
6.09284878e-01 -2.08605062e-02 -2.56883591e-01 5.21000102e-02
3.91270012e-01 -2.86295325e-01 -3.96075517e-01 -2.32609898e-01
-1.15306413e+00 7.63276875e-01 4.75661397e-01 1.00646746e+00
-2.51244940e-02 -1.92727700e-01 8.29636991e-01 2.93561339e-01
2.53551334e-01 5.16523182e-01 -2.21784189e-01 5.29649854e-02
-1.14566803e+00 -1.12077119e-02 6.50130391e-01 1.82062518e-02
1.62988096e-01 -5.24181783e-01 -3.18687052e-01 6.28938794e-01
6.74241930e-02 2.54144341e-01 1.23070312e+00 -2.22328603e-01
-1.09039821e-01 8.17429721e-01 -2.55844533e-01 -5.88269234e-01
-1.04163766e+00 -7.46368349e-01 -1.18026674e+00 3.73721898e-01
1.09678876e+00 2.97950655e-01 -1.17439926e+00 1.54732478e+00
1.60224259e-01 7.78021589e-02 -1.09928325e-01 8.10356140e-01
8.71365190e-01 4.37341146e-02 1.37543902e-01 -1.21590063e-01
1.57228518e+00 -6.07555449e-01 -3.55284154e-01 1.90766137e-02
1.25605440e+00 -6.52949810e-01 8.61575246e-01 5.77602148e-01
-7.18286753e-01 -2.04405412e-01 -1.16606140e+00 1.64252281e-01
-7.09226429e-01 2.32696682e-01 2.08430275e-01 8.58356774e-01
-1.35976565e+00 4.70671773e-01 -8.61675084e-01 -6.68994904e-01
5.72120428e-01 6.34582102e-01 -5.47283471e-01 2.38649771e-02
-1.02679598e+00 1.24644709e+00 5.25525331e-01 1.85287241e-02
-8.52429986e-01 -1.06635666e+00 -2.24726126e-01 -1.15164414e-01
6.15282804e-02 -6.26643658e-01 1.15983915e+00 -7.15130031e-01
-1.29589820e+00 1.62305450e+00 3.52508202e-02 -5.22230327e-01
9.93024230e-01 6.22124314e-01 -3.26749176e-01 9.95861813e-02
1.03494383e-01 4.33594406e-01 3.23904783e-01 -9.30549920e-01
-7.35775471e-01 -6.14075899e-01 -5.81309319e-01 -4.05174866e-02
-1.04863077e-01 -5.11901140e-01 -3.14664304e-01 -2.72627652e-01
-3.38814855e-01 -9.21126246e-01 -2.15552181e-01 -1.45761713e-01
-1.83567002e-01 -2.24199206e-01 4.74844486e-01 -5.01312852e-01
1.00529838e+00 -1.91766047e+00 -8.87265503e-02 4.37914342e-01
4.45866317e-01 4.53000575e-01 1.61718443e-01 -2.06016134e-02
-2.21226498e-01 2.83065736e-01 -4.20318283e-02 -1.28856316e-01
-1.96455806e-01 -7.54004344e-02 4.12958026e-01 9.07069802e-01
5.73461913e-02 7.34214723e-01 -9.04269099e-01 -6.48689151e-01
2.49871224e-01 4.33150917e-01 -9.77917165e-02 -4.35695387e-02
5.99882789e-02 2.96244711e-01 1.49907609e-02 5.89621127e-01
6.31911933e-01 -5.87966025e-01 3.44128609e-01 -1.73744142e-01
1.65566251e-01 -5.10058217e-02 -9.25384641e-01 1.48576677e+00
-1.52668968e-01 9.23617005e-01 -9.47753433e-03 -8.02803934e-01
8.30264091e-01 3.55663866e-01 3.98805737e-01 -6.19305313e-01
4.49694365e-01 5.99525094e-01 4.97055233e-01 -3.92542541e-01
3.00883129e-02 -5.16766310e-01 1.54244602e-01 1.38005048e-01
1.55635308e-02 -2.30555445e-01 3.81309509e-01 -4.02121916e-02
1.49658048e+00 -6.93431139e-01 5.65614581e-01 -6.03104532e-01
8.77748489e-01 1.50197044e-01 3.59093428e-01 5.41915596e-01
-6.28840566e-01 6.49434030e-01 8.52461874e-01 -5.14168561e-01
-9.68765795e-01 -9.45832908e-01 -5.74432373e-01 4.93658870e-01
-2.06238434e-01 1.12661973e-01 -6.25946999e-01 -7.92338789e-01
-5.10985851e-02 3.86263579e-01 -1.41051352e+00 -2.12902293e-01
-3.19488049e-01 -1.15660679e+00 7.95979023e-01 4.17377770e-01
1.12849668e-01 -7.13751495e-01 -8.59503984e-01 2.42290333e-01
7.83624277e-02 -7.45816350e-01 -8.26433301e-02 5.42951524e-01
-6.96470737e-01 -1.45948195e+00 -1.14147830e+00 -5.80117166e-01
6.15737677e-01 -1.96558684e-01 1.27543783e+00 1.90074399e-01
-6.37205005e-01 -1.40027136e-01 -2.22973302e-01 -5.79463005e-01
-9.05510902e-01 6.25155449e-01 -3.02194923e-01 -8.65488201e-02
6.64833844e-01 -7.30581656e-02 -6.91014171e-01 3.22173834e-01
-9.99058366e-01 -7.00044259e-02 1.12484789e+00 1.12650502e+00
7.64197648e-01 -2.05584168e-01 3.01535577e-01 -1.12504542e+00
2.20371976e-01 -2.88097769e-01 -4.61019307e-01 2.38967583e-01
-6.26961052e-01 -8.76030624e-02 6.07004941e-01 -4.19063836e-01
-5.80096900e-01 1.09886304e-01 -1.27973899e-01 -1.26013428e-01
-4.07421440e-01 5.82492352e-01 -1.42061021e-02 -1.81913763e-01
9.51770186e-01 -3.38123962e-02 3.40366870e-01 1.13007940e-01
-3.16992998e-01 3.97722632e-01 4.74367917e-01 -1.10874236e-01
4.76533532e-01 5.71014881e-01 4.84591037e-01 -8.16012859e-01
-5.89447081e-01 -7.32992947e-01 -4.19738710e-01 -2.97567546e-01
7.21691787e-01 -5.44729412e-01 -5.39180398e-01 7.15985000e-01
-7.64685452e-01 -6.47546351e-01 -2.82209039e-01 3.68119866e-01
-3.98217142e-01 3.70135973e-03 -6.06397152e-01 -1.96966484e-01
-5.09341359e-01 -1.28330123e+00 9.75660503e-01 4.22770381e-01
-5.28766274e-01 -1.08370674e+00 5.42829275e-01 2.25072980e-01
5.52003264e-01 5.76060593e-01 7.66126454e-01 -1.04035664e+00
-7.77441589e-03 -4.41311479e-01 -1.37207001e-01 -1.37620807e-01
6.69248849e-02 2.71003723e-01 -1.20463097e+00 -4.63267982e-01
-2.59882540e-01 -2.24185318e-01 1.04261339e+00 5.18472135e-01
1.03225720e+00 3.86202455e-01 -8.79960835e-01 5.55853963e-01
1.63999259e+00 1.08266704e-01 7.55983055e-01 6.91794693e-01
1.77144870e-01 6.42266750e-01 1.34381890e-01 1.32040773e-02
-1.47239178e-01 5.04752517e-01 3.66403252e-01 -1.97128668e-01
-1.75827742e-01 1.60222679e-01 -1.69042766e-01 2.16641992e-01
1.12769134e-01 -2.24720299e-01 -1.46819890e+00 7.51431823e-01
-1.23147464e+00 -7.22280681e-01 -3.87983650e-01 1.98693049e+00
7.31667280e-01 4.28473026e-01 8.80568102e-02 6.32843792e-01
6.99010491e-01 -2.29213387e-01 -4.58477437e-01 -2.47382730e-01
-2.50210404e-01 2.15919480e-01 6.44059360e-01 3.84170145e-01
-8.49159300e-01 4.06285793e-01 6.26291847e+00 9.87101257e-01
-1.60128462e+00 -7.25558549e-02 1.06893587e+00 -1.96284220e-01
2.42683589e-01 -6.11494958e-01 -7.71125495e-01 6.39962077e-01
1.31298971e+00 -3.09505641e-01 -4.01288629e-01 3.90762776e-01
1.38442323e-01 -4.98549938e-01 -1.27649021e+00 8.59457731e-01
1.73934221e-01 -1.56135201e+00 -1.19916812e-01 4.58401144e-01
4.97449189e-01 1.22742943e-01 1.94729477e-01 2.59468049e-01
-2.30889171e-02 -1.31131423e+00 7.17923865e-02 4.34708327e-01
9.95260894e-01 -4.61882055e-01 1.62391651e+00 8.31223130e-02
-7.41820276e-01 2.96252787e-01 -5.53405024e-02 3.91883135e-01
-1.81444660e-01 8.06591332e-01 -1.69939041e+00 3.48970681e-01
5.26659906e-01 3.46882790e-01 -9.08313811e-01 1.31599045e+00
1.61598623e-01 5.62713921e-01 -4.00570631e-01 -3.28969479e-01
1.79485321e-01 4.47697550e-01 2.35274181e-01 1.53208947e+00
3.32422167e-01 -2.22907171e-01 -4.79954600e-01 4.29170579e-01
1.20424926e-01 1.45910680e-01 -1.48730025e-01 -4.82305810e-02
2.41865575e-01 1.51965857e+00 -1.29788721e+00 -3.11986178e-01
-1.25962362e-01 4.63628381e-01 1.92409828e-01 -1.75274655e-01
-6.23876095e-01 -1.46344051e-01 6.91056177e-02 5.14358819e-01
7.58108571e-02 6.00954294e-01 -5.42778313e-01 -7.00637996e-01
-4.34018254e-01 -8.46811235e-01 8.39441717e-01 -3.40887368e-01
-1.40979385e+00 4.82246071e-01 -3.83480638e-01 -1.11522353e+00
-7.29695037e-02 -1.04727376e+00 -6.09086454e-01 9.77745771e-01
-1.46816838e+00 -9.29162443e-01 -6.30785406e-01 3.04611951e-01
2.01681495e-01 -1.81146964e-01 8.73372793e-01 2.14644223e-01
-2.75639892e-01 9.46281672e-01 2.20951065e-01 1.33191362e-01
1.09123218e+00 -1.57210422e+00 -2.39974171e-01 2.37341702e-01
-3.83253843e-01 1.15714192e-01 7.59933650e-01 -1.92654714e-01
-8.68496001e-01 -9.95102584e-01 8.75959337e-01 -5.47749877e-01
7.12073445e-01 1.32330745e-01 -1.11742640e+00 2.40575656e-01
-5.33910245e-02 3.31126153e-02 1.33382988e+00 -9.83741432e-02
-1.83017418e-01 -1.21073008e-01 -1.35164726e+00 1.88338369e-01
2.56981045e-01 -1.85320541e-01 -2.40754485e-01 5.79070970e-02
-1.65248483e-01 -6.57131553e-01 -8.60846519e-01 5.33600926e-01
5.79648376e-01 -1.01338136e+00 5.41108966e-01 -2.21619517e-01
3.80249798e-01 -3.00500900e-01 2.46794567e-01 -1.36828756e+00
-4.59945798e-01 9.59374383e-02 2.82005817e-01 8.34370255e-01
6.95387423e-01 -5.77138841e-01 1.40609598e+00 1.94144428e-01
-7.58214295e-02 -1.03314328e+00 -1.20086443e+00 -5.74880064e-01
5.83922267e-01 7.63301551e-02 2.69999444e-01 8.74086618e-01
-5.18101209e-04 2.02635109e-01 3.78657550e-01 -1.32294595e-01
3.35377723e-01 -3.49998027e-01 6.94560289e-01 -1.35610366e+00
-3.61090630e-01 -1.28051794e+00 -9.42638099e-01 -4.79279421e-02
-1.70403749e-01 -1.09089792e+00 -1.23686261e-01 -1.29328179e+00
4.71635461e-01 -2.19259962e-01 -6.61471069e-01 2.16539308e-01
-1.72157705e-01 4.37356323e-01 -1.00116707e-01 2.31917500e-01
-3.37422460e-01 -2.36628547e-01 1.03048193e+00 -4.15174454e-01
1.20814508e-02 -1.68322816e-01 -6.51866138e-01 4.93370682e-01
8.70247543e-01 -3.73459816e-01 9.01624486e-02 3.15820090e-02
2.28434041e-01 -1.94954380e-01 1.81090564e-01 -1.33842993e+00
3.52297693e-01 6.66142702e-02 8.47003520e-01 -7.60631144e-01
-3.15001234e-03 -6.01297796e-01 3.56889367e-01 1.01532662e+00
-3.07635367e-01 -1.16845965e-01 3.98885310e-01 3.67736161e-01
-1.31715223e-01 -3.20412666e-01 1.09892273e+00 -4.50108349e-02
-3.10239792e-01 1.40552118e-01 -5.98567963e-01 -7.85872787e-02
1.22468138e+00 -5.32283187e-01 -4.86549377e-01 9.20671523e-02
-8.12115252e-01 2.04042390e-01 7.23688483e-01 -2.13705767e-02
6.40506074e-02 -9.00719643e-01 -9.72942352e-01 -2.41305634e-01
3.84086818e-01 4.35903594e-02 5.69804609e-01 1.22668719e+00
-7.59737015e-01 5.48564494e-01 -2.76621401e-01 -9.45202529e-01
-1.44709539e+00 4.91519660e-01 8.46133113e-01 -9.96234238e-01
-2.17824593e-01 1.00004923e+00 6.62277825e-03 -1.35438621e-01
1.30370393e-01 -2.63931096e-01 -3.14547598e-01 4.22262162e-01
7.33987987e-01 3.55642110e-01 6.74813032e-01 -3.11613917e-01
-5.35398126e-01 4.51087475e-01 -5.30438423e-01 4.68510669e-03
9.72492158e-01 4.16061223e-01 1.19487844e-01 6.08572125e-01
1.17831528e+00 -3.00073057e-01 -7.51950622e-01 8.47201496e-02
1.32676676e-01 -2.04537839e-01 1.56187285e-02 -1.16077220e+00
-1.01652873e+00 8.95690739e-01 1.19951975e+00 2.29095161e-01
9.64689255e-01 -4.48014140e-02 2.77684212e-01 -1.66326910e-01
1.00553952e-01 -1.02770948e+00 -1.17959283e-01 1.89030722e-01
4.96945709e-01 -1.49589479e+00 -5.76446466e-02 -2.26633623e-01
-2.02470019e-01 1.34738171e+00 4.56667066e-01 1.07565992e-01
3.79290044e-01 5.71078300e-01 3.46386492e-01 -2.22626269e-01
-8.25734615e-01 7.08141625e-02 -2.75650676e-02 6.23589337e-01
7.92597532e-01 2.17333063e-01 -6.43192887e-01 4.26513672e-01
-1.45227879e-01 4.65098023e-01 6.31234109e-01 6.82474673e-01
-2.19987527e-01 -9.36814427e-01 -4.38256800e-01 7.06113696e-01
-5.58959007e-01 4.74244773e-01 -5.72046399e-01 1.15218949e+00
1.62173212e-01 4.21736062e-01 1.45949528e-01 -1.35495678e-01
2.07759574e-01 4.74743135e-02 3.66477758e-01 -3.23857218e-01
-7.75288939e-01 -7.30639249e-02 -9.13938731e-02 -1.55575350e-01
-2.93903679e-01 -7.73099661e-01 -1.06358457e+00 -3.47183913e-01
-1.92113772e-01 2.54673272e-01 4.46408600e-01 7.83452272e-01
5.53741446e-03 8.25706661e-01 1.51211247e-01 -6.14888072e-01
-3.98851544e-01 -1.09619343e+00 -6.74234688e-01 4.37706500e-01
3.23909909e-01 -6.91091657e-01 -7.16135383e-01 5.89339137e-02] | [15.126108169555664, -3.0650148391723633] |
d8e2407b-bcef-4925-94e5-6264bc1211c8 | forward-compatible-few-shot-class-incremental | 2203.06953 | null | https://arxiv.org/abs/2203.06953v1 | https://arxiv.org/pdf/2203.06953v1.pdf | Forward Compatible Few-Shot Class-Incremental Learning | Novel classes frequently arise in our dynamically changing world, e.g., new users in the authentication system, and a machine learning model should recognize new classes without forgetting old ones. This scenario becomes more challenging when new class instances are insufficient, which is called few-shot class-incremental learning (FSCIL). Current methods handle incremental learning retrospectively by making the updated model similar to the old one. By contrast, we suggest learning prospectively to prepare for future updates, and propose ForwArd Compatible Training (FACT) for FSCIL. Forward compatibility requires future new classes to be easily incorporated into the current model based on the current stage data, and we seek to realize it by reserving embedding space for future new classes. In detail, we assign virtual prototypes to squeeze the embedding of known classes and reserve for new ones. Besides, we forecast possible new classes and prepare for the updating process. The virtual prototypes allow the model to accept possible updates in the future, which act as proxies scattered among embedding space to build a stronger classifier during inference. FACT efficiently incorporates new classes with forward compatibility and meanwhile resists forgetting of old ones. Extensive experiments validate FACT's state-of-the-art performance. Code is available at: https://github.com/zhoudw-zdw/CVPR22-Fact | ['De-Chuan Zhan', 'ShiLiang Pu', 'Liang Ma', 'Han-Jia Ye', 'Fu-Yun Wang', 'Da-Wei Zhou'] | 2022-03-14 | null | http://openaccess.thecvf.com//content/CVPR2022/html/Zhou_Forward_Compatible_Few-Shot_Class-Incremental_Learning_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Zhou_Forward_Compatible_Few-Shot_Class-Incremental_Learning_CVPR_2022_paper.pdf | cvpr-2022-1 | ['few-shot-class-incremental-learning'] | ['methodology'] | [-2.21197516e-01 -2.34905839e-01 -5.92122912e-01 -3.78868699e-01
-3.57122183e-01 -4.72341061e-01 5.60189307e-01 2.84854621e-01
-5.16445041e-01 8.55505168e-01 2.79827439e-03 -3.40572625e-01
-1.16597831e-01 -1.12892914e+00 -4.45377856e-01 -6.29067838e-01
-1.83169991e-01 5.47151029e-01 4.88177210e-01 -1.52408704e-01
8.92988965e-02 3.95900011e-01 -1.56824040e+00 4.20835674e-01
6.95802808e-01 7.89095819e-01 1.39199300e-02 4.82702494e-01
-1.41841814e-01 2.74834901e-01 -4.62499171e-01 -5.81388712e-01
3.79832536e-01 2.23060418e-02 -7.24140942e-01 1.32363634e-02
5.59347384e-02 -6.78667188e-01 -5.42865932e-01 8.28395128e-01
4.97836024e-01 3.68196845e-01 1.91139311e-01 -1.41927612e+00
-7.45796680e-01 1.03125405e+00 -3.72482985e-01 2.67522097e-01
2.96450555e-01 3.32064986e-01 8.35442901e-01 -1.60968375e+00
5.63803494e-01 9.12901282e-01 6.06587052e-01 9.38989222e-01
-7.79379308e-01 -9.74208236e-01 9.39825535e-01 7.45661855e-01
-1.53910863e+00 -5.96464038e-01 9.37408090e-01 -1.11459963e-01
5.14345765e-01 4.73139197e-01 6.82303667e-01 1.08519435e+00
-2.44210944e-01 1.21860564e+00 6.03984118e-01 -2.07059368e-01
5.06433308e-01 5.70990980e-01 5.31004786e-01 5.90100944e-01
3.61698866e-01 9.59256738e-02 -5.42997718e-01 -3.20566773e-01
1.34987131e-01 8.16956043e-01 -2.60552704e-01 -2.45930716e-01
-1.13074815e+00 5.81581771e-01 2.12988317e-01 2.97820419e-01
-3.28102916e-01 -4.90133464e-01 2.23739341e-01 4.00771618e-01
3.88686985e-01 1.95931494e-01 -1.09072590e+00 -1.92086384e-01
-7.94808865e-01 -8.98437649e-02 5.32920480e-01 8.77743602e-01
1.09036803e+00 -7.64390603e-02 4.66856658e-02 7.33011365e-01
4.69640456e-02 3.20343018e-01 6.33609533e-01 -4.78163600e-01
3.69825006e-01 6.96545482e-01 -5.11576980e-02 -7.54659057e-01
-8.79426971e-02 -7.51416802e-01 -8.99466455e-01 -2.32949123e-01
1.62182108e-01 -1.66604429e-01 -7.71546066e-01 1.66087794e+00
6.87677681e-01 8.85987997e-01 4.77795601e-02 2.95340568e-01
3.83582771e-01 7.32496977e-01 -1.02361418e-01 -6.10859513e-01
1.13320732e+00 -6.83883548e-01 -3.99680495e-01 -1.80896103e-01
5.53340375e-01 -4.62980747e-01 1.07553363e+00 6.60568118e-01
-5.41505456e-01 -8.04499805e-01 -1.01380789e+00 6.53059304e-01
-5.21314561e-01 7.16095939e-02 7.08413959e-01 6.14402592e-01
-6.57169819e-01 6.57390177e-01 -6.57629490e-01 1.35027235e-02
5.19125760e-01 2.89541215e-01 -1.84769347e-01 -9.91230905e-02
-1.64997768e+00 5.15585601e-01 5.65235138e-01 3.23487967e-01
-7.22081602e-01 -6.93452537e-01 -5.66950023e-01 8.44210610e-02
7.01821387e-01 -2.70845056e-01 1.12625313e+00 -4.02367681e-01
-1.22389996e+00 1.95189789e-01 -2.94548392e-01 -2.93242037e-01
3.98750424e-01 -1.14167020e-01 -8.95923674e-01 -2.70641714e-01
-1.90988153e-01 3.16486746e-01 1.13729513e+00 -1.28863633e+00
-9.55253780e-01 -2.04523981e-01 2.02029780e-01 6.53502122e-02
-1.01875198e+00 -6.86004162e-01 -4.71723437e-01 -5.53150952e-01
1.96159244e-01 -8.60450327e-01 -2.25352883e-01 8.49382207e-02
-2.11470544e-01 -4.03702348e-01 1.08538079e+00 -3.53181899e-01
1.87841296e+00 -2.16030192e+00 -2.83909857e-01 3.08770418e-01
4.09676701e-01 6.83616161e-01 -9.48911160e-02 2.51688659e-01
-1.89170465e-01 -9.37873200e-02 -1.36265457e-01 -4.05052781e-01
-2.02287823e-01 3.47049952e-01 -6.38408840e-01 7.94373229e-02
-1.54731616e-01 6.23808801e-01 -1.08187187e+00 -1.29606098e-01
1.34551495e-01 1.67992562e-01 -6.07538402e-01 1.07286833e-01
-1.03376247e-01 1.62139103e-01 -4.76847619e-01 8.21856260e-01
1.01632524e+00 -2.89855450e-01 1.49349675e-01 -1.59588933e-01
1.70250446e-01 3.15102078e-02 -1.51189387e+00 1.21129382e+00
-3.51123333e-01 2.01523267e-02 -4.74841744e-01 -1.07809711e+00
9.23050106e-01 2.23457590e-01 3.41690004e-01 -2.44813025e-01
-1.18398577e-01 1.32439882e-01 -6.20413348e-02 -3.15043360e-01
3.60919803e-01 1.02184666e-02 1.04586601e-01 5.99930942e-01
-1.05124019e-01 4.69756156e-01 1.23545222e-01 3.90286088e-01
6.66478038e-01 -1.74865603e-01 2.56231934e-01 3.94650429e-01
9.02741432e-01 -6.44980133e-01 1.00274622e+00 8.63804579e-01
-3.49707007e-01 8.66424665e-02 -3.07488181e-02 -8.95580888e-01
-4.88388449e-01 -1.26422966e+00 -1.01527013e-01 1.04118037e+00
2.07886472e-01 -6.70126915e-01 -1.26647964e-01 -1.32291186e+00
1.39207408e-01 7.74324059e-01 -5.04234314e-01 -6.66731060e-01
-5.97165406e-01 -8.66108477e-01 2.55525392e-02 3.05085510e-01
4.30182636e-01 -8.64720285e-01 -2.15384349e-01 5.99618971e-01
-1.07093699e-01 -6.21076047e-01 -5.72032034e-01 1.38108032e-02
-8.51502717e-01 -1.28791654e+00 -4.47283179e-01 -5.01401544e-01
8.06909084e-01 3.91049474e-01 6.06163442e-01 4.39229041e-01
-9.02179256e-02 3.99625599e-01 -4.39859867e-01 -3.72573078e-01
-2.64662236e-01 2.15454608e-01 7.26535499e-01 4.41372901e-01
4.76063639e-01 -6.86094582e-01 -7.11099088e-01 2.44821295e-01
-7.54515052e-01 -8.89592916e-02 3.28026026e-01 1.07911551e+00
4.66946572e-01 4.24127728e-01 8.51699531e-01 -1.15740407e+00
4.73480552e-01 -8.02051365e-01 -3.40817243e-01 5.19728780e-01
-1.06981218e+00 -1.45444199e-01 1.00014973e+00 -1.05014765e+00
-9.85598087e-01 -1.33641124e-01 -3.19394112e-01 -4.00548041e-01
1.70926645e-01 4.76342052e-01 -2.40938783e-01 2.50381410e-01
5.42622447e-01 4.90036756e-01 -2.70180970e-01 -4.91090566e-01
4.15542036e-01 7.96602845e-01 3.43083948e-01 -6.76069856e-01
1.31593609e+00 3.31331432e-01 -4.86465007e-01 -3.81996274e-01
-8.22030246e-01 -3.78782481e-01 -7.39145756e-01 -2.93062776e-01
-9.50413272e-02 -7.94254541e-01 -5.54774523e-01 7.55716622e-01
-7.23257184e-01 -2.01149404e-01 -5.91999531e-01 4.17153627e-01
2.73744524e-01 6.26471043e-01 -5.83218455e-01 -8.45455348e-01
-3.49638402e-01 -9.21607673e-01 5.24680376e-01 4.52161402e-01
4.00049649e-02 -1.00647068e+00 -2.11147293e-02 -1.38561815e-01
4.33085769e-01 -4.88831311e-01 8.29587340e-01 -9.07689512e-01
-4.52021182e-01 -5.69803298e-01 2.70874709e-01 3.69323879e-01
4.72304016e-01 1.08633153e-01 -1.19334590e+00 -8.24390054e-01
3.55411731e-02 1.44488737e-01 9.00012732e-01 -1.93113744e-01
1.47244179e+00 -7.15035498e-01 -5.61367095e-01 3.61701488e-01
8.48809898e-01 4.29452538e-01 2.41863906e-01 2.68142849e-01
5.19076049e-01 2.56274551e-01 7.71472692e-01 8.91303837e-01
5.81272900e-01 6.00309193e-01 9.80504155e-02 1.89951658e-01
-5.78216873e-02 -4.39184546e-01 2.97220528e-01 1.23448491e+00
2.58273870e-01 -1.16103023e-01 -7.74962127e-01 2.57768422e-01
-1.81405747e+00 -1.07728171e+00 5.09166420e-01 2.53159857e+00
1.24849010e+00 6.06172204e-01 -9.62486267e-02 5.13043702e-01
6.80509567e-01 -1.29480502e-02 -8.53604138e-01 1.59379497e-01
1.21698596e-01 2.15912029e-01 -3.54753435e-03 6.77634120e-01
-1.10321343e+00 7.56620944e-01 4.84421587e+00 9.17706251e-01
-1.25692022e+00 3.00639480e-01 8.10242474e-01 -5.70132323e-02
-3.32667053e-01 3.51398975e-01 -1.27182150e+00 8.36021960e-01
8.15437973e-01 -4.04741108e-01 4.38363820e-01 1.12975240e+00
-2.59849906e-01 1.12171911e-01 -9.67283666e-01 9.58213210e-01
5.76873757e-02 -1.50511920e+00 3.01087767e-01 -2.88622975e-01
5.42090058e-01 -2.58589327e-01 2.64443099e-01 8.58486056e-01
2.49515712e-01 5.24346046e-02 4.14616376e-01 6.80246413e-01
8.20813656e-01 -6.07939243e-01 6.40318096e-01 6.99144900e-01
-1.50570464e+00 -5.93306839e-01 -3.80041838e-01 6.41048253e-02
-4.05278876e-02 8.55366826e-01 -9.91960049e-01 5.18566847e-01
5.02782166e-01 8.51786792e-01 -9.24832642e-01 1.07137954e+00
-1.74353123e-01 6.45381927e-01 -3.29599321e-01 1.63046643e-01
-1.74949646e-01 7.77102262e-02 3.84818614e-01 6.98239565e-01
3.66114587e-01 1.45238578e-01 4.95737672e-01 6.22401088e-02
-1.31092116e-01 -1.15487553e-01 -1.53031066e-01 2.96772301e-01
1.19518697e+00 1.01840353e+00 -4.52862948e-01 -9.49185848e-01
-2.91461170e-01 9.30623710e-01 1.59877956e-01 4.48983938e-01
-7.39454091e-01 -3.42094213e-01 6.76156580e-01 4.64152731e-02
9.93200913e-02 7.90010090e-05 1.19020641e-01 -1.74370050e+00
7.19557777e-02 -9.05394495e-01 8.80774736e-01 -3.77610534e-01
-1.56437457e+00 6.96994364e-01 9.02580563e-03 -1.71776974e+00
-1.25105768e-01 -3.49625260e-01 -7.13993013e-01 4.74159718e-01
-1.49108374e+00 -9.66623664e-01 -2.28724599e-01 7.49469519e-01
8.35533679e-01 -4.74718451e-01 8.78321469e-01 5.35974920e-01
-6.77389085e-01 1.17869270e+00 4.26099934e-02 -1.78302005e-02
8.98703516e-01 -7.57551789e-01 5.29567719e-01 1.08210397e+00
2.98539072e-01 1.09188044e+00 3.22737128e-01 -8.48435163e-01
-1.31133163e+00 -1.18833268e+00 9.24516320e-01 -4.59655255e-01
5.70358574e-01 -4.10313904e-01 -1.41433406e+00 5.39097428e-01
-6.02585912e-01 3.60123008e-01 7.68177986e-01 1.89346224e-01
-7.88158596e-01 -5.77440202e-01 -1.07208073e+00 6.25479817e-01
1.06055081e+00 -5.51725805e-01 -5.19581616e-01 2.73538142e-01
7.06754923e-01 -1.57709554e-01 -5.10247529e-01 5.44533253e-01
6.70100570e-01 -4.15779591e-01 9.95754898e-01 -7.13003814e-01
-5.87871134e-01 -6.10338628e-01 4.33152467e-02 -1.20290017e+00
-5.12327194e-01 -6.71949565e-01 -9.11644995e-01 1.10398948e+00
5.82410276e-01 -1.07683885e+00 9.66175139e-01 6.70597732e-01
-3.17225145e-04 -9.96714473e-01 -9.72818494e-01 -9.36452687e-01
-1.68466821e-01 -5.02891183e-01 1.13780522e+00 1.34249985e+00
9.56892148e-02 5.01390584e-02 -6.72667921e-01 2.67528176e-01
5.84890902e-01 1.35140732e-01 7.89793432e-01 -1.48175097e+00
-5.23061037e-01 -3.07850897e-01 -1.00720271e-01 -1.09387314e+00
-8.95390436e-02 -9.27313209e-01 -4.21339929e-01 -1.08440197e+00
3.08324605e-01 -1.11298144e+00 -7.89201617e-01 1.05770385e+00
-6.65223718e-01 -7.26398602e-02 2.16887757e-01 5.80266297e-01
-7.28951275e-01 6.20092392e-01 8.65819812e-01 -3.38284075e-01
-4.56813574e-01 6.20269835e-01 -8.23925555e-01 5.63800573e-01
9.13449943e-01 -3.58886659e-01 -6.62015080e-01 -7.39726285e-03
5.84583394e-02 -2.96312392e-01 -2.12122723e-01 -9.89873767e-01
6.02618217e-01 -2.24891573e-01 6.92918956e-01 -6.32005453e-01
3.47472697e-01 -7.83321023e-01 1.22906618e-01 7.98891425e-01
-1.32837832e-01 -5.43585494e-02 -8.26491937e-02 6.60771191e-01
1.70500159e-01 -1.35855600e-01 5.40222824e-01 7.04025775e-02
-1.01411450e+00 1.19684887e+00 -1.71113223e-01 -1.51899725e-01
1.09604406e+00 -3.55866730e-01 -4.01954234e-01 -2.15754494e-01
-1.07897437e+00 3.50977927e-01 1.56925410e-01 7.09629536e-01
1.13839543e+00 -1.34601700e+00 -5.43495655e-01 6.74257338e-01
2.92961717e-01 -1.89896956e-01 4.57527280e-01 4.77287143e-01
1.87301755e-01 -2.02982560e-01 2.00405508e-01 -3.06721479e-01
-1.28145635e+00 9.71562147e-01 3.84701677e-02 -2.07912385e-01
-5.21224678e-01 9.63067889e-01 -1.04282551e-01 -6.34337902e-01
4.39482003e-01 1.47846043e-01 -2.10940808e-01 2.62830883e-01
8.78244460e-01 1.24536969e-01 3.74930874e-02 -3.04667968e-02
-3.93214017e-01 1.05819561e-01 -7.19390392e-01 3.68353873e-01
1.27684832e+00 -2.74087548e-01 -1.68613531e-02 6.78078949e-01
9.39278722e-01 2.54211813e-01 -1.25580525e+00 -9.23690200e-01
1.80522799e-02 -6.85310721e-01 -3.88687909e-01 -6.45734906e-01
-1.09000146e+00 7.79744506e-01 7.81958282e-01 -4.69704233e-02
1.15153325e+00 -2.44926706e-01 1.05910110e+00 7.03104258e-01
7.57722557e-01 -1.03239274e+00 1.47998199e-01 5.89645922e-01
2.78180182e-01 -1.08232939e+00 2.63307661e-01 -1.95399985e-01
-4.57089990e-01 1.22939444e+00 9.15514171e-01 4.98412400e-01
1.21968710e+00 2.15540677e-02 -2.61342347e-01 3.29891086e-01
-1.04338348e+00 8.87902156e-02 -8.37785657e-03 4.93788719e-01
-1.93192169e-01 2.46276632e-01 1.84708065e-03 8.05368960e-01
1.10022321e-01 -1.77741364e-01 5.11814713e-01 9.47036147e-01
-4.61570144e-01 -1.54658151e+00 -7.78190643e-02 7.41834819e-01
1.34894028e-01 -1.14568800e-01 1.18290260e-01 3.17896277e-01
6.11014068e-01 6.68276727e-01 -1.27545759e-01 -9.21243548e-01
2.99945742e-01 2.48085156e-01 7.30850995e-02 -8.32400858e-01
-1.48158237e-01 -3.23001981e-01 -1.94637403e-01 -2.49854788e-01
7.59018213e-02 -7.14442849e-01 -1.11472380e+00 -4.00658935e-01
-5.88002980e-01 4.95923132e-01 1.65447846e-01 7.83876061e-01
4.81826842e-01 1.98033825e-01 1.33039534e+00 -4.27297711e-01
-7.02009678e-01 -7.33799338e-01 -2.91683972e-01 1.09353065e-01
3.45240474e-01 -7.81452119e-01 -4.59098667e-01 -5.88088892e-02] | [9.843206405639648, 3.3978559970855713] |
0d95698b-9a00-4429-91e1-25647540a628 | action-recognition-in-video-sequences-using | null | null | https://ieeexplore.ieee.org/abstract/document/8121994 | https://ieeexplore.ieee.org/document/8121994 | Action Recognition in Video Sequences using Deep Bi-Directional LSTM With CNN Features | Recurrent neural network (RNN) and long short-term memory (LSTM) have achieved great success in processing sequential multimedia data and yielded the state-of-the-art results in speech recognition, digital signal processing, video processing, and text data analysis. In this paper, we propose a novel action recognition method by processing the video data using convolutional neural network (CNN) and deep bidirectional LSTM (DB-LSTM) network. First, deep features are extracted from every sixth frame of the videos, which helps reduce the redundancy and complexity. Next, the sequential information among frame features is learnt using DB-LSTM network, where multiple layers are stacked together in both forward pass and backward pass of DB-LSTM to increase its depth. The proposed method is capable of learning long term sequences and can process lengthy videos by analyzing features for a certain time interval. Experimental results show significant improvements in action recognition using the proposed method on three benchmark data sets including UCF-101, YouTube 11 Actions, and HMDB51 compared with the state-of-the-art action recognition methods. | ['Sung Wook Baik', 'Muhammad Sajjad', 'Khan Muhammad', 'Jamil Ahmad', 'Amin Ullah'] | 2017-11-28 | null | null | null | ieee-access-2017-11 | ['action-recognition-in-videos'] | ['computer-vision'] | [ 4.17272687e-01 -6.68518484e-01 -2.57907957e-01 -2.34102592e-01
-3.86729240e-01 1.02532722e-01 4.80362415e-01 -3.21802527e-01
-7.24247873e-01 5.48043907e-01 5.36836743e-01 -1.66712075e-01
1.34751797e-01 -5.81825495e-01 -7.40113258e-01 -7.18937039e-01
-2.67659634e-01 -2.63072014e-01 6.29086077e-01 2.96268091e-02
3.65746498e-01 4.20152307e-01 -1.51892662e+00 1.10232270e+00
2.05642253e-01 1.47437310e+00 3.62721562e-01 9.10366416e-01
-3.58939737e-01 1.94169354e+00 -5.08773446e-01 -4.77953814e-02
-4.65985239e-02 -3.05972040e-01 -9.76432085e-01 1.48680076e-01
-1.91380139e-02 -7.72924662e-01 -1.12300205e+00 7.07196474e-01
5.16107559e-01 4.23603117e-01 7.91076347e-02 -9.08778071e-01
-5.94804406e-01 4.76119548e-01 -3.37872654e-01 7.69264758e-01
5.35688341e-01 8.47627148e-02 5.58507204e-01 -1.01191032e+00
2.65875012e-01 1.38492584e+00 5.50140262e-01 5.65440178e-01
-2.75085777e-01 -6.47078156e-01 2.00767249e-01 9.66709137e-01
-1.13861108e+00 -6.23790443e-01 5.33219934e-01 -2.21018106e-01
1.55238497e+00 -1.05266318e-01 6.54189169e-01 1.18848157e+00
4.64496911e-01 1.34782839e+00 3.53139162e-01 -3.44012380e-01
-1.36759505e-01 -6.55757606e-01 1.52901202e-01 8.48425508e-01
-5.33616662e-01 1.40250744e-02 -7.59233832e-01 3.89582813e-01
9.29012597e-01 6.17277920e-01 -1.87424675e-01 3.16189736e-01
-1.56039655e+00 4.69659835e-01 3.19931030e-01 5.43920398e-01
-6.11479819e-01 1.82020694e-01 1.14880836e+00 5.87376595e-01
4.50781554e-01 -2.97162801e-01 -3.88267428e-01 -7.51526356e-01
-7.20315695e-01 -3.16554368e-01 3.80776435e-01 6.65462911e-01
1.58281322e-03 4.48738694e-01 -3.88094157e-01 1.08048439e+00
1.40145734e-01 2.97862798e-01 1.28689945e+00 -6.78800881e-01
8.23463559e-01 4.91041303e-01 -1.52315915e-01 -1.06611037e+00
-2.36284807e-01 1.55419141e-01 -9.87081349e-01 -1.64285824e-01
7.77617544e-02 -1.71275437e-01 -1.15685678e+00 1.06378078e+00
-7.66009539e-02 5.42808175e-01 4.41604942e-01 7.58616090e-01
9.57619071e-01 1.27002680e+00 -1.03980890e-02 -3.41855466e-01
9.84960258e-01 -1.34231365e+00 -1.00081301e+00 -1.10580593e-01
7.72937536e-01 -5.79371214e-01 7.62006283e-01 4.12788332e-01
-1.03458703e+00 -9.86756325e-01 -1.05115461e+00 -1.82677448e-01
-4.84743237e-01 3.22075397e-01 2.51980007e-01 1.89963162e-01
-7.98078954e-01 6.69115484e-01 -1.12935877e+00 -3.75422090e-01
5.25202990e-01 4.62251037e-01 -3.17363411e-01 -1.12825260e-01
-1.35213971e+00 5.28913379e-01 4.37737167e-01 5.44279099e-01
-1.05450320e+00 -2.06082404e-01 -9.15500820e-01 1.37861356e-01
4.66449767e-01 -5.98916896e-02 1.33716130e+00 -1.32735085e+00
-1.93911862e+00 1.86158001e-01 -2.75385886e-01 -8.25767815e-01
2.50794798e-01 -3.29793602e-01 -7.75498271e-01 4.32276666e-01
-4.84073848e-01 5.61536252e-01 9.77426052e-01 -2.25834221e-01
-8.80264878e-01 -3.75252694e-01 -4.74839099e-02 2.07761630e-01
-6.60047829e-01 5.48643649e-01 -3.99001867e-01 -7.68061996e-01
8.18881169e-02 -6.45737588e-01 8.85446817e-02 -1.63103506e-01
-3.27706933e-02 -3.55120093e-01 1.47822750e+00 -8.26370716e-01
1.34336996e+00 -2.41437888e+00 3.31892595e-02 -2.83296317e-01
-1.86970696e-01 8.89995933e-01 -3.94095987e-01 3.17380846e-01
-1.62785888e-01 -1.25373930e-01 3.82067859e-01 -1.78101063e-01
-3.80430013e-01 4.52671915e-01 -4.54453588e-01 3.69516343e-01
-2.55792830e-02 9.16708767e-01 -6.35714412e-01 -4.20270324e-01
5.94431043e-01 6.56762421e-01 1.26468930e-02 2.78950036e-01
-6.03766995e-04 2.36795112e-01 -3.53854746e-01 5.58237553e-01
5.12164608e-02 -1.30752489e-01 -3.83428633e-02 -1.45814463e-01
-2.15889543e-01 2.85458565e-01 -8.73182774e-01 1.67342961e+00
-4.41613168e-01 1.03758097e+00 -4.14902806e-01 -1.23566329e+00
7.36207545e-01 6.79444432e-01 6.06693685e-01 -1.02148902e+00
3.07531208e-01 -4.06935997e-02 -1.38535559e-01 -9.62576628e-01
3.68971944e-01 6.83397278e-02 2.82433838e-01 2.59548008e-01
2.28921309e-01 7.42809057e-01 3.25429946e-01 -2.26070628e-01
1.12068248e+00 -7.23933801e-02 3.23372930e-02 4.69578594e-01
1.05302060e+00 -5.08187711e-01 6.45331621e-01 4.10505146e-01
-3.06773305e-01 3.43510956e-01 2.54986644e-01 -8.90442550e-01
-9.14944470e-01 -4.68676001e-01 4.62849855e-01 1.25540483e+00
-2.33071491e-01 -3.59185517e-01 -5.21964371e-01 -5.00328362e-01
-2.76627511e-01 3.44932377e-01 -6.08039677e-01 -2.56669074e-01
-8.20896685e-01 -2.61958063e-01 8.07936132e-01 1.00648952e+00
1.11281645e+00 -1.63454783e+00 -7.58717179e-01 4.36077744e-01
-1.63430512e-01 -1.38902247e+00 -4.37991530e-01 -2.35513106e-01
-9.80336905e-01 -9.20709491e-01 -8.76255572e-01 -1.00679195e+00
3.39593560e-01 4.03937519e-01 3.50536823e-01 -8.28084871e-02
-6.03426471e-02 1.45597428e-01 -7.23626435e-01 3.87247019e-02
-4.82410081e-02 -2.43783429e-01 7.97885507e-02 3.82900298e-01
4.64861393e-01 -3.39917749e-01 -3.44370991e-01 4.54686821e-01
-8.84482145e-01 -4.47761863e-02 6.41286373e-01 1.02811027e+00
5.51439524e-01 2.27282479e-01 4.19897079e-01 -3.28263760e-01
3.88136536e-01 -2.19339058e-01 -2.58234411e-01 4.47360516e-01
-2.67762281e-02 -1.39380649e-01 9.27189469e-01 -5.02096951e-01
-1.24622965e+00 -9.26177874e-02 -2.42337510e-01 -8.50341260e-01
-6.51225597e-02 8.29838753e-01 1.14537984e-01 -1.30224004e-01
-5.46650849e-02 8.17137063e-01 6.22419976e-02 -5.14031053e-01
-1.66697890e-01 1.02667201e+00 5.09539783e-01 1.09392039e-01
-1.57266036e-01 3.47204685e-01 -2.67347962e-01 -1.12289405e+00
-7.29590774e-01 -4.22317892e-01 -7.84763455e-01 -4.59558308e-01
1.20010388e+00 -8.93767893e-01 -9.06481922e-01 1.24397886e+00
-1.16335785e+00 -4.28487152e-01 -2.73535475e-02 8.47377002e-01
-5.60319483e-01 4.66549963e-01 -1.20091832e+00 -6.46766782e-01
-4.04101163e-01 -1.13095331e+00 7.18165934e-01 9.42417532e-02
1.68833449e-01 -8.68661642e-01 -3.04259032e-01 4.22160715e-01
2.97207713e-01 -3.91252600e-02 5.46032488e-01 -6.10764742e-01
-3.44828010e-01 -4.29274976e-01 -2.04457387e-01 7.37413347e-01
2.68391728e-01 8.97878706e-02 -5.65095425e-01 -2.68675357e-01
2.75321361e-02 -5.61559319e-01 1.18697560e+00 5.33619821e-01
1.71480608e+00 -4.93385017e-01 -7.80403316e-02 5.29304147e-01
1.01722372e+00 8.70150685e-01 9.91412640e-01 2.73278803e-01
8.92564535e-01 1.50958896e-01 5.76001942e-01 4.98806208e-01
1.80031881e-01 4.25067216e-01 1.52075961e-01 7.28236586e-02
-7.36428350e-02 -1.04854129e-01 9.39472914e-01 1.20473886e+00
-2.37728283e-01 -3.17396790e-01 -8.06428075e-01 4.24944907e-01
-2.10043263e+00 -1.48249972e+00 1.62768960e-01 1.94401479e+00
3.64264131e-01 3.33650827e-01 6.85801078e-03 4.55813706e-01
6.82894588e-01 7.13635206e-01 -5.30928671e-01 -4.59071100e-01
2.60410411e-03 2.01253151e-03 3.80498171e-01 1.50092602e-01
-1.26648760e+00 9.15205121e-01 6.16196299e+00 8.99215817e-01
-1.56165862e+00 2.68203225e-02 6.32388592e-01 -4.12843853e-01
6.56363666e-01 -6.58523142e-01 -5.86755216e-01 5.27552903e-01
1.42123294e+00 4.75588180e-02 2.24051714e-01 6.89106703e-01
5.68319559e-01 -7.26422369e-02 -9.07519460e-01 1.30610645e+00
3.29751730e-01 -1.48473918e+00 3.46055031e-01 -2.82511860e-01
4.56415057e-01 2.25523800e-01 1.20276526e-01 4.94413257e-01
-7.59467930e-02 -1.05077791e+00 5.21384239e-01 7.35861540e-01
6.85106874e-01 -9.79633272e-01 9.77681160e-01 2.25102410e-01
-1.47598922e+00 -7.22768843e-01 -4.37412649e-01 -3.01874012e-01
2.01067448e-01 2.12439105e-01 -3.60657156e-01 4.16863143e-01
9.69320595e-01 1.50441933e+00 -2.95943916e-01 7.06600010e-01
8.56445879e-02 6.05895698e-01 4.26791199e-02 -1.27677038e-01
6.65150642e-01 1.56735957e-01 2.70613998e-01 1.23865330e+00
3.46873075e-01 5.38730860e-01 3.40769649e-01 -1.44975871e-01
-1.50961965e-01 -8.68015215e-02 -5.03887892e-01 -6.25465214e-01
-4.66696434e-02 7.25060105e-01 -5.04088819e-01 -6.98213041e-01
-7.96420753e-01 1.02378213e+00 -2.28455607e-02 3.54319394e-01
-8.37199390e-01 -5.12486219e-01 4.42741960e-01 -2.65511394e-01
7.59458601e-01 -4.24785912e-01 8.52674693e-02 -1.15576530e+00
7.55954832e-02 -8.32786739e-01 6.98848903e-01 -9.40764606e-01
-8.40299368e-01 6.63250864e-01 -3.50220442e-01 -1.48692334e+00
-3.27264279e-01 -7.11498618e-01 -5.30182958e-01 2.87780821e-01
-1.41182387e+00 -9.89978135e-01 -1.66121289e-01 1.06997156e+00
1.36001098e+00 -5.08342743e-01 5.31129122e-01 5.52032411e-01
-8.83419335e-01 3.84822458e-01 2.82453805e-01 5.30066490e-01
4.25792038e-01 -4.60047066e-01 2.25474298e-01 9.50093329e-01
6.99178129e-02 2.60951072e-01 1.24452606e-01 -5.92131674e-01
-1.53415179e+00 -1.27019143e+00 7.24147558e-01 3.65519553e-01
6.77320480e-01 3.50227091e-03 -7.37363160e-01 8.37130427e-01
2.24973008e-01 3.11679900e-01 6.20012045e-01 -4.70660478e-01
-1.07020088e-01 -3.34508002e-01 -6.63186908e-01 3.25737447e-01
8.74576330e-01 -7.43119419e-01 -5.13830662e-01 1.79944396e-01
7.15650141e-01 -3.85890841e-01 -7.57814050e-01 4.89177674e-01
8.35916221e-01 -9.21759963e-01 8.62228572e-01 -7.08952427e-01
5.78829706e-01 -8.72885436e-02 -3.09663147e-01 -1.03517330e+00
-1.12274058e-01 -3.87548953e-01 -3.39838147e-01 8.26916099e-01
1.92507133e-01 -2.93179125e-01 6.01922452e-01 1.68844700e-01
-3.35446090e-01 -1.01160538e+00 -9.68295574e-01 -7.96577156e-01
-4.62318093e-01 -4.89379942e-01 3.46075386e-01 5.68457842e-01
1.33011073e-01 4.43072096e-02 -8.41330945e-01 -1.89476356e-01
6.69450387e-02 -1.23939618e-01 4.19692695e-01 -6.01033270e-01
5.82800433e-03 -1.82104915e-01 -7.62954891e-01 -1.57228935e+00
3.08021754e-01 -3.06696326e-01 5.06033376e-02 -1.58259511e+00
7.16358516e-03 5.25187731e-01 -7.31356859e-01 6.64888740e-01
2.80339003e-01 2.17769723e-02 1.81999102e-01 1.29636854e-01
-9.24810886e-01 8.09577644e-01 1.15345430e+00 -4.07155275e-01
-1.61994368e-01 -7.71381706e-02 2.38472149e-01 7.80031204e-01
6.51725054e-01 -1.93991095e-01 -3.47332388e-01 -7.33321130e-01
-2.66178161e-01 4.37258571e-01 2.40689054e-01 -1.46084869e+00
6.11212313e-01 -2.09289845e-02 7.81266689e-01 -8.87060821e-01
5.63258469e-01 -7.10500777e-01 -2.68631160e-01 7.38279223e-01
-6.98842049e-01 1.58707142e-01 1.41886383e-01 4.71921712e-01
-7.27451026e-01 1.56988218e-01 5.93255222e-01 -1.38143644e-01
-1.40093970e+00 4.50642288e-01 -8.71836722e-01 -4.87144321e-01
9.43666875e-01 -4.68890190e-01 -2.71317571e-01 -5.01921535e-01
-8.22430968e-01 2.26900131e-01 -3.25674295e-01 7.94475317e-01
1.24843192e+00 -1.50896609e+00 -4.34397489e-01 4.09537137e-01
-2.08623454e-01 -4.84597504e-01 6.35322928e-01 8.75127077e-01
-4.52237129e-01 7.41129160e-01 -6.00623906e-01 -4.47531402e-01
-1.71901834e+00 5.28982460e-01 2.41848007e-01 -2.23629355e-01
-6.91779375e-01 9.25335705e-01 -2.42139220e-01 1.11614361e-01
6.63442016e-01 -5.15580535e-01 -6.11777067e-01 7.52810314e-02
9.93915737e-01 5.34770787e-01 -4.48051207e-02 -8.82254243e-01
-3.23478490e-01 3.80504012e-01 -3.52674991e-01 1.26874685e-01
1.44325578e+00 -2.25804955e-01 -6.15870915e-02 7.67027259e-01
1.49396610e+00 -8.68629515e-01 -1.31183887e+00 -3.34192991e-01
-7.66495913e-02 -4.77619737e-01 3.02479953e-01 -4.14504409e-01
-1.50812316e+00 1.13056540e+00 6.28382683e-01 -3.38649005e-02
1.25758493e+00 -6.05880618e-01 1.55473077e+00 7.82329440e-01
1.31342396e-01 -1.42602181e+00 5.31084776e-01 1.02100277e+00
8.65428448e-01 -1.05992198e+00 -2.95613617e-01 1.31495163e-01
-6.09492779e-01 1.60208559e+00 6.12531602e-01 -1.38157085e-01
7.37764955e-01 4.50478420e-02 -5.22464141e-02 2.18023602e-02
-1.07626677e+00 -9.65555198e-03 9.37824175e-02 8.60716775e-02
3.12031865e-01 -3.59518200e-01 -3.51986848e-02 3.42934877e-01
3.92357379e-01 4.83715534e-01 4.47162837e-01 1.18609798e+00
-5.81204653e-01 -7.67529070e-01 -1.77425131e-01 5.32712936e-01
-6.77991927e-01 -3.54910307e-02 -1.80649757e-01 3.23561341e-01
-2.03840971e-01 1.07425880e+00 3.31817895e-01 -8.14613938e-01
1.66379020e-01 1.76320940e-01 3.84839952e-01 -2.33640686e-01
-4.03869629e-01 2.77598072e-02 1.65620774e-01 -9.77851510e-01
-7.64065027e-01 -6.11591101e-01 -1.35799003e+00 -2.09890649e-01
-7.85657912e-02 1.83583871e-02 5.07602811e-01 1.34486103e+00
4.25986707e-01 7.75032938e-01 7.24394083e-01 -8.76292169e-01
-3.11780542e-01 -1.13333809e+00 -4.14364219e-01 1.50459364e-01
5.51306844e-01 -5.72216094e-01 -6.01755790e-02 4.32331949e-01] | [8.44186782836914, 0.3907858729362488] |
1a7acfcb-b3d7-4324-a771-51af3988ee42 | in-n-out-towards-good-initialization-for | 2106.13953 | null | https://arxiv.org/abs/2106.13953v3 | https://arxiv.org/pdf/2106.13953v3.pdf | In-N-Out: Towards Good Initialization for Inpainting and Outpainting | In computer vision, recovering spatial information by filling in masked regions, e.g., inpainting, has been widely investigated for its usability and wide applicability to other various applications: image inpainting, image extrapolation, and environment map estimation. Most of them are studied separately depending on the applications. Our focus, however, is on accommodating the opposite task, e.g., image outpainting, which would benefit the target applications, e.g., image inpainting. Our self-supervision method, In-N-Out, is summarized as a training approach that leverages the knowledge of the opposite task into the target model. We empirically show that In-N-Out -- which explores the complementary information -- effectively takes advantage over the traditional pipelines where only task-specific learning takes place in training. In experiments, we compare our method to the traditional procedure and analyze the effectiveness of our method on different applications: image inpainting, image extrapolation, and environment map estimation. For these tasks, we demonstrate that In-N-Out consistently improves the performance of the recent works with In-N-Out self-supervision to their training procedure. Also, we show that our approach achieves better results than an existing training approach for outpainting. | ['Sung-Eui Yoon', 'Woobin Im', 'Changho Jo'] | 2021-06-26 | null | null | null | null | ['image-outpainting'] | ['computer-vision'] | [ 4.98257220e-01 5.32425344e-02 -1.14925377e-01 -2.91037291e-01
-8.67483735e-01 -2.73106843e-01 4.07914221e-01 -3.18513811e-01
-3.92875403e-01 5.03049433e-01 1.85967073e-01 -3.20608526e-01
3.56258929e-01 -3.84748131e-01 -1.22726429e+00 -7.20358729e-01
2.39104107e-01 6.66568950e-02 5.18966734e-01 -1.82574049e-01
3.75306517e-01 3.32361549e-01 -1.54816794e+00 4.01069611e-01
9.89684939e-01 8.50929260e-01 8.27399373e-01 5.31587839e-01
3.10461987e-02 9.22854722e-01 -5.84570527e-01 -1.00377416e-02
4.93368924e-01 -4.86911237e-01 -5.81467092e-01 2.95484811e-01
5.66552699e-01 -6.83393180e-01 -3.88742745e-01 1.02952111e+00
3.78057808e-01 2.28414521e-01 3.94515693e-01 -9.94178653e-01
-5.13124287e-01 2.56770521e-01 -1.17791700e+00 2.93070357e-02
1.82591528e-01 2.73632795e-01 3.17327142e-01 -1.02822137e+00
6.78527296e-01 1.11387432e+00 8.29867661e-01 3.65075648e-01
-1.19852364e+00 -5.93809187e-01 1.87262028e-01 1.41252145e-01
-1.18687499e+00 -5.01846373e-01 8.67022634e-01 -4.25818443e-01
8.06198895e-01 1.68812945e-02 2.27028385e-01 8.99584055e-01
1.22495815e-01 1.10940194e+00 1.34557176e+00 -6.42712772e-01
2.15675980e-01 1.20093465e-01 -2.84481585e-01 5.73707879e-01
-2.03645170e-01 3.55596095e-01 -6.00557029e-01 3.31773937e-01
1.01725471e+00 4.87875976e-02 -3.28203082e-01 -2.48204544e-01
-1.20046210e+00 4.59905863e-01 6.42679989e-01 -1.15701512e-01
-2.99616396e-01 1.33728206e-01 2.56246686e-01 2.78811038e-01
7.63601899e-01 3.34343731e-01 -4.34004456e-01 1.04387075e-01
-1.39574730e+00 2.33950675e-01 3.95626128e-01 1.24336517e+00
9.97971117e-01 -2.62400806e-02 -2.38700718e-01 8.35000634e-01
-1.10283554e-01 3.24697256e-01 4.54865843e-01 -1.10400164e+00
5.54772377e-01 3.26267242e-01 3.91805887e-01 -4.85253364e-01
-1.40282676e-01 -4.08830643e-01 -8.44773233e-01 4.51376826e-01
2.86412090e-01 -1.74266100e-01 -1.18916261e+00 1.70961988e+00
4.56287026e-01 5.56563854e-01 2.38093883e-02 8.35308552e-01
5.84912300e-01 6.74273193e-01 -4.86729443e-02 -6.11271858e-02
1.30731642e+00 -1.71340132e+00 -6.03145540e-01 -6.78087711e-01
4.23762172e-01 -8.82080615e-01 1.25588596e+00 4.08104658e-01
-1.07018387e+00 -8.70729625e-01 -9.95637953e-01 -4.15806502e-01
-1.33046493e-01 2.77108014e-01 4.50144380e-01 2.19183043e-01
-1.18458116e+00 7.10927784e-01 -7.06486583e-01 -4.83454615e-01
4.87572074e-01 9.55079496e-02 -6.23211205e-01 -2.45901167e-01
-8.58319759e-01 9.48484004e-01 4.54412997e-01 -1.91642120e-02
-9.11413252e-01 -6.08975351e-01 -9.95471478e-01 -1.26653120e-01
5.29656291e-01 -6.61074102e-01 1.33315194e+00 -1.02332473e+00
-1.39329100e+00 8.25810730e-01 -3.38140845e-01 -6.50727391e-01
7.58634746e-01 -5.93214989e-01 -8.24366789e-03 1.19275637e-01
3.59507143e-01 1.12772632e+00 1.08039129e+00 -1.41591120e+00
-6.08843625e-01 -3.63251448e-01 1.07120275e-01 5.06968319e-01
-2.80861855e-01 5.86381666e-02 -8.60157490e-01 -9.11933661e-01
4.54628021e-02 -8.35102081e-01 -2.46990621e-01 1.58620343e-01
-4.22646821e-01 1.45524889e-01 1.20804226e+00 -9.66728151e-01
1.03461528e+00 -2.38894272e+00 3.06839421e-02 -3.74423027e-01
4.84744646e-02 3.19888175e-01 -1.51125014e-01 5.07963479e-01
1.45836025e-02 -2.00306505e-01 -5.04028916e-01 -8.72929394e-01
-4.05468374e-01 3.41517955e-01 -3.89348418e-01 3.85108948e-01
3.90877038e-01 7.39306748e-01 -9.24013436e-01 -6.54366314e-01
4.75664198e-01 5.21670401e-01 -4.88179147e-01 4.23105329e-01
-3.38021338e-01 8.23466420e-01 -5.43797798e-02 4.40958828e-01
1.02550554e+00 -1.28980905e-01 -1.49843916e-01 -2.56380230e-01
-2.83746302e-01 -2.95465849e-02 -8.87926698e-01 2.10960627e+00
-7.23290324e-01 8.24614346e-01 2.75371104e-01 -9.01539028e-01
6.04847908e-01 4.11889218e-02 1.57779858e-01 -8.28696251e-01
-1.86207920e-01 3.09285432e-01 -6.25954270e-02 -6.90337479e-01
5.55190027e-01 1.16866447e-01 2.16166005e-01 2.66274452e-01
4.47103046e-02 -2.73783039e-02 1.26130849e-01 1.48588315e-01
9.24886525e-01 8.51190507e-01 5.58902621e-01 -3.97744365e-02
1.93833038e-01 -6.67145429e-03 4.50037867e-01 8.91222835e-01
-1.84287056e-01 1.02340448e+00 2.40813375e-01 -1.82954684e-01
-1.19716418e+00 -7.12347448e-01 -6.03627749e-02 1.06710517e+00
4.24787551e-01 -2.11933002e-01 -1.06358409e+00 -7.28038430e-01
-9.45775360e-02 6.70829356e-01 -7.49558866e-01 6.29601628e-03
-6.62230432e-01 -4.43152905e-01 2.59508133e-01 7.80724823e-01
9.98437881e-01 -1.28193820e+00 -8.94935668e-01 1.33397385e-01
-2.12807789e-01 -1.32288885e+00 -6.82842076e-01 2.77509838e-01
-1.07587564e+00 -7.81803370e-01 -9.37398672e-01 -7.74598897e-01
6.41040087e-01 9.06922936e-01 1.05202508e+00 -6.39064908e-02
-2.32693721e-02 5.74074872e-02 -4.15197134e-01 -4.57469940e-01
-2.93624163e-01 1.23233005e-01 -3.62441391e-01 5.11727668e-02
6.31115660e-02 -8.07068706e-01 -9.21089828e-01 5.25122285e-01
-1.15581763e+00 5.25003612e-01 1.05610716e+00 8.87159526e-01
5.45418561e-01 -1.31727979e-01 3.42070013e-01 -1.23520708e+00
2.00541347e-01 -4.33141917e-01 -3.89921963e-01 1.75193697e-01
-3.95662278e-01 1.03683621e-01 6.01132691e-01 -4.28610057e-01
-1.16692185e+00 4.46862161e-01 -2.27207795e-01 -7.58052528e-01
-2.27711201e-01 1.66628391e-01 -3.20665061e-01 -2.15395465e-01
7.27348864e-01 5.48575521e-01 1.54090941e-01 -5.75333536e-01
3.15497577e-01 7.24152088e-01 8.71366084e-01 -3.23722243e-01
8.19894195e-01 6.37547135e-01 -1.13200285e-01 -7.99941242e-01
-1.00372362e+00 -6.19483352e-01 -5.31152189e-01 2.94822939e-02
7.50530303e-01 -1.16997111e+00 -1.89282209e-01 5.70294678e-01
-1.27156126e+00 -6.03401363e-01 -2.85538793e-01 1.86841488e-01
-6.50581956e-01 4.82900620e-01 -4.83150810e-01 -7.87824988e-01
-1.10936701e-01 -1.37182522e+00 1.55031407e+00 2.21931547e-01
1.58550829e-01 -8.97423267e-01 -1.61238313e-01 3.73598397e-01
5.29695332e-01 1.13587342e-01 3.63165796e-01 -3.42697382e-01
-6.91618443e-01 1.07169211e-01 -3.91968668e-01 5.44234633e-01
2.00567275e-01 -6.44663453e-01 -1.32815289e+00 -2.54696071e-01
2.74057508e-01 -2.95268267e-01 1.12176096e+00 3.15870345e-01
1.36686671e+00 -2.75861681e-01 -3.46353173e-01 9.24986362e-01
1.44816196e+00 4.72715683e-02 1.01294959e+00 5.27135015e-01
8.35050762e-01 7.15553164e-01 9.54010844e-01 1.95405930e-01
3.39192003e-01 9.73115563e-01 5.65167904e-01 -5.07858217e-01
-4.38073397e-01 -4.51758921e-01 3.88938129e-01 3.21666837e-01
1.39711592e-02 -3.81911509e-02 -5.52042305e-01 5.75432360e-01
-2.14561915e+00 -7.37508714e-01 2.10630402e-01 2.20700860e+00
9.73975182e-01 -1.41575024e-01 1.08290792e-01 -4.60662916e-02
7.73114920e-01 2.92818487e-01 -8.09285343e-01 3.51895206e-02
2.23612878e-03 9.38042253e-02 6.05878949e-01 5.92087567e-01
-1.25586700e+00 1.11235952e+00 5.81588697e+00 9.36911464e-01
-1.08748090e+00 4.11249101e-01 8.78551126e-01 5.76004460e-02
-2.69768611e-02 1.72949195e-01 -6.78076923e-01 6.48617029e-01
4.21161234e-01 5.25299549e-01 4.31091964e-01 9.02059495e-01
3.42842728e-01 -5.35784841e-01 -1.28539062e+00 1.04821658e+00
3.02541047e-01 -1.17163014e+00 -1.60523072e-01 -1.96491584e-01
8.10549676e-01 -3.11128385e-02 5.61154820e-02 2.69760609e-01
-2.92463005e-02 -7.75510848e-01 9.54157352e-01 2.14243010e-01
9.42923009e-01 -5.53519130e-01 5.94994009e-01 7.22505867e-01
-9.26837325e-01 -6.31748214e-02 -2.51902252e-01 -1.21575736e-01
2.80487061e-01 7.45734751e-01 -6.46789432e-01 5.75649917e-01
7.04106688e-01 6.82417512e-01 -6.16638660e-01 1.07758498e+00
-4.14306045e-01 5.30731618e-01 -1.75527081e-01 5.81251144e-01
1.82524130e-01 -2.35236064e-01 4.17373329e-01 1.21883559e+00
3.64382207e-01 -2.34347433e-01 3.33196998e-01 7.64215469e-01
-1.55044897e-02 -9.43464488e-02 -7.49515355e-01 5.60594559e-01
4.35398728e-01 1.08994615e+00 -6.40985131e-01 -3.77993077e-01
-4.57508087e-01 1.41485059e+00 4.30212200e-01 3.37405682e-01
-9.02700722e-01 -5.53541422e-01 2.83884495e-01 4.44273174e-01
5.38535893e-01 -1.10771395e-01 -2.57293254e-01 -1.05365503e+00
3.56476873e-01 -9.08460796e-01 -3.83911803e-02 -1.00579810e+00
-9.32825327e-01 6.71961665e-01 2.05052063e-01 -1.33844304e+00
-4.20858681e-01 -4.31492239e-01 -6.44052684e-01 9.75904107e-01
-1.87650824e+00 -1.46699119e+00 -5.77462733e-01 5.10168850e-01
1.08408749e+00 7.16099665e-02 3.64871532e-01 3.92087400e-01
-6.13160133e-01 4.90648299e-01 1.08215131e-01 -8.01080316e-02
9.92838681e-01 -1.17791057e+00 5.04653215e-01 1.20640922e+00
9.37586054e-02 5.90812206e-01 7.68705130e-01 -6.21632636e-01
-1.26700032e+00 -1.32883096e+00 5.32486558e-01 -1.78821549e-01
2.72745520e-01 -5.39573908e-01 -7.26855278e-01 9.71873879e-01
4.51990277e-01 3.85452151e-01 8.38064179e-02 -2.99133182e-01
-4.07447606e-01 -7.52488822e-02 -1.19362962e+00 6.36612475e-01
1.17447102e+00 -4.28025693e-01 -2.80375928e-01 5.34364581e-01
7.96470642e-01 -8.81849349e-01 -3.17503422e-01 2.11713180e-01
4.38673466e-01 -1.33268785e+00 1.08900774e+00 -2.29173660e-01
8.38160813e-01 -6.42236292e-01 -1.37791410e-01 -1.34842992e+00
-1.35247916e-01 -8.30073118e-01 -2.87272662e-01 1.11212182e+00
3.10971849e-02 -5.79711914e-01 7.63072908e-01 1.58938199e-01
-3.78225684e-01 -8.06874096e-01 -5.18714964e-01 -7.62409925e-01
-1.66182309e-01 -2.32410446e-01 5.50205112e-01 7.09246814e-01
-5.77933252e-01 3.09584826e-01 -8.14409077e-01 2.78556198e-01
7.14218974e-01 1.87111259e-01 1.07701743e+00 -7.68548727e-01
-7.28151202e-01 1.11741222e-01 -1.10445671e-01 -1.68961680e+00
6.11387938e-02 -4.02771920e-01 4.08348054e-01 -1.53318655e+00
2.14979023e-01 -3.19224834e-01 -4.10282984e-03 4.73058313e-01
-5.34619451e-01 5.71117282e-01 3.05024177e-01 4.85985935e-01
-5.36484540e-01 2.88454711e-01 1.40684068e+00 -9.54464227e-02
-1.98462725e-01 7.91718289e-02 -1.00229275e+00 7.63185441e-01
7.19365954e-01 -3.50210816e-01 -6.70186102e-01 -6.80781245e-01
-1.83009058e-01 1.11712113e-01 4.83206242e-01 -1.16271079e+00
2.91124046e-01 -4.97272871e-02 3.04740340e-01 -5.79178095e-01
4.07112986e-01 -7.70608842e-01 -5.09800576e-02 1.55736327e-01
-1.74295902e-01 1.09930485e-01 3.25318605e-01 5.60170352e-01
-4.51122105e-01 -3.91984969e-01 7.05871284e-01 -3.37443203e-01
-1.00571394e+00 1.43328398e-01 1.03690311e-01 -1.73586980e-02
9.86284256e-01 -5.00949919e-01 -1.90385282e-01 -5.28391182e-01
-5.13142228e-01 1.43471181e-01 6.38482809e-01 1.51064128e-01
6.84496880e-01 -1.09024131e+00 -7.22354829e-01 2.98463941e-01
1.69294015e-01 5.12466013e-01 3.96335900e-01 9.55939174e-01
-5.86088061e-01 4.11876738e-02 -1.42535731e-01 -8.49049985e-01
-1.10308373e+00 6.86472297e-01 1.04268722e-01 -4.19577181e-01
-7.78031528e-01 7.29969144e-01 7.66554236e-01 -2.97360420e-01
3.36492091e-01 -2.01855719e-01 5.37673905e-02 -3.81333828e-01
6.42466486e-01 1.43236384e-01 1.51212052e-01 -3.91495168e-01
-7.35466033e-02 5.70810318e-01 -3.86470020e-01 -1.37280777e-01
1.32056677e+00 -2.01008677e-01 -1.08186468e-01 2.29964197e-01
1.02538383e+00 8.19735043e-03 -2.02560687e+00 -2.84963876e-01
-1.55120820e-01 -7.97752678e-01 7.14174286e-02 -7.34166861e-01
-1.05924821e+00 1.02701938e+00 6.43073082e-01 -1.08777449e-01
1.59865904e+00 -1.85585931e-01 7.84142852e-01 1.74110532e-01
4.30471420e-01 -9.84840035e-01 3.71365994e-02 3.58528316e-01
9.19041395e-01 -1.51642036e+00 8.60060230e-02 -6.82702839e-01
-7.34227359e-01 8.20140183e-01 8.14492643e-01 -2.70800710e-01
4.84009206e-01 5.14255345e-01 6.76550269e-02 2.52989501e-01
-4.56098884e-01 -1.50468037e-01 6.16909638e-02 8.13029528e-01
2.69550204e-01 -2.47766510e-01 1.72805786e-02 8.18424746e-02
3.33959647e-02 1.91308320e-01 3.29016536e-01 1.03716695e+00
-3.09992760e-01 -1.11766589e+00 -5.15109181e-01 1.96950898e-01
-1.99373081e-01 -3.31142277e-01 -2.28707671e-01 9.50659335e-01
3.54623854e-01 7.55629241e-01 -8.20274800e-02 -2.64426857e-01
2.03662485e-01 -2.40450189e-01 6.24180257e-01 -6.25323057e-01
-6.46164596e-01 1.93680167e-01 -1.80391550e-01 -7.24088013e-01
-5.42531013e-01 -5.10811031e-01 -9.15677547e-01 3.56191099e-02
-3.01499277e-01 -2.63416857e-01 6.58586025e-01 9.34069395e-01
6.75523341e-01 6.20483816e-01 5.60603142e-01 -1.30551910e+00
-6.36385739e-01 -1.15819883e+00 -4.11799699e-01 4.79007423e-01
5.53053141e-01 -5.93083739e-01 -1.94935232e-01 8.63051713e-02] | [11.090998649597168, -0.8745765089988708] |
7df981a2-7bfc-454d-ab2e-0e159304275b | disco-distilled-student-models-co-training | 2305.12074 | null | https://arxiv.org/abs/2305.12074v1 | https://arxiv.org/pdf/2305.12074v1.pdf | DisCo: Distilled Student Models Co-training for Semi-supervised Text Mining | Many text mining models are constructed by fine-tuning a large deep pre-trained language model (PLM) in downstream tasks. However, a significant challenge is maintaining performance when we use a lightweight model with limited labeled samples. We present DisCo, a semi-supervised learning (SSL) framework for fine-tuning a cohort of small student models generated from a large PLM using knowledge distillation. Our key insight is to share complementary knowledge among distilled student cohorts to promote their SSL effectiveness. DisCo employs a novel co-training technique to optimize multiple small student models by promoting knowledge sharing among students under diversified views: model views produced by different distillation strategies and data views produced by various input augmentations. We evaluate DisCo on both semi-supervised text classification and extractive summarization tasks. Experimental results show that DisCo can produce student models that are 7.6 times smaller and 4.8 times faster in inference than the baseline PLMs while maintaining comparable performance. We also show that DisCo-generated student models outperform the similar-sized models elaborately tuned in distinct tasks. | ['Zheng Wang', 'Ting Deng', 'Weiyi Yang', 'Chenghua Lin', 'JianXin Li', 'Qianren Mao', 'Weifeng Jiang'] | 2023-05-20 | null | null | null | null | ['semi-supervised-text-classification-1', 'extractive-summarization'] | ['natural-language-processing', 'natural-language-processing'] | [ 2.65069574e-01 6.35252833e-01 -6.14782155e-01 -5.23134410e-01
-1.18646324e+00 -6.68121338e-01 7.52261162e-01 4.92413282e-01
-4.49801981e-01 9.20672119e-01 7.21053421e-01 -3.89296114e-01
1.83050662e-01 -8.49187016e-01 -9.19402003e-01 -2.44564697e-01
2.93641090e-01 9.23611224e-01 1.14393428e-01 -2.98539490e-01
2.09625274e-01 -6.60626441e-02 -1.39869189e+00 7.47750759e-01
1.49586844e+00 4.32545006e-01 1.45959571e-01 1.10906351e+00
-4.37796503e-01 9.39351737e-01 -8.40346515e-01 -6.89410210e-01
1.37698695e-01 -4.51425403e-01 -1.10242569e+00 -1.56956345e-01
1.00766158e+00 -2.71589667e-01 3.59563828e-02 6.36008382e-01
6.96498394e-01 3.00359398e-01 6.93550408e-01 -9.74568963e-01
-8.98537099e-01 1.61132812e+00 -4.30372864e-01 5.63751683e-02
-1.88672394e-02 -9.11266655e-02 1.24630511e+00 -9.08764184e-01
5.43387890e-01 1.36560643e+00 6.89588428e-01 7.41850913e-01
-1.53222847e+00 -7.22322583e-01 2.91347444e-01 -1.22162163e-01
-7.53881037e-01 -5.00406623e-01 3.99154246e-01 -1.60599738e-01
9.96220171e-01 3.12222421e-01 4.60282892e-01 1.25196159e+00
-9.79253799e-02 1.48517776e+00 9.38758314e-01 -5.15428901e-01
3.64351198e-02 5.57285488e-01 4.36361462e-01 9.75790381e-01
5.20202041e-01 -4.94059891e-01 -1.30881286e+00 -2.78663188e-01
2.21188515e-01 -1.52877331e-01 -2.34526426e-01 -2.61719972e-01
-1.12909544e+00 9.80093420e-01 -6.68638125e-02 -1.92449078e-01
1.88488483e-01 -2.32213605e-02 5.31042993e-01 5.12915373e-01
8.12637329e-01 9.73534882e-01 -1.03235948e+00 -2.93408632e-01
-1.37628579e+00 2.75112122e-01 1.11401975e+00 1.06352115e+00
4.79216635e-01 2.02649012e-01 -4.59856510e-01 9.54548955e-01
2.01941133e-01 3.99032116e-01 1.04039705e+00 -7.98497856e-01
9.21786964e-01 8.83376300e-01 -3.40034068e-01 -2.36254588e-01
-1.50153652e-01 -6.68416500e-01 -7.83962905e-01 -1.01783685e-01
2.42229670e-01 -3.93447191e-01 -9.15001094e-01 1.56889403e+00
2.02384546e-01 6.28364235e-02 5.13531089e-01 1.19467182e-02
1.31677401e+00 7.66432464e-01 2.55435817e-02 3.43842283e-02
1.01466107e+00 -1.53974640e+00 -4.29954946e-01 -4.21263218e-01
1.15034318e+00 -6.08311951e-01 1.13579452e+00 5.51805556e-01
-1.55149496e+00 -5.84499359e-01 -8.64593685e-01 -4.28362757e-01
-2.82548100e-01 2.95449525e-01 3.35428387e-01 5.36598742e-01
-1.01037025e+00 6.73363388e-01 -7.72027671e-01 -4.50736936e-03
6.73429489e-01 2.41220340e-01 -2.24099636e-01 2.91954011e-01
-8.72289956e-01 7.34030426e-01 5.65367281e-01 -5.06638050e-01
-8.32269013e-01 -1.47735536e+00 -9.84782219e-01 4.53957856e-01
2.12588623e-01 -9.03079331e-01 1.60103154e+00 -5.54876506e-01
-1.56878531e+00 8.68416846e-01 -2.59434313e-01 -8.05936515e-01
6.74212277e-01 -6.71446264e-01 1.51614636e-01 -7.68289939e-02
2.32106045e-01 6.41262949e-01 5.38298368e-01 -1.20833910e+00
-7.08537519e-01 -3.19846809e-01 -3.37694108e-01 4.62747425e-01
-6.87787533e-01 -2.58075148e-01 -2.51097977e-01 -8.45008194e-01
-3.22653025e-01 -5.50963819e-01 -3.65752041e-01 -5.62565982e-01
-7.60321081e-01 -4.63806480e-01 6.06416702e-01 -5.60575426e-01
1.55159581e+00 -1.34951484e+00 3.30949098e-01 -9.67090875e-02
3.63218278e-01 7.67127335e-01 -3.00526828e-01 4.04512674e-01
2.34469011e-01 3.30351114e-01 -2.19715357e-01 -9.12962794e-01
-3.24075036e-02 1.12891130e-01 -6.55820549e-01 -3.22728783e-01
8.49603564e-02 1.07492852e+00 -1.09934616e+00 -6.31744683e-01
-2.03265756e-01 -2.16188468e-02 -7.69269645e-01 3.91046375e-01
-5.25895953e-01 -1.05992975e-02 -4.41888839e-01 3.43302816e-01
3.60829681e-01 -3.08702320e-01 1.92111954e-01 3.00891131e-01
6.65250719e-02 7.31285870e-01 -1.11105239e+00 1.75068212e+00
-6.67894661e-01 6.07129991e-01 -8.49598367e-03 -9.74115968e-01
9.60080504e-01 1.50815323e-01 7.04683736e-02 -2.87299842e-01
-2.02368498e-01 1.25351056e-01 -4.43391979e-01 -3.92917633e-01
8.68409157e-01 6.24875948e-02 -8.55127200e-02 1.11636055e+00
5.15788674e-01 -3.34387153e-01 2.79981524e-01 7.72843838e-01
9.92083490e-01 1.12270370e-01 3.34005535e-01 -3.73712510e-01
2.87110329e-01 7.02224225e-02 4.96853054e-01 1.07906556e+00
1.99896321e-01 4.14650261e-01 5.59853137e-01 -2.12930948e-01
-7.49646068e-01 -1.03351939e+00 2.69294102e-02 1.83637559e+00
-5.02370834e-01 -9.15723801e-01 -6.21671736e-01 -1.12445772e+00
2.39348277e-01 9.96865094e-01 -7.21996605e-01 -3.06288511e-01
-6.12701118e-01 -9.37858164e-01 8.40834916e-01 8.96965563e-01
3.45110238e-01 -9.04073834e-01 -3.31847608e-01 1.43831968e-01
-1.48287099e-02 -8.17975879e-01 -5.79501271e-01 4.77164000e-01
-1.19274247e+00 -8.11144888e-01 -4.58249778e-01 -9.92733598e-01
7.62180626e-01 2.40989342e-01 1.63357162e+00 -2.19289199e-01
1.44684687e-02 2.91154832e-01 -2.05052778e-01 -9.62777555e-01
-8.76916528e-01 7.93422997e-01 1.73515361e-02 -4.08609658e-01
3.37509573e-01 -5.30389488e-01 -2.83276290e-01 -2.36269489e-01
-7.82981873e-01 6.57415926e-01 5.30477881e-01 1.01076341e+00
3.80195767e-01 -2.97171623e-01 9.11730587e-01 -1.65516257e+00
8.99043202e-01 -3.67333829e-01 -2.12612465e-01 5.88326931e-01
-9.43034828e-01 7.26907253e-01 8.08582604e-01 -5.39694667e-01
-1.53101921e+00 -3.29380453e-01 1.35641918e-01 -3.97888124e-02
4.70072888e-02 4.86341476e-01 4.09794934e-02 5.70726991e-01
9.68341470e-01 3.73461872e-01 -2.70098858e-02 -5.89771211e-01
6.43699765e-01 6.83201730e-01 3.78936857e-01 -8.79761815e-01
7.37437189e-01 3.54359478e-01 -4.34752524e-01 -5.98070443e-01
-1.64016700e+00 -3.71656418e-01 -7.20586658e-01 3.65448475e-01
4.60392535e-01 -1.21808684e+00 -3.61324370e-01 4.57841575e-01
-8.15412343e-01 -8.73485446e-01 -6.30446553e-01 1.59977287e-01
-2.53333032e-01 1.53327912e-01 -7.23950803e-01 -4.37246591e-01
-1.04751015e+00 -8.47233891e-01 1.14209068e+00 4.09696162e-01
-3.85333329e-01 -1.44442391e+00 4.14722711e-01 9.64669526e-01
3.55038434e-01 -3.62690687e-01 9.05121326e-01 -1.41780579e+00
-2.25535035e-01 1.82094127e-01 1.39577046e-01 4.71877158e-01
-5.87431109e-03 -2.48300694e-02 -1.11175096e+00 -2.46029362e-01
-4.51760679e-01 -1.04763043e+00 1.48809242e+00 2.81285256e-01
1.29337335e+00 -5.29268265e-01 -2.76247025e-01 6.86378896e-01
8.55501711e-01 -4.98684645e-01 -1.44501716e-01 3.06497633e-01
1.00774658e+00 7.43433058e-01 3.45500261e-01 3.55895519e-01
6.51043236e-01 1.34828284e-01 -8.45875740e-02 3.24288569e-02
-6.33989647e-02 -6.06950700e-01 6.86007023e-01 1.08296597e+00
1.95123449e-01 -1.32995158e-01 -7.85602987e-01 7.20897734e-01
-1.96799636e+00 -1.02923071e+00 -1.49763092e-01 1.96600366e+00
1.76966083e+00 2.91188806e-01 7.14072213e-02 -3.13670963e-01
2.26647273e-01 2.78166831e-01 -5.92062354e-01 -6.99523807e-01
-5.58201484e-02 7.17395008e-01 4.17861253e-01 7.18027234e-01
-8.82782340e-01 1.07174933e+00 6.24921179e+00 9.44469273e-01
-7.36588776e-01 -4.79037985e-02 6.62621081e-01 -7.28107333e-01
-6.74582660e-01 -2.22101092e-01 -1.53580642e+00 1.82711869e-01
1.25066650e+00 -4.99166399e-01 -1.78372815e-01 9.91693616e-01
1.08019724e-01 -8.28191638e-04 -1.23120499e+00 3.56339306e-01
4.70146716e-01 -1.76976764e+00 6.05295122e-01 -9.86015946e-02
1.43734908e+00 4.09470163e-02 9.37041044e-02 8.53070199e-01
1.07409894e+00 -1.13167334e+00 4.78556842e-01 2.67550766e-01
4.27028358e-01 -8.72063696e-01 3.97086173e-01 7.72562683e-01
-8.47890973e-01 2.07229573e-02 -2.67799139e-01 7.48108253e-02
-1.54048696e-01 5.26691079e-01 -1.36811507e+00 3.50660235e-01
5.63360989e-01 6.67867780e-01 -9.98675346e-01 3.99456918e-01
-2.83228189e-01 1.19099581e+00 -9.54627618e-02 -4.42993455e-02
2.62126356e-01 -5.18427826e-02 3.35524470e-01 1.56427705e+00
-5.08485921e-02 -2.72295564e-01 2.32807755e-01 7.92087257e-01
-5.75820327e-01 5.12572341e-02 -4.40219969e-01 4.41115163e-02
6.76094830e-01 1.30518317e+00 -1.50993854e-01 -8.79396260e-01
-3.26936454e-01 8.75153005e-01 8.51089597e-01 8.17060322e-02
-3.50352675e-01 -3.20728093e-01 4.38574582e-01 -1.49032176e-02
2.50174254e-01 1.17603339e-01 -5.03039420e-01 -1.40505087e+00
-1.71793565e-01 -1.05216300e+00 6.32242084e-01 -4.77110237e-01
-1.16629791e+00 1.60704643e-01 8.68705958e-02 -6.90549850e-01
-4.38552707e-01 -4.66703743e-01 -1.02523410e+00 8.85186017e-01
-1.54729176e+00 -1.37310433e+00 -4.14492115e-02 2.22574919e-01
1.10471523e+00 -4.85760599e-01 8.52080703e-01 -3.07977229e-01
-8.85613978e-01 1.02363276e+00 4.19591933e-01 1.28448531e-01
1.17743003e+00 -2.00921774e+00 7.40584075e-01 6.62541807e-01
2.65316099e-01 6.88208401e-01 5.30109942e-01 -7.73524106e-01
-1.07240570e+00 -1.45400929e+00 1.15204501e+00 -9.90648746e-01
5.89717329e-01 -3.18493098e-01 -1.16382945e+00 1.08231556e+00
5.18607438e-01 -3.83643001e-01 1.27393603e+00 4.40091193e-01
-3.86454582e-01 1.59091756e-01 -7.30882525e-01 6.59564316e-01
8.25726688e-01 -2.69203335e-01 -1.11571515e+00 2.57777601e-01
9.10720706e-01 -4.76571292e-01 -8.88350129e-01 2.63312161e-01
3.85690838e-01 -8.04028630e-01 9.18717504e-01 -1.27445626e+00
9.55700576e-01 3.71073961e-01 4.00697261e-01 -1.81693673e+00
1.34123107e-02 -7.15732574e-01 -7.72000611e-01 1.63587976e+00
7.56932139e-01 -3.34031910e-01 1.10769713e+00 4.32571441e-01
-1.90737128e-01 -1.11137974e+00 -2.33027443e-01 -4.43378031e-01
6.19655073e-01 -7.25023821e-02 4.68543947e-01 1.08137858e+00
8.67150575e-02 9.49631631e-01 -7.56324381e-02 -1.95741698e-01
7.96519816e-01 3.91775608e-01 1.05781019e+00 -1.47815728e+00
-3.98580492e-01 -6.33460224e-01 5.17053246e-01 -1.15729141e+00
3.88888776e-01 -1.41246092e+00 -1.75789103e-01 -1.59412873e+00
5.43577075e-01 -4.39912289e-01 -1.07109018e-01 4.84504700e-01
-9.00661051e-01 -2.10135445e-01 6.25756662e-03 -5.94829656e-02
-8.08661580e-01 6.96848452e-01 1.29413068e+00 -6.63499981e-02
-4.87405926e-01 2.22886011e-01 -1.35704148e+00 9.52316940e-01
9.18859124e-01 -4.10807133e-01 -6.06183231e-01 -5.78405499e-01
4.59813118e-01 -3.16635996e-01 -1.60333570e-02 -5.01521289e-01
3.32403630e-01 -1.58426687e-01 5.34891248e-01 -8.03204119e-01
-4.24679043e-03 3.12890895e-02 -7.62800753e-01 3.47437114e-01
-1.09539711e+00 -1.44409299e-01 2.67331928e-01 5.21460474e-01
-2.08303309e-03 -5.34382999e-01 6.09781802e-01 -3.76404375e-01
-4.02707875e-01 2.61233542e-02 -2.95343369e-01 7.52097964e-01
6.21794343e-01 1.48755580e-01 -5.81383228e-01 -2.19310626e-01
-5.15893459e-01 8.50946367e-01 4.84639183e-02 3.63382488e-01
4.21509057e-01 -1.08982849e+00 -1.10066497e+00 3.51215154e-01
4.27638963e-02 7.85082281e-01 6.55126646e-02 3.90198231e-01
-3.51490229e-01 5.21796823e-01 1.39595479e-01 -3.46533835e-01
-1.60801435e+00 3.34640965e-02 4.44907183e-03 -1.08669400e+00
-4.26997662e-01 1.34602547e+00 3.69927548e-02 -1.23831093e+00
3.89038414e-01 -3.60505044e-01 -2.27302924e-01 4.28179532e-01
6.76874578e-01 7.44715989e-01 5.58085553e-02 1.23057403e-01
2.51333207e-01 -4.56794165e-02 -5.90225697e-01 -2.86438465e-02
1.62743783e+00 1.73653170e-01 1.47346288e-01 4.39301878e-01
8.91320288e-01 4.37961251e-01 -1.19294679e+00 -7.01969326e-01
-9.40996259e-02 1.08266532e-01 5.66353574e-02 -9.35844541e-01
-6.47252858e-01 9.17526782e-01 -2.60945857e-01 -5.56160733e-02
6.65884018e-01 1.62205026e-02 8.70933115e-01 8.50968897e-01
-2.63512313e-01 -1.34364808e+00 4.58298892e-01 7.36661613e-01
6.20172083e-01 -1.53081048e+00 2.49962702e-01 -8.78752023e-02
-9.11182404e-01 9.94292080e-01 1.10673630e+00 2.22395524e-01
3.89441639e-01 4.26515788e-01 1.25044892e-02 -7.35674351e-02
-1.45823264e+00 1.25507444e-01 5.33403814e-01 1.39194012e-01
7.98819840e-01 1.24151811e-01 1.87400058e-01 9.82770801e-01
-8.14358234e-01 -3.53789747e-01 5.93508124e-01 1.02213323e+00
-7.48710811e-01 -1.20265114e+00 -2.94488162e-01 1.01974559e+00
-5.10160387e-01 -4.37728554e-01 -5.89852333e-01 4.68756229e-01
-2.74576515e-01 7.82986820e-01 2.50008958e-03 -9.21451524e-02
1.88140035e-01 5.66265762e-01 4.43433672e-01 -1.35673225e+00
-1.11894023e+00 -4.53280956e-01 1.95233032e-01 -1.39091149e-01
-1.75239295e-01 -5.55055618e-01 -1.20070207e+00 -2.34681845e-01
-4.25164521e-01 4.01156962e-01 2.43398562e-01 8.20888519e-01
5.53446710e-01 7.28504181e-01 2.87793338e-01 -5.65763235e-01
-1.11679554e+00 -1.17964172e+00 -2.26163968e-01 2.46023104e-01
2.64393598e-01 -3.02120186e-02 -2.20263690e-01 4.21513826e-01] | [11.899808883666992, 8.99242115020752] |
522cecca-ad41-4ef4-9fd8-c29fba7ca728 | reference-limited-compositional-zero-shot | 2208.10046 | null | https://arxiv.org/abs/2208.10046v2 | https://arxiv.org/pdf/2208.10046v2.pdf | Reference-Limited Compositional Zero-Shot Learning | Compositional zero-shot learning (CZSL) refers to recognizing unseen compositions of known visual primitives, which is an essential ability for artificial intelligence systems to learn and understand the world. While considerable progress has been made on existing benchmarks, we suspect whether popular CZSL methods can address the challenges of few-shot and few referential compositions, which is common when learning in real-world unseen environments. To this end, we study the challenging reference-limited compositional zero-shot learning (RL-CZSL) problem in this paper, i.e., given limited seen compositions that contain only a few samples as reference, unseen compositions of observed primitives should be identified. We propose a novel Meta Compositional Graph Learner (MetaCGL) that can efficiently learn the compositionality from insufficient referential information and generalize to unseen compositions. Besides, we build a benchmark with two new large-scale datasets that consist of natural images with diverse compositional labels, providing more realistic environments for RL-CZSL. Extensive experiments in the benchmarks show that our method achieves state-of-the-art performance in recognizing unseen compositions when reference is limited for compositional learning. | ['Donglin Wang', 'Qiyao Wei', 'Siteng Huang'] | 2022-08-22 | null | null | null | null | ['compositional-zero-shot-learning'] | ['computer-vision'] | [ 6.46532118e-01 -2.39940181e-01 -1.97133899e-01 1.33520458e-02
-5.60311437e-01 -3.76206934e-01 7.37073779e-01 -1.15131773e-01
7.65088350e-02 3.06063086e-01 3.35733622e-01 1.09159306e-01
1.87404200e-01 -9.01446640e-01 -9.68067169e-01 -8.58690321e-01
1.91414401e-01 6.44855738e-01 7.29616880e-01 -2.66857475e-01
-4.37085889e-03 1.72386557e-01 -2.13861418e+00 6.68573022e-01
5.87453365e-01 8.62061024e-01 2.57167697e-01 6.01780295e-01
-4.33701873e-01 1.50307715e+00 -3.82845849e-01 -4.71948206e-01
3.42320621e-01 -6.09848022e-01 -6.56213701e-01 4.62637186e-01
1.10039532e+00 -2.57416010e-01 -5.50795138e-01 1.23548019e+00
4.67192292e-01 3.66167486e-01 8.42280924e-01 -1.63019061e+00
-1.25214660e+00 7.34764755e-01 -3.71778905e-01 1.17593698e-01
4.36662614e-01 5.64079046e-01 1.25382137e+00 -9.79058206e-01
1.05635810e+00 1.45967031e+00 4.07692343e-01 1.02546144e+00
-1.30032504e+00 -9.22431767e-01 5.57565868e-01 5.39410174e-01
-1.40457940e+00 -4.80679810e-01 1.10705984e+00 -6.68786526e-01
6.86567903e-01 1.90762177e-01 5.18939972e-01 1.53604603e+00
-1.51875228e-01 9.95859623e-01 8.36357713e-01 -2.33336017e-01
3.83488566e-01 -3.13434571e-01 1.10726587e-01 8.10731769e-01
3.36220711e-01 6.63780048e-02 -7.12897420e-01 -1.02805540e-01
3.89368713e-01 3.32069993e-01 -5.62249482e-01 -9.45523620e-01
-1.35556281e+00 5.82298517e-01 5.25220990e-01 5.12442030e-02
5.73821180e-02 2.07423866e-01 5.00489771e-01 5.54091752e-01
3.36127698e-01 3.21888894e-01 6.16519451e-02 4.03744400e-01
-4.36026990e-01 1.11704901e-01 7.40062773e-01 1.47615588e+00
9.82552469e-01 4.12448108e-01 -3.02673072e-01 8.45640898e-01
-7.42256194e-02 4.87646043e-01 6.71778977e-01 -4.84768629e-01
3.43224436e-01 9.95618224e-01 -4.15465772e-01 -5.65038204e-01
1.22417718e-01 -6.50443211e-02 -1.06423903e+00 8.32938477e-02
1.49148285e-01 2.29324728e-01 -1.35079801e+00 1.67142129e+00
3.67722541e-01 8.92834306e-01 4.82072756e-02 1.02956164e+00
1.35432541e+00 5.79169452e-01 8.94066542e-02 -1.13318197e-01
1.05269587e+00 -1.17599702e+00 -4.81617242e-01 -1.41990557e-01
2.11077318e-01 -5.42013407e-01 1.67420757e+00 1.78933367e-01
-6.15351617e-01 -6.38769865e-01 -1.19443750e+00 -1.22034267e-01
-6.52647555e-01 -6.46662772e-01 5.29058576e-01 3.31818998e-01
-7.36648142e-01 4.67784315e-01 -5.38794398e-01 -2.64458895e-01
4.37907159e-01 -2.64400207e-02 -3.66315186e-01 -6.35681570e-01
-8.21196437e-01 2.76719064e-01 5.03340960e-01 -4.17493582e-01
-1.69730508e+00 -1.03360212e+00 -1.19428110e+00 -6.48698807e-02
1.04477155e+00 -6.55162573e-01 1.20857060e+00 -1.12503469e+00
-1.08636403e+00 7.70321488e-01 -9.28737223e-02 -1.43119082e-01
4.24692720e-01 2.07617655e-01 -5.46380043e-01 -1.01978509e-02
1.39822766e-01 3.61425728e-01 1.12810016e+00 -1.71585000e+00
-4.71366823e-01 -2.13806421e-01 -8.47258326e-03 1.32216990e-01
-2.25658447e-01 -3.56236458e-01 -3.20668876e-01 -7.13886023e-01
1.76109254e-01 -6.60065353e-01 6.31800219e-02 7.21405149e-02
-3.26390296e-01 -3.85413945e-01 9.73552585e-01 -2.78806128e-02
7.88334012e-01 -2.07300091e+00 3.29091489e-01 -1.92790091e-01
3.87468785e-01 2.81945430e-02 -5.14122546e-01 6.76339090e-01
-3.13870870e-02 -5.81734926e-02 -4.39754814e-01 -1.52536869e-01
1.07133254e-01 3.42823684e-01 -6.50076330e-01 3.03891927e-01
1.44076720e-01 1.08411014e+00 -1.34673262e+00 -4.76775378e-01
2.79406190e-01 1.78877577e-01 -1.86066881e-01 5.53084493e-01
-7.37490296e-01 3.80645543e-01 9.32560023e-03 1.08003354e+00
3.74662787e-01 -4.23082799e-01 -7.58257136e-02 -2.56356508e-01
3.45945358e-01 -3.93772125e-01 -1.15843785e+00 1.76031995e+00
-1.01217374e-01 3.04586977e-01 -4.82115507e-01 -7.75663853e-01
6.52378619e-01 1.90283000e-01 1.54355079e-01 -6.85088038e-01
5.08275535e-03 6.03462607e-02 -2.84662880e-02 -5.99512815e-01
9.93982852e-02 -4.91931915e-01 -6.44765655e-03 5.55927277e-01
5.55154145e-01 -2.28568986e-01 3.50915939e-01 2.88804352e-01
1.27447987e+00 1.24732248e-01 5.75684369e-01 -2.48422787e-01
6.71379924e-01 -1.15266651e-01 6.89005494e-01 7.88110316e-01
-1.90538302e-01 8.63735855e-01 1.93398625e-01 -8.03466022e-01
-1.18677258e+00 -1.45540524e+00 5.16554475e-01 1.37994671e+00
6.68828905e-01 -4.41208720e-01 -3.27482671e-01 -7.50011981e-01
-7.92609975e-02 6.30306959e-01 -7.53014147e-01 -3.07184041e-01
-5.89790463e-01 -2.68168867e-01 1.63902536e-01 4.58864212e-01
3.57799202e-01 -1.36147225e+00 -3.81501704e-01 -8.14589411e-02
6.89701214e-02 -1.06355429e+00 -7.16556191e-01 -1.46596253e-01
-4.28140759e-01 -1.50103533e+00 -6.60107851e-01 -1.17181981e+00
7.68840015e-01 7.28349805e-01 1.32441425e+00 1.18154719e-01
-6.10995054e-01 5.23774445e-01 -5.57239115e-01 -2.93991417e-01
-3.79768103e-01 -2.78486311e-01 -1.56838298e-01 3.46994191e-01
4.03288543e-01 -6.61485732e-01 -4.36771363e-01 1.63588524e-01
-1.13463116e+00 3.15193772e-01 3.86902869e-01 9.14889395e-01
9.29102957e-01 -1.79585870e-02 4.63290989e-01 -1.28366590e+00
5.54287136e-01 -5.49253345e-01 -3.10582131e-01 7.29952037e-01
-3.29508752e-01 2.11299866e-01 1.06875706e+00 -9.89566147e-01
-9.39260840e-01 1.07010333e-02 2.42267311e-01 -1.15962350e+00
-3.03284615e-01 -3.77813727e-02 -4.63755429e-01 -1.82177544e-01
6.50632203e-01 5.78553021e-01 -3.55726689e-01 -2.33155087e-01
6.86219692e-01 3.91418517e-01 8.47584426e-01 -9.40621495e-01
1.06676853e+00 7.73563147e-01 -1.69777069e-02 -8.76108170e-01
-1.04813910e+00 -8.67518842e-01 -5.18257618e-01 -2.20092118e-01
7.43847609e-01 -9.35174108e-01 -5.68217814e-01 4.70944643e-01
-6.39499187e-01 -6.04404569e-01 -5.95873475e-01 -6.42956048e-02
-7.97254086e-01 4.94401246e-01 -5.26541889e-01 -6.54369771e-01
-5.50795138e-01 -1.09742606e+00 1.22135556e+00 -2.95717847e-02
-1.63756292e-02 -7.73560107e-01 2.25075901e-01 1.35262758e-01
1.08775904e-03 5.92834294e-01 1.27851176e+00 -5.22357166e-01
-9.40947711e-01 1.69740140e-01 -1.09029360e-01 1.97721124e-01
4.48636800e-01 -1.19214043e-01 -9.98277605e-01 -4.18614089e-01
-6.40521646e-02 -7.13781953e-01 1.02229798e+00 -3.38550240e-01
1.32160449e+00 -3.00191104e-01 1.57119818e-02 7.01172471e-01
1.60746658e+00 2.02721655e-01 4.13013905e-01 -1.11457303e-01
1.24107921e+00 4.75615114e-01 3.70672017e-01 2.09916875e-01
3.44075918e-01 1.56855255e-01 5.89576602e-01 3.10691535e-01
-8.09144795e-01 -8.48150134e-01 8.23821574e-02 1.10267484e+00
1.11931734e-01 -2.77651906e-01 -9.57175195e-01 7.89285004e-01
-2.04928660e+00 -1.13241422e+00 2.76746929e-01 2.14321256e+00
7.53426611e-01 -1.04240842e-01 -2.65839379e-02 1.17694937e-01
8.23159456e-01 5.56148767e-01 -9.09798563e-01 2.91953385e-01
-3.22246939e-01 1.76283106e-01 1.42026797e-01 1.99341908e-01
-1.04531372e+00 1.16722155e+00 5.43286085e+00 9.28390741e-01
-9.56015229e-01 1.86639160e-01 2.63956666e-01 -1.34741627e-02
-5.07900000e-01 -4.34272736e-02 -5.85673213e-01 5.80419898e-01
3.89737785e-01 -3.82439584e-01 8.89066935e-01 1.05948019e+00
-5.85072994e-01 2.63404429e-01 -1.49163675e+00 1.26808524e+00
7.38134503e-01 -1.42402935e+00 8.33238125e-01 -3.70450050e-01
1.28099644e+00 -7.13219419e-02 4.00710255e-02 5.64210951e-01
5.55792391e-01 -8.05754483e-01 7.46410966e-01 4.61961716e-01
8.87037218e-01 -5.83466947e-01 4.38314646e-01 5.67640305e-01
-1.75417233e+00 -3.13345402e-01 -7.43351877e-01 -2.68736272e-03
2.34956015e-03 5.72533384e-02 -6.31200194e-01 6.47686899e-01
4.99133497e-01 9.04703498e-01 -6.46623075e-01 1.03091288e+00
-3.95564646e-01 3.63526165e-01 2.85287738e-01 -6.12038411e-02
2.17655182e-01 -1.32815078e-01 4.41305429e-01 7.31657445e-01
1.94582805e-01 1.34178355e-01 4.56181109e-01 9.21934485e-01
-5.21412313e-01 6.40996173e-02 -8.89756739e-01 -2.26727709e-01
3.40568095e-01 1.00193584e+00 -7.01562881e-01 -6.43371642e-01
-5.73968470e-01 9.22468722e-01 5.89593887e-01 4.66203272e-01
-5.39438069e-01 -4.01044206e-04 7.30381429e-01 6.54815882e-02
2.75590777e-01 1.23839304e-01 1.44843414e-01 -1.64924133e+00
-1.27586365e-01 -1.23739707e+00 7.87242234e-01 -8.08951914e-01
-2.00896025e+00 4.12069917e-01 -1.45731494e-01 -1.43495917e+00
4.14520800e-02 -5.62078834e-01 -8.68529260e-01 1.91653579e-01
-1.37729537e+00 -1.60350251e+00 -8.82683575e-01 8.74743521e-01
1.26139605e+00 -2.95324534e-01 6.32668078e-01 1.59200445e-01
-4.12047148e-01 3.62680465e-01 -2.55526841e-01 1.43468201e-01
7.81059504e-01 -1.32476401e+00 6.25957012e-01 9.77408111e-01
6.20818853e-01 4.75290090e-01 5.68641901e-01 -9.02441382e-01
-1.75339031e+00 -1.45817709e+00 2.29146197e-01 -3.41772676e-01
9.80729640e-01 -8.01309407e-01 -1.01702940e+00 7.99892902e-01
1.48230195e-01 4.84771162e-01 7.78976917e-01 -2.63689250e-01
-9.48277771e-01 -9.66091007e-02 -7.43806005e-01 1.02577603e+00
1.73549807e+00 -8.56610179e-01 -8.95765722e-01 3.14724177e-01
1.22555697e+00 -2.01717943e-01 -5.42144418e-01 6.99803889e-01
3.64862233e-01 -7.61159658e-01 9.94313598e-01 -7.46082187e-01
2.84691602e-01 -5.46703577e-01 -3.76747012e-01 -1.33125460e+00
-2.89835632e-01 -4.95355070e-01 -5.90421081e-01 1.01632404e+00
-8.27419162e-02 -2.78881460e-01 8.57238114e-01 3.28969181e-01
-2.37981826e-01 -4.80433345e-01 -5.18302083e-01 -1.19526601e+00
-1.26708850e-01 -3.15346748e-01 9.58670795e-01 1.01980495e+00
-3.62188369e-01 4.94561583e-01 -5.43903947e-01 5.20139188e-02
1.05250823e+00 6.05078101e-01 1.02611351e+00 -1.20345485e+00
-1.92073882e-01 -2.98894376e-01 -7.37251222e-01 -7.02122271e-01
5.86847782e-01 -1.00909579e+00 1.25896618e-01 -1.53634465e+00
6.07561707e-01 5.51254451e-02 -3.88892204e-01 2.20533252e-01
-4.54977453e-01 2.80621558e-01 2.64077932e-01 2.46144623e-01
-9.55521107e-01 7.80799031e-01 1.51207757e+00 -9.33423996e-01
2.01687515e-01 -4.35284913e-01 -4.65050668e-01 6.33505404e-01
4.34111893e-01 -1.89166218e-01 -1.06038582e+00 -3.88817877e-01
-7.70724416e-02 -4.90011305e-01 2.89103478e-01 -1.09345734e+00
4.61321652e-01 -4.20504659e-01 1.48592219e-01 -6.97732866e-01
3.48913074e-01 -8.61867726e-01 4.63396072e-01 3.40235382e-01
-3.30543399e-01 -2.67649472e-01 -3.03581417e-01 9.99264300e-01
-1.28501445e-01 -1.64613605e-01 6.43009365e-01 -5.56006253e-01
-1.50717616e+00 9.53465164e-01 2.79579073e-01 4.96920079e-01
1.35034263e+00 -2.64642507e-01 -8.19222152e-01 -2.59427577e-01
-5.04237950e-01 2.50022352e-01 8.40839148e-01 6.59202039e-01
9.99704301e-01 -1.53746402e+00 -5.03825903e-01 3.55350703e-01
9.10893559e-01 2.15827152e-01 4.81424958e-01 2.17216700e-01
-6.30006790e-01 -7.20962882e-02 -2.45575473e-01 -4.29945260e-01
-1.21117020e+00 1.52614439e+00 1.31500050e-01 2.87801862e-01
-1.09898996e+00 1.07924175e+00 8.04473519e-01 -5.25186479e-01
4.76388723e-01 -1.73901096e-01 2.26527274e-01 -4.09864299e-02
7.76452184e-01 3.82241815e-01 -3.33684385e-01 -7.48239994e-01
1.47763249e-02 6.52938962e-01 1.96797401e-02 3.34079236e-01
1.17593253e+00 9.79710296e-02 -1.44672439e-01 1.07074618e+00
1.14263701e+00 -2.27843776e-01 -1.38373315e+00 -6.34187520e-01
2.02293873e-01 -6.36832535e-01 -5.45034885e-01 -2.95694441e-01
-9.70097423e-01 1.07802486e+00 2.91651219e-01 -2.69841794e-02
8.99317801e-01 1.86078593e-01 8.43655705e-01 5.97299695e-01
6.85407698e-01 -6.85679138e-01 6.00348353e-01 2.59427398e-01
9.38770950e-01 -1.39900625e+00 -1.47788614e-01 -4.89484727e-01
-7.19600022e-01 9.02812481e-01 1.03010798e+00 -2.98895776e-01
4.22349274e-01 1.15868665e-01 -4.62650023e-02 -4.26175147e-01
-1.06612277e+00 -7.30681837e-01 3.60513687e-01 7.28601933e-01
-4.47656661e-02 1.70969576e-01 2.96695381e-01 2.42167145e-01
-3.95769998e-03 -4.44343299e-01 3.79558295e-01 7.66467273e-01
-4.93122220e-01 -9.46749330e-01 -3.24164480e-02 4.43209618e-01
9.98298377e-02 -7.91881531e-02 -8.50158095e-01 6.74686432e-01
2.82016367e-01 7.75871813e-01 -9.77560133e-03 -5.46573102e-01
1.96438670e-01 2.26167679e-01 5.56891739e-01 -1.12181354e+00
-2.05867827e-01 -1.51228070e-01 -2.28746414e-01 -6.26947224e-01
-3.68560910e-01 -2.19070762e-01 -9.52056706e-01 -4.23903205e-02
-6.53767586e-02 -2.13176951e-01 -1.47571504e-01 6.68201029e-01
-2.22486958e-01 6.83124661e-01 4.42378104e-01 -7.18115985e-01
-4.17572945e-01 -7.44605601e-01 -6.15689754e-01 1.05457997e+00
4.11815315e-01 -7.74615884e-01 -5.55757463e-01 2.52436340e-01] | [10.238746643066406, 2.258432388305664] |
827bafa3-e62c-4c24-a53e-b068c6d53234 | video-violence-recognition-and-localization | 2202.02212 | null | https://arxiv.org/abs/2202.02212v4 | https://arxiv.org/pdf/2202.02212v4.pdf | Video Violence Recognition and Localization Using a Semi-Supervised Hard Attention Model | The significant growth of surveillance camera networks necessitates scalable AI solutions to efficiently analyze the large amount of video data produced by these networks. As a typical analysis performed on surveillance footage, video violence detection has recently received considerable attention. The majority of research has focused on improving existing methods using supervised methods, with little, if any, attention to the semi-supervised learning approaches. In this study, a reinforcement learning model is introduced that can outperform existing models through a semi-supervised approach. The main novelty of the proposed method lies in the introduction of a semi-supervised hard attention mechanism. Using hard attention, the essential regions of videos are identified and separated from the non-informative parts of the data. A model's accuracy is improved by removing redundant data and focusing on useful visual information in a higher resolution. Implementing hard attention mechanisms using semi-supervised reinforcement learning algorithms eliminates the need for attention annotations in video violence datasets, thus making them readily applicable. The proposed model utilizes a pre-trained I3D backbone to accelerate and stabilize the training process. The proposed model achieved state-of-the-art accuracy of 90.4% and 98.7% on RWF and Hockey datasets, respectively. | ['Ehsan Nazerfard', 'Hamid Mohammadi'] | 2022-02-04 | null | null | null | null | ['hard-attention'] | ['methodology'] | [ 2.08768100e-01 3.48616153e-01 -2.99642235e-01 -2.20574021e-01
-3.55583847e-01 -1.85784519e-01 4.42019641e-01 -2.75772680e-02
-5.45021713e-01 5.65466106e-01 2.15654522e-01 1.16793767e-01
-7.64580145e-02 -5.52906454e-01 -6.48540676e-01 -6.87387466e-01
-5.72974123e-02 9.09412950e-02 4.88239944e-01 -8.49998593e-02
2.55770564e-01 3.60550255e-01 -1.55794156e+00 1.83935940e-01
8.18914115e-01 1.01895952e+00 3.14892113e-01 6.65877998e-01
1.49606496e-01 1.35502315e+00 -6.61484182e-01 -4.36573476e-01
3.36551696e-01 -4.70094353e-01 -7.97472656e-01 3.34822774e-01
4.37842041e-01 -8.02506745e-01 -5.27270019e-01 9.71714854e-01
2.92422324e-01 7.83854946e-02 5.05515695e-01 -1.44258225e+00
-4.92998481e-01 3.03339958e-01 -8.29093575e-01 5.55784523e-01
4.53396648e-01 8.11672285e-02 7.22271740e-01 -6.49142861e-01
6.44553304e-01 1.07977474e+00 5.38089931e-01 6.47077084e-01
-7.80452192e-01 -7.11963832e-01 1.23027273e-01 5.42407274e-01
-1.23458421e+00 -2.52170086e-01 1.18054545e+00 -4.89096254e-01
1.00315559e+00 -3.03556863e-02 7.67763138e-01 1.08819783e+00
2.04555124e-01 8.98538649e-01 6.98561192e-01 -4.00240868e-01
1.68147817e-01 1.63512170e-01 9.19564813e-02 8.94310117e-01
2.36818925e-01 9.76256374e-03 -4.66191232e-01 5.21595292e-02
5.99624932e-01 3.56059641e-01 -1.67478174e-01 -5.01219749e-01
-7.39834130e-01 8.21303487e-01 3.91387194e-01 8.11755806e-02
-5.75711370e-01 -2.40533173e-01 5.68735898e-01 -6.58895522e-02
5.34685194e-01 2.24498987e-01 -2.55009811e-02 -3.42464417e-01
-1.18231618e+00 -1.31913990e-01 2.67118394e-01 9.11090970e-01
5.73710501e-01 2.06513211e-01 -4.62311655e-02 5.97166061e-01
3.05299222e-01 5.14421999e-01 2.23845214e-01 -1.06890452e+00
4.26803470e-01 9.88994062e-01 -9.04811323e-02 -1.35993111e+00
-3.50514382e-01 -1.92676678e-01 -9.75942850e-01 2.61155874e-01
2.29172021e-01 -2.48443604e-01 -8.51830125e-01 1.57646489e+00
3.72586280e-01 2.84660876e-01 -1.23702232e-02 9.40202117e-01
6.62173271e-01 7.07065642e-01 1.71572834e-01 -4.81437266e-01
8.61153364e-01 -1.22707808e+00 -8.35126519e-01 1.60492528e-02
3.40571254e-01 -4.32431012e-01 6.15691900e-01 5.01587927e-01
-1.09161246e+00 -7.11672127e-01 -9.07680452e-01 8.51219520e-02
-1.18036628e-01 8.71248171e-02 2.55409002e-01 5.66819489e-01
-9.77918208e-01 3.05483520e-01 -8.31566095e-01 -5.74811757e-01
8.20178747e-01 2.63399541e-01 -4.19153094e-01 -1.87832177e-01
-9.38147128e-01 8.27151477e-01 2.99395114e-01 1.66951478e-01
-1.26687849e+00 -2.99752325e-01 -8.30332577e-01 4.90041859e-02
7.07492650e-01 -1.24180138e-01 9.07925248e-01 -1.60048842e+00
-1.45557785e+00 7.57082522e-01 7.45908543e-02 -5.47008991e-01
4.15435940e-01 -4.51410562e-01 -2.82409072e-01 9.34674561e-01
-5.69226667e-02 6.90343499e-01 1.30924904e+00 -1.01979613e+00
-6.46502197e-01 -2.96525449e-01 5.43448366e-02 1.33392856e-01
-6.23935878e-01 4.08994049e-01 -4.92679745e-01 -7.30778694e-01
-4.22690094e-01 -9.01000619e-01 2.29118913e-02 1.50422394e-01
-1.20004721e-01 -1.64838046e-01 1.35832584e+00 -9.11941826e-01
1.25511539e+00 -2.13030863e+00 3.17488134e-01 8.98158550e-02
4.32459563e-01 8.98265719e-01 -3.50313932e-02 2.22917408e-01
-2.94006802e-02 -1.30367801e-01 -2.71070123e-01 -1.22252993e-01
-4.76767570e-01 1.62281305e-01 -8.39147493e-02 5.02665758e-01
6.37452006e-01 5.35235703e-01 -9.99516189e-01 -8.33393335e-01
5.02767265e-01 4.57053214e-01 -7.78610647e-01 6.20708346e-01
7.69269839e-02 3.52759153e-01 -2.97679126e-01 7.69656479e-01
5.50913572e-01 -7.69674778e-02 8.94872919e-02 -4.48296070e-02
-1.06760956e-01 -3.58582646e-01 -7.59043992e-01 1.49328291e+00
1.81952268e-01 6.66951239e-01 1.89662009e-01 -1.48786581e+00
9.10202980e-01 2.39763424e-01 7.61491835e-01 -6.47190273e-01
3.51690114e-01 -3.50096673e-01 -2.93571465e-02 -9.79735494e-01
3.00791800e-01 4.24719423e-01 1.82394251e-01 1.92392766e-01
2.24766135e-01 3.58251691e-01 2.86568642e-01 3.68803412e-01
1.17973948e+00 2.82748073e-01 3.49628240e-01 -1.18315138e-01
6.50965273e-01 2.61176139e-01 7.73669422e-01 5.83600581e-01
-6.27116621e-01 4.35992718e-01 4.97370541e-01 -6.77608490e-01
-9.11052167e-01 -6.78439379e-01 2.51438767e-01 1.00461102e+00
2.41508797e-01 -4.42727298e-01 -1.12752485e+00 -9.86442685e-01
-3.67735296e-01 3.25817406e-01 -7.87035346e-01 -5.10948062e-01
-4.64740336e-01 -4.29735750e-01 5.23427427e-01 5.40883720e-01
8.11454058e-01 -1.33478212e+00 -1.13784599e+00 -5.09216962e-03
-2.15320960e-01 -1.23091865e+00 -1.56294093e-01 -1.74243987e-01
-7.88629353e-01 -1.39093196e+00 -7.16602981e-01 -6.39618576e-01
9.39422190e-01 6.61103070e-01 7.61822522e-01 2.33236134e-01
-4.03945178e-01 4.31158453e-01 -7.81364501e-01 -3.50512207e-01
-2.56008565e-01 -3.06557631e-03 1.03008896e-01 2.62396842e-01
4.48976547e-01 -1.69520497e-01 -3.97579491e-01 3.80173057e-01
-8.75291944e-01 1.91319153e-01 7.54770160e-01 8.34405959e-01
3.48246485e-01 -8.34212266e-03 4.59042400e-01 -6.18055046e-01
2.14069694e-01 -6.00635946e-01 -5.08939385e-01 1.46991298e-01
-2.41711825e-01 -1.85479850e-01 8.21121633e-01 -4.67988133e-01
-1.17343986e+00 1.60160691e-01 1.93533272e-01 -7.12758422e-01
-2.66370803e-01 1.94062233e-01 -8.99086893e-02 3.60593610e-02
3.29827487e-01 -2.13721246e-02 1.75383970e-01 -6.60612881e-02
-1.93181038e-01 7.60041893e-01 3.97048384e-01 -9.02054235e-02
6.48875296e-01 4.86466199e-01 -6.31721839e-02 -1.18580341e+00
-7.78403521e-01 -4.10734594e-01 -6.87787294e-01 -6.55047894e-01
1.29503727e+00 -7.53182352e-01 -5.09786606e-01 6.29132867e-01
-8.84263158e-01 -2.41268769e-01 -1.73198164e-01 5.48843443e-01
-3.42278630e-01 6.50240183e-01 -4.29251134e-01 -9.68949795e-01
-3.59420598e-01 -1.12549603e+00 8.18931818e-01 3.29086542e-01
-1.38974100e-01 -5.15585065e-01 1.26729041e-01 6.51986778e-01
2.61099577e-01 2.29319185e-01 3.46916735e-01 -5.29914498e-01
-4.18579370e-01 -2.47413576e-01 -3.73816788e-01 6.53283536e-01
9.13140476e-02 3.65490794e-01 -8.43351543e-01 -2.01720148e-01
-2.00563669e-01 -6.30144358e-01 8.12730730e-01 4.25273627e-01
1.20076132e+00 -1.40260354e-01 -1.21011846e-01 5.00744104e-01
9.09591377e-01 2.95555770e-01 6.39379621e-01 2.56567478e-01
7.67728984e-01 5.92557967e-01 8.33375037e-01 5.92579067e-01
2.61714131e-01 5.57468891e-01 8.50577295e-01 -2.91289210e-01
-6.04621656e-02 -1.79078445e-01 5.65759957e-01 5.02009869e-01
-4.74742711e-01 -7.63244405e-02 -9.66484010e-01 4.70219702e-01
-2.06935430e+00 -1.48553383e+00 1.80371001e-01 1.96088541e+00
3.70074600e-01 2.26781383e-01 3.74935627e-01 2.96029240e-01
8.63912344e-01 1.73039824e-01 -4.16395754e-01 -3.16316485e-01
-2.70235874e-02 -1.68159679e-01 2.76222259e-01 2.46340021e-01
-1.40087676e+00 8.29111457e-01 5.76091719e+00 5.04422188e-01
-9.98935401e-01 -2.02205092e-01 5.23136616e-01 -2.79975533e-01
4.89759207e-01 -2.98297346e-01 -5.61328292e-01 5.44236243e-01
8.00720334e-01 2.85654485e-01 2.84168303e-01 1.00685990e+00
5.06625056e-01 -3.63465697e-01 -6.93966091e-01 1.07131493e+00
6.14126325e-01 -1.08572805e+00 8.63156915e-02 -8.53344705e-03
5.59172690e-01 -3.17008346e-01 -1.43314540e-01 2.21123651e-01
-6.91579431e-02 -7.54909456e-01 6.08292401e-01 5.76703370e-01
4.76698995e-01 -1.00072169e+00 9.79712129e-01 3.92239094e-01
-9.40343916e-01 -4.40654606e-01 -2.87818432e-01 -2.43653283e-01
-1.07263252e-01 1.69947088e-01 -6.94485486e-01 3.81644249e-01
1.12459910e+00 9.59200144e-01 -7.63909876e-01 8.97222042e-01
-2.33418941e-01 8.82303238e-01 -1.24286450e-01 3.08496524e-02
3.73719186e-01 -9.23291445e-02 5.26817024e-01 1.25318456e+00
-1.77711789e-02 2.18039855e-01 4.03627247e-01 1.86428294e-01
-2.77115963e-03 -6.10639751e-02 -6.84952497e-01 -5.10956421e-02
4.69060279e-02 1.15583599e+00 -7.16977596e-01 -4.05444294e-01
-6.75048590e-01 8.39840531e-01 4.37396586e-01 8.47982690e-02
-1.20363235e+00 -3.67296785e-01 2.24843368e-01 2.02615529e-01
4.30406660e-01 -8.08004662e-02 2.58639753e-01 -1.05748153e+00
-1.33066867e-02 -9.52451348e-01 5.48520803e-01 -7.34469593e-01
-8.61502111e-01 6.80874169e-01 2.99586356e-01 -1.23807752e+00
-4.25361872e-01 -3.59054804e-01 -5.09467840e-01 2.14858755e-01
-1.41789258e+00 -1.14700246e+00 -7.86747158e-01 7.65444517e-01
8.62263143e-01 -4.47979212e-01 6.31051660e-01 2.96670586e-01
-1.08137465e+00 4.78743196e-01 -1.34456903e-01 5.12864649e-01
6.52953625e-01 -7.94618845e-01 -2.00519204e-01 1.23585570e+00
-1.60313040e-01 1.30339429e-01 5.27519941e-01 -7.45954156e-01
-1.32922518e+00 -1.04568028e+00 3.71682942e-01 -1.18151575e-01
3.04166943e-01 -1.53404176e-01 -8.93860579e-01 5.31092286e-01
4.86295283e-01 2.55972981e-01 6.61711395e-01 -4.01736677e-01
-2.03712538e-01 -1.43388569e-01 -1.22978723e+00 3.92905414e-01
1.00794387e+00 -1.78413123e-01 -8.58121395e-01 -2.04557832e-02
2.72480965e-01 -3.71983685e-02 -2.70590544e-01 2.90903598e-01
4.07799482e-01 -1.04928398e+00 6.86805964e-01 -7.61524081e-01
7.66290963e-01 -2.29657069e-01 9.68964547e-02 -9.79960263e-01
-5.08873105e-01 -5.98424613e-01 -5.81889510e-01 1.07218802e+00
-9.05808359e-02 -5.41024804e-02 8.67406189e-01 5.38962662e-01
-1.14289537e-01 -4.99379784e-01 -7.95113742e-01 -3.75887424e-01
-6.27223253e-01 -1.57573327e-01 -1.02154709e-01 9.61833179e-01
2.38039345e-01 1.62020758e-01 -8.73852372e-01 2.08105713e-01
8.19185078e-01 -1.93076462e-01 8.60574722e-01 -1.15107477e+00
-4.60080057e-02 -1.82377577e-01 -7.65826225e-01 -6.48451447e-01
1.86343327e-01 -4.00922239e-01 -1.72248572e-01 -1.31381297e+00
4.39736158e-01 1.72033370e-01 -2.80940980e-01 5.99732101e-01
-1.97744846e-01 4.55837071e-01 2.39028245e-01 1.05743557e-01
-1.13191640e+00 6.49899840e-01 8.94269168e-01 -8.26397762e-02
-1.55295193e-01 -1.59670308e-01 -4.55464631e-01 1.06036186e+00
9.75743294e-01 -4.84914422e-01 -5.28136253e-01 -5.45735478e-01
-3.00379485e-01 -4.54708524e-02 3.60525548e-01 -1.17826915e+00
3.32155198e-01 -1.83361411e-01 5.73621988e-01 -5.30865371e-01
2.63177216e-01 -1.12604487e+00 -2.59073049e-01 5.26710272e-01
-2.68756002e-01 2.04615757e-01 1.92677945e-01 6.43451214e-01
-2.60322273e-01 -2.00376049e-01 8.87711942e-01 -9.98495705e-03
-9.30510819e-01 8.11478719e-02 -6.67870104e-01 1.31809786e-01
1.65879548e+00 -3.53839099e-01 -1.30544558e-01 -3.20830256e-01
-4.88351226e-01 2.65251309e-01 2.82032937e-01 5.99057376e-01
9.20160353e-01 -9.90632534e-01 -5.75844288e-01 3.17741960e-01
-1.24821197e-02 -2.33881585e-02 4.72272843e-01 6.91655755e-01
-6.15380704e-01 2.80404150e-01 -7.84953237e-01 -4.98760849e-01
-1.55621743e+00 8.38400543e-01 -2.11634524e-02 -2.02929035e-01
-6.07851863e-01 4.07033026e-01 -4.94972654e-02 2.53410220e-01
5.65838456e-01 -3.27203050e-02 -5.58557212e-01 6.54749619e-03
9.02641237e-01 6.26599014e-01 -3.45351934e-01 -9.60334837e-01
-5.17517269e-01 5.23177743e-01 -2.34447598e-01 3.04603219e-01
1.58133817e+00 1.34115602e-04 3.13943356e-01 -5.36775924e-02
9.53995943e-01 -2.16369927e-01 -1.64022398e+00 -5.75663038e-02
-2.92159289e-01 -5.69300294e-01 2.89304078e-01 -4.96250093e-01
-1.35005701e+00 9.34733152e-01 5.36370039e-01 1.97628692e-01
1.47341681e+00 -2.35094115e-01 5.38970590e-01 3.69280845e-01
1.53833270e-01 -1.29680347e+00 4.09732372e-01 2.82469004e-01
7.91277289e-01 -1.52902281e+00 1.74350813e-01 -2.88090855e-01
-9.61948931e-01 1.03589141e+00 9.44969654e-01 -3.50730121e-01
4.06984717e-01 2.22735584e-01 -1.69831768e-01 -3.56780738e-02
-5.44313848e-01 -4.09653932e-02 1.65899798e-01 8.49956989e-01
3.64532918e-02 -4.93223011e-01 -3.96446250e-02 5.54572761e-01
4.91683215e-01 1.89242572e-01 5.00028670e-01 1.23797119e+00
-4.91907746e-01 -4.58396882e-01 -4.40741152e-01 3.48530948e-01
-5.63296854e-01 2.40375817e-01 -5.11771441e-01 7.68778682e-01
1.44035742e-01 1.06954670e+00 1.06292911e-01 -4.54513490e-01
2.41249681e-01 -1.35695443e-01 3.31343859e-01 -3.87168050e-01
-5.57188809e-01 -8.99314359e-02 -1.12455666e-01 -6.63948655e-01
-8.99572730e-01 -6.30774021e-01 -9.93772805e-01 -1.57067329e-01
-1.90697908e-01 4.20838222e-03 3.13315660e-01 8.20580482e-01
5.06536603e-01 4.81931061e-01 6.45571709e-01 -1.09113038e+00
-2.91702449e-01 -7.81432331e-01 -1.14111625e-01 6.03431284e-01
3.25222731e-01 -8.99421573e-01 -1.58525601e-01 2.42582351e-01] | [8.00833797454834, 0.8244011402130127] |
c902686e-05a7-4524-8f5f-94fbf57604cd | 190406197 | 1904.06197 | null | https://arxiv.org/abs/1904.06197v2 | https://arxiv.org/pdf/1904.06197v2.pdf | Simulation of hyperelastic materials in real-time using Deep Learning | The finite element method (FEM) is among the most commonly used numerical methods for solving engineering problems. Due to its computational cost, various ideas have been introduced to reduce computation times, such as domain decomposition, parallel computing, adaptive meshing, and model order reduction. In this paper we present U-Mesh: a data-driven method based on a U-Net architecture that approximates the non-linear relation between a contact force and the displacement field computed by a FEM algorithm. We show that deep learning, one of the latest machine learning methods based on artificial neural networks, can enhance computational mechanics through its ability to encode highly non-linear models in a compact form. Our method is applied to two benchmark examples: a cantilever beam and an L-shape subject to moving punctual loads. A comparison between our method and proper orthogonal decomposition (POD) is done through the paper. The results show that U-Mesh can perform very fast simulations on various geometries, mesh resolutions and number of input forces with very small errors. | ['Stéphane Cotin', 'Pablo Márquez-Neila', 'Andrea Mendizabal'] | 2019-04-10 | null | null | null | null | ['cantilever-beam'] | ['miscellaneous'] | [ 8.00615847e-02 -2.80112207e-01 3.17919701e-01 1.13649599e-01
-2.38293499e-01 -3.04141995e-02 2.26073861e-01 1.15552463e-01
-1.10522367e-01 9.34364498e-01 -1.48988232e-01 -1.14202693e-01
-3.64791542e-01 -1.27421308e+00 -9.22381461e-01 -6.01637602e-01
-2.47188285e-01 8.22433114e-01 1.49982035e-01 -6.14752352e-01
2.95560598e-01 6.97997451e-01 -1.59322524e+00 2.98947543e-01
6.20403230e-01 1.35889888e+00 -1.46519318e-01 5.85933566e-01
2.48467699e-01 4.57369834e-01 2.68940013e-02 1.69466794e-01
8.47131312e-02 -1.84752747e-01 -9.81848300e-01 -2.03231350e-01
3.13619673e-01 -4.10656750e-01 -1.64960504e-01 6.55497193e-01
7.61509895e-01 4.45962608e-01 9.11931932e-01 -7.29445279e-01
-5.19439042e-01 3.62117648e-01 -4.08336222e-01 -2.51077443e-01
2.72933304e-01 -4.92785387e-02 5.57273805e-01 -1.22698998e+00
7.82656133e-01 1.24320054e+00 1.59144735e+00 4.16280597e-01
-1.66025567e+00 -3.26199949e-01 -7.84471750e-01 1.02626435e-01
-1.14771831e+00 1.39908195e-01 1.23350668e+00 -7.40369678e-01
1.04294205e+00 3.19747180e-01 6.08985066e-01 7.20310330e-01
1.02153277e+00 1.50239691e-01 1.02546167e+00 -5.38952887e-01
5.75586855e-01 -4.63327706e-01 3.61043960e-02 5.06059766e-01
4.00620587e-02 4.78666365e-01 -3.87798324e-02 -5.43698967e-01
1.21729803e+00 -1.92486018e-01 -1.04698405e-01 -3.59367490e-01
-8.67465436e-01 8.02601755e-01 3.44533086e-01 3.99743229e-01
-4.92547184e-01 4.48913366e-01 6.77884579e-01 1.23827130e-01
7.57162750e-01 6.99494720e-01 -7.28706598e-01 -1.37524098e-01
-8.09742093e-01 8.45233321e-01 7.90341258e-01 1.21543311e-01
5.82031250e-01 3.80105406e-01 2.73748308e-01 8.35854888e-01
1.80051610e-01 2.32803419e-01 2.90472031e-01 -1.47828245e+00
-2.45938182e-01 4.93444532e-01 1.34670407e-01 -1.58204734e+00
-7.21310556e-01 -1.92386255e-01 -1.46031404e+00 8.19256544e-01
1.97717309e-01 -5.19556880e-01 -5.20629883e-01 1.26526678e+00
5.42589188e-01 1.51245728e-01 -3.62128139e-01 8.90366256e-01
5.76624572e-01 8.00147295e-01 -1.30786240e-01 -5.74992336e-02
9.58094835e-01 -4.49216068e-01 -7.37239838e-01 3.89290273e-01
7.21938491e-01 -6.74068928e-01 5.41662872e-01 5.03285229e-01
-1.43138015e+00 -8.94408286e-01 -9.54815090e-01 -1.17883077e-02
-4.58362520e-01 -3.52385119e-02 4.84875113e-01 1.21654205e-01
-9.03374135e-01 1.80031574e+00 -9.28823173e-01 1.82655931e-01
4.02756304e-01 4.67391223e-01 -2.72335023e-01 2.41779119e-01
-1.47557211e+00 9.51993465e-01 1.29695445e-01 4.54869345e-02
-4.87293750e-01 -1.13656878e+00 -7.00612307e-01 1.51815057e-01
-6.26325309e-02 -8.26001406e-01 8.40086102e-01 -4.00385976e-01
-1.67567551e+00 5.09057820e-01 2.78679669e-01 -3.22267175e-01
4.24955368e-01 -1.87277332e-01 1.98361557e-02 -2.05153719e-01
-2.01306447e-01 3.00715834e-01 5.17952800e-01 -1.27587235e+00
8.31251293e-02 -2.36419424e-01 -2.95450151e-01 -4.06407803e-01
-2.86188573e-01 -3.93609166e-01 3.71061772e-01 -9.68412817e-01
2.26720914e-01 -8.49849641e-01 -6.23566806e-01 1.48353353e-01
-1.82250977e-01 -4.56699461e-01 9.41853166e-01 -8.78707349e-01
1.14104056e+00 -1.79396915e+00 6.39562845e-01 3.73738766e-01
2.82849818e-01 1.62125677e-01 3.56316954e-01 8.43410492e-01
-2.94549108e-01 -6.68534338e-02 -9.03216541e-01 -5.50127290e-02
-1.21135011e-01 3.59136134e-01 -4.30810787e-02 4.19633955e-01
3.63728255e-02 6.73896015e-01 -6.15217090e-01 -2.18447119e-01
2.40798175e-01 7.01107442e-01 -9.76233363e-01 5.02179787e-02
-4.16912168e-01 5.76081932e-01 -9.71755236e-02 2.42712140e-01
8.89069438e-01 -2.78253183e-02 1.41937032e-01 -6.18422866e-01
-3.75582039e-01 -2.47077420e-01 -1.49576867e+00 1.69685519e+00
-5.88531733e-01 3.71375918e-01 5.99937975e-01 -1.37732077e+00
9.81833100e-01 5.81019938e-01 8.75020325e-01 -6.26560211e-01
5.27767777e-01 4.75593895e-01 -3.46620902e-02 -4.98950422e-01
1.41211987e-01 -2.18910769e-01 -3.29555944e-02 4.64163333e-01
-1.38240993e-01 -3.59532267e-01 2.09799349e-01 -3.44034404e-01
8.98192346e-01 3.76221567e-01 -9.58765671e-02 -7.94206142e-01
6.22920990e-01 2.10469112e-01 5.67311227e-01 2.58312911e-01
3.39769751e-01 4.11661893e-01 3.31780493e-01 -1.10657001e+00
-1.50260401e+00 -6.22016072e-01 -6.30632341e-01 7.04987824e-01
-1.64633721e-01 -8.46610442e-02 -9.07828152e-01 3.59584838e-01
2.89165974e-01 4.29778457e-01 -7.58027852e-01 -1.22419834e-01
-1.14747512e+00 -6.75128460e-01 1.96763724e-01 8.20355296e-01
2.07748428e-01 -1.27421784e+00 -6.87302887e-01 7.11066127e-01
2.88887054e-01 -7.29876995e-01 1.99951977e-01 2.26266995e-01
-1.29717398e+00 -9.53235924e-01 -5.69263160e-01 -9.94202673e-01
4.83540535e-01 -6.72486007e-01 1.27724683e+00 2.63258517e-01
-5.42460084e-01 5.19414991e-02 3.46971452e-02 -9.99225676e-02
-6.69193149e-01 -2.35092819e-01 4.35169905e-01 -3.49074274e-01
-2.91325897e-01 -1.16330791e+00 -5.13339877e-01 1.58733755e-01
-1.04838479e+00 2.36237943e-01 4.54619415e-02 1.13730061e+00
9.14793909e-01 4.36503649e-01 5.77187598e-01 -8.31137300e-01
7.56258786e-01 -3.94480854e-01 -5.01293778e-01 -4.31387782e-01
-3.88107866e-01 2.53053172e-03 9.08595860e-01 -4.44262207e-01
-9.21016812e-01 9.04692858e-02 -7.20252812e-01 -5.04459441e-01
-2.10552648e-01 7.33978391e-01 1.94295555e-01 -6.11480772e-01
8.37105751e-01 8.69583711e-02 1.76810741e-01 -7.71722436e-01
7.76756704e-02 2.49909267e-01 4.19789612e-01 -1.09708679e+00
4.07613546e-01 4.10716474e-01 7.70461619e-01 -9.40963209e-01
-1.50096923e-01 1.94698200e-01 -8.45545769e-01 -3.31890494e-01
5.77602267e-01 -1.37937307e-01 -1.14863455e+00 7.60646045e-01
-1.42398715e+00 -4.02858913e-01 -3.59882116e-01 2.64470249e-01
-7.84787893e-01 2.98837692e-01 -1.09767640e+00 -5.12068808e-01
-5.53014159e-01 -1.00105715e+00 8.71909916e-01 -1.49949789e-01
-4.31414604e-01 -1.18620002e+00 5.10406911e-01 -1.48450196e-01
8.47353518e-01 1.10531759e+00 1.36885405e+00 3.38917114e-02
8.82793888e-02 -3.58286798e-01 7.94565231e-02 4.44156617e-01
-6.48531094e-02 7.68742189e-02 -5.45777321e-01 -3.20517153e-01
2.39926308e-01 -3.73617619e-01 5.00683129e-01 6.77649796e-01
1.60535789e+00 -1.89441517e-01 -3.87639046e-01 5.12994587e-01
1.81221592e+00 1.91152319e-01 5.24249136e-01 -1.88922897e-01
8.94091189e-01 4.60021973e-01 -5.82914241e-03 6.62576377e-01
-1.63339332e-01 6.85878277e-01 6.55776441e-01 -1.30470872e-01
1.23698838e-01 4.24009383e-01 -2.94934034e-01 1.12026060e+00
-5.70941389e-01 2.30686784e-01 -1.15930974e+00 4.41043198e-01
-1.69528913e+00 -7.26166308e-01 -4.81044948e-01 1.71305096e+00
8.98587942e-01 1.67782292e-01 -5.57722822e-02 6.74962044e-01
4.67421830e-01 -2.95876950e-01 -5.83602488e-01 -9.16197240e-01
1.66707411e-01 7.10462332e-01 7.81929791e-02 5.53956032e-01
-1.02897835e+00 4.04647976e-01 6.48112392e+00 7.74857521e-01
-1.37752056e+00 -7.63590261e-02 7.22145617e-01 4.21864718e-01
-3.22368145e-02 -5.59121907e-01 -2.49835730e-01 5.58830738e-01
1.09244728e+00 1.15697486e-02 3.54611158e-01 9.52566803e-01
1.77264050e-01 1.72211990e-01 -1.13029277e+00 7.44578302e-01
-3.88638526e-01 -2.01396084e+00 -1.53962880e-01 -6.54266328e-02
1.00416434e+00 -1.81804880e-01 -2.25905523e-01 1.71295062e-01
2.65662931e-02 -1.11287856e+00 4.05989558e-01 7.92614102e-01
7.96084642e-01 -9.68131602e-01 8.13836813e-01 4.27105576e-01
-1.19048393e+00 7.44485781e-02 -4.83243555e-01 -7.65822828e-01
3.76744270e-01 8.80258083e-01 -1.17341178e-02 5.20876646e-01
7.88215160e-01 6.40451849e-01 3.21993232e-01 6.47199810e-01
6.60519958e-01 5.28974831e-01 -4.28885520e-01 6.71730861e-02
4.66513000e-02 -4.17293727e-01 5.78348100e-01 8.23682249e-01
4.49031413e-01 2.86185473e-01 1.87158391e-01 1.11239195e+00
4.37258221e-02 1.16039425e-01 -6.49676442e-01 2.88216084e-01
1.39534980e-01 9.02090251e-01 -5.35763204e-01 -1.13922529e-01
4.55971770e-02 3.66340071e-01 3.52636934e-03 1.10979237e-01
-6.64030254e-01 -3.31984371e-01 5.63051164e-01 5.64023495e-01
1.01725474e-01 -5.54082572e-01 -7.31802464e-01 -4.63159889e-01
-1.95002109e-01 -6.12710893e-01 7.43051693e-02 -7.54762292e-01
-1.40327847e+00 5.44356942e-01 -4.50325944e-02 -1.08288717e+00
-5.47247827e-02 -1.17897558e+00 -6.06750250e-01 8.65967214e-01
-1.01968634e+00 -7.76940346e-01 -8.94639082e-03 2.97555625e-01
2.41482511e-01 3.10632568e-02 1.12788165e+00 6.33006990e-01
-2.99517870e-01 -2.62739900e-02 4.80713785e-01 2.08758354e-01
1.68374091e-01 -1.07517493e+00 4.45292711e-01 2.01282635e-01
-7.91978478e-01 3.48867714e-01 9.45293307e-01 -7.76373148e-01
-1.61605585e+00 -8.04051518e-01 6.10996246e-01 6.76754192e-02
6.21626079e-01 -4.41663042e-02 -1.31720269e+00 2.06287026e-01
1.80965364e-01 4.23218280e-01 3.96443486e-01 -2.84491688e-01
4.25218463e-01 -1.40672242e-02 -1.26683593e+00 3.77151519e-01
6.46146595e-01 -1.26509845e-01 -3.39902490e-01 2.57702708e-01
4.51899618e-01 -7.54833460e-01 -1.50342941e+00 9.64030445e-01
6.73226535e-01 -9.07274842e-01 1.21492255e+00 -6.10999048e-01
1.22555637e+00 6.86131269e-02 1.65499449e-02 -1.37375021e+00
-7.05651402e-01 -3.71333390e-01 -5.25837123e-01 7.62529016e-01
-2.44751498e-01 -5.15597701e-01 7.61982977e-01 4.93094981e-01
-3.63927543e-01 -1.60350883e+00 -1.19619489e+00 -3.93217236e-01
8.23090494e-01 -4.17219400e-01 4.64460731e-01 1.05424786e+00
-1.97825342e-01 -4.27447446e-02 -3.61990035e-01 -2.37136066e-01
7.25295722e-01 1.50326431e-01 2.52339274e-01 -1.87535608e+00
-7.74947405e-02 -3.86470288e-01 -1.38716593e-01 -5.44971883e-01
9.39218998e-02 -5.91188431e-01 -1.25313193e-01 -1.51352668e+00
-4.94665235e-01 -6.10817194e-01 -9.14200395e-02 2.79381156e-01
4.39649999e-01 3.51008743e-01 -4.44126785e-01 1.03315204e-01
2.33504504e-01 5.82723975e-01 1.66292870e+00 -8.50678757e-02
-4.03897278e-02 -4.13705826e-01 2.46878131e-03 9.18552876e-01
7.81349063e-01 -4.01721030e-01 2.12581277e-01 -4.57537085e-01
5.37517011e-01 4.69955444e-01 2.59138703e-01 -1.48888779e+00
1.50868922e-01 -3.35472114e-02 5.66946805e-01 -5.78756988e-01
2.16497913e-01 -1.06559622e+00 3.76452893e-01 9.07088935e-01
-1.86987206e-01 2.44751200e-01 6.68557584e-01 1.95817888e-01
-2.22301155e-01 -2.59726774e-02 1.11817849e+00 -1.82365432e-01
-4.12556708e-01 2.33138636e-01 -6.01606071e-01 -2.60876715e-01
8.23409081e-01 -1.60797596e-01 1.85010940e-01 1.01744622e-01
-9.46521580e-01 -1.26662552e-01 3.15431476e-01 -1.60498306e-01
6.08676016e-01 -1.74306881e+00 -8.13495338e-01 5.58694243e-01
-6.32139325e-01 2.02198505e-01 6.59616590e-01 6.55279398e-01
-1.19158947e+00 2.65337944e-01 -6.70078039e-01 -6.20433629e-01
-8.50280762e-01 3.45420033e-01 6.23564124e-01 -5.42838693e-01
-5.39454341e-01 9.37471688e-01 -3.18503976e-01 -7.22536206e-01
-1.59949854e-01 -3.70321065e-01 -2.27888942e-01 -9.32972431e-02
1.65588453e-01 1.11464429e+00 4.09943551e-01 -6.42629206e-01
-1.47319809e-01 9.70608175e-01 5.15676618e-01 3.81888866e-01
1.68129349e+00 3.90159428e-01 -7.07794607e-01 2.55891055e-01
1.26066911e+00 -4.89752114e-01 -1.09954941e+00 -1.10909142e-01
-3.84617060e-01 -6.61930665e-02 4.10937935e-01 -5.09409547e-01
-1.20305526e+00 9.70545471e-01 6.16916537e-01 4.16079521e-01
9.21597958e-01 -6.20068789e-01 1.46464491e+00 3.13361347e-01
4.40617889e-01 -1.24362230e+00 -5.32501638e-02 9.29437399e-01
1.23317337e+00 -9.24310148e-01 2.21572503e-01 -3.67490381e-01
3.37720931e-01 1.56758273e+00 6.42418027e-01 -6.92345321e-01
1.16576862e+00 7.74237454e-01 -1.89931333e-01 -2.70233154e-01
-3.90719384e-01 6.60174549e-01 2.66008228e-01 2.88579136e-01
6.47918642e-01 -1.81823403e-01 -5.26076138e-01 5.13163567e-01
-2.82923728e-02 3.00398231e-01 2.53836215e-01 1.08793485e+00
-2.71759838e-01 -1.09594893e+00 -7.08055019e-01 5.44295132e-01
-5.47884703e-01 2.55006790e-01 -1.84676200e-02 5.56023300e-01
4.24604058e-01 2.58245707e-01 1.64080754e-01 -4.27438587e-01
2.98037350e-01 5.09839840e-02 5.66260278e-01 -2.85580635e-01
-6.37770474e-01 -2.31423199e-01 -2.54736692e-01 -5.64603567e-01
-3.03711951e-01 -2.61767179e-01 -1.41795278e+00 -7.55060732e-01
-1.06559610e-02 1.67687669e-01 4.60219622e-01 9.58258510e-01
5.15769184e-01 1.11507380e+00 3.38403106e-01 -1.87149644e+00
-2.06642598e-01 -8.94085169e-01 -6.40815794e-01 5.83502412e-01
1.04503296e-01 -1.10243809e+00 -2.53082842e-01 2.80257668e-02] | [6.320991516113281, 3.3872530460357666] |
32fd29c5-ee46-4f0a-99e0-51d6d248eb5e | objective-assessment-of-spatial-audio-quality | 2212.01451 | null | https://arxiv.org/abs/2212.01451v1 | https://arxiv.org/pdf/2212.01451v1.pdf | Objective Assessment of Spatial Audio Quality using Directional Loudness Maps | This work introduces a feature extracted from stereophonic/binaural audio signals aiming to represent a measure of perceived quality degradation in processed spatial auditory scenes. The feature extraction technique is based on a simplified stereo signal model considering auditory events positioned towards a given direction in the stereo field using amplitude panning (AP) techniques. We decompose the stereo signal into a set of directional signals for given AP values in the Short-Time Fourier Transform domain and calculate their overall loudness to form a directional loudness representation or maps. Then, we compare directional loudness maps of a reference signal and a deteriorated version to derive a distortion measure aiming to describe the associated perceived degradation scores reported in listening tests. The measure is then tested on an extensive listening test database with stereo signals processed by state-of-the-art perceptual audio codecs using non waveform-preserving techniques such as bandwidth extension and joint stereo coding, known for presenting a challenge to existing quality predictors. Results suggest that the derived distortion measure can be incorporated as an extension to existing automated perceptual quality assessment algorithms for improving prediction on spatially coded audio signals. | ['Jürgen Herre', 'Pablo M. Delgado'] | 2022-12-02 | null | null | null | null | ['bandwidth-extension', 'bandwidth-extension'] | ['audio', 'speech'] | [ 5.70299685e-01 -3.27966273e-01 6.81750953e-01 -2.79973030e-01
-1.19180751e+00 -5.24074256e-01 3.59532386e-01 6.14639878e-01
-2.83339322e-01 5.02347350e-01 7.68033385e-01 4.62595187e-02
-7.29618013e-01 -5.45053184e-01 -1.69661954e-01 -6.96224153e-01
-2.00574532e-01 4.56243344e-02 6.58265412e-01 -3.35528612e-01
4.31075662e-01 5.33787906e-01 -2.29238343e+00 5.80347657e-01
7.25731134e-01 1.32455277e+00 2.52028614e-01 1.18141079e+00
3.84749055e-01 3.61808956e-01 -9.66384411e-01 -7.88416043e-02
2.06880599e-01 -4.96118039e-01 -5.73792756e-01 -2.29600854e-02
3.45368117e-01 3.29727456e-02 2.33300120e-01 1.06379426e+00
9.57952380e-01 2.42918268e-01 6.53289199e-01 -9.06134427e-01
8.11521783e-02 3.82576197e-01 1.99114919e-01 5.14726341e-01
9.58374023e-01 -1.31981105e-01 8.89597476e-01 -8.71266067e-01
2.46581361e-01 1.02400649e+00 5.46382844e-01 -1.39119640e-01
-1.48880708e+00 -3.44516814e-01 -5.84368944e-01 7.20534384e-01
-1.17063546e+00 -6.68113053e-01 1.01679575e+00 -5.50818324e-01
9.18916643e-01 7.17346311e-01 8.80873799e-01 3.66746664e-01
2.36082613e-01 2.67951220e-01 1.34787750e+00 -7.22904921e-01
5.67803681e-01 -1.44078150e-01 -8.60046074e-02 -1.05252028e-01
-3.79225791e-01 3.86876345e-01 -8.00926387e-01 -1.52444974e-01
8.57160389e-02 -8.19703341e-01 -7.41303980e-01 -3.80144179e-01
-1.00212336e+00 3.65187794e-01 2.22068682e-01 6.11028433e-01
-4.57056373e-01 -4.00790870e-01 4.01164740e-01 5.97846925e-01
5.86960971e-01 4.69570518e-01 -3.06098491e-01 -5.70099115e-01
-1.24324942e+00 4.54328716e-01 6.86229110e-01 8.69616047e-02
4.64198053e-01 3.67146999e-01 -1.61956131e-01 1.26801598e+00
2.13416263e-01 3.76743138e-01 5.31903982e-01 -1.02427685e+00
8.89136791e-02 3.75266187e-02 7.76645094e-02 -1.11591077e+00
-3.85721087e-01 -6.53069317e-01 -4.09017920e-01 5.75863302e-01
3.19345653e-01 3.08514893e-01 -3.68343174e-01 1.41434956e+00
2.22419634e-01 4.36478592e-02 -1.59924552e-01 9.06797111e-01
2.71825105e-01 7.10339308e-01 -4.83499110e-01 -5.63566208e-01
1.26412117e+00 -3.37533385e-01 -7.58355618e-01 1.75389007e-01
-4.60549537e-03 -1.26910365e+00 1.03791225e+00 1.32655692e+00
-1.42901230e+00 -1.08507109e+00 -1.24185145e+00 2.81016260e-01
-2.64144808e-01 -2.59427637e-01 -3.42852563e-01 1.06868267e+00
-1.38097477e+00 6.04483485e-01 -3.29153925e-01 1.05391592e-01
-4.65012342e-01 7.14996532e-02 -1.10835306e-01 1.59641162e-01
-1.13217151e+00 7.75132358e-01 3.52035314e-02 -2.28351191e-01
-8.90130758e-01 -9.92011547e-01 -7.99630582e-01 1.71897605e-01
-1.94309875e-01 -2.38341987e-01 1.33962393e+00 -6.00334346e-01
-1.54483819e+00 6.79544866e-01 -1.89798295e-01 -5.56801677e-01
3.14349979e-01 -2.92438328e-01 -1.18855262e+00 3.99658203e-01
-5.25761843e-02 1.51909187e-01 1.25404024e+00 -1.29448259e+00
-6.43246889e-01 -3.16829771e-01 -4.42528665e-01 2.72055537e-01
-7.27764890e-02 1.87566847e-01 -7.40747377e-02 -8.23608994e-01
3.69807661e-01 -2.54622936e-01 2.74328709e-01 -2.24514365e-01
-2.72183150e-01 2.37782717e-01 5.82061231e-01 -1.14190173e+00
1.58020139e+00 -2.28486180e+00 1.86067164e-01 2.62584418e-01
-1.52496472e-01 3.12071174e-01 -4.22836468e-02 4.07695860e-01
-4.91848618e-01 -4.80300635e-01 -2.53353745e-01 -5.71146496e-02
4.63381633e-02 -4.10069585e-01 -2.89912641e-01 3.17658991e-01
-9.90864858e-02 -1.38820887e-01 -8.59825432e-01 -2.76865512e-01
5.30620396e-01 6.44213438e-01 -7.33257234e-01 3.33613187e-01
2.52036780e-01 3.23353052e-01 3.46015632e-01 4.66408342e-01
8.22689116e-01 1.05232382e+00 -3.92348975e-01 -4.68066394e-01
-4.62148398e-01 4.28139955e-01 -1.48458183e+00 1.50063598e+00
-5.93324423e-01 6.95921719e-01 3.09499323e-01 -7.15376318e-01
1.44759464e+00 5.47516286e-01 2.87581831e-01 -8.19010794e-01
-2.40383651e-02 4.99594301e-01 2.34397322e-01 -3.64501923e-01
5.93685746e-01 -3.86465669e-01 2.54805267e-01 1.32813733e-02
2.24111170e-01 -8.11256349e-01 1.23500831e-01 -3.30610245e-01
9.18706417e-01 -2.21781626e-01 2.38558635e-01 -3.49239409e-01
1.01235580e+00 -7.89470136e-01 4.63065505e-02 3.03713500e-01
-4.09416169e-01 8.87009859e-01 1.89103499e-01 1.02293432e-01
-1.12935185e+00 -1.49779141e+00 -5.15891910e-01 1.09383619e+00
-3.87258679e-01 -4.53707606e-01 -9.01391923e-01 3.79513919e-01
-3.28443021e-01 9.90837157e-01 -1.08180605e-01 -4.76519495e-01
-2.79561669e-01 -2.18901932e-01 4.36906010e-01 8.72537959e-03
2.62769550e-01 -9.17633355e-01 -5.46885490e-01 5.34627080e-01
-2.62186855e-01 -4.57177818e-01 -2.30216771e-01 2.44974047e-01
-6.34152651e-01 -6.87370360e-01 -8.25644672e-01 -7.29504645e-01
-2.58016497e-01 3.69574651e-02 1.11312914e+00 -6.34246647e-01
-3.70777637e-01 5.83553076e-01 -6.25148118e-01 -4.23587292e-01
-7.13457763e-01 -6.97017610e-01 2.12483779e-01 3.87349337e-01
5.54947332e-02 -1.03300488e+00 -8.64090741e-01 5.63556790e-01
-1.00900364e+00 -4.49902982e-01 2.08370954e-01 4.40441757e-01
5.27209997e-01 5.49945533e-01 4.67953295e-01 2.67166942e-01
1.11313343e+00 -1.22476503e-01 -3.60450715e-01 -2.25483164e-01
-4.55763459e-01 -2.51617223e-01 5.61570227e-01 -1.86624721e-01
-1.14827240e+00 -3.55136037e-01 -5.40966511e-01 -3.61141376e-02
-4.11679447e-01 3.05356264e-01 -3.36207539e-01 -8.37820489e-03
9.90072668e-01 5.15176535e-01 -2.36724824e-01 -7.91860998e-01
4.53682169e-02 9.88712490e-01 9.97914791e-01 -3.53342474e-01
5.51691413e-01 2.07406148e-01 5.86131513e-02 -1.08822167e+00
-1.90875605e-01 -8.87751281e-01 -5.95326722e-01 -6.18282676e-01
5.43754637e-01 -5.70026577e-01 -4.29486781e-01 5.42794287e-01
-8.71247709e-01 4.34691831e-02 -5.02834976e-01 7.03453541e-01
-9.82350945e-01 5.14380157e-01 -3.32336456e-01 -1.19479370e+00
-6.71764463e-02 -1.15962422e+00 1.15355897e+00 -2.28300616e-01
-4.10298675e-01 -7.40406036e-01 6.06317878e-01 2.51668513e-01
6.19698346e-01 -1.13337345e-01 9.59582508e-01 -3.64418447e-01
1.71382263e-01 -2.60424405e-01 3.70683223e-01 8.69001687e-01
1.81692705e-01 -1.37730435e-01 -1.24948239e+00 -2.32008219e-01
5.13599753e-01 -8.47637728e-02 5.62233686e-01 7.03853071e-01
1.03864372e+00 -1.00406399e-02 3.79052758e-01 3.37853879e-01
1.40073037e+00 5.67750454e-01 8.29356611e-01 1.77899450e-01
-1.91871956e-01 8.20338488e-01 8.80193233e-01 5.82793176e-01
-3.12289119e-01 1.11367166e+00 4.23797965e-01 1.11356519e-01
-4.60973114e-01 -1.83090419e-01 3.73092085e-01 1.13934791e+00
-1.03975309e-03 2.82459687e-02 -7.99347937e-01 7.81526983e-01
-9.13449049e-01 -9.63662803e-01 -1.99433342e-01 2.53382850e+00
8.59988570e-01 3.50197226e-01 3.71554613e-01 1.46689284e+00
8.26295853e-01 9.66820270e-02 -2.57249065e-02 -1.02518117e+00
-1.87323421e-01 7.15849102e-01 -3.91634554e-02 8.54878545e-01
-6.84627354e-01 1.57498285e-01 6.12900114e+00 1.11503661e+00
-1.06890500e+00 -1.36542246e-02 1.80593833e-01 -9.22950953e-02
-1.73274338e-01 -2.14999378e-01 -1.99722260e-01 4.06395048e-01
1.42525923e+00 -2.63002336e-01 4.65421379e-01 5.46360612e-01
6.31024241e-01 -3.39531958e-01 -8.74706864e-01 1.16019785e+00
2.66661555e-01 -5.75623274e-01 -2.11844761e-02 -1.39169291e-01
3.48411381e-01 -4.65654463e-01 4.75047469e-01 -1.66349649e-01
-4.29518342e-01 -7.72823334e-01 1.09089088e+00 8.14498484e-01
7.44208097e-01 -8.94120991e-01 4.72456604e-01 4.99019958e-02
-1.31212246e+00 -2.05025405e-01 -3.23081881e-01 -1.89388677e-01
3.89269650e-01 7.46840000e-01 -1.00871110e+00 5.63930392e-01
7.85268188e-01 3.26866567e-01 -5.75490713e-01 1.85793865e+00
1.54265225e-01 7.45427728e-01 -1.46779388e-01 3.64467442e-01
-1.95502952e-01 -1.21354703e-02 1.19796634e+00 1.26605690e+00
9.28635538e-01 -3.92905474e-01 -6.11182034e-01 4.86230254e-01
6.62851214e-01 3.64983410e-01 -3.87213379e-01 3.74936402e-01
5.14048696e-01 8.72949839e-01 -3.79076779e-01 6.40015258e-03
-3.25162560e-02 8.00203145e-01 -7.42000401e-01 2.15168729e-01
-4.34492767e-01 -6.18331015e-01 5.88277936e-01 1.88099712e-01
2.50415623e-01 1.02330148e-01 -1.46963239e-01 -4.75028038e-01
3.48188244e-02 -9.90000308e-01 1.01103298e-01 -1.40686178e+00
-8.80020916e-01 8.21029305e-01 5.08815460e-02 -1.82657933e+00
-4.85331327e-01 -5.70376456e-01 -5.24768889e-01 1.01997530e+00
-1.11035693e+00 -4.00334567e-01 -2.55751181e-02 8.16864848e-01
6.66282415e-01 -1.89559251e-01 9.58756804e-01 4.16215807e-01
3.64612550e-01 3.44287574e-01 1.96523950e-01 -6.07104123e-01
7.42255449e-01 -1.50814545e+00 -1.49814650e-01 8.28237891e-01
-4.37667556e-02 1.25153661e-01 1.37800813e+00 -1.88892603e-01
-6.51872218e-01 -7.86100090e-01 9.77538228e-01 -3.39557156e-02
7.01769233e-01 -2.27009594e-01 -8.32080662e-01 -2.15624914e-01
1.60999879e-01 -3.83124650e-01 8.50483179e-01 -2.13385239e-01
-1.51371658e-01 -5.36265135e-01 -1.15667844e+00 2.44512916e-01
6.15692139e-01 -9.14298654e-01 -1.06039524e+00 -1.67586312e-01
5.44799685e-01 1.43117541e-02 -1.01392627e+00 4.48282450e-01
5.03070831e-01 -1.60670900e+00 1.19134867e+00 1.14432909e-01
2.63980895e-01 -5.80524027e-01 -5.61991751e-01 -1.87992120e+00
-3.17457736e-01 -8.07305157e-01 3.41568500e-01 1.25312495e+00
1.90506056e-01 -3.49807441e-01 1.98085770e-01 -1.68767616e-01
-3.98501307e-01 -1.80016324e-01 -1.39021719e+00 -7.85284102e-01
-1.54563904e-01 -1.07852256e+00 4.32145506e-01 3.32310736e-01
2.44709894e-01 1.07144229e-01 -1.99461922e-01 2.15358511e-01
6.11915827e-01 -1.74096122e-01 2.88671494e-01 -1.31764054e+00
-5.97129643e-01 -6.23297095e-01 -9.49762762e-01 -5.84476411e-01
-4.52136666e-01 -4.73275244e-01 1.35225207e-01 -1.15441644e+00
-5.13699293e-01 3.90765518e-02 -5.44376314e-01 -3.93529624e-01
3.00608665e-01 4.27822053e-01 1.58731282e-01 -1.67454436e-01
-6.32756576e-02 5.81532538e-01 1.02974057e+00 -2.29690254e-01
-2.61204332e-01 4.86680537e-01 -2.17979282e-01 6.99483991e-01
4.81919140e-01 -2.08803520e-01 -5.47484934e-01 1.64295569e-01
3.12904060e-01 6.12851918e-01 4.78548408e-01 -1.83332312e+00
2.00932875e-01 4.23782408e-01 2.49433890e-02 -7.70039022e-01
7.60460734e-01 -9.27751243e-01 3.88883740e-01 4.72484887e-01
-5.19676328e-01 -1.46842644e-01 2.70353645e-01 4.54199821e-01
-8.71093929e-01 -2.56282002e-01 9.95863140e-01 4.26647604e-01
-5.22044420e-01 -5.03739536e-01 -6.39093757e-01 -4.07028884e-01
5.95527887e-01 -5.44431210e-01 1.28287330e-01 -8.14543009e-01
-1.21082008e+00 -6.73842371e-01 2.50389814e-01 3.99657972e-02
8.39246213e-01 -1.24931121e+00 -7.58269966e-01 3.96043658e-01
9.39629152e-02 -5.91748416e-01 6.37452841e-01 7.46714473e-01
-5.54948688e-01 5.74599087e-01 -6.01790488e-01 -7.46502221e-01
-1.65185058e+00 4.70390141e-01 5.49011350e-01 6.35728762e-02
-1.35844136e-02 8.80615115e-01 2.12445650e-02 8.98519456e-02
2.89141387e-01 -6.15385473e-01 -5.96806586e-01 2.61082590e-01
7.87733376e-01 8.13238621e-01 7.56703317e-01 -8.63635302e-01
-1.97316274e-01 7.44783819e-01 6.53765380e-01 -7.77454197e-01
1.26848686e+00 -4.51818496e-01 -1.15671521e-02 7.74022400e-01
1.36950600e+00 5.76895177e-01 -1.01124930e+00 -2.65014824e-02
-1.28115520e-01 -6.24566197e-01 3.44018579e-01 -1.08550346e+00
-5.51831365e-01 1.08709431e+00 1.25123978e+00 8.03086519e-01
1.90023017e+00 -2.43582249e-01 3.53932112e-01 -2.04950899e-01
4.23865676e-01 -1.10951924e+00 1.90969825e-01 2.16540307e-01
1.32925045e+00 -4.14871246e-01 -3.59397292e-01 -3.75826329e-01
-2.98298419e-01 1.26832378e+00 -2.21535787e-01 1.21845528e-01
9.00811672e-01 1.81912735e-01 7.78848454e-02 1.25007614e-01
-4.76182610e-01 -2.80728638e-01 6.79727197e-01 9.88000751e-01
4.26115006e-01 5.48315011e-02 -2.65335768e-01 4.03231442e-01
-1.03626513e+00 -3.51534188e-01 3.75467569e-01 4.44325179e-01
-8.90439153e-01 -9.89298403e-01 -9.67346489e-01 1.51321456e-01
-3.50026190e-01 -2.97752023e-01 -2.27542594e-01 1.16325602e-01
2.82475322e-01 1.44838667e+00 1.10733479e-01 -6.50185585e-01
7.58950710e-01 3.92873973e-01 3.98838848e-01 -2.29497418e-01
-5.39456010e-01 4.95914131e-01 8.47862884e-02 -4.54586208e-01
-3.48783314e-01 -7.58832633e-01 -4.75005955e-01 6.09642044e-02
-5.24871238e-02 2.08194301e-01 8.89075518e-01 5.54473162e-01
-4.05557714e-02 8.85692418e-01 8.21305096e-01 -1.16240489e+00
-3.04142803e-01 -1.17690086e+00 -1.05869699e+00 5.04924953e-01
6.34148479e-01 -4.86076146e-01 -6.93327487e-01 1.57074526e-01] | [15.339091300964355, 5.614477634429932] |
24da084e-b47c-4591-be91-a2ad3adeb65d | meta-neural-network-for-realtime-and-passive | 1909.07122 | null | https://arxiv.org/abs/1909.07122v1 | https://arxiv.org/pdf/1909.07122v1.pdf | Meta-neural-network for Realtime and Passive Deep-learning-based Object Recognition | Deep-learning recently show great success across disciplines yet conventionally require time-consuming computer processing or bulky-sized diffractive elements. Here we theoretically propose and experimentally demonstrate a purely-passive "meta-neural-network" with compactness and high-resolution for real-time recognizing complicated objects by analyzing acoustic scattering. We prove our meta-neural-network mimics standard neural network despite its small footprint, thanks to unique capability of its metamaterial unit cells, dubbed "meta-neurons", to produce deep-subwavelength-distribution of discrete phase shift as learnable parameters during training. The resulting device exhibits the "intelligence" to perform desired tasks with potential to address the current trade-off between reducing device's size, cost and energy consumption and increasing recognition speed and accuracy, showcased by an example of handwritten digit recognition. Our mechanism opens the route to new metamaterial-based deep-learning paradigms and enable conceptual devices such as smart transducers automatically analyzing signals, with far-reaching implications for acoustics, optics and related fields. | ['Xue-Feng Zhu', 'Yujiang Ding', 'Jingkai Weng', 'Bin Liang', 'Jing Yang', 'Jianchun Cheng', 'Chengbo Hu'] | 2019-09-16 | null | null | null | null | ['handwritten-digit-recognition'] | ['computer-vision'] | [ 5.41662574e-01 5.66487491e-01 6.35237992e-01 -6.76097050e-02
-5.76639056e-01 -3.39588314e-01 4.55333769e-01 -4.38270867e-01
-4.37855572e-01 4.29383993e-01 -2.27603674e-01 -2.81173080e-01
-4.56129909e-01 -1.13355362e+00 -8.91666055e-01 -1.69750643e+00
5.35727339e-03 2.57404804e-01 1.74952224e-01 -1.39637187e-01
5.16640604e-01 4.12657678e-01 -1.74210525e+00 5.25871754e-01
6.05595410e-01 1.54193771e+00 3.08971375e-01 3.21898192e-01
1.07543997e-01 6.79146409e-01 -3.38229746e-01 -1.42963767e-01
1.67540237e-01 -1.81777269e-01 -2.49069944e-01 -4.73233491e-01
6.29441738e-02 -6.80315271e-02 -3.99571866e-01 8.59702289e-01
5.55858374e-01 9.58982576e-03 8.91795814e-01 -3.43404323e-01
-1.13636827e+00 6.55858755e-01 -3.98560584e-01 -6.78604469e-02
-3.20138603e-01 2.15659901e-01 6.39345884e-01 -7.49774277e-01
1.77826658e-01 7.58196294e-01 8.98235619e-01 1.08774877e+00
-9.04771686e-01 -6.82015240e-01 -7.30266988e-01 1.90725908e-01
-7.45762408e-01 -7.04822719e-01 9.75650907e-01 -3.66321087e-01
1.32067323e+00 2.62341946e-01 4.89719838e-01 1.24695623e+00
6.84351563e-01 -6.11781003e-03 9.34893847e-01 -6.97701991e-01
4.08564627e-01 1.14010610e-01 -1.95661411e-01 1.11788988e+00
3.55517596e-01 1.18326992e-01 -4.99692172e-01 2.59479135e-01
8.54761064e-01 -1.02751769e-01 1.25580683e-01 2.64561474e-02
-1.09167147e+00 3.08535874e-01 4.14755523e-01 7.76802778e-01
-3.49680692e-01 6.83855653e-01 -1.82750642e-01 3.42275172e-01
4.15285230e-01 1.07282436e+00 -1.04392782e-01 9.91912857e-02
-4.91565645e-01 -3.29735130e-01 6.52160347e-01 4.29406166e-01
5.88962376e-01 6.26332819e-01 1.58278912e-01 8.82254064e-01
3.47057700e-01 8.93190622e-01 6.16522789e-01 -1.22027409e+00
3.75335254e-02 5.56964517e-01 1.71672195e-01 -8.04966807e-01
-7.73399532e-01 -7.19255924e-01 -9.89606440e-01 4.22515303e-01
2.70800412e-01 -1.81221887e-01 -7.33193636e-01 1.34593415e+00
1.67722851e-01 -1.55480102e-01 1.97305575e-01 8.31342220e-01
9.12189960e-01 8.67057502e-01 -3.48207444e-01 1.42699167e-01
1.43268931e+00 -5.55113018e-01 -9.24664885e-02 -1.13240309e-01
5.72202146e-01 -1.73906624e-01 1.10347283e+00 6.90038741e-01
-1.11352015e+00 -2.82819599e-01 -1.29186881e+00 1.05365157e-01
-3.13091010e-01 1.14202917e-01 8.50732863e-01 9.15378511e-01
-8.06803644e-01 8.19751441e-01 -9.86401737e-01 9.01606027e-03
6.63113773e-01 7.91250646e-01 9.48743448e-02 4.66233820e-01
-5.88768244e-01 3.64600509e-01 4.44519445e-02 1.00222513e-01
-6.06701732e-01 -1.10986948e+00 7.30878711e-02 9.07245949e-02
-1.89114362e-01 -9.71514106e-01 1.12645555e+00 -7.83412635e-01
-2.38898802e+00 1.16318202e+00 4.05417770e-01 -4.70794141e-01
-1.64989587e-02 -7.97289014e-02 -4.88500357e-01 3.01957637e-01
-5.33738792e-01 3.06621283e-01 1.16497779e+00 -1.11353111e+00
-2.90140629e-01 -5.09471953e-01 -1.20781764e-01 -6.44778728e-01
-8.77418816e-01 -1.39070034e-01 6.24399364e-01 -4.94440198e-01
6.00670397e-01 -8.76291871e-01 -1.14356667e-01 -1.59967113e-02
-2.43098095e-01 -3.29640061e-01 9.52329755e-01 7.84857869e-02
3.98987323e-01 -2.04621458e+00 -1.94096237e-01 1.84510991e-01
5.72215199e-01 3.35644573e-01 -3.24622989e-01 5.62745392e-01
1.86209261e-01 -2.00826198e-01 -3.11326161e-02 1.32832348e-01
-1.52535960e-01 -2.93970615e-01 -3.23344439e-01 4.89474684e-01
9.34583843e-02 1.08278120e+00 -5.89828134e-01 5.14977984e-02
3.10853630e-01 6.35194242e-01 -6.88951075e-01 9.89076793e-02
-4.50325727e-01 8.52418005e-01 -8.26102972e-01 7.73515165e-01
6.47838712e-01 -5.34474432e-01 -6.11582249e-02 -4.33090270e-01
-4.48090047e-01 1.55534849e-01 -5.41003466e-01 1.48139298e+00
-7.32040107e-01 8.94939780e-01 5.29262364e-01 -1.70490599e+00
1.05396247e+00 5.82700111e-02 8.59564006e-01 -1.70618975e+00
3.77725333e-01 7.03658044e-01 1.37795703e-02 -1.01418638e+00
-4.87074032e-02 -3.67831558e-01 1.36297822e-01 6.88392401e-01
6.46134838e-02 1.27344429e-01 -5.59795201e-01 -6.79268360e-01
1.33317184e+00 -1.31717607e-01 -4.64117378e-01 -7.03685105e-01
3.67181927e-01 -1.69009432e-01 -2.59411514e-01 1.05708373e+00
4.84793633e-01 2.78765440e-01 -6.08839653e-02 -8.26778233e-01
-1.07001495e+00 -1.11160803e+00 -2.31970236e-01 1.32090938e+00
3.29769328e-02 4.05355543e-01 -8.66957963e-01 4.19767886e-01
-2.15033665e-01 2.07856208e-01 -4.94462579e-01 -3.18300366e-01
-9.48314667e-01 -8.88666093e-01 6.71249151e-01 9.36850235e-02
6.75245762e-01 -1.17933714e+00 -1.32281268e+00 5.95639467e-01
3.17805350e-01 -1.19402945e+00 8.70892346e-01 1.95187107e-01
-9.10037220e-01 -6.22231126e-01 -6.81847215e-01 -1.16009891e+00
4.74567860e-01 2.64496207e-02 8.61176670e-01 -1.08183458e-01
-7.66549349e-01 5.45417368e-01 -8.97001475e-02 -6.80276036e-01
-4.61925626e-01 2.15636075e-01 2.74852246e-01 1.35974482e-01
6.72683641e-02 -1.47710502e+00 -1.27523506e+00 5.85619807e-02
-5.42048275e-01 1.60446286e-01 9.86552477e-01 4.69613105e-01
3.25920105e-01 -3.23722333e-01 1.01182485e+00 -6.78558230e-01
3.17351848e-01 -2.87278533e-01 -6.27332807e-01 2.53802389e-01
-3.63437086e-01 1.56827107e-01 8.54650497e-01 -5.85394561e-01
-9.48028147e-01 -6.02941632e-01 -5.19046485e-01 3.06843102e-01
-1.73420727e-01 5.99157289e-02 1.24850735e-01 -8.08225751e-01
8.26263845e-01 3.57080102e-01 -1.08627997e-01 -2.27884337e-01
-3.75592709e-02 7.91505635e-01 7.44394302e-01 -7.35854030e-01
7.43107557e-01 7.71494508e-01 4.60818172e-01 -1.57728088e+00
-6.44563735e-01 8.52349997e-02 -1.75635833e-02 -2.36368373e-01
8.23863745e-01 -6.49355650e-01 -1.41199195e+00 8.68865192e-01
-1.17839754e+00 -4.04561371e-01 -4.61087912e-01 3.56588274e-01
-5.15727222e-01 -2.46367455e-01 -6.05314076e-01 -8.96712899e-01
-8.90835881e-01 -5.42897761e-01 1.05378664e+00 3.11465263e-01
2.55954385e-01 -8.84189844e-01 -4.86676861e-03 3.82098585e-01
9.71362352e-01 3.75806332e-01 1.49455369e+00 -1.40441015e-01
-1.04513597e+00 -6.66136593e-02 -2.56208837e-01 1.30496264e-01
8.02934691e-02 -4.21834707e-01 -1.50289416e+00 -1.92575976e-01
4.76388395e-01 -4.68489617e-01 1.10695541e+00 5.14640093e-01
1.65194082e+00 -5.50095737e-01 -4.67153192e-01 7.80472398e-01
1.64947248e+00 2.94215947e-01 6.29318476e-01 1.79520920e-02
6.82300687e-01 7.34587193e-01 -4.42276210e-01 4.26664203e-01
-2.80473441e-01 3.39969784e-01 3.40310097e-01 8.56286213e-02
-6.61550164e-02 2.71343261e-01 1.41414851e-01 1.38361692e+00
-5.80583692e-01 -3.37660760e-01 -9.05913651e-01 4.84244376e-02
-1.24846721e+00 -9.60853517e-01 6.01021945e-02 1.81614459e+00
3.07194710e-01 4.38412130e-02 -4.48475629e-01 5.40768690e-02
1.63650826e-01 2.35063452e-02 -7.79242516e-01 -3.48276734e-01
-1.02530979e-01 9.50380862e-01 2.65340596e-01 5.91299720e-02
-6.96775436e-01 5.00637770e-01 5.81525087e+00 7.62442768e-01
-1.57405674e+00 3.57881218e-01 2.47222468e-01 -2.14381069e-01
-6.51864588e-01 -6.93956554e-01 -3.07946146e-01 2.97542125e-01
1.19338202e+00 5.68153977e-01 3.70769978e-01 2.30893821e-01
-2.03006580e-01 4.23242114e-02 -1.09136403e+00 1.15740848e+00
-3.87766391e-01 -1.99226403e+00 1.90848216e-01 1.85065389e-01
7.41269350e-01 3.16382051e-01 5.94635785e-01 -1.20555818e-01
-3.01274687e-01 -1.04863060e+00 6.09104335e-01 6.74534798e-01
9.56679523e-01 -4.92261618e-01 -4.56014015e-02 3.34632993e-01
-7.07692683e-01 -7.23026931e-01 -5.94541073e-01 -9.23704763e-04
-5.32601118e-01 1.00644982e+00 -3.55093658e-01 -5.82158156e-02
9.36070025e-01 2.60078400e-01 -2.05987282e-02 5.37692547e-01
7.00058401e-01 7.57405639e-01 -3.86224836e-01 -8.30945730e-01
3.69584322e-01 -4.57865417e-01 5.59047341e-01 1.05277932e+00
8.09644163e-01 4.09638435e-01 -5.91665268e-01 1.38192689e+00
-1.93356439e-01 -3.00391406e-01 -5.52962542e-01 -3.34815294e-01
3.71369898e-01 1.30247796e+00 -9.66603816e-01 9.53712985e-02
-2.11307704e-01 4.84738320e-01 7.99380466e-02 2.65756696e-01
-4.72038567e-01 -7.03029513e-01 3.91655147e-01 4.30204839e-01
3.44841510e-01 -5.82320869e-01 -5.83476305e-01 -7.52016068e-01
1.08725540e-01 -1.99824497e-01 -5.05600929e-01 -6.63075209e-01
-9.53688800e-01 2.04693958e-01 -6.18747413e-01 -9.35135543e-01
1.55867502e-01 -1.33561075e+00 -5.09266913e-01 2.36059800e-01
-1.21128953e+00 -1.30210459e+00 -2.96677083e-01 2.60220051e-01
9.81038138e-02 -4.58220631e-01 1.01726794e+00 3.34703356e-01
-2.71813273e-01 5.57911098e-01 8.63779962e-01 1.78258285e-01
1.13535762e-01 -7.18068480e-01 3.64006013e-01 1.27426594e-01
2.15293050e-01 3.43574911e-01 7.06202209e-01 1.29011884e-01
-2.08236527e+00 -8.51019204e-01 1.14533521e-01 -4.01090145e-01
6.56219006e-01 -7.37362802e-01 -7.03805208e-01 -9.47914496e-02
2.11941585e-01 -2.58315802e-01 7.98452556e-01 -5.40648997e-02
-5.28230071e-01 -7.41400421e-01 -1.20367098e+00 4.03103799e-01
1.48705912e+00 -5.91238379e-01 -2.61959046e-01 5.40884137e-01
6.67623818e-01 -1.51119372e-02 -8.05057228e-01 5.24345636e-01
1.18704832e+00 -1.10615027e+00 1.13704264e+00 -3.64548296e-01
8.55009675e-01 5.03701925e-01 -1.07451119e-01 -9.47582245e-01
-4.94106025e-01 -8.66050422e-01 -2.67524600e-01 8.63920331e-01
5.46155095e-01 -1.18546331e+00 1.14654613e+00 1.30060807e-01
-6.25436127e-01 -1.00026536e+00 -1.43079913e+00 -6.94231749e-01
3.64816457e-01 -2.25483313e-01 6.03061676e-01 6.15532756e-01
-2.55324781e-01 -3.56940143e-02 1.43287554e-01 2.04871893e-01
9.08536196e-01 3.64751011e-01 3.04078776e-02 -1.59957552e+00
-3.86344910e-01 -7.94437885e-01 -2.85021365e-01 -8.53000283e-01
6.65262621e-03 -5.98118603e-01 8.14941451e-02 -1.06640947e+00
-2.83822358e-01 -8.40898514e-01 -4.28176999e-01 7.31803402e-02
9.77694035e-01 4.76575166e-01 -5.19012392e-01 1.30867109e-01
-6.24949813e-01 3.78064275e-01 1.21986163e+00 -3.28636169e-01
-2.53447145e-01 5.12050502e-02 -5.14171064e-01 4.94044363e-01
7.88596630e-01 -4.55196142e-01 -1.97076783e-01 -9.20983195e-01
8.35995853e-01 -8.04073885e-02 6.27535582e-01 -1.99207163e+00
5.25461376e-01 -1.36874998e-02 3.83954078e-01 1.92370266e-01
4.72874343e-01 -7.40003228e-01 3.14028114e-02 6.79836750e-01
-3.80742669e-01 -5.48764050e-01 -2.70013124e-01 3.33405852e-01
3.46879065e-01 -1.56983554e-01 9.49991107e-01 -3.94833833e-01
-5.08430958e-01 -2.08790433e-02 -8.00768018e-01 4.65763509e-02
9.12415862e-01 -6.22456312e-01 -7.60265470e-01 2.04015970e-01
-2.50153124e-01 -6.86601460e-01 -2.75793346e-03 1.56823367e-01
6.40640616e-01 -1.01080024e+00 -3.47492486e-01 4.24247622e-01
-1.14579782e-01 -1.37321636e-01 5.01160741e-01 5.05252481e-01
-5.43909609e-01 5.29706895e-01 -5.44287026e-01 -8.18905354e-01
-5.45715272e-01 6.81306869e-02 7.79279113e-01 1.93633705e-01
-8.66532862e-01 1.11445475e+00 2.31651552e-02 -3.05351615e-01
1.61452815e-02 -5.19777119e-01 1.68939486e-01 -2.81238854e-01
3.58157665e-01 4.28013891e-01 4.20049191e-01 1.33269995e-01
-1.22969642e-01 1.06207967e+00 3.75161856e-01 1.56777546e-01
1.85879517e+00 2.62368679e-01 -3.11331987e-01 4.32694286e-01
1.08431613e+00 -1.72561482e-01 -1.13883626e+00 -2.71605879e-01
1.72660630e-02 3.63602251e-01 -6.93775341e-02 -5.61384797e-01
-1.12740195e+00 1.22760034e+00 9.53209877e-01 4.23686683e-01
9.33036089e-01 3.35218251e-01 1.10876250e+00 1.34553885e+00
7.51678705e-01 -1.36606824e+00 6.57244980e-01 3.87150884e-01
6.50288820e-01 -1.00937450e+00 -3.52395296e-01 9.63468403e-02
5.50825655e-01 1.21922934e+00 5.14353693e-01 -4.62726474e-01
1.05112207e+00 5.71394801e-01 -8.55186507e-02 -5.35478175e-01
-5.64796567e-01 5.00512600e-01 2.22047135e-01 9.93583202e-01
3.12687337e-01 2.94348300e-01 5.78353405e-01 6.70598447e-01
-3.40137035e-01 -1.91733450e-01 1.34052038e-01 7.97380209e-01
-1.10553384e+00 -7.70829201e-01 8.98487419e-02 7.23547816e-01
-5.49815297e-01 2.85564691e-01 7.01927245e-02 2.24360645e-01
2.55609542e-01 6.60251737e-01 4.43404138e-01 -5.81555188e-01
1.66796282e-01 -1.25265390e-01 1.11390626e+00 -4.24333632e-01
-4.47218627e-01 -3.57198417e-01 -3.02956283e-01 -3.10433000e-01
-7.32751131e-01 1.62180967e-03 -1.28213680e+00 -2.85559952e-01
1.79232713e-02 -1.03552127e-02 8.34559143e-01 1.25284719e+00
6.70387208e-01 4.84405071e-01 6.12649083e-01 -1.17441607e+00
-5.62814951e-01 -8.20994198e-01 -4.54548299e-01 -1.78914323e-01
6.87522054e-01 -3.88150334e-01 -3.54777463e-03 -3.85133028e-01] | [8.191153526306152, 2.4575695991516113] |
4782a326-0755-455c-944d-4594c37ae53a | neural-modular-control-for-embodied-question | 1810.11181 | null | https://arxiv.org/abs/1810.11181v2 | https://arxiv.org/pdf/1810.11181v2.pdf | Neural Modular Control for Embodied Question Answering | We present a modular approach for learning policies for navigation over long planning horizons from language input. Our hierarchical policy operates at multiple timescales, where the higher-level master policy proposes subgoals to be executed by specialized sub-policies. Our choice of subgoals is compositional and semantic, i.e. they can be sequentially combined in arbitrary orderings, and assume human-interpretable descriptions (e.g. 'exit room', 'find kitchen', 'find refrigerator', etc.). We use imitation learning to warm-start policies at each level of the hierarchy, dramatically increasing sample efficiency, followed by reinforcement learning. Independent reinforcement learning at each level of hierarchy enables sub-policies to adapt to consequences of their actions and recover from errors. Subsequent joint hierarchical training enables the master policy to adapt to the sub-policies. On the challenging EQA (Das et al., 2018) benchmark in House3D (Wu et al., 2018), requiring navigating diverse realistic indoor environments, our approach outperforms prior work by a significant margin, both in terms of navigation and question answering. | ['Stefan Lee', 'Dhruv Batra', 'Georgia Gkioxari', 'Devi Parikh', 'Abhishek Das'] | 2018-10-26 | null | null | null | null | ['embodied-question-answering'] | ['computer-vision'] | [-9.77998227e-03 3.31665277e-01 -7.47483410e-03 -4.14727539e-01
-8.53808403e-01 -1.02124858e+00 6.56891406e-01 3.41966808e-01
-6.12022340e-01 1.01315010e+00 4.28133398e-01 -8.40636015e-01
-2.23222226e-01 -9.11879003e-01 -1.14302266e+00 -4.15438741e-01
-3.42039078e-01 7.00135469e-01 4.30249512e-01 -4.48304564e-01
2.78362125e-01 1.73616707e-01 -1.51048839e+00 1.79063573e-01
7.93618858e-01 7.43442535e-01 5.23551047e-01 1.04933488e+00
-7.51107261e-02 1.02680135e+00 -3.42301697e-01 1.40794337e-01
2.32933804e-01 -1.88566372e-01 -1.28013027e+00 -5.07540926e-02
2.89239138e-01 -6.08962178e-01 -3.23898822e-01 8.04483414e-01
2.72225827e-01 6.60178483e-01 4.72997308e-01 -1.03531718e+00
-4.46706295e-01 4.88774955e-01 2.58033186e-01 -7.95036852e-02
8.10873747e-01 6.77171946e-01 1.13551176e+00 -2.19372064e-01
3.77619475e-01 1.56823409e+00 3.43167245e-01 6.02675438e-01
-1.33361518e+00 -2.38220111e-01 1.02215838e+00 1.35590807e-01
-7.88679481e-01 -3.34317058e-01 -3.14003564e-02 -3.35884064e-01
1.27069843e+00 8.76168087e-02 5.78520060e-01 1.39934576e+00
2.55052507e-01 9.32244122e-01 1.20222557e+00 -8.80241767e-02
7.83431113e-01 -2.79447854e-01 -1.37643233e-01 1.02387750e+00
-1.80594966e-01 4.12381142e-01 -3.64179939e-01 -2.95328349e-01
7.11303055e-01 -1.14596784e-01 8.32934864e-03 -5.89164436e-01
-1.28092670e+00 5.43400586e-01 5.16106248e-01 -1.91879481e-01
-5.86947322e-01 4.79868740e-01 3.59722942e-01 5.07469475e-01
-2.60420948e-01 7.97612906e-01 -1.02454495e+00 -2.48711690e-01
-2.51988411e-01 7.83333242e-01 1.04347944e+00 1.38531435e+00
7.77367115e-01 -2.36727953e-01 -4.67988282e-01 4.61336374e-01
2.66998917e-01 4.88301307e-01 3.22736561e-01 -1.65443814e+00
6.43455207e-01 1.56254321e-01 8.71026397e-01 -2.53746867e-01
-7.36184835e-01 -8.36015791e-02 -5.59386313e-01 2.67787844e-01
6.23741508e-01 -2.77151525e-01 -1.20668018e+00 1.85646701e+00
5.38982391e-01 9.86005962e-02 2.52492666e-01 7.51589894e-01
2.98900843e-01 7.34509587e-01 3.84728640e-01 1.29259393e-01
1.39284968e+00 -1.40005660e+00 -2.21419752e-01 -4.79484916e-01
6.48029625e-01 -2.03174472e-01 1.49698067e+00 4.72665161e-01
-9.75739479e-01 -6.82860136e-01 -6.67728066e-01 -1.06354013e-01
-4.18484598e-01 -2.58520246e-01 5.86715996e-01 4.42990512e-02
-1.32223117e+00 7.50229239e-01 -1.22770095e+00 -3.95602733e-01
-7.48150563e-03 2.49521300e-01 -9.53248143e-02 -9.47409943e-02
-1.26362467e+00 8.44469726e-01 5.73821723e-01 -1.11679815e-01
-1.68870461e+00 -3.78243387e-01 -9.58410561e-01 4.72995304e-02
7.08136201e-01 -9.44217920e-01 2.02341986e+00 -4.17326003e-01
-2.00411463e+00 3.53107125e-01 -2.58440763e-01 -6.67481840e-01
4.85524476e-01 -4.97768044e-01 1.13756426e-01 -4.61781509e-02
4.95463639e-01 8.93562615e-01 6.91284537e-01 -1.20542014e+00
-1.19461751e+00 -1.55075774e-01 7.82851994e-01 6.15485907e-01
4.31139916e-01 -6.36223614e-01 -4.86631960e-01 -3.25899720e-02
1.87810063e-01 -1.14999092e+00 -7.19963849e-01 -2.45619580e-01
-2.86245137e-01 -4.37979549e-01 1.43356174e-01 -5.61287880e-01
9.52198207e-01 -1.71747339e+00 3.93750638e-01 -1.84760299e-02
-4.80790250e-02 -2.39398092e-01 -1.19136535e-01 5.24899900e-01
4.10574704e-01 -2.03491971e-02 -3.68426204e-01 -4.99206126e-01
2.67360985e-01 6.97892666e-01 -4.41225171e-01 1.12036154e-01
-2.26349309e-01 9.08903956e-01 -1.49466264e+00 -1.37632534e-01
4.42616999e-01 -1.41131490e-01 -8.66190255e-01 4.07818019e-01
-9.74006832e-01 9.93674278e-01 -7.60806441e-01 6.05400980e-01
3.53866955e-03 -2.64834940e-01 2.11956337e-01 4.05816436e-01
-2.54228711e-01 8.46650958e-01 -1.10465109e+00 2.05515218e+00
-8.27706814e-01 7.46932030e-02 1.53252721e-01 -7.14607120e-01
4.09201473e-01 1.72565714e-01 1.28840521e-01 -9.67108905e-01
-3.24099600e-01 1.99872658e-01 -1.48182258e-01 -6.46477401e-01
5.44805825e-01 1.91474006e-01 -5.65484583e-01 2.61482537e-01
-2.31485330e-02 -3.23807091e-01 1.29495785e-01 5.67811690e-02
1.21018648e+00 9.51493561e-01 4.45774227e-01 -1.84935480e-01
6.42249465e-01 1.39648557e-01 4.13082153e-01 1.58217692e+00
-4.36944157e-01 6.25280440e-02 5.26988685e-01 -5.63326836e-01
-1.03555596e+00 -1.22146440e+00 3.11612725e-01 1.66187787e+00
3.29598010e-01 -2.49804094e-01 -5.37491202e-01 -9.37680304e-01
1.51277095e-01 1.16315258e+00 -6.19811475e-01 6.27623349e-02
-8.12813699e-01 -3.09233274e-02 3.31927776e-01 5.29235780e-01
5.16498268e-01 -1.41399646e+00 -1.21865642e+00 5.42002797e-01
-2.36981720e-01 -1.01610410e+00 -2.76238889e-01 5.65494776e-01
-8.62171233e-01 -9.91422534e-01 -3.27892393e-01 -6.10353470e-01
5.64866066e-01 8.09441805e-02 1.35919213e+00 1.83011778e-02
2.82791227e-01 7.14188874e-01 -3.23248565e-01 7.70789199e-03
-4.03246969e-01 2.51644135e-01 1.64579511e-01 -5.53848743e-01
-3.40077847e-01 -5.33114254e-01 -8.16511393e-01 1.57791242e-01
-4.36176777e-01 1.80420488e-01 4.97701496e-01 7.66189635e-01
7.46879816e-01 -1.88350320e-01 2.37310901e-01 -7.89869726e-01
6.97296858e-01 -5.74263096e-01 -8.88831198e-01 1.99633732e-01
-5.27266383e-01 6.26634479e-01 1.06610119e+00 -1.78138182e-01
-1.02269566e+00 -5.68182170e-02 -2.79275775e-01 1.45431571e-02
-7.00944006e-01 1.57566890e-01 -1.16805352e-01 3.48753005e-01
6.23464465e-01 5.45871079e-01 -4.11134094e-01 -4.31162834e-01
6.47435546e-01 1.93447277e-01 7.06445634e-01 -1.32885683e+00
4.80801821e-01 2.19592229e-01 -1.22845666e-02 -3.61920804e-01
-1.09238017e+00 -3.52369457e-01 -3.66003543e-01 6.44183457e-02
8.84277701e-01 -9.23143446e-01 -1.25643885e+00 2.15714037e-01
-9.54488873e-01 -1.18118167e+00 -1.23757191e-01 1.27298057e-01
-1.04386389e+00 2.10420620e-02 -4.84686285e-01 -8.38721454e-01
5.51705174e-02 -1.36869144e+00 1.23777819e+00 4.12788630e-01
-2.41276845e-01 -8.56554806e-01 -5.86943850e-02 7.27773160e-02
3.40510219e-01 7.75827393e-02 9.95149493e-01 -5.24308920e-01
-8.63854349e-01 3.33546430e-01 1.71666816e-01 4.97298203e-02
1.16791427e-01 -5.75392663e-01 -5.50875902e-01 -5.06364107e-01
-2.86619991e-01 -6.28904521e-01 6.73302352e-01 1.75126001e-01
1.39481461e+00 -7.38016367e-01 -3.33539873e-01 4.61769342e-01
1.13049018e+00 4.38278139e-01 3.22878987e-01 9.55226958e-01
3.86877984e-01 5.84438384e-01 8.39995205e-01 2.35837087e-01
1.03040075e+00 4.19106215e-01 7.46846914e-01 4.80053931e-01
3.14439893e-01 -6.52724504e-01 5.02740085e-01 1.30652472e-01
2.37077489e-01 -1.29199758e-01 -9.62204158e-01 4.40684885e-01
-2.18705535e+00 -7.34443545e-01 3.43578756e-01 2.23905444e+00
6.93152487e-01 5.69387317e-01 1.39484987e-01 -5.71691811e-01
1.37384767e-02 1.95156142e-01 -1.05905247e+00 -3.93263817e-01
5.39626360e-01 -1.23150256e-02 6.05284989e-01 1.19508696e+00
-1.19474328e+00 1.34445369e+00 6.03488398e+00 2.18921810e-01
-5.77196181e-01 -6.20066635e-02 3.90465230e-01 3.99874989e-03
-2.89930642e-01 2.69723237e-01 -8.49184155e-01 4.12245393e-01
1.03025997e+00 3.23584974e-01 9.99552310e-01 8.12785804e-01
1.65029854e-01 -4.40526426e-01 -1.32283533e+00 3.48044842e-01
-6.51937783e-01 -1.04666722e+00 -2.44951956e-02 -1.92329884e-01
6.04811966e-01 2.47818157e-01 -2.31363401e-02 1.09444475e+00
1.14947510e+00 -9.49623287e-01 9.73457396e-01 5.64527988e-01
2.08398074e-01 -7.74624884e-01 2.45478421e-01 9.90050733e-01
-1.09787440e+00 -4.20474499e-01 -1.95048064e-01 -2.64035404e-01
3.15480709e-01 -3.44449222e-01 -8.16791773e-01 5.10630846e-01
8.47242296e-01 3.15365583e-01 -2.05871552e-01 6.26607001e-01
-9.26901042e-01 4.29197550e-01 -3.90826970e-01 -3.47866118e-01
7.37381041e-01 -4.04873699e-01 4.39036876e-01 9.15557384e-01
8.40745866e-02 3.28445196e-01 7.02769399e-01 5.59287429e-01
2.68749177e-01 -3.70689988e-01 -3.76749784e-01 3.44216883e-01
6.90761507e-01 8.79115999e-01 -4.10592377e-01 -6.28345490e-01
-5.66182882e-02 1.19982207e+00 6.67742491e-01 7.31125891e-01
-8.38410020e-01 -1.69976018e-02 8.65132093e-01 -1.16881907e-01
4.55751985e-01 -5.34845412e-01 -3.67132574e-02 -9.78979111e-01
-1.19919851e-01 -1.02354932e+00 4.35097039e-01 -7.36937284e-01
-8.11180472e-01 3.16595823e-01 1.36953622e-01 -8.32617104e-01
-4.55716282e-01 -5.01952529e-01 -2.70948946e-01 8.80954683e-01
-1.63459575e+00 -7.78523982e-01 -1.33082807e-01 4.48324054e-01
8.08849156e-01 1.29599959e-01 1.10005224e+00 -3.13265294e-01
-2.95336217e-01 1.84093237e-01 -2.70295776e-02 -2.67342478e-01
3.18395257e-01 -1.72647083e+00 9.34803724e-01 6.50543094e-01
-1.47557661e-01 7.43787050e-01 8.67309451e-01 -5.97910762e-01
-1.36270022e+00 -9.75516260e-01 5.41730821e-01 -7.44257331e-01
6.93656743e-01 -4.73401845e-01 -8.41869831e-01 9.93117273e-01
9.14845020e-02 -4.12418962e-01 2.38668606e-01 3.04265559e-01
-2.68014371e-01 1.15491226e-01 -1.03511226e+00 1.04236042e+00
1.17943895e+00 -4.86629874e-01 -7.55590022e-01 4.63776916e-01
1.12727416e+00 -1.08405650e+00 -6.61798000e-01 2.85088867e-01
4.54254329e-01 -1.06005406e+00 1.06744134e+00 -9.92290318e-01
3.28611463e-01 -4.44214344e-01 -2.40815476e-01 -1.35368013e+00
-3.70064080e-01 -7.40069151e-01 -5.19475639e-01 4.89083052e-01
4.36916739e-01 -7.52016068e-01 5.55684090e-01 6.00173295e-01
-3.03858846e-01 -1.06539047e+00 -7.98376143e-01 -6.74208164e-01
3.70108299e-02 -4.69791353e-01 7.15922236e-01 2.11008251e-01
-2.85693407e-01 4.57580209e-01 -3.60012621e-01 6.13751471e-01
6.95840061e-01 1.61213636e-01 8.78934860e-01 -5.73432803e-01
-7.41494536e-01 -4.77408916e-01 5.36765516e-01 -1.95289588e+00
1.62185237e-01 -6.60265326e-01 5.67936718e-01 -1.92249286e+00
-4.79486376e-01 -6.65567160e-01 -2.95081496e-01 6.57435894e-01
-3.35105449e-01 -8.00024927e-01 2.04820201e-01 1.06310092e-01
-1.07109869e+00 6.53346419e-01 1.51835799e+00 -9.98262241e-02
-5.05173862e-01 2.92361706e-01 -7.04269111e-01 7.57340908e-01
9.31578636e-01 -2.65769899e-01 -5.92784286e-01 -6.79223120e-01
2.03856856e-01 5.74709594e-01 3.94355416e-01 -1.04221642e+00
3.60543936e-01 -4.55792278e-01 3.18653464e-01 -3.73904586e-01
2.21001223e-01 -6.08487725e-01 -4.36828911e-01 7.12449670e-01
-6.95284307e-01 2.15886652e-01 3.07828337e-01 8.70091796e-01
2.14292094e-01 3.91154364e-02 5.26646137e-01 -5.88947475e-01
-9.66745794e-01 3.63238156e-01 -6.17035389e-01 1.53558612e-01
9.05918062e-01 1.89921912e-02 -5.91124501e-03 -6.03471398e-01
-9.72576499e-01 1.19701767e+00 2.73805916e-01 5.18708050e-01
4.03749287e-01 -9.55359101e-01 -3.00828516e-01 -1.02985986e-01
9.74630378e-03 7.50537872e-01 6.12573586e-02 5.72894931e-01
-3.26728463e-01 5.94145179e-01 -5.99818192e-02 -5.79364479e-01
-5.47651291e-01 6.39349461e-01 4.89571065e-01 -6.07954502e-01
-6.68884218e-01 9.55694377e-01 1.31178424e-01 -1.03527737e+00
6.57313704e-01 -6.90063477e-01 -8.57599154e-02 -3.79957289e-01
2.12590531e-01 2.33381361e-01 -3.97105426e-01 5.43030091e-02
-3.29516292e-01 1.48434088e-01 -1.51519373e-01 -4.83913988e-01
1.00442433e+00 -4.23605561e-01 1.23943999e-01 6.50923431e-01
7.49960661e-01 -4.86024708e-01 -1.77650845e+00 -2.19956756e-01
2.37269536e-01 -1.47254705e-01 -5.26955724e-01 -1.23635101e+00
5.30944318e-02 5.88029921e-01 1.81356847e-01 2.02292055e-01
7.19680727e-01 4.66453210e-02 5.72980464e-01 1.08828223e+00
1.08223093e+00 -1.11904073e+00 1.40697733e-01 1.22009873e+00
9.29939508e-01 -1.19222665e+00 -2.43062377e-01 3.82965714e-01
-6.43816292e-01 8.64919543e-01 9.52142715e-01 -1.54069096e-01
1.06627978e-01 -1.85950577e-01 -1.21201210e-01 1.08008802e-01
-1.08156359e+00 -3.43470424e-01 -3.14959176e-02 4.92090791e-01
-4.81087677e-02 3.65869492e-01 2.16017500e-01 1.93966523e-01
-3.22083056e-01 -3.47927600e-01 2.40556464e-01 1.14436936e+00
-7.81396806e-01 -1.04525411e+00 -3.35011452e-01 3.76381010e-01
2.56166369e-01 1.20883416e-02 5.36070913e-02 7.41082370e-01
1.29413456e-01 8.17243218e-01 -1.23789534e-01 -1.36018485e-01
6.76930010e-01 7.83980414e-02 5.78740299e-01 -7.50898719e-01
-4.43165421e-01 -2.03709334e-01 2.17890024e-01 -1.18974268e+00
-2.21128855e-02 -8.64697695e-01 -1.70991635e+00 6.06465079e-02
4.78719354e-01 2.14876294e-01 3.78517181e-01 1.16182959e+00
4.41161215e-01 8.23666155e-01 4.37290579e-01 -8.56828749e-01
-1.17516208e+00 -7.75266111e-01 8.72042030e-02 1.88976064e-01
8.84939432e-01 -6.32640123e-01 -1.83239326e-01 -1.77784517e-01] | [4.226853370666504, 1.253157377243042] |
ddc34d61-a7c3-46c4-8222-7b2291223a6c | noise-blind-image-deblurring | null | null | http://openaccess.thecvf.com/content_cvpr_2017/html/Jin_Noise-Blind_Image_Deblurring_CVPR_2017_paper.html | http://openaccess.thecvf.com/content_cvpr_2017/papers/Jin_Noise-Blind_Image_Deblurring_CVPR_2017_paper.pdf | Noise-Blind Image Deblurring | We present a novel approach to noise-blind deblurring, the problem of deblurring an image with known blur, but unknown noise level. We introduce an efficient and robust solution based on a Bayesian framework using a smooth generalization of the 0-1 loss. A novel bound allows the calculation of very high-dimensional integrals in closed form. It avoids the degeneracy of Maximum a-Posteriori (MAP) estimates and leads to an effective noise-adaptive scheme. Moreover, we drastically accelerate our algorithm by using Majorization Minimization (MM) without introducing any approximation or boundary artifacts. We further speed up convergence by turning our algorithm into a neural network termed GradNet, which is highly parallelizable and can be efficiently trained. We demonstrate that our noise-blind formulation can be integrated with different priors and significantly improves existing deblurring algorithms in the noise-blind and in the known-noise case. Furthermore, GradNet leads to state-of-the-art performance across different noise levels, while retaining high computational efficiency.
| ['Paolo Favaro', 'Stefan Roth', 'Meiguang Jin'] | 2017-07-01 | null | null | null | cvpr-2017-7 | ['blind-image-deblurring'] | ['computer-vision'] | [ 2.12485030e-01 -4.56120819e-01 3.28941166e-01 6.08956395e-03
-1.04433000e+00 -3.45599294e-01 5.21800339e-01 -6.03718817e-01
-4.85675573e-01 7.30728805e-01 5.02171159e-01 -2.10946992e-01
-2.74826974e-01 -3.43962908e-01 -6.56011999e-01 -1.02694440e+00
4.29715291e-02 3.69577259e-02 1.11129120e-01 1.76694989e-01
2.68272102e-01 4.06718194e-01 -1.21256280e+00 -4.12131816e-01
9.15954709e-01 8.56389403e-01 3.91269952e-01 9.59343851e-01
5.03494859e-01 6.86335206e-01 -3.39681685e-01 -1.37285009e-01
4.91782427e-01 -4.64447737e-01 -8.07495117e-01 2.19902113e-01
8.68742168e-01 -6.93083465e-01 -7.55684078e-01 1.43489993e+00
7.95332968e-01 2.81884491e-01 6.01591587e-01 -3.92322153e-01
-9.53989446e-01 2.28877187e-01 -8.02335501e-01 3.45514834e-01
1.19782403e-01 9.30662907e-04 6.46640301e-01 -1.12796593e+00
4.16979641e-01 1.14143610e+00 1.21396005e+00 3.36911947e-01
-1.36222208e+00 -2.03745499e-01 -2.58664280e-01 3.41665447e-01
-1.48463678e+00 -8.19895029e-01 4.72833723e-01 -4.46849108e-01
5.34049332e-01 3.25918764e-01 3.40446979e-01 8.35425436e-01
-1.18582398e-02 7.50662625e-01 1.36435783e+00 -4.95011151e-01
3.20336372e-01 -5.61865568e-01 2.56319106e-01 5.36898613e-01
4.11173493e-01 2.25439131e-01 -3.69040251e-01 -3.10990304e-01
1.01888478e+00 -9.64738950e-02 -1.10645473e+00 -3.46201748e-01
-1.37319732e+00 5.92103541e-01 3.95883441e-01 2.97937661e-01
-4.78610247e-01 5.28500378e-01 4.45408188e-02 1.93873495e-01
7.20119476e-01 3.86144936e-01 -3.29642624e-01 -1.28103957e-01
-1.56152534e+00 1.84348971e-01 8.00748706e-01 5.51539361e-01
8.00796151e-01 7.06069469e-02 -2.72764117e-01 9.91577268e-01
3.47005159e-01 8.41616094e-01 2.19259351e-01 -1.45369124e+00
-8.08179826e-02 -4.79186207e-01 5.08617282e-01 -1.02568448e+00
-2.71958262e-01 -7.54836440e-01 -1.22977054e+00 2.15103880e-01
4.39509064e-01 -2.07123682e-01 -1.00654590e+00 1.68664205e+00
8.23666155e-02 6.78017199e-01 -8.58827233e-02 1.27142000e+00
3.27199847e-01 6.19662642e-01 -4.55765933e-01 -5.41294575e-01
1.25989854e+00 -1.11712086e+00 -1.05010152e+00 -3.38097602e-01
1.23759486e-01 -8.70521963e-01 5.81601381e-01 5.08174539e-01
-1.21487856e+00 -2.61005640e-01 -1.00018406e+00 -2.21698731e-01
1.38908967e-01 9.93751436e-02 4.19032961e-01 7.68466055e-01
-1.63109529e+00 9.04485703e-01 -9.53258634e-01 -2.30364159e-01
4.44092482e-01 1.93894267e-01 -1.01507351e-01 -4.08876628e-01
-9.28725660e-01 1.05381227e+00 -9.58303064e-02 3.87975007e-01
-9.97081280e-01 -8.21398258e-01 -8.66430819e-01 1.43480496e-02
3.02511245e-01 -1.08094215e+00 1.25820816e+00 -7.83289731e-01
-1.67591774e+00 6.21975541e-01 -5.09547651e-01 -4.05033499e-01
6.95058584e-01 -7.15180874e-01 -1.74602896e-01 3.07261407e-01
-3.08382064e-02 2.79273182e-01 1.39454150e+00 -1.31120372e+00
-1.86071557e-03 -1.43774509e-01 -4.01798755e-01 1.46955460e-01
-2.05751404e-01 2.02927917e-01 -6.00508094e-01 -1.07972288e+00
3.77315491e-01 -8.59689176e-01 -4.53713953e-01 2.14163303e-01
-1.90149814e-01 5.14414966e-01 5.26711464e-01 -1.20393097e+00
1.24499595e+00 -2.01468658e+00 4.11489397e-01 -2.53973827e-02
4.82079268e-01 2.67344505e-01 -1.03891253e-01 6.43029734e-02
-1.09949671e-01 -2.39407375e-01 -8.93025219e-01 -4.77435291e-01
-1.83604851e-01 -4.91960160e-02 -2.02111378e-01 9.51879919e-01
-1.70181856e-01 8.15255940e-01 -9.85728204e-01 -4.24726382e-02
2.22907260e-01 6.76154733e-01 -5.23565769e-01 2.32165575e-01
2.61082470e-01 3.50154430e-01 -1.35139097e-02 3.19936574e-01
1.19316411e+00 -4.99499738e-01 -1.13334380e-01 -4.57143843e-01
-1.62857845e-01 -1.75162464e-01 -1.30038595e+00 1.83868027e+00
-5.28665781e-01 1.04631460e+00 8.75450253e-01 -9.01552677e-01
3.93810511e-01 3.53233516e-01 2.84354627e-01 -9.07658488e-02
1.59874633e-02 4.04756606e-01 -4.58223909e-01 -4.90658373e-01
6.62822723e-01 -2.10683063e-01 4.96615827e-01 5.36614895e-01
-1.07074538e-02 -1.72523603e-01 -1.07110232e-01 4.77807447e-02
1.13961720e+00 5.29706590e-02 2.08702579e-01 -7.25025654e-01
4.42289621e-01 -3.90697271e-01 2.21906200e-01 1.14082801e+00
-1.73340112e-01 1.04577684e+00 -8.27549994e-02 -2.39309505e-01
-8.89334440e-01 -9.69125748e-01 -4.02680784e-01 7.14871109e-01
3.01548183e-01 -9.23448950e-02 -1.01229155e+00 -2.16081530e-01
-2.85331994e-01 4.77386534e-01 -3.75892162e-01 1.67915583e-01
-6.25663936e-01 -1.38002360e+00 3.72917503e-01 2.48774290e-01
7.28713632e-01 -3.44801396e-01 -2.51207799e-01 2.06718519e-01
-4.93841648e-01 -1.06955802e+00 -8.49567235e-01 -5.58631644e-02
-9.86410558e-01 -8.48517120e-01 -1.55873311e+00 -7.27653861e-01
7.13375568e-01 7.58763194e-01 9.78596032e-01 -1.41725793e-01
-1.29279181e-01 5.99090636e-01 3.71422917e-02 3.91821653e-01
-3.59550655e-01 -4.42837715e-01 2.33620897e-01 8.79538879e-02
9.30090547e-02 -6.36094093e-01 -9.23543394e-01 3.67765695e-01
-8.38362277e-01 -1.45975545e-01 4.96299565e-01 1.05059862e+00
2.94222295e-01 9.09534693e-02 2.92849541e-01 -2.44512096e-01
6.67400718e-01 -2.71239460e-01 -7.05495954e-01 1.25541747e-01
-6.54153466e-01 1.91553786e-01 2.47340113e-01 -4.40247357e-01
-1.17365050e+00 -5.53498305e-02 3.05484384e-02 -4.83756810e-01
1.46986982e-02 4.18825001e-01 1.96717367e-01 -5.79303086e-01
9.20990407e-01 3.55314910e-01 1.27103715e-03 -8.34715903e-01
8.08229864e-01 8.64639223e-01 1.03062665e+00 -5.24307966e-01
9.26367939e-01 6.34059489e-01 8.06135312e-02 -1.04925394e+00
-7.63644993e-01 -6.72100306e-01 -3.62844259e-01 -7.79712424e-02
6.11666203e-01 -1.23695529e+00 -4.82207268e-01 1.10479677e+00
-1.30741274e+00 -3.69939655e-01 2.10611075e-02 6.79877102e-01
-6.27352059e-01 1.12603772e+00 -1.03516102e+00 -7.61737883e-01
-3.67025048e-01 -1.13313079e+00 8.09478164e-01 1.45691618e-01
1.09820627e-02 -9.90646482e-01 1.13933533e-01 3.10073644e-01
9.21608567e-01 -3.37342650e-01 2.92536974e-01 -5.14801173e-03
-7.88970828e-01 3.14262733e-02 -8.18138719e-01 5.57906568e-01
1.59298003e-01 -5.63643157e-01 -1.15316343e+00 -6.07039511e-01
4.67094541e-01 -4.78759557e-02 1.35624802e+00 1.00347459e+00
9.70085323e-01 -4.37608868e-01 -1.88343793e-01 9.73324716e-01
1.51752412e+00 -5.03961265e-01 6.94346130e-01 2.73553729e-01
7.46824265e-01 2.99442321e-01 -5.93267530e-02 2.63064146e-01
1.12753324e-01 7.16563821e-01 1.62472695e-01 -1.52932853e-01
-6.03654861e-01 3.84811759e-01 3.12983453e-01 8.58473778e-01
-1.90334573e-01 -1.33531451e-01 -7.36394346e-01 8.62521112e-01
-1.84830236e+00 -8.85626674e-01 -3.82496089e-01 2.22727013e+00
1.24336362e+00 -4.80779022e-01 -3.49674582e-01 -1.22016266e-01
9.67748344e-01 3.79889160e-01 -3.51163685e-01 2.60869652e-01
-2.89969414e-01 2.40787894e-01 8.20473135e-01 1.08981371e+00
-1.44455147e+00 8.21013689e-01 7.61964655e+00 9.10678089e-01
-7.65966296e-01 4.24025357e-01 3.73825938e-01 1.45482376e-01
-8.09318125e-02 -1.74815834e-01 -3.55873436e-01 3.99936080e-01
7.47714877e-01 -1.12410523e-02 8.89118612e-01 5.38152754e-01
4.91991132e-01 -2.18719423e-01 -7.81780541e-01 1.29255080e+00
1.94727048e-01 -1.33773470e+00 -3.71459305e-01 -2.56406870e-02
9.86698925e-01 3.59230727e-01 -6.25824779e-02 -4.66980577e-01
4.09218520e-01 -9.10061002e-01 5.87542653e-01 6.74190462e-01
8.54836345e-01 -2.12820396e-01 6.51244223e-01 3.09954584e-02
-6.76171422e-01 8.72590765e-02 -3.71537209e-01 -4.69398014e-02
4.56697404e-01 1.23771298e+00 7.61020603e-03 5.45554578e-01
7.15975821e-01 9.22210097e-01 -9.70806628e-02 1.54199696e+00
-2.16120213e-01 6.04212046e-01 -4.03009593e-01 4.83985513e-01
-2.03370117e-02 -3.75850916e-01 9.91341710e-01 1.42589867e+00
6.65605962e-01 1.69893846e-01 -2.25862458e-01 8.43055367e-01
-1.00225382e-01 -3.37130964e-01 -8.49019364e-02 4.37353045e-01
2.38187417e-01 9.18514907e-01 -4.70548272e-01 -3.76799822e-01
-3.12407047e-01 1.41502237e+00 -7.76964054e-02 8.61578345e-01
-3.52373391e-01 -4.00026381e-01 7.43560195e-01 -1.64289996e-01
5.96242189e-01 -4.67104375e-01 -3.51906508e-01 -1.54395461e+00
3.77053134e-02 -8.90881538e-01 -7.81527907e-02 -9.85939145e-01
-1.54863179e+00 4.22850400e-01 -9.72213894e-02 -1.10180056e+00
2.80703623e-02 -6.59289658e-01 -3.73025328e-01 1.17518520e+00
-1.89208269e+00 -7.71174133e-01 -4.01644230e-01 4.22593653e-01
3.70273560e-01 2.37473398e-01 5.83942294e-01 3.91337067e-01
-4.54737037e-01 3.04491550e-01 6.19583189e-01 -1.38051465e-01
9.33229685e-01 -1.22928786e+00 4.64822352e-01 1.50885797e+00
-2.74335682e-01 7.60377884e-01 1.02532327e+00 -5.34168243e-01
-1.28354895e+00 -9.43527579e-01 6.47117496e-01 -2.79033244e-01
8.81573021e-01 9.30761695e-02 -1.09580839e+00 4.30166841e-01
3.03826064e-01 2.16926098e-01 2.47470945e-01 -1.56643331e-01
-4.28935438e-01 1.89529195e-01 -1.01817465e+00 5.01042783e-01
1.02377594e+00 -5.83139896e-01 -6.63334310e-01 3.70123178e-01
4.89493340e-01 -5.08804679e-01 -6.52579069e-01 2.15932488e-01
3.94032031e-01 -9.60615396e-01 1.22956598e+00 -7.79585838e-02
3.65933567e-01 -5.47317982e-01 -1.28489226e-01 -1.42251480e+00
-7.55783737e-01 -1.23160124e+00 -5.59657574e-01 6.81878328e-01
6.88631088e-02 -6.40144587e-01 3.66879702e-01 4.05847073e-01
-2.16218576e-01 -1.94108143e-01 -9.60244238e-01 -8.05254221e-01
-3.76391374e-02 -4.60463077e-01 1.18915409e-01 9.21440721e-01
-2.36553684e-01 -1.63593456e-01 -9.93407607e-01 5.11045992e-01
1.36225855e+00 -9.72274691e-02 2.51772523e-01 -9.72333312e-01
-5.79297066e-01 -5.35044491e-01 -2.26935133e-01 -1.88444662e+00
-5.77246360e-02 -5.72819114e-01 4.83310163e-01 -1.54432690e+00
4.05424625e-01 -1.19638331e-01 -7.69400597e-02 2.45758072e-01
-4.81767029e-01 5.42936563e-01 -1.15888588e-01 6.28604650e-01
-3.42387259e-01 5.28509200e-01 1.22834623e+00 -1.14242718e-01
-1.04369270e-02 -9.16558057e-02 -6.20950520e-01 7.83016503e-01
3.75566036e-01 -4.95987803e-01 1.24215141e-01 -7.78925538e-01
1.55243322e-01 1.89193618e-02 6.31400049e-01 -8.66671622e-01
5.19868255e-01 1.93574697e-01 3.07812363e-01 -2.13086978e-01
2.30983466e-01 -5.89303672e-01 9.46718827e-02 2.64361709e-01
-9.11004990e-02 -4.92898673e-01 1.09028921e-01 8.19671214e-01
-1.20726191e-01 -4.04838592e-01 1.10063064e+00 -4.98355646e-03
-4.53333050e-01 9.62159187e-02 -5.44718444e-01 -1.08067058e-01
2.76387364e-01 -5.63921407e-02 -4.94827956e-01 -6.54905796e-01
-6.83329284e-01 -1.28463730e-01 6.87760293e-01 -7.78540298e-02
4.22911078e-01 -1.18454766e+00 -9.63062048e-01 1.02601424e-01
-3.68680239e-01 -2.23838076e-01 4.14687306e-01 1.16287792e+00
-7.61544645e-01 1.45750977e-02 2.18777210e-01 -5.34653068e-01
-9.05170321e-01 4.13347304e-01 6.79298460e-01 -2.34968066e-02
-6.98484600e-01 9.33773220e-01 1.62050679e-01 -6.08941279e-02
1.42316431e-01 -1.04883642e-04 1.99375659e-01 -3.23304832e-01
9.48162198e-01 8.08960259e-01 -1.09552080e-02 -6.49664044e-01
-1.69669196e-01 9.11185205e-01 1.42252564e-01 -3.49712014e-01
1.27610469e+00 -5.57460725e-01 -6.66721702e-01 -1.87320456e-01
1.18320811e+00 3.37535828e-01 -1.59615958e+00 -5.00321031e-01
-1.27772644e-01 -6.45890653e-01 8.31522882e-01 -7.30757296e-01
-1.09726262e+00 6.75795436e-01 7.71614671e-01 3.81748863e-02
1.27064502e+00 -1.63577542e-01 8.32561612e-01 3.44272226e-01
3.47295366e-02 -7.02561677e-01 -1.31454438e-01 5.57662845e-01
9.28560734e-01 -1.12066734e+00 2.96174496e-01 -3.69780898e-01
1.24200106e-01 1.15906346e+00 -9.96069089e-02 -6.00007316e-03
8.58227015e-01 3.50784600e-01 2.45967224e-01 5.60522564e-02
-5.51086478e-03 -1.83593124e-01 4.41622674e-01 7.00016081e-01
1.67551324e-01 -1.25844017e-01 -3.01938951e-01 2.14334771e-01
2.64613688e-01 2.08137825e-01 7.60568142e-01 5.60942411e-01
-5.85898817e-01 -7.79793203e-01 -7.76265919e-01 1.42166048e-01
-6.44240260e-01 -6.31165206e-01 2.42435440e-01 5.48144244e-02
-3.19165289e-01 1.10578549e+00 -1.57379240e-01 6.60461634e-02
-2.35551417e-01 -2.22147822e-01 6.33170843e-01 -1.64316401e-01
-1.36039453e-02 3.97571981e-01 -5.86241931e-02 -5.56355476e-01
-7.89702475e-01 -5.76907337e-01 -5.06159723e-01 -4.00217891e-01
-6.08641624e-01 6.67392612e-02 5.14507890e-01 9.58037138e-01
4.85520571e-01 1.21730722e-01 5.13877451e-01 -1.33385324e+00
-7.12560534e-01 -1.13823044e+00 -5.22385180e-01 7.97151551e-02
9.86186564e-01 -3.34031016e-01 -9.10833716e-01 3.60547692e-01] | [11.664325714111328, -2.642866849899292] |
544aba5d-cd86-4582-87ef-1d3109d21990 | single-anchor-uwb-localization-using-channel | 2211.04246 | null | https://arxiv.org/abs/2211.04246v1 | https://arxiv.org/pdf/2211.04246v1.pdf | Single-anchor UWB Localization using Channel Impulse Response Distributions | Ultra-wideband (UWB) devices are widely used in indoor localization scenarios. Single-anchor UWB localization shows advantages because of its simple system setup compared to conventional two-way ranging (TWR) and trilateration localization methods. In this work, we focus on single-anchor UWB localization methods that learn statistical features of the channel impulse response (CIR) in different location areas using a Gaussian mixture model (GMM). We show that by learning the joint distributions of the amplitudes of different delay components, we achieve a more accurate location estimate compared to considering each delay bin independently. Moreover, we develop a similarity metric between sets of CIRs. With this set-based similarity metric, we can further improve the estimation performance, compared to treating each snapshot separately. We showcase the advantages of the proposed methods in multiple application scenarios. | ['Andreas Burg', 'Alexios Balatsoukas-Stimming', 'Sitian Li'] | 2022-11-08 | null | null | null | null | ['indoor-localization'] | ['computer-vision'] | [-1.82943732e-01 -5.62264025e-01 8.34225118e-02 -3.31671029e-01
-1.38768935e+00 -5.19511700e-01 3.48003715e-01 1.62367791e-01
-4.29049075e-01 9.61211145e-01 -7.78844506e-02 -3.74067128e-01
-4.37283337e-01 -6.67558789e-01 -4.20250744e-01 -8.84234488e-01
-4.17468488e-01 9.89334732e-02 -1.38641670e-01 1.74843237e-01
-3.33814099e-02 4.16165411e-01 -6.17364526e-01 -6.34716392e-01
7.73975551e-01 1.25096738e+00 1.32412076e-01 7.16993392e-01
2.20110208e-01 3.03322464e-01 -1.12287104e+00 3.57123241e-02
1.01314701e-01 -1.51726931e-01 -1.76118705e-02 -4.52589989e-01
3.50652565e-03 -2.33117789e-01 -6.11410797e-01 6.69148028e-01
8.55918348e-01 3.69203806e-01 6.58118188e-01 -1.22789562e+00
-1.57000095e-01 6.36903584e-01 -8.24184716e-01 -2.35639904e-02
6.10530734e-01 -8.14792275e-01 4.92697954e-01 -5.49135566e-01
-3.82235311e-02 9.09789860e-01 1.08975410e+00 7.79290730e-03
-8.73560846e-01 -8.83217335e-01 -1.99515000e-01 1.32440403e-01
-2.04793835e+00 -5.17580330e-01 6.71988189e-01 1.12880662e-01
1.91496104e-01 1.39504775e-01 7.63949752e-02 1.15723896e+00
2.78527111e-01 3.99789631e-01 7.97425926e-01 -4.83590990e-01
4.68818933e-01 -2.07112357e-01 1.93044528e-01 4.18000817e-01
6.31262481e-01 -3.80702585e-01 -3.23738068e-01 -4.67100501e-01
4.95027959e-01 8.70021153e-03 -4.18647945e-01 -8.75423774e-02
-1.04136598e+00 3.75481635e-01 3.72825891e-01 7.42304504e-01
-3.05148929e-01 7.65945554e-01 -2.56050438e-01 -4.87683006e-02
2.01638997e-01 1.02827027e-01 -1.37454629e-01 -2.17761591e-01
-1.21959305e+00 -2.78338671e-01 9.06464994e-01 1.12670553e+00
7.54625738e-01 2.60858625e-01 -1.10140823e-01 8.05212379e-01
7.44802296e-01 1.23343480e+00 9.90006924e-02 -8.39662910e-01
3.73423755e-01 -6.91794455e-01 4.87780064e-01 -1.05833888e+00
-5.21999836e-01 -8.26263964e-01 -9.62971091e-01 -6.08010054e-01
6.05566680e-01 -7.92234600e-01 -7.57883370e-01 1.63217342e+00
1.00647017e-01 7.81580985e-01 2.77079910e-01 3.99658442e-01
3.31103384e-01 8.07492375e-01 -1.97179541e-01 -7.77947009e-02
1.07988000e+00 -4.93808508e-01 -8.69493961e-01 -1.92077190e-01
4.09314573e-01 -9.62630928e-01 -6.98624030e-02 4.43581522e-01
-5.93570709e-01 -5.21483123e-01 -1.24273050e+00 5.33876359e-01
-2.55818605e-01 2.48953953e-01 3.93532336e-01 1.27068007e+00
-8.94310653e-01 1.66752309e-01 -8.86744440e-01 -2.61123836e-01
4.11889739e-02 2.23989323e-01 -2.41988271e-01 -5.68230033e-01
-1.08646679e+00 6.09999180e-01 -1.45634651e-01 2.96448976e-01
-3.87455612e-01 -6.94262981e-01 -1.07010520e+00 1.31947011e-01
8.75182003e-02 -2.26311073e-01 1.23776758e+00 5.42590506e-02
-1.20634079e+00 -2.61470884e-01 -5.81566036e-01 -5.42165458e-01
2.86780775e-01 -1.51127309e-01 -7.94055283e-01 5.42467088e-02
2.30067596e-01 -9.34107974e-02 6.42324924e-01 -1.40682006e+00
-5.45747519e-01 -1.25648320e-01 -2.71776289e-01 -4.80783194e-01
-1.51404947e-01 -5.41775286e-01 -2.57076502e-01 -5.06199956e-01
5.21002889e-01 -8.38864803e-01 -4.52799708e-01 -2.81083643e-01
-4.00988728e-01 3.80520910e-01 6.84840798e-01 -5.78724980e-01
1.14562809e+00 -2.25284839e+00 -5.53708792e-01 7.58368075e-01
-2.66917735e-01 -1.30965471e-01 -5.51376231e-02 6.30950451e-01
5.19977927e-01 -2.09634736e-01 8.06081966e-02 -7.15145767e-01
4.24240157e-02 1.29475757e-01 -2.71154910e-01 9.55407560e-01
-5.84979117e-01 1.94948584e-01 -1.05041373e+00 -3.66404116e-01
3.05768847e-01 8.88339639e-01 -3.99164334e-02 -2.16326445e-01
6.33772016e-01 7.14220047e-01 -6.41257942e-01 5.59639454e-01
1.19210649e+00 2.13747293e-01 3.05649132e-01 -3.65886927e-01
-2.28355572e-01 -2.79473394e-01 -1.42868674e+00 1.73316467e+00
-1.36558759e+00 7.07794785e-01 1.91757470e-01 -9.63288188e-01
1.16376519e+00 5.27877629e-01 5.37993908e-01 -6.99301422e-01
2.40610823e-01 5.53702593e-01 -9.25444961e-02 2.07522027e-02
1.46247178e-01 2.43947595e-01 -5.54007053e-01 4.06321794e-01
2.85367548e-01 1.83433518e-01 1.08788069e-02 7.67491460e-02
1.27699924e+00 -3.13818038e-01 5.21261215e-01 -2.44524479e-01
5.53071082e-01 -7.81737924e-01 4.35725629e-01 1.31070089e+00
1.11088514e-01 4.04233038e-01 -2.18950450e-01 8.58831480e-02
-4.61600631e-01 -1.33740902e+00 -2.98999637e-01 4.43286568e-01
5.05890369e-01 -3.12883288e-01 -2.96907246e-01 -1.98030606e-01
1.62062362e-01 7.78453469e-01 -2.82770991e-01 -1.64964776e-02
-5.50939858e-01 -5.62795758e-01 1.09962678e+00 3.35003912e-01
7.62679577e-01 3.01254958e-01 -3.32266122e-01 4.12325710e-01
-4.17147815e-01 -1.48075557e+00 -2.72257179e-01 3.04870248e-01
-3.48564237e-01 -5.95699489e-01 -1.02536535e+00 -4.14680392e-01
3.87580872e-01 9.77900028e-01 3.87819618e-01 -2.50859678e-01
-3.43080372e-01 1.04037988e+00 -4.46480572e-01 -3.27592313e-01
2.14207873e-01 -7.09079131e-02 4.41808850e-01 2.71486878e-01
-2.87983537e-01 -7.58686364e-01 -7.03871489e-01 6.12221658e-01
-3.92864913e-01 -8.37516248e-01 5.84855735e-01 3.99693072e-01
1.30004644e-01 3.47089887e-01 6.93268657e-01 -1.38764396e-01
4.33207273e-01 -8.05630028e-01 -6.51580095e-01 4.65954781e-01
-1.79715663e-01 -1.10257603e-01 5.43023169e-01 -4.22997713e-01
-1.32785511e+00 -1.44551113e-01 -1.79624096e-01 2.07832128e-01
-1.92670867e-01 4.30866957e-01 -3.21996510e-01 -6.52921617e-01
2.67796397e-01 2.31897056e-01 -6.55960441e-01 -3.85343730e-01
4.72342998e-01 9.13057864e-01 5.77915311e-01 -9.08699751e-01
1.09015751e+00 6.32702529e-01 3.81198615e-01 -1.39968562e+00
-5.08992314e-01 -8.79736960e-01 -4.48126465e-01 -3.46290767e-02
2.50604540e-01 -1.00108242e+00 -6.70040250e-01 2.51069695e-01
-1.15596223e+00 -3.80394645e-02 4.29878056e-01 9.50222731e-01
-1.29626289e-01 6.64607167e-01 -3.17737073e-01 -1.40623748e+00
3.81962471e-02 -7.47849584e-01 9.67046618e-01 4.16973919e-01
-2.08895460e-01 -1.18160689e+00 2.60091841e-01 -1.58603713e-01
7.90073216e-01 1.83267668e-01 2.61804879e-01 -5.93540609e-01
-2.71984786e-01 -7.36560643e-01 -1.50835171e-01 -1.54579237e-01
3.79144579e-01 -2.95973927e-01 -9.07637119e-01 -2.50509977e-01
-1.10202134e-01 3.84219080e-01 4.84300703e-01 8.89982402e-01
7.88117528e-01 -6.18605912e-02 -9.10175323e-01 5.84610999e-01
1.51436090e+00 4.64544922e-01 6.68476224e-01 8.48086476e-02
3.48583937e-01 -2.71133274e-01 9.14622009e-01 7.71680176e-01
3.45015049e-01 8.47882748e-01 3.58188637e-02 -9.44450796e-02
1.17151245e-01 1.21192992e-01 -5.36024123e-02 2.62642950e-01
-1.02012485e-01 -6.38412178e-01 -7.93254137e-01 3.45015556e-01
-1.72648823e+00 -7.45780110e-01 -2.26984754e-01 2.32401395e+00
1.00330010e-01 -1.38489172e-01 -2.52673417e-01 -2.18481384e-02
6.61836565e-01 2.81939596e-01 2.47476045e-02 1.80598523e-03
1.17005870e-01 4.18588817e-01 1.05400777e+00 6.59978867e-01
-9.97252762e-01 2.01610968e-01 5.70957947e+00 1.00051844e+00
-9.17947471e-01 2.67267853e-01 1.10592477e-01 4.99164075e-01
-4.71306443e-02 -1.94954917e-01 -7.31209159e-01 4.69302446e-01
1.04716003e+00 -7.59767219e-02 3.86638870e-03 6.75198913e-01
1.83828905e-01 -6.37339532e-01 -9.67598498e-01 1.21491420e+00
1.24720640e-01 -8.91655147e-01 -6.30876422e-01 2.12503731e-01
5.59054136e-01 -2.85042316e-01 2.28511631e-01 1.48816466e-01
1.13099873e-01 -6.14684939e-01 6.10066414e-01 6.28147960e-01
4.73802745e-01 -7.46486783e-01 1.04906821e+00 1.05506398e-01
-1.66322148e+00 -1.85812786e-01 -4.33594525e-01 9.99284461e-02
7.14510441e-01 1.21941841e+00 -8.27330291e-01 1.12613261e+00
3.43215555e-01 4.61555310e-02 -2.64036715e-01 1.92440450e+00
-2.32909158e-01 5.59882998e-01 -6.43937171e-01 -1.38696786e-02
2.02155843e-01 -1.02845080e-01 3.37280512e-01 1.35795343e+00
1.34387481e+00 2.89844926e-02 8.71004090e-02 3.15608829e-01
1.12380445e-01 -2.74919748e-01 -5.68560600e-01 6.48755372e-01
1.15714407e+00 1.30580771e+00 -7.84904480e-01 -3.72196804e-03
-3.67942363e-01 7.73662686e-01 -6.31116569e-01 6.85921967e-01
-1.12734842e+00 -9.14466739e-01 4.29053128e-01 -2.17260674e-01
4.38806713e-01 -9.57341790e-01 -4.62656952e-02 -9.38462377e-01
-4.30013567e-01 -7.69304335e-02 6.28907084e-02 -3.79096478e-01
-8.82058084e-01 2.50927031e-01 7.71294907e-02 -1.41323602e+00
-1.21695563e-01 -2.37939313e-01 -8.13280761e-01 7.05470860e-01
-1.41358209e+00 -1.15374374e+00 -3.83669585e-01 4.61859703e-01
1.91989154e-01 3.47856849e-01 1.02992570e+00 7.73162186e-01
-3.19720775e-01 8.18095148e-01 8.21789861e-01 3.12474400e-01
8.29445243e-01 -1.06822109e+00 1.02224305e-01 8.00824404e-01
3.07290226e-01 9.40283298e-01 7.74286270e-01 -4.09985214e-01
-1.27351236e+00 -8.24360907e-01 4.92232859e-01 6.07695468e-02
6.07158363e-01 -3.76321554e-01 -3.53695035e-01 5.03000140e-01
9.77043658e-02 1.63562939e-01 1.21277857e+00 2.09321767e-01
-3.23464364e-01 -5.01663506e-01 -1.27356315e+00 2.25395069e-01
4.44891691e-01 -1.33172780e-01 -8.86715949e-02 -3.16516566e-03
1.11155048e-01 -1.13060907e-01 -9.20821071e-01 2.56690085e-01
1.07709563e+00 -4.42804843e-01 1.44671893e+00 4.57069486e-01
-6.93628728e-01 -4.09996927e-01 -6.90949857e-01 -1.35701227e+00
-1.01677120e-01 -6.00292385e-01 -2.77433813e-01 1.66279161e+00
3.02522063e-01 -8.98410499e-01 2.66817629e-01 1.80539757e-01
1.58189356e-01 -1.81668885e-02 -1.30952656e+00 -1.14379406e+00
-4.57357138e-01 -6.06519341e-01 5.81392407e-01 4.83236939e-01
-1.06687412e-01 7.12909549e-02 -6.92362487e-01 1.04304051e+00
1.30084932e+00 -1.96591556e-01 6.89005077e-01 -1.24838853e+00
-3.76121998e-01 2.17704177e-01 -4.78655457e-01 -1.27819991e+00
-2.57410035e-02 -1.69489890e-01 5.14959633e-01 -1.63718522e+00
-3.99872094e-01 -9.09053981e-01 -4.94030803e-01 4.69875373e-02
2.47855440e-01 5.02279818e-01 -3.28389965e-02 -2.78127491e-01
-9.63208616e-01 3.38377148e-01 3.67391855e-01 -1.00634962e-01
-4.38739620e-02 7.51646996e-01 -3.49373907e-01 6.99445188e-01
9.18356836e-01 -4.55659688e-01 -1.70713991e-01 -4.72438991e-01
-2.71411330e-01 3.23419869e-01 2.37493485e-01 -1.68936193e+00
4.20511663e-01 2.55311340e-01 7.77934730e-01 -5.93137741e-01
6.79712951e-01 -1.12123299e+00 2.48463929e-01 5.53483248e-01
2.49087423e-01 -5.88479042e-01 3.74132879e-02 9.42069650e-01
6.10112399e-02 -3.74339968e-01 6.05286539e-01 2.05809668e-01
-6.33075893e-01 7.23688602e-02 -7.52849102e-01 -6.04141235e-01
9.61703002e-01 -1.08190048e-02 -2.22137496e-01 -1.07936990e+00
-4.38766927e-01 1.35244951e-01 -1.06044769e-01 -1.01569764e-01
5.01838803e-01 -1.39150858e+00 -2.17886522e-01 -3.04199845e-01
-1.02839313e-01 -3.23374391e-01 4.56574500e-01 1.09152424e+00
-2.02167541e-01 6.92931473e-01 1.95512608e-01 -5.42648792e-01
-1.16919506e+00 3.73629737e-03 1.16540253e-01 3.75213521e-03
-1.15139885e-02 6.46892011e-01 -2.93615818e-01 -2.05779821e-01
4.60635692e-01 -1.00893632e-01 -1.00118453e-02 -7.70342350e-02
7.20723212e-01 8.04667175e-01 -3.48497555e-02 -5.99684358e-01
-7.09904552e-01 1.03288805e+00 4.86073822e-01 -4.32543069e-01
1.07591951e+00 -5.14136851e-01 9.22614709e-02 2.23659217e-01
1.25033510e+00 6.20363533e-01 -8.33139598e-01 -1.99521333e-01
-4.77739461e-02 -6.17545664e-01 -1.53222650e-01 -4.26388055e-01
-5.50594151e-01 7.06549406e-01 8.58621716e-01 2.50973552e-01
7.90833950e-01 -6.32309020e-02 8.49834800e-01 6.90436840e-01
1.18327296e+00 -6.57223225e-01 -8.47322941e-02 3.06536555e-01
9.86831188e-02 -8.57716620e-01 4.57307836e-03 -2.57657826e-01
1.90189838e-01 1.27745211e+00 -1.24458328e-03 -3.30180600e-02
8.82190883e-01 5.11344194e-01 1.77843437e-01 5.48717558e-01
2.89197385e-01 -3.15514594e-01 1.22328568e-02 1.11139107e+00
3.17232370e-01 3.66058876e-03 -2.53470480e-01 7.79617310e-01
7.11259842e-02 -3.29924941e-01 5.34595609e-01 9.96436179e-01
-6.90758288e-01 -1.30189908e+00 -1.01877034e+00 1.38810247e-01
-4.94622916e-01 2.45082647e-01 5.77385783e-01 7.98361897e-01
-7.59216724e-03 1.51389372e+00 -2.49025479e-01 -2.38119632e-01
2.05824673e-01 -2.20880479e-01 6.81858540e-01 -1.93635806e-01
6.32246196e-01 3.70223016e-01 1.63202435e-02 -2.89758533e-01
-2.58968621e-01 -4.61162150e-01 -1.22853684e+00 -2.04740316e-01
-4.94503081e-01 7.53774941e-01 1.16395116e+00 9.31708634e-01
4.87209111e-02 6.70396626e-01 7.88334370e-01 -1.09201086e+00
-3.21205378e-01 -7.44367063e-01 -9.05611694e-01 -4.68543321e-01
4.76404190e-01 -7.24396467e-01 -3.16446066e-01 -4.62304801e-01] | [6.3155598640441895, 1.048790693283081] |
9a3dd177-4c61-4664-89dc-c8283625f513 | android-malware-detection-using-autoencoder | 1901.07315 | null | http://arxiv.org/abs/1901.07315v1 | http://arxiv.org/pdf/1901.07315v1.pdf | Android Malware Detection Using Autoencoder | Smartphones have become an intrinsic part of human's life. The smartphone
unifies diverse advanced characteristics. It enables users to store various
data such as photos, health data, credential bank data, and personal
information. The Android operating system is the prevalent mobile operating
system and, in the meantime, the most targeted operating system by malware
developers. Recently the unparalleled development of Android malware put
pressure on researchers to propose effective methods to suppress the spread of
the malware. In this paper, we propose a deep learning approach for Android
malware detection. The proposed approach investigates five different feature
sets and applies Autoencoder to identify malware. The experimental results show
that the proposed approach can identify malware with high accuracy. | ['Abdelmonim Naway', 'Yuancheng Li'] | 2019-01-14 | null | null | null | null | ['android-malware-detection'] | ['miscellaneous'] | [-2.29543261e-02 -3.46259296e-01 -5.91886878e-01 1.09632201e-01
8.14393014e-02 -3.08655977e-01 7.54086077e-01 1.09021841e-02
-3.74218315e-01 5.53158700e-01 -1.48245737e-01 -6.00044549e-01
1.35443166e-01 -6.68549478e-01 -4.79064316e-01 -6.28326416e-01
1.78873479e-01 -4.96433526e-02 1.72490761e-01 -2.27103010e-01
5.08114219e-01 4.33751374e-01 -1.67058837e+00 7.96653107e-02
7.38894403e-01 1.13727164e+00 1.93994820e-01 4.26118881e-01
-1.72291115e-01 7.20600486e-01 -9.99685168e-01 -3.14135313e-01
-4.26614322e-02 -1.74213618e-01 -3.71880352e-01 -2.87277639e-01
-3.09754193e-01 -6.49739385e-01 -5.01296878e-01 1.20419669e+00
1.57778859e-01 -2.88549095e-01 4.08511192e-01 -1.32938385e+00
-6.63798630e-01 2.85818636e-01 -6.60205901e-01 4.73917782e-01
3.22975844e-01 2.10459437e-03 3.54864508e-01 -3.91665965e-01
2.16310218e-01 8.18630219e-01 5.96349478e-01 3.68947417e-01
-4.66480821e-01 -6.40394092e-01 -2.13892177e-01 3.96599114e-01
-1.34709489e+00 -3.45910072e-01 6.74436450e-01 -5.59158802e-01
9.07198906e-01 2.12262601e-01 8.19726646e-01 1.43853176e+00
9.37735081e-01 4.09291834e-01 9.62156117e-01 -2.76046902e-01
1.18416764e-01 5.77548921e-01 3.16026419e-01 7.71244228e-01
6.82263494e-01 2.50835344e-02 -2.19351817e-02 -3.88568610e-01
2.52440184e-01 6.63826227e-01 -1.66951790e-01 1.61227226e-01
-6.75729692e-01 8.90438855e-01 4.65260930e-02 7.04751134e-01
-3.16268802e-01 -1.37146175e-01 5.77175260e-01 -2.27201078e-03
2.51705557e-01 3.10263205e-02 -4.81658816e-01 -5.52233875e-01
-4.94107217e-01 -2.57446855e-01 8.12982917e-01 2.42684305e-01
3.98900896e-01 2.23245159e-01 7.12493002e-01 6.51338577e-01
5.17595470e-01 5.42905390e-01 1.22141683e+00 -5.52584887e-01
-1.08550593e-01 9.18873727e-01 -2.33147934e-01 -1.31939387e+00
-6.93975240e-02 -2.67288268e-01 -9.33282495e-01 -2.42460027e-01
-1.68987550e-02 -8.35258886e-02 -3.86238217e-01 1.18370569e+00
2.61724323e-01 3.00119966e-01 -1.22628085e-01 1.07906580e-01
5.85384846e-01 8.83794427e-01 5.32847345e-02 -1.76511303e-01
1.27840173e+00 -4.16397691e-01 -7.62593448e-01 2.80967616e-02
2.74103016e-01 -5.09248197e-01 1.02945924e+00 5.38638055e-01
-3.49929601e-01 -4.48090971e-01 -1.18265748e+00 3.07090342e-01
-8.17629158e-01 1.16642185e-01 5.86123407e-01 1.08837748e+00
-7.89102256e-01 3.98475498e-01 -1.03800917e+00 -3.94760907e-01
6.32793903e-01 4.66246784e-01 -2.56599456e-01 2.96303302e-01
-1.02932465e+00 6.75429642e-01 4.37611848e-01 -7.53939897e-02
-7.30525792e-01 5.44121377e-02 -6.97305262e-01 2.07326204e-01
4.29585576e-01 -3.16589683e-01 1.08836579e+00 -8.81259322e-01
-1.50027823e+00 7.07092524e-01 1.04662813e-01 -6.69466376e-01
-1.42056808e-01 -3.01456511e-01 -8.53742421e-01 -4.79851440e-02
-1.03222199e-01 5.55728339e-02 1.50791287e+00 -7.63161063e-01
-5.62072039e-01 -5.97047865e-01 -3.66038829e-02 -3.07299972e-01
-1.00236368e+00 3.95020805e-02 -2.05902502e-01 -3.03688169e-01
-5.85153401e-01 -1.02203381e+00 2.29130924e-01 -8.99661124e-01
-4.01627570e-01 -1.32486284e-01 1.64638007e+00 -7.95578718e-01
1.45554531e+00 -2.28642035e+00 -2.46882644e-02 3.95951159e-02
3.52891713e-01 6.85034096e-01 5.75863361e-01 2.50015080e-01
7.42946118e-02 1.52844772e-01 -1.23459257e-01 -8.99839923e-02
-2.24931151e-01 2.37907127e-01 -4.04441595e-01 3.34536344e-01
-1.87216505e-01 6.31789923e-01 -4.55907136e-01 -3.41731578e-01
2.34012663e-01 8.33782792e-01 -1.77440777e-01 4.56275083e-02
1.90625221e-01 4.03331846e-01 -6.58065557e-01 7.89334953e-01
4.15090352e-01 -4.86093760e-01 2.18838841e-01 1.36573419e-01
-4.64322418e-03 1.23065293e-01 -5.48116863e-01 7.79591978e-01
-4.31034356e-01 7.89545894e-01 5.56742698e-02 -9.88158882e-01
5.53297579e-01 2.36090627e-02 4.91547674e-01 -4.33310777e-01
8.14616740e-01 2.05926318e-03 2.30691582e-01 -6.98700070e-01
4.72984582e-01 3.32375735e-01 -5.87332509e-02 3.79728079e-01
-3.32728028e-01 3.41562957e-01 -1.53645635e-01 1.95767969e-01
1.01164401e+00 -3.00421923e-01 7.94802010e-01 4.56758477e-02
1.04383087e+00 -2.58006454e-01 2.51592189e-01 1.91385701e-01
-5.40155947e-01 -4.18708861e-01 4.13399011e-01 -4.38620627e-01
-4.59278435e-01 -7.19386160e-01 -5.59127890e-02 9.32470381e-01
1.03205226e-01 -3.55964035e-01 -1.19439507e+00 -9.38767731e-01
-1.11191742e-01 3.04772764e-01 -4.14524734e-01 -4.23726469e-01
-5.55583060e-01 -5.01235306e-01 6.77949727e-01 7.87079930e-02
8.44705701e-01 -1.14427233e+00 -8.76291633e-01 -1.50622562e-01
5.39969839e-02 -1.10447371e+00 -2.27978960e-01 -9.38662589e-02
-7.94676244e-01 -1.23386526e+00 -3.66103888e-01 -7.36222863e-01
3.92756492e-01 3.04865152e-01 5.42604029e-01 3.78552616e-01
-2.40613997e-01 3.43530893e-01 -2.06918612e-01 -4.73877311e-01
-6.93285167e-01 2.76982278e-01 4.75378007e-01 2.39306107e-01
5.73720276e-01 -3.60542297e-01 -2.98242807e-01 3.35581638e-02
-9.17619705e-01 -5.82232058e-01 5.39022803e-01 5.23698449e-01
3.25689286e-01 7.74859071e-01 5.84726393e-01 -7.21444607e-01
9.28278506e-01 -8.13908637e-01 -6.11745596e-01 -1.92500085e-01
-6.14227653e-01 -3.78036618e-01 9.23394144e-01 -4.46241915e-01
-8.29108953e-01 -1.29133716e-01 -4.31949049e-01 -1.10896356e-01
-4.37026620e-01 2.99705863e-01 -4.22523767e-01 -2.60220557e-01
3.04825723e-01 4.97626722e-01 4.09745753e-01 -3.87756228e-01
-1.12989098e-01 1.19600427e+00 3.01981121e-01 1.40530482e-01
4.73716050e-01 4.49488819e-01 -2.29090124e-01 -1.57045233e+00
-1.78934723e-01 -1.76302761e-01 -8.82230699e-02 -2.86633998e-01
7.72833884e-01 -5.58346272e-01 -1.16060269e+00 7.61711895e-01
-1.02625811e+00 5.05113266e-02 2.66439825e-01 2.12251678e-01
2.31394414e-02 6.57173812e-01 -6.53810382e-01 -7.17093766e-01
-3.06030840e-01 -1.40015888e+00 8.00589085e-01 3.18990916e-01
-1.45463087e-02 -8.91622603e-01 1.05096422e-01 4.05995399e-01
6.16922081e-01 1.36790454e-01 7.21417785e-01 -7.33868182e-01
-3.99784476e-01 -5.06954670e-01 -8.47808197e-02 6.45576835e-01
5.28217614e-01 1.78788573e-01 -8.31542313e-01 -1.18747331e-01
5.96667349e-01 1.91722691e-01 5.68399012e-01 4.48177159e-01
1.53753710e+00 -5.40758669e-01 -7.41618156e-01 5.46030939e-01
1.07505655e+00 7.69705832e-01 7.32906044e-01 2.55957782e-01
8.20084870e-01 4.62604612e-01 4.27771837e-01 3.23340744e-01
4.25806403e-01 3.75732541e-01 7.84227908e-01 4.25702482e-01
4.53158259e-01 -3.40339132e-02 5.23114979e-01 9.50120926e-01
6.25312105e-02 -1.46540970e-01 -8.84416699e-01 6.83917403e-02
-1.26032817e+00 -1.06585205e+00 2.99155731e-02 2.14483428e+00
2.58698076e-01 3.49893123e-01 2.58448780e-01 3.70477498e-01
7.69873619e-01 1.05941877e-01 -4.93375480e-01 -4.19055820e-01
3.31353545e-01 1.95962608e-01 3.99890989e-01 3.33729714e-01
-1.29663599e+00 7.77724981e-01 6.18010998e+00 6.90108716e-01
-1.48252404e+00 4.57728684e-01 6.62271142e-01 2.87801057e-01
2.69196272e-01 -5.30721962e-01 -8.03541541e-01 1.09053981e+00
1.36439335e+00 -6.62380531e-02 3.99462849e-01 1.22923315e+00
1.03645705e-01 -2.97906220e-01 -4.03188378e-01 1.22214043e+00
2.90359557e-01 -1.23068595e+00 -1.88605532e-01 7.22946167e-01
2.97391683e-01 -8.54889452e-02 6.18314147e-01 3.70181233e-01
-3.56700808e-01 -9.02785540e-01 -2.11311597e-02 3.45856398e-01
6.06822491e-01 -9.03710604e-01 8.34203362e-01 5.13875425e-01
-8.77788126e-01 -6.02482855e-01 5.89006245e-02 -3.15408915e-01
3.61482762e-02 4.79110181e-01 -8.37191880e-01 -5.15346676e-02
8.36173117e-01 8.14893603e-01 -7.26462483e-01 4.28137153e-01
1.80528581e-01 7.75393069e-01 -4.78750542e-02 -4.09295201e-01
-1.90441664e-02 -3.66568565e-01 4.60194647e-01 8.13507915e-01
4.15476948e-01 -3.13174754e-01 -1.06540538e-01 3.02459121e-01
-1.68260559e-02 2.84781624e-02 -1.23596835e+00 -7.16048181e-01
3.86912763e-01 1.17763698e+00 -8.14745665e-01 -3.74737233e-01
-3.62643480e-01 9.75642741e-01 -1.64605051e-01 -2.91601643e-02
-1.10896254e+00 -5.35074472e-01 8.01171124e-01 3.83493483e-01
1.64110795e-01 -3.18797678e-01 -1.06831446e-01 -1.04030490e+00
-2.69825578e-01 -1.19796860e+00 -1.65819714e-03 -2.67391831e-01
-6.94602907e-01 7.01583207e-01 -2.29239359e-01 -9.90716457e-01
-4.24073130e-01 -9.01797295e-01 -4.88784313e-01 2.15903208e-01
-9.45286691e-01 -7.01874316e-01 -2.74691224e-01 6.29772902e-01
4.07327741e-01 -8.07371259e-01 8.13566029e-01 4.61261928e-01
-1.03689837e+00 2.84013689e-01 3.01411867e-01 3.04624550e-02
2.90901273e-01 -8.11641097e-01 2.12523460e-01 7.74617076e-01
1.66030210e-02 1.13485813e+00 3.71566415e-01 -8.59534919e-01
-1.47088754e+00 -7.63562620e-01 5.48033714e-01 -4.60456103e-01
6.67074680e-01 -1.40357152e-01 -8.51190150e-01 6.74746275e-01
3.63773882e-01 -4.07143474e-01 8.23004484e-01 -3.84642392e-01
-1.17499620e-01 -3.93673539e-01 -1.25097764e+00 4.13337827e-01
2.94881582e-01 -6.61874413e-01 -3.41171205e-01 1.43069118e-01
6.62043154e-01 -1.15024552e-01 -5.44224083e-01 -1.59559455e-02
6.72750771e-01 -1.33507204e+00 7.80339301e-01 -8.84712711e-02
2.75328010e-01 -3.65013182e-02 -1.00061759e-01 -8.22896600e-01
1.93117559e-01 -5.64247072e-01 -9.00311172e-01 9.18098629e-01
-1.87364593e-01 -1.10281706e+00 9.27564979e-01 6.13536919e-03
1.45412877e-01 -9.12655532e-01 -8.42875600e-01 -5.39431870e-01
-5.16547680e-01 -2.84363687e-01 7.58285761e-01 9.14659142e-01
-2.29561880e-01 3.51341695e-01 -3.03913832e-01 -4.08483185e-02
4.66132641e-01 -3.96991074e-01 7.25255609e-01 -1.42016935e+00
-5.53814709e-01 -6.28255427e-01 -4.84425664e-01 -7.53858924e-01
3.14871252e-01 -5.04237533e-01 -5.35512388e-01 -8.49992216e-01
1.66075304e-01 -2.35703476e-02 -2.43893623e-01 2.31481463e-01
2.10604742e-01 2.27534220e-01 -1.74024969e-01 1.55934036e-01
-4.23119426e-01 3.42943072e-01 5.43945968e-01 -1.56835094e-01
-3.20519716e-01 2.86081374e-01 -7.83023715e-01 1.04784822e+00
1.14148688e+00 -1.91548064e-01 -4.55458283e-01 2.09382195e-02
7.78304636e-02 -2.89625943e-01 2.64547437e-01 -1.05016041e+00
-1.65635034e-01 1.71160903e-02 3.03547651e-01 -5.04347384e-01
3.90836418e-01 -1.14381838e+00 3.62478867e-02 1.05806947e+00
3.32190841e-01 1.61830008e-01 1.67475432e-01 7.68615067e-01
-1.47428706e-01 -2.14882568e-02 8.30853105e-01 1.28117010e-01
-7.23871350e-01 1.85128689e-01 -7.57946134e-01 -4.54027385e-01
1.28354657e+00 -4.23472226e-01 -2.94122159e-01 -2.82727629e-01
-2.60463595e-01 -4.38038111e-01 5.24602771e-01 6.92334116e-01
7.77213633e-01 -1.05408728e+00 1.12191141e-01 4.03993964e-01
-2.64997125e-01 -6.97240591e-01 2.25797981e-01 8.31007659e-01
-5.99155068e-01 5.37832618e-01 -3.76621425e-01 -4.10830587e-01
-1.58008945e+00 7.31504381e-01 8.15448985e-02 -1.62628904e-01
-3.45516652e-01 4.30661440e-01 -8.40981603e-02 -3.57815661e-02
2.27006063e-01 -1.10763073e-01 -6.26522303e-01 -1.30302966e-01
9.47873652e-01 6.96335495e-01 -8.28814227e-03 -1.06272936e+00
-5.73262632e-01 2.80393839e-01 -3.24448079e-01 3.30591023e-01
1.03970861e+00 8.10395274e-03 -3.57922286e-01 3.09942782e-01
1.30214310e+00 3.00664812e-01 -6.79441631e-01 5.58742046e-01
1.77582782e-02 -4.77930844e-01 1.09652774e-02 -5.19313931e-01
-9.94168222e-01 8.44074547e-01 8.59110415e-01 9.01785374e-01
1.11325932e+00 -3.40738416e-01 1.24027073e+00 3.77812177e-01
5.45078695e-01 -9.24063206e-01 2.26121470e-01 4.89858806e-01
4.05175269e-01 -1.29021037e+00 -1.62172377e-01 -1.31970853e-01
-4.54209954e-01 7.65717387e-01 5.34079432e-01 2.43648142e-02
9.65484858e-01 3.84637803e-01 -2.82085329e-01 -9.21106413e-02
-1.11501507e-01 2.07044840e-01 2.59231776e-02 8.52729201e-01
3.63548398e-01 9.46644992e-02 -2.58518934e-01 6.94284260e-01
-6.78330883e-02 -1.45392463e-01 5.13966680e-01 8.38707149e-01
-7.10339069e-01 -1.13179350e+00 -4.87966090e-01 7.40597546e-01
-1.07112586e+00 1.00913860e-01 -6.43708646e-01 7.82137156e-01
3.74969453e-01 9.64579701e-01 -7.20195547e-02 -7.23690808e-01
-2.77033269e-01 9.63501483e-02 9.47819948e-02 -3.42857391e-01
-3.84546906e-01 -4.20263559e-01 -4.22345489e-01 -4.60033864e-01
-6.45491993e-04 -3.25091571e-01 -1.12196910e+00 -5.11221647e-01
-1.91138491e-01 3.23106125e-02 1.05868936e+00 8.84678602e-01
7.27893174e-01 3.94452602e-01 7.15529025e-01 -6.70274854e-01
-4.88040894e-01 -9.18509126e-01 -3.35183501e-01 3.33422348e-02
4.44471687e-01 -7.36359596e-01 -2.49817714e-01 -4.45115864e-02] | [14.426189422607422, 9.682908058166504] |
08db6033-c2a6-45d0-be5d-cb939bffb281 | the-nni-query-by-example-system-for-mediaeval | null | null | http://www.npu-aslp.org/lxie/papers/2014QUESST-NNI.pdf | http://www.npu-aslp.org/lxie/papers/2014QUESST-NNI.pdf | The NNI Query-by-Example System for MediaEval 2014 | In this paper we describe the system proposed by NNI (NWPU-NTU-I2R) team for the QUESST task within the Mediaeval 2014 evaluation. To solve the problem, we used both dynamic time warping (DTW) and symbolic search (SS) based approaches. The DTW system performs template matching using subsequence DTW algorithm and posterior representations. The symbolic search is performed on phone sequences generated by phone recognizers. For both symbolic and DTW search, partial sequence matching is performed to reduce missing rate, especially for query type 2 and 3. After fusing 9 DTW systems, 7 symbolic systems, and query length side information, we obtained 0.6023 actual normalized cross entropy (actCnxe) for all queries combined. For type 3 complex queries, we achieved 0.7252 actCnxe. | ['Haizhou Li', 'Eng Siong Chng', 'Bin Ma', 'Su Jun Leow', 'Lei Wang', 'Hang Lv', 'JIA YU', 'Hongjie Chen', 'Cheung-Chi Leung', 'Lei Xie', 'Xiong Xiao', 'HaiHua Xu', 'Peng Yang'] | 2014-10-16 | null | null | null | null | ['template-matching'] | ['computer-vision'] | [ 5.31822324e-01 -5.06446123e-01 -4.55056787e-01 -2.24359959e-01
-1.21378112e+00 -1.05567002e+00 6.45679533e-01 -1.09545738e-01
-7.99174249e-01 5.76468349e-01 2.50005722e-01 -4.25042123e-01
-3.68781269e-01 -5.83199978e-01 -4.70516741e-01 -2.06966221e-01
-1.00363925e-01 5.31068563e-01 5.18740714e-01 -2.92989463e-01
3.59431803e-01 6.13944948e-01 -1.67414880e+00 5.01506805e-01
5.41808784e-01 1.17475951e+00 6.28900453e-02 9.19936180e-01
-1.72715932e-01 3.67057651e-01 -6.37978017e-01 -4.95899081e-01
3.76692593e-01 -1.15279719e-01 -6.05461538e-01 -9.22458827e-01
7.33361468e-02 4.08852436e-02 -4.00200546e-01 6.65346265e-01
8.75104785e-01 6.25517309e-01 5.32054245e-01 -1.17335665e+00
-5.96149638e-02 5.76034844e-01 -1.87545136e-01 4.40475106e-01
8.55853319e-01 6.92322999e-02 1.15968335e+00 -1.10041213e+00
8.58354807e-01 1.27598405e+00 7.55142033e-01 9.96157229e-02
-1.13916779e+00 -8.81566405e-01 -2.97672749e-01 7.55278945e-01
-1.75759125e+00 -3.04502457e-01 3.04881990e-01 -2.13464707e-01
1.72040224e+00 8.55466485e-01 4.30033028e-01 1.46171927e+00
2.02029478e-02 7.85409272e-01 7.57306218e-01 -4.04275060e-01
7.28132352e-02 -3.36286634e-01 2.52188444e-01 1.00319549e-01
-3.42171013e-01 5.08426964e-01 -7.83017397e-01 -6.21360183e-01
1.42022461e-01 -1.63005978e-01 -3.97113897e-02 3.34002048e-01
-1.36023068e+00 8.00441146e-01 -2.01107293e-01 3.72301668e-01
-3.97540152e-01 -1.52352944e-01 5.38070321e-01 4.03667212e-01
4.27463278e-02 4.30772007e-01 -6.49932802e-01 -8.89815390e-01
-1.22063804e+00 3.89703125e-01 8.82021606e-01 1.03681314e+00
4.01784569e-01 -1.88967273e-01 -7.64297426e-01 9.61925268e-01
-4.21782658e-02 8.23688686e-01 8.90019000e-01 -9.53106940e-01
6.43033326e-01 3.00251812e-01 9.14689228e-02 -9.34800625e-01
-2.10577965e-01 -2.17780828e-01 -4.29844767e-01 -5.40005267e-01
3.56666833e-01 5.79996146e-02 -1.03566611e+00 1.50745595e+00
-8.17311257e-02 2.49582738e-01 7.55736530e-02 3.83816242e-01
4.11292911e-01 8.68126750e-01 -2.30494246e-01 -2.71794587e-01
1.25651598e+00 -8.16753864e-01 -9.31001067e-01 1.50027916e-01
7.69622803e-01 -1.26699281e+00 9.18305337e-01 3.77339721e-01
-1.03661656e+00 -4.99635458e-01 -1.15852988e+00 7.22081363e-02
-6.37129128e-01 -6.70963377e-02 -1.47628084e-01 6.43315017e-01
-8.75722706e-01 6.52396202e-01 -7.84341872e-01 -2.93111891e-01
-2.91684985e-01 4.69674438e-01 3.91632877e-02 2.77822204e-02
-1.67383611e+00 8.63286376e-01 4.72755402e-01 -3.45228612e-02
-4.66076374e-01 -6.70261443e-01 -6.00241959e-01 5.45468405e-02
5.09270668e-01 -1.19085081e-01 1.31655371e+00 -2.41028033e-02
-1.47278976e+00 5.61427414e-01 -6.77245200e-01 -8.33433270e-01
5.66457689e-01 -2.49548405e-01 -9.55597997e-01 -1.26291923e-02
-2.62664437e-01 3.07235450e-01 6.34890437e-01 -5.28325319e-01
-3.34229618e-01 -1.42175658e-02 -7.09258616e-01 7.21258223e-02
-4.45507951e-02 2.84157842e-01 -7.42679834e-01 -8.72958243e-01
-3.40682245e-03 -1.08342004e+00 7.70189166e-02 -7.43221939e-01
-4.91466552e-01 -2.19959274e-01 8.51328611e-01 -1.14876831e+00
2.00163198e+00 -2.18368578e+00 -1.01889350e-01 9.67010498e-01
-3.65108937e-01 5.53350270e-01 -5.22358119e-01 8.65389109e-01
-1.02988236e-01 4.66274209e-02 -2.16216341e-01 -2.12746993e-01
7.34646842e-02 3.49107414e-01 -7.44024932e-01 -8.24249610e-02
-2.10435882e-01 1.16795397e+00 -5.36242187e-01 -4.27576661e-01
-3.49532701e-02 3.53115290e-01 -4.84683037e-01 1.42574459e-01
-2.15243205e-01 -2.69913226e-01 -1.80099547e-01 7.78637111e-01
4.85257238e-01 2.75804192e-01 3.38203818e-01 -1.85344204e-01
-2.12157056e-01 4.29739833e-01 -1.12381840e+00 1.53877282e+00
-5.75879037e-01 5.47303081e-01 -4.11761343e-01 -5.76225400e-01
9.79926765e-01 3.08662802e-01 6.90094531e-01 -9.95070100e-01
-1.93482772e-01 4.42027628e-01 -7.34843537e-02 -5.64310849e-01
7.03356862e-01 5.49081028e-01 -9.92878154e-02 4.50532079e-01
-1.13573126e-01 2.77576279e-02 1.34827569e-01 -3.38765830e-02
1.35131752e+00 -7.67158763e-03 1.69048712e-01 1.99771419e-01
5.03506958e-01 -7.87161440e-02 4.67293531e-01 8.93901765e-01
-8.52955598e-03 6.59184158e-01 1.28687471e-01 -9.73334536e-02
-9.71540093e-01 -1.06040907e+00 9.78731811e-02 1.07998610e+00
-2.30660096e-01 -6.67693138e-01 -4.45187300e-01 -4.09506381e-01
1.34511784e-01 1.08542681e+00 -2.84567684e-01 -1.82705849e-01
-8.66592348e-01 -3.82795423e-01 1.16541731e+00 3.05804312e-01
2.95835376e-01 -1.21646750e+00 -6.61100447e-01 4.54365402e-01
-6.67800665e-01 -1.23542535e+00 -1.09947968e+00 1.01622254e-01
-5.49547791e-01 -7.53813088e-01 -6.95754111e-01 -2.65457869e-01
-3.29131871e-01 4.96167094e-02 8.32583487e-01 -3.11030954e-01
-4.96152043e-01 2.08506376e-01 -5.58920622e-01 -9.33993682e-02
-2.47097015e-01 3.70058984e-01 1.07721835e-01 -1.62609503e-01
4.87353384e-01 -4.80498821e-01 -3.92111361e-01 8.30736995e-01
-7.95034230e-01 -5.98677099e-01 3.82189065e-01 7.41387427e-01
7.65280485e-01 -3.30508262e-01 1.54149145e-01 -4.16278899e-01
7.58667171e-01 -3.63797069e-01 -8.36292624e-01 7.47184455e-01
-8.65792692e-01 1.94697171e-01 9.61872861e-02 -9.19292629e-01
-7.75482714e-01 -1.34218827e-01 -5.09906769e-01 -5.62160254e-01
2.21882015e-01 7.28301227e-01 1.26142234e-01 1.80639565e-01
5.78894138e-01 5.31334460e-01 -8.83315280e-02 -5.38551807e-01
2.86424235e-02 9.24695134e-01 6.67318523e-01 -2.82310456e-01
5.49563706e-01 5.06416336e-02 -2.98348188e-01 -8.11896682e-01
-3.47243041e-01 -5.79634428e-01 -2.12967098e-01 6.26373962e-02
8.28798831e-01 -6.20560408e-01 -9.45666790e-01 3.73423576e-01
-1.11499763e+00 -3.41398001e-01 -1.83877289e-01 6.91922486e-01
-5.24395287e-01 4.91042525e-01 -2.29031205e-01 -1.11525559e+00
-5.20129144e-01 -1.10831618e+00 9.59550083e-01 -2.24074438e-01
-7.95910001e-01 -6.48585856e-01 3.38725269e-01 2.97488034e-01
6.93528652e-01 6.91597983e-02 8.08642089e-01 -1.37505817e+00
-2.87497759e-01 -5.38895011e-01 3.71908247e-02 1.70268156e-02
-2.31782675e-01 1.15905344e-01 -9.15023088e-01 -7.39506632e-02
8.93703625e-02 7.77560622e-02 6.63079739e-01 1.99838012e-01
1.18069410e+00 -4.25982356e-01 -4.08390015e-01 5.51124632e-01
9.98565733e-01 7.33846426e-01 7.43663549e-01 2.25005537e-01
2.55815804e-01 4.83414918e-01 1.00995052e+00 5.57937801e-01
1.46507591e-01 1.45301282e+00 -3.44018154e-02 5.17349243e-01
9.77412760e-02 -3.56718123e-01 6.40093386e-01 6.27351344e-01
2.52074040e-02 -4.36091334e-01 -1.20048499e+00 2.75974035e-01
-1.79940689e+00 -1.15109301e+00 5.48563600e-02 2.38348150e+00
8.24490845e-01 2.29665190e-01 1.04389496e-01 1.54021636e-01
6.14443898e-01 6.92465454e-02 -5.67917645e-01 -6.21431708e-01
-3.08731824e-01 4.84616756e-01 6.35407090e-01 7.02343524e-01
-5.19386053e-01 7.88397968e-01 6.46908569e+00 1.58159375e+00
-6.93999052e-01 9.27592143e-02 7.50286132e-02 -4.09840107e-01
-2.92308003e-01 -1.86960205e-01 -9.90750730e-01 6.18474722e-01
1.54170132e+00 -2.98875123e-01 8.69664907e-01 4.02074069e-01
-1.30877215e-02 -2.61639953e-01 -8.34908545e-01 1.21589482e+00
-1.09074853e-01 -1.24063694e+00 -2.40148783e-01 1.13515146e-01
4.03385460e-01 3.70231658e-01 1.60313979e-01 5.44183195e-01
3.47442068e-02 -9.59890485e-01 4.67417270e-01 6.70736492e-01
1.02772474e+00 -7.92708218e-01 7.88055062e-01 3.95715714e-01
-1.35387814e+00 1.85456276e-02 1.50621116e-01 4.93825644e-01
3.47008586e-01 4.49925363e-01 -9.01579857e-01 8.63229275e-01
6.01606250e-01 3.52812827e-01 -3.46760720e-01 8.74217212e-01
2.22986892e-01 6.49843276e-01 -8.31121564e-01 -1.10076135e-02
1.03634298e-02 -2.35876381e-01 7.42662668e-01 1.36217189e+00
7.99500227e-01 -4.58704680e-02 -9.66960266e-02 3.19553107e-01
5.68813309e-02 -5.10439277e-03 -5.06265163e-01 -2.07324058e-01
1.21474886e+00 9.52919304e-01 -5.07957935e-01 -4.17160839e-01
-3.57794911e-02 1.09379053e+00 -2.76858211e-01 4.87149715e-01
-9.95020509e-01 -7.32266307e-01 6.44211054e-01 -1.99400052e-01
4.94183838e-01 -4.48203504e-01 -6.54487833e-02 -1.01920009e+00
2.57877391e-02 -1.03807962e+00 6.52360678e-01 -4.77484107e-01
-8.37877929e-01 6.01673663e-01 2.54260421e-01 -1.15040088e+00
-9.60896075e-01 -3.22483659e-01 -4.96163785e-01 1.09375727e+00
-1.10073936e+00 -5.21694839e-01 2.38931537e-01 7.24966168e-01
5.09537399e-01 -3.47123474e-01 1.01950443e+00 5.38980663e-01
-4.07660067e-01 1.18435144e+00 2.76479244e-01 -1.64890096e-01
6.92178249e-01 -6.99848711e-01 8.13807786e-01 5.70469439e-01
3.35067004e-01 8.08847845e-01 6.68723822e-01 -9.01289284e-01
-1.48451114e+00 -8.11115682e-01 1.47707343e+00 -4.77160245e-01
7.12127864e-01 -4.43897933e-01 -8.85352671e-01 3.76033157e-01
-2.21510436e-02 -4.79988009e-01 6.56374633e-01 -1.01508237e-02
-5.79753995e-01 1.73345253e-01 -1.13385975e+00 5.00957429e-01
1.04228413e+00 -8.92423689e-01 -4.21560735e-01 2.26491898e-01
1.04084802e+00 -5.27059019e-01 -1.11415994e+00 6.55957580e-01
9.36339796e-01 -7.69272983e-01 1.26722932e+00 -1.88404456e-01
-3.35913956e-01 -2.80805547e-02 -6.42117918e-01 -7.22406745e-01
4.58703727e-01 -1.01780891e+00 -4.40215886e-01 1.02139866e+00
8.04800987e-01 -8.02825153e-01 6.59064770e-01 4.58783358e-01
5.51602803e-02 -7.87289023e-01 -1.23335433e+00 -1.42456913e+00
-4.84725595e-01 -9.95395184e-01 7.83374786e-01 6.86275125e-01
-1.77594945e-01 -4.41759899e-02 -3.30806524e-01 -1.85396507e-01
1.68137610e-01 9.42632779e-02 3.72029960e-01 -7.30870485e-01
-4.87792343e-01 -3.63476306e-01 -5.37749715e-02 -7.60832787e-01
6.26649335e-02 -9.33168709e-01 -2.73434997e-01 -8.29617202e-01
-2.71117449e-01 2.02571973e-01 -2.18592599e-01 4.70824569e-01
2.58459359e-01 -7.84888044e-02 2.97699541e-01 3.21441770e-01
-2.35404640e-01 4.86951053e-01 5.52806854e-01 1.43759280e-01
-5.28822124e-01 3.34567815e-01 2.16726795e-01 1.88384876e-01
8.96522760e-01 -5.30430377e-01 -1.99259415e-01 -5.46167083e-02
2.65609115e-01 8.98752138e-02 6.07143342e-02 -8.23447347e-01
3.06443721e-01 4.18377742e-02 3.24707814e-02 -1.40323710e+00
6.29477620e-01 -7.54766762e-01 5.44736564e-01 7.91783333e-01
-3.78684253e-01 5.72304964e-01 2.54525810e-01 3.87760758e-01
-3.64246488e-01 -1.24178052e-01 2.85403818e-01 4.59610522e-01
-3.26795191e-01 3.53350379e-02 -5.65843463e-01 4.75022867e-02
7.14297771e-01 -3.31523567e-01 -7.82946572e-02 -4.87558454e-01
-8.50804687e-01 5.42039216e-01 -1.61872786e-02 4.62905079e-01
8.03808331e-01 -1.39106500e+00 -3.41197550e-01 1.98879614e-01
1.51115909e-01 -6.88133657e-01 2.11864203e-01 8.68471026e-01
-1.69797286e-01 7.30641365e-01 1.73773065e-01 -3.32711130e-01
-1.53266299e+00 4.92463350e-01 4.37852107e-02 -6.77123547e-01
-8.12137052e-02 7.06855714e-01 -8.86926830e-01 -4.76244748e-01
6.15705311e-01 -3.17097247e-01 -5.44023374e-03 5.86959422e-01
3.58910561e-01 8.94471467e-01 3.63730818e-01 -3.45938623e-01
-5.72274387e-01 6.78551733e-01 -1.19517431e-01 -8.63032460e-01
1.03206444e+00 -7.17547759e-02 1.86535239e-01 3.60827059e-01
1.63106644e+00 -1.06397560e-02 -5.11441529e-01 -4.77555335e-01
6.33347392e-01 -2.37548873e-01 -5.22391200e-01 -8.31843734e-01
-4.08686012e-01 7.99696028e-01 8.06131542e-01 2.38929048e-01
1.16088879e+00 -3.22536319e-01 1.18500268e+00 7.14735091e-01
1.79728031e-01 -1.25869274e+00 -1.92498446e-01 1.14118505e+00
9.03800130e-01 -6.87442243e-01 -2.51642048e-01 -9.24118683e-02
-7.32267618e-01 9.52251256e-01 7.21950233e-02 3.33314210e-01
5.06697476e-01 1.83727369e-01 -1.64556861e-01 1.25998482e-01
-9.29447234e-01 -2.24972010e-01 5.23125529e-01 3.21609765e-01
1.76005200e-01 -9.08384994e-02 -3.68105859e-01 5.55071056e-01
-5.12327850e-01 -2.76171733e-02 -9.58355814e-02 8.46344352e-01
-1.32030159e-01 -1.49776149e+00 -3.54811341e-01 3.52302074e-01
-4.86770421e-01 -3.48072171e-01 -4.58407193e-01 7.24496484e-01
-3.49413574e-01 8.95270288e-01 8.15469678e-03 -1.00402176e+00
6.53013885e-01 5.14703751e-01 2.10302293e-01 -1.87381729e-01
-9.17616546e-01 2.51196742e-01 3.22752833e-01 -9.98161018e-01
4.32330668e-02 -8.76872241e-01 -1.23334908e+00 -1.34451106e-01
-1.84846058e-01 1.56423077e-01 8.05474162e-01 6.16270661e-01
6.63747072e-01 2.27687940e-01 7.24675059e-01 -4.16438937e-01
-7.39084423e-01 -1.05776155e+00 -2.25741103e-01 5.29706888e-02
9.90974531e-02 -3.41882885e-01 -2.86684811e-01 -2.77941734e-01] | [14.301230430603027, 6.448639869689941] |
56149a83-5c92-4192-a9f0-5108ebae38e1 | evolutionary-deep-nets-for-non-intrusive-load | 2303.03538 | null | https://arxiv.org/abs/2303.03538v1 | https://arxiv.org/pdf/2303.03538v1.pdf | Evolutionary Deep Nets for Non-Intrusive Load Monitoring | Non-Intrusive Load Monitoring (NILM) is an energy efficiency technique to track electricity consumption of an individual appliance in a household by one aggregated single, such as building level meter readings. The goal of NILM is to disaggregate the appliance from the aggregated singles by computational method. In this work, deep learning approaches are implemented to operate the desegregations. Deep neural networks, convolutional neural networks, and recurrent neural networks are employed for this operation. Additionally, sparse evolutionary training is applied to accelerate training efficiency of each deep learning model. UK-Dale dataset is used for this work. | ['Kenneth A. Loparo', 'Jinsong Wang'] | 2023-03-06 | null | null | null | null | ['non-intrusive-load-monitoring', 'non-intrusive-load-monitoring', 'non-intrusive-load-monitoring'] | ['knowledge-base', 'miscellaneous', 'time-series'] | [-5.47226295e-02 -4.31222767e-02 5.36635071e-02 -6.53885186e-01
-4.07607317e-01 -7.72928447e-02 4.20240641e-01 -2.64222831e-01
7.34109357e-02 9.02086496e-01 2.16739908e-01 -5.93062155e-02
-8.52527469e-02 -1.18552220e+00 -2.12711811e-01 -9.90820527e-01
-2.95444112e-02 3.91582757e-01 -5.94422281e-01 1.44105712e-02
-1.74109131e-01 4.37295109e-01 -1.25801194e+00 1.15592115e-01
9.20677841e-01 1.33702195e+00 1.04640290e-01 7.47573495e-01
2.52203315e-01 1.32347596e+00 -9.19766784e-01 4.28562239e-03
8.08813702e-03 -4.82519597e-01 -6.48118556e-01 -1.22602537e-01
-3.77976805e-01 -9.60789084e-01 -4.23575670e-01 9.79210079e-01
8.36420000e-01 2.06775397e-01 6.85961545e-01 -1.47993886e+00
-6.50000632e-01 1.15703344e+00 -4.37180519e-01 3.05715621e-01
2.08480373e-01 1.12647295e-01 5.59829354e-01 -5.65258190e-02
-4.01755929e-01 7.97039807e-01 8.65126133e-01 2.06783656e-02
-1.06156838e+00 -7.19490588e-01 -1.68891788e-01 6.26078725e-01
-1.38637447e+00 -4.08093452e-01 1.10878980e+00 7.05733448e-02
1.87740803e+00 5.43009818e-01 8.69388700e-01 1.06748128e+00
4.83029783e-01 9.76033568e-01 8.43745172e-01 -1.54749826e-01
5.62165141e-01 -1.28275603e-01 1.83567420e-01 1.77807212e-01
3.72552335e-01 1.05420671e-01 -1.21149801e-01 -1.69588491e-01
1.66172132e-01 3.87627006e-01 6.57597780e-02 4.91280019e-01
-3.99935544e-01 9.50635552e-01 5.76903105e-01 5.70884645e-01
-9.51783776e-01 3.91686082e-01 8.29503715e-01 2.22279564e-01
3.93110424e-01 3.57346863e-01 -5.70028067e-01 -3.20799500e-01
-1.07262027e+00 -1.41406432e-01 9.52623904e-01 8.10345709e-01
6.32083714e-01 8.54104102e-01 -1.41563937e-01 4.73686010e-01
2.41303831e-01 2.79181927e-01 1.19600713e+00 -8.95653129e-01
6.13429882e-02 9.17948008e-01 -1.51976392e-01 -6.10514820e-01
-8.90222311e-01 -3.00539546e-02 -1.73218262e+00 8.81707519e-02
-6.91133440e-01 -5.17262220e-01 -6.79668963e-01 1.36063409e+00
2.52027899e-01 2.73862869e-01 1.65792942e-01 1.96877405e-01
7.03417003e-01 9.71204937e-01 2.50989377e-01 -4.60859060e-01
1.20773923e+00 -8.26979876e-01 -1.09975684e+00 1.80442169e-01
6.10786498e-01 -1.75124183e-01 4.29110616e-01 2.57213444e-01
-9.79004264e-01 -5.26665032e-01 -1.19534099e+00 -1.06745541e-01
-9.09657955e-01 2.14141309e-01 5.61204195e-01 6.15486324e-01
-7.34597027e-01 9.07006204e-01 -1.11690545e+00 -1.28789008e-01
7.45034695e-01 6.35516286e-01 2.01050073e-01 5.61158955e-01
-1.24876809e+00 9.94010150e-01 1.01294577e+00 4.32497680e-01
-5.04004717e-01 -6.53165638e-01 -8.66975307e-01 3.88151556e-01
-4.57111388e-01 -3.82301897e-01 1.35118258e+00 -7.69975126e-01
-1.92571557e+00 3.22443604e-01 3.65262806e-01 -1.10513675e+00
3.99428308e-01 -2.16322303e-01 -7.41373241e-01 -3.26964289e-01
-1.90768257e-01 1.84482321e-01 7.06469774e-01 -6.45923316e-01
-8.70649517e-01 -4.45109367e-01 -5.79770029e-01 -2.64903512e-02
-3.63801450e-01 -4.70332026e-01 5.17588615e-01 -5.72522938e-01
-3.67235243e-01 -5.73106527e-01 -3.52279544e-02 -1.09760880e+00
-6.93944931e-01 -8.75856578e-01 1.47553349e+00 -1.27328610e+00
1.35744190e+00 -1.86121011e+00 -4.00554836e-01 2.07321301e-01
-2.56862566e-02 4.80882525e-01 3.11041623e-01 2.86321729e-01
-1.81427002e-01 -3.98751736e-01 -2.32392237e-01 -6.96686625e-01
5.78789353e-01 3.92615288e-01 -2.01207008e-02 4.94435519e-01
3.09595745e-02 1.51527834e+00 -5.77998459e-01 -2.29575932e-01
1.09266317e+00 7.49672115e-01 1.16922043e-01 5.13926029e-01
-1.75573692e-01 7.91058764e-02 -2.10854352e-01 5.61893404e-01
7.22532332e-01 -2.94383615e-01 4.16760802e-01 -6.19763255e-01
-5.45708314e-02 5.35718858e-01 -9.67837036e-01 1.48384428e+00
-7.26355910e-01 4.85631973e-01 -2.36764655e-01 -1.39697433e+00
8.48706961e-01 6.30868912e-01 8.24094713e-01 -9.46194291e-01
6.34021819e-01 -1.07093066e-01 -2.70574987e-01 -4.42810118e-01
4.45907027e-01 5.09263098e-01 -3.42883319e-01 6.83894455e-01
-3.91077027e-02 5.10010645e-02 -4.14304212e-02 -5.88963568e-01
1.05408907e+00 -1.00220360e-01 7.35180438e-01 -2.07976863e-01
2.67436087e-01 -4.32502151e-01 5.61161637e-01 4.69833553e-01
-1.33485422e-01 -3.78864296e-02 -2.36591369e-01 -8.40315819e-01
-1.12878299e+00 -6.67945445e-01 -1.93346009e-01 1.06738448e+00
-4.57446605e-01 -3.56393568e-02 -8.35996628e-01 -4.15654778e-01
1.35631397e-01 1.39005816e+00 -5.51734924e-01 -4.76567894e-01
-8.03228498e-01 -1.22728312e+00 3.33564192e-01 9.21376944e-01
1.19867587e+00 -1.47696710e+00 -1.43283796e+00 8.23174953e-01
2.97522638e-03 -6.44812644e-01 -2.44924068e-01 9.56034720e-01
-4.46423471e-01 -1.00466657e+00 -1.22317798e-01 -7.66966939e-01
2.51145601e-01 -4.40166384e-01 1.35539055e+00 -3.53277475e-01
-2.92658448e-01 -1.52229980e-01 4.69409153e-02 -5.16628325e-01
-4.15569365e-01 7.07139134e-01 1.15826271e-01 -1.58187196e-01
1.13699532e+00 -1.25404620e+00 -7.83184767e-01 -8.80365074e-02
-5.01895666e-01 -1.19232528e-01 4.37640339e-01 5.81071675e-01
4.93503302e-01 8.29419971e-01 8.62794161e-01 -6.21762812e-01
7.89401829e-01 -6.91043258e-01 -8.45212400e-01 1.53227881e-01
-1.06141329e+00 -1.55954793e-01 9.98279512e-01 -4.80775028e-01
-9.00611699e-01 1.54248774e-01 -2.28584722e-01 -3.25426698e-01
-3.33869457e-01 3.15829925e-02 -4.93720442e-01 4.94117737e-01
8.83337110e-02 4.53428149e-01 -6.46644175e-01 -6.77751124e-01
1.27948180e-01 9.83922958e-01 1.02637720e+00 7.87811279e-02
3.45443547e-01 3.55559774e-02 -1.53674349e-01 -5.24427474e-01
-4.51606333e-01 -1.39658153e-01 -5.74318945e-01 9.18042436e-02
8.47794473e-01 -9.61304426e-01 -1.47250450e+00 7.69885242e-01
-1.03943312e+00 -5.08470535e-01 -6.95313871e-01 2.29183778e-01
-4.72439080e-01 -2.64064461e-01 -6.70696557e-01 -1.15365648e+00
-1.27153909e+00 -8.29587162e-01 9.23182964e-01 6.32968962e-01
-7.24990249e-01 -1.01951635e+00 9.95501354e-02 7.14519322e-02
5.96806288e-01 6.36066735e-01 7.29585826e-01 -8.61096203e-01
-1.81104735e-01 -6.03016376e-01 1.28474802e-01 6.64383113e-01
7.31417835e-01 -2.38873899e-01 -1.21479321e+00 -5.70459247e-01
4.90120322e-01 -2.86785960e-01 4.36747849e-01 7.02818215e-01
1.54262280e+00 -9.11122382e-01 -3.11221540e-01 8.09217453e-01
1.65384150e+00 7.54585862e-01 6.66327417e-01 2.36442402e-01
8.72239769e-01 -1.39551759e-01 -3.70370120e-01 7.34470785e-01
6.67218626e-01 1.84362188e-01 3.52379709e-01 -8.52728263e-03
3.08093578e-01 -2.68242121e-01 3.06432903e-01 1.05270386e+00
2.96015620e-01 -3.34457815e-01 -3.38236302e-01 4.75078136e-01
-2.10128093e+00 -9.93330598e-01 2.00386643e-01 1.41781688e+00
6.55129135e-01 -7.35738352e-02 9.13746282e-02 7.54320323e-01
4.76987749e-01 1.36656985e-01 -1.27059960e+00 -7.10064650e-01
-6.08112626e-02 2.42296234e-01 6.23756647e-01 6.02556095e-02
-1.13633859e+00 1.80038705e-01 6.47326612e+00 6.49004757e-01
-8.63049090e-01 3.53676796e-01 8.63397777e-01 -2.19975889e-01
2.78926939e-01 -9.09148037e-01 -6.53951883e-01 1.18240404e+00
1.44654810e+00 -1.86064616e-01 7.97538161e-01 1.11838651e+00
5.02241015e-01 -2.42717892e-01 -1.40520191e+00 1.25383294e+00
-2.14534551e-01 -1.08326244e+00 -4.14970040e-01 6.28741924e-03
1.10876000e+00 2.70977795e-01 -1.26184896e-01 4.41043496e-01
1.15896416e+00 -1.23947561e+00 1.28965735e-01 6.61034405e-01
4.83355522e-01 -1.32249629e+00 1.07319903e+00 2.05784038e-01
-1.64194870e+00 -5.28382838e-01 -1.55114671e-02 -6.86854497e-02
1.40694678e-01 6.28839970e-01 -5.64719081e-01 6.48040056e-01
1.12988305e+00 5.65321684e-01 -2.87375540e-01 3.84557217e-01
7.25979032e-03 6.22077644e-01 -6.89132929e-01 -7.10646808e-02
6.07155561e-02 -3.51485908e-01 -2.48009339e-01 1.05413830e+00
1.14113495e-01 7.20960200e-02 1.40928224e-01 9.57854271e-01
-3.97119492e-01 -5.49477279e-01 -6.22024536e-01 2.12138548e-01
6.27845645e-01 1.40497220e+00 -2.29969606e-01 -4.46784079e-01
-1.62534580e-01 1.39404368e+00 9.08068269e-02 1.75152615e-01
-8.95222306e-01 -5.78417361e-01 7.76014626e-01 -4.93093640e-01
4.52654481e-01 3.18401575e-01 -5.19022465e-01 -8.13036501e-01
-2.74626672e-01 -3.96065354e-01 4.66573238e-01 -7.30483174e-01
-1.48768926e+00 1.23714305e-01 -2.52500623e-01 -5.80972791e-01
-8.96681905e-01 1.94070674e-03 -1.31916845e+00 8.74170303e-01
-1.42587233e+00 -1.11111891e+00 -4.97958958e-01 7.61277437e-01
5.60405910e-01 -2.41603509e-01 1.16237378e+00 3.93219769e-01
-1.08884680e+00 6.13495946e-01 6.21889710e-01 4.26948994e-01
-4.14791852e-01 -1.49353242e+00 7.73328841e-01 5.12280285e-01
-2.60370225e-01 -2.82047868e-01 4.12587702e-01 -5.16187847e-01
-1.30534065e+00 -1.80347466e+00 6.85744107e-01 -1.64701387e-01
3.71169090e-01 -2.92207569e-01 -6.64715767e-01 1.07880294e+00
7.29316771e-01 -3.43276203e-01 6.24364614e-01 -2.77209491e-01
3.62770677e-01 -4.39278662e-01 -1.62526727e+00 -8.58747512e-02
3.90237182e-01 -5.59823632e-01 -6.58466578e-01 3.33549201e-01
3.07791084e-01 -2.40770385e-01 -1.25551224e+00 2.99667597e-01
3.46366286e-01 -7.83711672e-01 7.21977055e-01 -9.05515403e-02
-1.00452960e-01 -1.68146864e-01 -3.67390811e-01 -1.53059936e+00
-5.98441184e-01 -7.74575949e-01 -1.26455986e+00 1.56224036e+00
-2.43482739e-01 -8.31041217e-01 8.06442618e-01 6.49542093e-01
8.46736729e-02 -5.85385382e-01 -1.08088160e+00 -5.29889822e-01
-1.30736947e-01 -8.87308046e-02 1.72131813e+00 9.51230407e-01
-2.45738506e-01 5.07790864e-01 -3.51370245e-01 3.32251400e-01
6.10924184e-01 1.11716479e-01 2.14016810e-01 -1.11959517e+00
1.36054933e-01 -5.04700005e-01 -2.94255376e-01 -5.05209744e-01
2.53553212e-01 -6.64355099e-01 -1.32579073e-01 -1.59569848e+00
-9.93412137e-02 1.94194913e-01 -7.43730843e-01 6.38495386e-01
3.22654158e-01 1.85284361e-01 -1.49646223e-01 -2.32193783e-01
-3.40239555e-01 9.22275484e-01 1.85223505e-01 -6.65756941e-01
-3.99426550e-01 2.18667358e-01 -5.04619002e-01 6.88210309e-01
1.57474470e+00 -1.89437494e-01 -4.16705489e-01 -3.03780317e-01
-2.12198734e-01 -3.74479949e-01 2.18924239e-01 -1.00277030e+00
2.03755125e-01 2.49853283e-01 1.20701241e+00 -1.32557535e+00
9.11332518e-02 -1.12656903e+00 7.90687025e-01 8.25253665e-01
-2.96601467e-02 5.10014236e-01 3.24564911e-02 8.57250616e-02
2.78704762e-01 4.17108461e-02 8.10447991e-01 -1.54714718e-01
-5.25503039e-01 3.27847183e-01 -4.11753297e-01 -5.85041344e-01
1.06171644e+00 -1.29481539e-01 -7.32312500e-02 5.66714304e-03
-4.28935528e-01 6.75994456e-01 -1.14154667e-01 2.98949867e-01
2.11896524e-01 -2.01843190e+00 -5.56892931e-01 3.48888516e-01
-5.82475066e-01 3.24142873e-01 1.42413378e-01 3.96145642e-01
-1.80207402e-01 5.25621712e-01 1.02552981e-03 -4.85795677e-01
-8.87377679e-01 5.62609673e-01 8.52560639e-01 -4.56428200e-01
-9.09071743e-01 3.17759752e-01 -5.02863288e-01 -2.98287123e-01
4.37276602e-01 -6.78490281e-01 -2.11740211e-01 3.73332322e-01
6.79585338e-01 1.03569961e+00 3.96847665e-01 -2.84596145e-01
-3.75337869e-01 1.18232779e-01 8.47601593e-02 6.92831516e-01
1.74101543e+00 -3.12600523e-01 -2.01554835e-01 6.63918436e-01
1.76490426e+00 -1.15324175e+00 -1.12221968e+00 1.12675577e-01
7.34282881e-02 4.25572455e-01 4.25289333e-01 -9.06384051e-01
-1.37414920e+00 3.72979969e-01 1.03778458e+00 7.04172194e-01
1.63103151e+00 -4.24109489e-01 1.24468815e+00 4.43410456e-01
5.83967566e-02 -1.71586788e+00 -5.50084352e-01 6.71494976e-02
4.42623019e-01 -1.22100687e+00 1.03538455e-02 7.83491313e-01
3.91559489e-03 9.27402318e-01 4.26329225e-01 -1.92546487e-01
9.77722049e-01 6.17476225e-01 -4.34091061e-01 -7.82878250e-02
-6.71660304e-01 1.40006199e-01 -3.24026138e-01 7.17919171e-01
6.43085241e-02 5.19103646e-01 3.07454944e-01 7.43415892e-01
-6.02614760e-01 3.91723186e-01 -4.71107997e-02 1.02634406e+00
3.54893580e-02 -3.11118424e-01 -2.00361893e-01 9.77779448e-01
-4.09704447e-01 7.86476508e-02 1.70284703e-01 3.69922727e-01
2.90326834e-01 1.19336987e+00 6.33073509e-01 -3.59272331e-01
2.83799380e-01 2.33606756e-01 2.20380187e-01 1.91684932e-01
-9.00219142e-01 -3.28244567e-01 -2.87935585e-01 -4.04013246e-01
-2.18767747e-01 -5.05641401e-01 -1.14171922e+00 -1.09045362e+00
-2.19566748e-01 6.78886427e-04 5.63985229e-01 8.92575979e-01
1.69920415e-01 1.10030937e+00 1.44940746e+00 -1.01131690e+00
-5.61029673e-01 -1.33214247e+00 -9.04421806e-01 2.19609335e-01
5.90770543e-01 -4.02721688e-02 -2.25182161e-01 -6.14508577e-02] | [16.061491012573242, 7.576128959655762] |
930932f3-6953-490c-99d2-6e404a20badb | scalable-discovery-of-time-series-shapelets | 1503.03238 | null | http://arxiv.org/abs/1503.03238v1 | http://arxiv.org/pdf/1503.03238v1.pdf | Scalable Discovery of Time-Series Shapelets | Time-series classification is an important problem for the data mining
community due to the wide range of application domains involving time-series
data. A recent paradigm, called shapelets, represents patterns that are highly
predictive for the target variable. Shapelets are discovered by measuring the
prediction accuracy of a set of potential (shapelet) candidates. The candidates
typically consist of all the segments of a dataset, therefore, the discovery of
shapelets is computationally expensive. This paper proposes a novel method that
avoids measuring the prediction accuracy of similar candidates in Euclidean
distance space, through an online clustering pruning technique. In addition,
our algorithm incorporates a supervised shapelet selection that filters out
only those candidates that improve classification accuracy. Empirical evidence
on 45 datasets from the UCR collection demonstrate that our method is 3-4
orders of magnitudes faster than the fastest existing shapelet-discovery
method, while providing better prediction accuracy. | ['Lars Schmidt-Thieme', 'Josif Grabocka', 'Martin Wistuba'] | 2015-03-11 | null | null | null | null | ['online-clustering'] | ['computer-vision'] | [-2.13995501e-02 -4.01689261e-01 -3.80128205e-01 -1.76433846e-01
-5.13302505e-01 -6.50574505e-01 3.45712125e-01 6.37299895e-01
8.19280818e-02 4.76942390e-01 -1.07900761e-01 -2.61175603e-01
-7.80796885e-01 -9.43616390e-01 -2.92505771e-01 -6.36083543e-01
-7.73977757e-01 5.52551031e-01 4.15643334e-01 6.78885803e-02
5.36970437e-01 6.42981291e-01 -1.69953251e+00 4.58586305e-01
8.31412673e-01 1.53195965e+00 -2.28997901e-01 1.20836698e-01
-8.10516700e-02 1.40214771e-01 -4.87524539e-01 -2.74864376e-01
6.00107074e-01 -1.80654526e-01 -3.24023694e-01 -1.03782207e-01
1.08242966e-02 4.12701160e-01 -3.27871852e-02 7.34938383e-01
1.17125757e-01 2.44613171e-01 5.07126093e-01 -1.38539195e+00
-2.04298958e-01 2.76140660e-01 -6.47105873e-01 3.84957641e-01
2.12514862e-01 -5.49491584e-01 1.09146810e+00 -1.01307344e+00
4.56461787e-01 6.76272988e-01 1.15838301e+00 -1.84880093e-01
-1.25062025e+00 -6.22963309e-01 -3.82194251e-01 4.38803375e-01
-1.54567182e+00 -1.61058664e-01 1.09984171e+00 -4.25558597e-01
8.29277754e-01 6.62162185e-01 7.71089017e-01 4.21759993e-01
1.93686321e-01 4.50147808e-01 7.95349121e-01 -2.72309303e-01
5.40327370e-01 -2.01691076e-01 2.95023531e-01 4.12700862e-01
4.78274256e-01 1.55303165e-01 -4.26588088e-01 -9.36094463e-01
4.46805120e-01 2.79356271e-01 5.09618083e-03 -5.07858098e-01
-9.15769577e-01 8.13177586e-01 -2.33664259e-01 4.81674552e-01
-5.29866099e-01 -3.83147985e-01 5.18458784e-01 6.10220730e-01
7.82565951e-01 5.13072133e-01 -6.57532513e-01 -3.62738103e-01
-1.10257375e+00 2.40512136e-02 8.03388298e-01 8.30650389e-01
3.79421443e-01 -1.17206424e-01 -2.54103057e-02 8.02554190e-01
-7.99140930e-02 1.57677382e-01 4.59720224e-01 -5.18962204e-01
2.26257145e-01 1.11443162e+00 -1.15128011e-01 -1.43507981e+00
-4.52653319e-01 -6.05639338e-01 -7.97433078e-01 -8.02402049e-02
2.56453395e-01 2.19522968e-01 -3.29299569e-01 9.06579673e-01
6.04898274e-01 4.29877460e-01 -3.46303523e-01 4.94048655e-01
5.46879947e-01 6.11011744e-01 -2.49933541e-01 -8.19503367e-01
1.02612996e+00 -3.18261296e-01 -1.87877163e-01 5.88601708e-01
4.78318244e-01 -8.55761945e-01 5.16334116e-01 6.92874193e-01
-7.53810823e-01 -6.63594186e-01 -8.37690830e-01 7.25128889e-01
-8.67643654e-02 1.07000053e-01 6.32090211e-01 5.26967585e-01
-6.55278563e-01 1.04301810e+00 -7.99632132e-01 -2.51128197e-01
3.15856159e-01 2.74139553e-01 -4.60159630e-02 2.85506785e-01
-6.30106807e-01 3.40644240e-01 3.65552634e-01 -5.98230958e-01
8.57246295e-02 -1.04906952e+00 -4.40682262e-01 7.49979913e-02
1.35839015e-01 -3.83247510e-02 6.93283081e-01 -6.98712051e-01
-8.62599671e-01 5.43814778e-01 -3.64973009e-01 -8.12732458e-01
3.17393214e-01 1.12371318e-01 -8.02282333e-01 2.46928528e-01
7.34033510e-02 -3.03696483e-01 1.18687403e+00 -8.02084744e-01
-6.93292022e-01 -3.93862098e-01 -7.14066565e-01 -2.39476278e-01
-4.82918680e-01 1.56110227e-02 -7.52833635e-02 -9.54641163e-01
5.20729482e-01 -8.15184116e-01 -2.63502091e-01 -1.03397779e-01
-1.08972844e-02 -7.78260827e-01 1.05077863e+00 -3.43290895e-01
1.66662598e+00 -2.37710547e+00 -4.18360382e-01 8.22552800e-01
3.59839022e-01 1.84529960e-01 -5.93443140e-02 8.66962373e-01
-2.64002770e-01 2.35222057e-02 -1.74092054e-01 2.05929533e-01
-4.18181688e-01 9.37692523e-02 -7.54518509e-01 6.69393361e-01
-7.76083246e-02 5.25380909e-01 -8.85358751e-01 -3.52326930e-01
1.78132728e-01 2.80931685e-02 -2.33628109e-01 -2.13888317e-01
3.48766595e-02 3.05137157e-01 -5.41481018e-01 5.74647784e-01
6.36546075e-01 -3.60583901e-01 2.15411652e-03 -2.92075008e-01
-4.11331385e-01 6.34263903e-02 -1.09320033e+00 9.06684041e-01
1.65790424e-01 4.76122230e-01 -5.63804388e-01 -1.52120531e+00
1.67816710e+00 3.60621125e-01 1.60592055e+00 -7.12278068e-01
-2.72977054e-02 3.69961470e-01 7.97914639e-02 -3.47871155e-01
2.13407651e-01 1.26291513e-01 9.02754217e-02 6.11270547e-01
-4.09532517e-01 2.57595032e-01 5.23037493e-01 -2.83521622e-01
1.36061990e+00 -2.46859476e-01 5.65121830e-01 -2.83640146e-01
4.71764088e-01 3.19108129e-01 7.37979770e-01 4.20024306e-01
-1.79241881e-01 4.77675438e-01 7.57298097e-02 -1.11023235e+00
-1.30088544e+00 -8.43632579e-01 -5.10189891e-01 8.60027254e-01
-5.03802262e-02 -8.86584938e-01 -1.53981507e-01 -4.95929152e-01
4.68589395e-01 4.90153700e-01 -5.96825898e-01 -1.81338727e-01
-7.29452670e-01 -6.44442379e-01 4.12925422e-01 3.83743972e-01
1.23531662e-01 -7.78259456e-01 -8.86992872e-01 3.22660893e-01
-2.00100057e-02 -8.14920247e-01 -3.85066986e-01 -2.46360917e-02
-1.46505690e+00 -1.42196751e+00 -2.71155149e-01 -7.26411045e-01
6.69637799e-01 5.40945351e-01 1.08966076e+00 1.67273179e-01
-4.29672927e-01 3.94006282e-01 -9.16622102e-01 -6.31953120e-01
-2.16868520e-01 -3.13769132e-01 1.84063300e-01 2.58002520e-01
7.85318434e-01 -7.76255548e-01 -4.84293908e-01 4.46827441e-01
-3.23005378e-01 -3.21017027e-01 3.31705660e-01 6.69580042e-01
1.19919491e+00 6.94772243e-01 7.13407695e-01 -5.23533165e-01
7.96037555e-01 -6.83548033e-01 -5.76624334e-01 3.14296573e-01
-8.53934288e-01 -2.01976448e-01 1.02104104e+00 -6.71211421e-01
-3.25305134e-01 2.44578764e-01 4.30223882e-01 -6.81637645e-01
-3.21984179e-02 6.82625055e-01 6.61957979e-01 -2.91735888e-01
7.39197493e-01 6.23862267e-01 -1.97618939e-02 -6.54638886e-01
-1.44757628e-01 3.81152511e-01 2.20064655e-01 -5.06640911e-01
8.65247250e-01 6.56165719e-01 4.36062932e-01 -1.14088702e+00
-3.69472295e-01 -1.04616892e+00 -8.64469290e-01 -1.89298004e-01
9.92226452e-02 -3.21129650e-01 -6.92999363e-01 -1.85287818e-01
-8.35529685e-01 4.79652584e-01 -5.12530625e-01 5.06277382e-01
-5.23465812e-01 5.12534678e-01 -6.72942474e-02 -9.39333618e-01
-7.09268868e-01 -2.88127780e-01 7.52918005e-01 -1.56964123e-01
-7.38997042e-01 -7.66315341e-01 3.48378658e-01 -8.62211362e-02
7.55100623e-02 6.08119011e-01 1.07119358e+00 -1.17764676e+00
-1.14019051e-01 -5.27500808e-01 7.60378912e-02 -2.14250892e-01
2.23499730e-01 1.16616912e-01 -4.45704073e-01 -4.59594995e-01
2.32073262e-01 3.49822879e-01 4.61505324e-01 6.09700680e-01
1.53427196e+00 -4.28296089e-01 -6.22646689e-01 4.61797297e-01
1.19703305e+00 6.96198523e-01 3.19272578e-01 2.71064192e-01
1.01020344e-01 4.43608820e-01 9.34442937e-01 1.00113571e+00
-2.07433719e-02 5.69778085e-01 -8.02742690e-02 9.80970189e-02
2.04529509e-01 -1.70037106e-01 -8.13783929e-02 1.03415978e+00
-4.26323354e-01 3.13547254e-01 -9.69207406e-01 5.25354266e-01
-1.94877541e+00 -1.41971219e+00 -4.44539100e-01 2.61819601e+00
6.90775931e-01 1.03426225e-01 6.61791801e-01 9.58602726e-01
6.50011182e-01 -3.22854847e-01 -5.50240636e-01 -3.11654389e-01
2.38642041e-02 5.62409222e-01 3.74329656e-01 -2.40627035e-01
-1.29329550e+00 1.27195209e-01 6.61398602e+00 7.94686139e-01
-9.80560899e-01 -3.16630453e-01 4.68488455e-01 5.48951477e-02
2.55359951e-02 -1.73427954e-01 -4.26045418e-01 5.73771000e-01
8.01634610e-01 -8.63679945e-01 6.74574971e-02 1.04135787e+00
4.90761310e-01 1.02026895e-01 -9.61892486e-01 1.25013852e+00
-7.03485981e-02 -1.34299219e+00 -8.04612860e-02 5.25543243e-02
7.79817879e-01 -8.84592310e-02 4.84549068e-02 -6.85576946e-02
-1.98016673e-01 -6.74501121e-01 4.53779250e-01 4.55878019e-01
6.25243545e-01 -1.04283738e+00 3.52850497e-01 4.36337084e-01
-1.75087404e+00 -3.80350679e-01 -5.33661246e-01 -8.08594599e-02
-1.05282791e-01 1.04072130e+00 -1.03949523e+00 5.38490236e-01
8.38927865e-01 1.02619088e+00 -3.81727576e-01 1.82472777e+00
5.03708720e-01 9.66179490e-01 -4.96029019e-01 -7.12286532e-02
-1.30234182e-01 -4.81585681e-01 7.30622888e-01 9.54582274e-01
7.61493981e-01 4.37539935e-01 4.60307688e-01 4.83332187e-01
4.33791339e-01 5.21522582e-01 -7.79934168e-01 -2.39073541e-02
8.09922874e-01 1.02094674e+00 -9.69613075e-01 -1.74004421e-01
-4.64316130e-01 3.93431932e-01 -1.11504972e-01 -1.96602587e-02
-5.75418174e-01 -4.22196031e-01 3.45612228e-01 3.40122938e-01
5.24462223e-01 -5.19724488e-01 -8.61353159e-01 -6.82380676e-01
2.51804411e-01 -7.58826256e-01 8.19125414e-01 -1.18770622e-01
-1.57799661e+00 3.50572258e-01 -2.19360396e-01 -2.14123011e+00
-2.20231824e-02 -2.77471662e-01 -8.16725314e-01 3.57639909e-01
-9.51577365e-01 -6.71311021e-01 -7.69945085e-02 6.53875589e-01
5.63338816e-01 -2.51417905e-01 9.49376047e-01 3.01145911e-01
-6.90503567e-02 5.46022654e-01 5.05311191e-01 -8.42423551e-03
3.17196280e-01 -7.73908794e-01 4.41538364e-01 5.80116510e-01
4.82971400e-01 4.94592935e-01 5.68528950e-01 -9.44257915e-01
-1.26597404e+00 -1.28013659e+00 9.46343243e-01 -2.55281180e-01
6.79295421e-01 1.74356520e-01 -1.17900038e+00 1.99570674e-02
-5.72377861e-01 5.61347492e-02 1.18842804e+00 9.98229533e-02
-4.43518907e-01 -3.04504573e-01 -1.11893415e+00 3.05580318e-01
1.12015784e+00 -1.51902542e-01 -6.34250760e-01 1.09214589e-01
1.99002042e-01 9.87580419e-02 -1.25082934e+00 7.68212795e-01
6.86923921e-01 -8.80542815e-01 1.16724622e+00 -5.90272009e-01
1.02682769e-01 -3.31996709e-01 -3.77145559e-02 -1.12813950e+00
-6.84107900e-01 -7.28967309e-01 -6.52523756e-01 7.97185838e-01
2.76096404e-01 -5.73108435e-01 9.83365119e-01 2.98557013e-01
7.03468472e-02 -1.03618991e+00 -1.16794479e+00 -1.33452833e+00
-3.05639535e-01 -4.05579984e-01 7.43315697e-01 1.19576240e+00
3.73211175e-01 -1.57506820e-02 -1.90563247e-01 -1.08181782e-01
9.94660616e-01 9.25207913e-01 5.31691670e-01 -2.10998821e+00
-1.43747225e-01 -6.22105360e-01 -7.00871646e-01 -5.63700259e-01
-2.72068590e-01 -1.05109608e+00 -4.35770869e-01 -1.10595191e+00
-8.91994163e-02 -9.11829889e-01 -2.38434091e-01 4.13536638e-01
2.05070302e-01 1.10558517e-01 -1.90021902e-01 7.34048307e-01
-4.56502974e-01 3.89492035e-01 1.13986194e+00 7.20928609e-02
-5.30292809e-01 5.45260131e-01 -1.47440687e-01 6.42244160e-01
9.63320255e-01 -5.24958849e-01 -5.21098077e-01 1.45657703e-01
1.21698920e-02 5.71227111e-02 7.65912700e-03 -1.21674836e+00
3.80152881e-01 -3.67779970e-01 6.96092606e-01 -1.08048034e+00
1.07064925e-01 -1.02165031e+00 4.22709018e-01 7.26843357e-01
-5.74459024e-02 2.83491552e-01 1.45219415e-01 6.84751511e-01
-2.06306845e-01 1.04469523e-01 6.95811987e-01 3.16026211e-01
-7.15024769e-01 5.62726498e-01 -1.54925212e-01 -4.79906425e-02
1.38549244e+00 -5.52538097e-01 1.78488698e-02 -9.17060003e-02
-5.04796386e-01 -7.17936009e-02 1.87622845e-01 4.74809378e-01
1.03362143e+00 -1.57344139e+00 -6.70860291e-01 4.82091069e-01
2.94464916e-01 -5.66532791e-01 -4.21775937e-01 7.82497346e-01
-1.77720815e-01 4.38998699e-01 -9.44626406e-02 -6.52032316e-01
-1.63705587e+00 7.04010963e-01 -9.64008719e-02 -1.88124180e-02
-1.08920217e+00 4.20064956e-01 -4.16153550e-01 6.46278262e-02
-3.42355259e-02 -4.68182527e-02 -4.33887213e-01 8.97245184e-02
5.96824408e-01 9.17815149e-01 2.70528167e-01 -5.05131006e-01
-3.23246360e-01 8.41504872e-01 3.22538495e-01 4.01259899e-01
1.60592425e+00 1.95906818e-01 -2.19660386e-01 4.07587886e-01
1.11894262e+00 3.81023660e-02 -6.22915566e-01 -3.57503682e-01
6.45024002e-01 -7.25529790e-01 -4.85936195e-01 -2.89431125e-01
-6.69390500e-01 4.02009398e-01 4.04523283e-01 8.24430645e-01
1.30990577e+00 -3.48054357e-02 9.03677523e-01 3.15289021e-01
6.81804121e-01 -1.13788378e+00 3.42115723e-02 4.92787004e-01
9.12287176e-01 -9.11703646e-01 1.15703441e-01 -5.50814211e-01
-1.33049965e-01 1.47679603e+00 2.02902764e-01 -3.72221500e-01
1.06279171e+00 1.94185108e-01 -3.16563457e-01 -2.99821913e-01
-8.29534709e-01 7.19136447e-02 7.50213563e-01 5.16974688e-01
3.93063664e-01 3.51085633e-01 -9.49372709e-01 8.29574406e-01
-4.13931161e-01 -2.05449075e-01 -6.12569116e-02 7.76598394e-01
-6.79461241e-01 -1.15523434e+00 -5.69185734e-01 1.10921955e+00
-3.19979906e-01 1.95508555e-01 -4.93022025e-01 2.40650475e-01
1.09441072e-01 1.07595873e+00 2.76909858e-01 -6.46263301e-01
3.56160104e-01 4.19274457e-02 1.60552487e-01 -3.37524116e-01
-4.08521682e-01 4.64683808e-02 -1.77985534e-01 -4.70340908e-01
-2.77792484e-01 -8.79590571e-01 -1.31522048e+00 -4.05675530e-01
-1.91416368e-01 4.81045872e-01 4.36903179e-01 6.94809675e-01
6.32072628e-01 -1.32030516e-03 1.25219274e+00 -2.17034414e-01
-5.66894472e-01 -5.06318808e-01 -5.15927196e-01 6.25347376e-01
-7.26850033e-02 -7.34513938e-01 -6.67285398e-02 1.54809222e-01] | [7.292686462402344, 3.346958637237549] |
ad03a8b2-7379-4ff5-8e70-18de1a020efb | moore-model-based-offline-to-online | 2201.10070 | null | https://arxiv.org/abs/2201.10070v1 | https://arxiv.org/pdf/2201.10070v1.pdf | MOORe: Model-based Offline-to-Online Reinforcement Learning | With the success of offline reinforcement learning (RL), offline trained RL policies have the potential to be further improved when deployed online. A smooth transfer of the policy matters in safe real-world deployment. Besides, fast adaptation of the policy plays a vital role in practical online performance improvement. To tackle these challenges, we propose a simple yet efficient algorithm, Model-based Offline-to-Online Reinforcement learning (MOORe), which employs a prioritized sampling scheme that can dynamically adjust the offline and online data for smooth and efficient online adaptation of the policy. We provide a theoretical foundation for our algorithms design. Experiment results on the D4RL benchmark show that our algorithm smoothly transfers from offline to online stages while enabling sample-efficient online adaption, and also significantly outperforms existing methods. | ['Chongjie Zhang', 'Bin Wang', 'Chao Wang', 'Yihuan Mao'] | 2022-01-25 | null | null | null | null | ['d4rl'] | ['robots'] | [-4.85744804e-01 -1.11260287e-01 -7.89734840e-01 -1.19130298e-01
-7.96613634e-01 -7.43163049e-01 3.96004736e-01 -3.23226303e-02
-5.11766255e-01 9.83085752e-01 -8.35183114e-02 -9.53836441e-01
-1.88998058e-01 -4.34625894e-01 -8.45936179e-01 -4.86392021e-01
-5.04929543e-01 5.54815292e-01 2.08265096e-01 -3.39668185e-01
2.90655810e-02 4.87952650e-01 -1.22129738e+00 -2.79488593e-01
8.55477989e-01 1.18181944e+00 4.54099327e-01 8.48309338e-01
6.62288861e-04 9.91932750e-01 -4.10295814e-01 1.06555559e-01
6.89835608e-01 -1.96090162e-01 -6.02181315e-01 6.19309358e-02
-1.66627318e-01 -1.06566274e+00 -4.63433832e-01 6.17472768e-01
6.92315161e-01 3.52502555e-01 1.75212789e-02 -1.02580440e+00
-3.58867466e-01 6.90746963e-01 -6.10658765e-01 1.97786361e-01
2.57865369e-01 4.98115212e-01 9.68710721e-01 -5.26019931e-01
3.06597680e-01 1.14163077e+00 3.05988759e-01 5.78247964e-01
-9.91850257e-01 -6.09819233e-01 6.06076837e-01 2.80921429e-01
-9.37138438e-01 -4.50492144e-01 4.56670582e-01 1.47439107e-01
9.73872066e-01 -1.31038442e-01 9.18998003e-01 8.40366900e-01
1.25954255e-01 1.21972036e+00 1.25080264e+00 -3.47354323e-01
5.97925425e-01 -1.14676334e-01 -7.00448215e-01 6.37185276e-01
-2.53402954e-03 5.49643040e-01 -3.72067928e-01 -1.73506156e-01
9.99777436e-01 -2.97670271e-02 4.24284190e-02 -4.19860452e-01
-8.08890104e-01 5.49549878e-01 2.71250486e-01 -2.48940736e-01
-6.12668693e-01 4.99640107e-01 4.88135368e-01 7.04693735e-01
2.61027575e-01 4.23186004e-01 -8.16873908e-01 -8.27373147e-01
-7.44816363e-01 3.86995852e-01 7.64677763e-01 1.07242954e+00
5.96837342e-01 3.36599499e-01 -3.03858191e-01 7.03346312e-01
5.18221036e-02 7.33055294e-01 5.08610904e-01 -1.33850682e+00
4.62726384e-01 2.37723917e-01 7.63763845e-01 -4.02682275e-01
-2.18939051e-01 -3.56326789e-01 -2.81672359e-01 1.25393137e-01
3.03897262e-01 -4.80198771e-01 -4.62248832e-01 1.60378897e+00
7.22836614e-01 3.59041929e-01 -8.43711644e-02 7.40712345e-01
-2.31421918e-01 4.75543648e-01 -7.14586899e-02 -4.69046772e-01
6.97711945e-01 -1.04583526e+00 -7.31612206e-01 -2.99274504e-01
3.67030770e-01 -4.74499255e-01 1.22490633e+00 4.26380724e-01
-1.18449092e+00 -3.58994573e-01 -1.00681591e+00 5.40706813e-01
1.76936522e-01 8.51411149e-02 8.13593507e-01 4.42506135e-01
-9.60399389e-01 7.13422179e-01 -1.08111477e+00 -3.88114853e-03
5.03501475e-01 5.99574208e-01 1.10017911e-01 1.44015513e-02
-8.76831234e-01 7.47921288e-01 4.66497719e-01 -7.80789107e-02
-1.37374222e+00 -6.15557373e-01 -3.91580582e-01 -2.91112158e-02
1.10573268e+00 -3.09140593e-01 2.27300453e+00 -1.11198294e+00
-2.47584414e+00 8.98600519e-02 -4.78036748e-03 -8.68360996e-01
6.88856542e-01 -4.50114846e-01 -2.61010468e-01 2.21940637e-01
-3.23557854e-01 3.85362595e-01 1.16783440e+00 -1.10685372e+00
-1.27888846e+00 6.71065226e-02 4.52987134e-01 4.18967664e-01
-6.04910612e-01 -1.86067581e-01 -4.43991303e-01 -2.30601117e-01
-6.29372299e-01 -9.42977607e-01 -5.84793210e-01 -2.17792019e-01
3.62616032e-01 -3.09511721e-01 8.24324012e-01 -6.08001828e-01
1.36307156e+00 -1.92111111e+00 -2.33034760e-01 2.30179712e-01
1.60999596e-01 4.49021548e-01 -2.79445976e-01 5.84252357e-01
4.71644014e-01 -2.84553945e-01 2.65806437e-01 -1.35845691e-01
2.85977840e-01 5.56750059e-01 -6.40542328e-01 4.37073052e-01
2.51881368e-02 9.76596534e-01 -1.39671993e+00 -1.14650920e-01
1.59419894e-01 -2.00157896e-01 -8.33483934e-01 6.83406711e-01
-5.43170393e-01 6.16958737e-01 -6.43593371e-01 6.40454233e-01
3.47157449e-01 -1.49643749e-01 6.32368803e-01 4.64608163e-01
-1.68059319e-01 2.93418020e-01 -8.61211896e-01 1.35939658e+00
-9.70402002e-01 2.86417544e-01 4.80476826e-01 -9.53874946e-01
8.30872536e-01 1.38387591e-01 8.16042066e-01 -1.23162448e+00
8.32929537e-02 1.83217779e-01 -1.19881798e-02 -3.01166475e-01
5.78041255e-01 3.00356537e-01 9.72779244e-02 5.79666853e-01
-1.77972585e-01 1.22877084e-01 1.84611499e-01 1.03855789e-01
1.20765305e+00 4.50020999e-01 5.78359246e-01 -8.54043439e-02
1.86707720e-01 -1.31606638e-01 5.94546497e-01 9.95600343e-01
-6.19852841e-01 -4.71053153e-01 3.90949637e-01 -5.45511067e-01
-9.31683958e-01 -8.41543794e-01 4.03086394e-01 1.49519694e+00
4.84145842e-02 -2.57974327e-01 -5.95413506e-01 -1.05093122e+00
5.02033889e-01 6.17256522e-01 -3.59882265e-01 -1.24726191e-01
-7.32983232e-01 -1.69915542e-01 1.93060607e-01 6.07006788e-01
4.43470955e-01 -9.30697858e-01 -8.77118707e-01 6.77502692e-01
2.91730195e-01 -1.21660686e+00 -6.83789432e-01 1.88304409e-01
-9.25319433e-01 -9.04834509e-01 -3.74138594e-01 -3.68295550e-01
6.02266908e-01 5.02574444e-01 1.00998044e+00 7.52099380e-02
1.93936512e-01 9.24078286e-01 -5.02625167e-01 -4.37902421e-01
-5.16190350e-01 7.22132325e-02 4.22882676e-01 -7.15178326e-02
-1.83821887e-01 -5.07063270e-01 -8.45475912e-01 3.77782643e-01
-5.93653798e-01 -2.74733752e-01 6.56613410e-01 1.02548230e+00
7.04163194e-01 -6.35154396e-02 9.18399096e-01 -7.32285023e-01
9.37332451e-01 -5.69533408e-01 -1.22156310e+00 4.13196057e-01
-1.05357492e+00 2.30942309e-01 1.36588681e+00 -6.70886755e-01
-1.00037491e+00 -4.19999026e-02 -1.66595444e-01 -6.96226776e-01
1.71013072e-01 2.44641811e-01 2.92538583e-01 -2.52037883e-01
3.62150103e-01 3.72121334e-01 2.15785131e-01 -2.87488401e-01
6.38756990e-01 6.17239296e-01 3.21612984e-01 -1.11044216e+00
8.10761690e-01 3.66104066e-01 -2.62784421e-01 -4.51831102e-01
-8.12955797e-01 -3.87294412e-01 -1.24740057e-01 -3.41500640e-01
-3.64774242e-02 -9.43684936e-01 -1.20036161e+00 -3.38685280e-03
-4.50525165e-01 -1.29879212e+00 -5.96597552e-01 2.50066757e-01
-9.64976907e-01 1.33602574e-01 -4.84020799e-01 -1.14493322e+00
-3.82988304e-01 -1.02805638e+00 7.29615092e-01 5.20085156e-01
2.57755488e-01 -1.20340991e+00 1.69527337e-01 -1.08742759e-01
5.89600742e-01 -3.09446961e-01 4.60689992e-01 -4.22941267e-01
-6.31905437e-01 3.81482467e-02 2.78497115e-02 3.31673145e-01
2.11602405e-01 -1.23663589e-01 -5.67753434e-01 -1.06018639e+00
-3.27335566e-01 -6.75672770e-01 3.03561181e-01 2.96415955e-01
1.51811516e+00 -7.76824534e-01 -4.95522544e-02 6.88346088e-01
1.37784672e+00 2.85869300e-01 1.95814267e-01 4.76074517e-01
2.95989931e-01 -1.55246779e-01 1.33514225e+00 1.22596729e+00
5.32545567e-01 5.73702097e-01 4.45453703e-01 6.22226223e-02
1.62699208e-01 -8.22095752e-01 8.35437000e-01 9.39749777e-01
-4.19730833e-03 1.49838924e-01 -6.12256110e-01 4.51321393e-01
-2.09384942e+00 -8.40744436e-01 1.03002286e+00 2.52888918e+00
1.13143289e+00 2.82256991e-01 5.39918840e-01 -3.54844600e-01
1.68322176e-01 -3.68723311e-02 -1.25083899e+00 -5.71778297e-01
3.84624600e-01 3.69084537e-01 9.99773383e-01 4.89675283e-01
-7.83115268e-01 1.26630354e+00 7.43316317e+00 1.13263035e+00
-1.21845376e+00 1.50062412e-01 4.82350886e-01 -2.67422527e-01
-7.39298537e-02 1.44881740e-01 -8.89696538e-01 4.46535528e-01
1.33188355e+00 -3.95802528e-01 1.34858215e+00 1.25097108e+00
7.49431610e-01 -1.57584384e-01 -7.96706498e-01 7.95564175e-01
-6.02616012e-01 -1.29629850e+00 -2.96768129e-01 4.01053461e-04
9.56141949e-01 1.22066289e-01 8.84830952e-02 8.48532319e-01
8.87924969e-01 -5.97866476e-01 7.37667024e-01 2.69219667e-01
9.01373684e-01 -1.25299788e+00 2.13327497e-01 6.64317310e-01
-1.27265322e+00 -7.40099311e-01 -5.21376669e-01 -2.56540567e-01
-9.16361809e-02 7.64686838e-02 -1.25714731e+00 2.91190028e-01
4.90194082e-01 6.08280540e-01 -1.42555058e-01 8.82461071e-01
-5.23788452e-01 1.05310452e+00 -2.69125551e-01 -3.24323922e-01
4.32982236e-01 -1.33006558e-01 3.47415954e-01 8.64575803e-01
8.95922706e-02 -4.22271267e-02 8.66064668e-01 3.64440978e-01
-4.57590595e-02 8.83018598e-02 -3.85885596e-01 -3.76480699e-01
8.04473221e-01 1.26363850e+00 -3.65668863e-01 -1.43900827e-01
-3.48514974e-01 7.51101673e-01 7.15465844e-01 4.86332119e-01
-1.03915811e+00 -1.35416359e-01 7.99077630e-01 -1.70613870e-01
6.87501848e-01 -6.73400998e-01 2.91749716e-01 -7.44479895e-01
-1.27702534e-01 -1.24900424e+00 2.10189134e-01 -5.34471050e-02
-1.10603869e+00 6.75203055e-02 -1.58403888e-01 -1.29799521e+00
-5.62550366e-01 -4.36441422e-01 -4.12189037e-01 2.38770038e-01
-1.89993572e+00 -6.91621840e-01 1.84065446e-01 5.80722570e-01
7.44470656e-01 -3.26513261e-01 5.02136528e-01 1.52930364e-01
-6.32715404e-01 9.39772129e-01 6.87087059e-01 -3.37638319e-01
7.51950026e-01 -1.33930886e+00 3.93267721e-01 6.06004119e-01
-6.31648256e-03 4.28759098e-01 4.96766120e-01 -4.68927473e-01
-2.08714557e+00 -1.10124433e+00 -1.18126631e-01 5.75649040e-03
1.00557113e+00 -2.08796263e-01 -4.23530787e-01 6.22775912e-01
1.59476280e-01 8.77360478e-02 3.99445087e-01 1.27066439e-02
1.56895313e-02 -5.56968093e-01 -8.19670200e-01 7.27054775e-01
9.21393037e-01 -4.00528431e-01 8.97412971e-02 3.74423027e-01
9.48943317e-01 -9.29850221e-01 -9.10933018e-01 9.48082879e-02
6.25081241e-01 -7.41564691e-01 7.40099370e-01 -8.92872393e-01
-1.12752140e-01 -3.67426575e-04 1.18142739e-01 -1.58810139e+00
-2.01109707e-01 -1.53516710e+00 -9.70968962e-01 6.51875317e-01
1.81494251e-01 -8.97917330e-01 5.76740801e-01 3.77934277e-01
-3.84171233e-02 -1.28216231e+00 -8.14005435e-01 -1.14802933e+00
-5.73039465e-02 -4.02712524e-01 8.08467090e-01 4.51215029e-01
1.73103425e-03 -1.67108640e-01 -6.77392066e-01 1.10754415e-01
5.49159467e-01 3.09151292e-01 1.06288660e+00 -4.36052263e-01
-9.47977901e-01 -4.22554880e-01 3.08977932e-01 -1.90509593e+00
2.89743423e-01 -4.37905371e-01 1.35723621e-01 -1.19797039e+00
-9.72027704e-02 -7.92279363e-01 -6.07689798e-01 6.26914203e-01
-2.26531059e-01 -5.00899374e-01 2.13241115e-01 8.64259526e-02
-1.31006956e+00 9.78666961e-01 1.58946860e+00 2.28984147e-01
-4.00310397e-01 3.66056919e-01 -5.54275274e-01 3.82948130e-01
1.01031888e+00 -3.77727181e-01 -6.79973245e-01 -2.47180179e-01
1.11255296e-01 4.12100732e-01 -2.49894843e-01 -6.89531565e-01
1.49301827e-01 -7.55167365e-01 -3.49101275e-02 -2.30397433e-01
-4.66760099e-02 -8.61072779e-01 -6.54257894e-01 5.99915087e-01
-3.47445875e-01 3.34358126e-01 2.79208899e-01 8.63892078e-01
8.12835619e-02 1.61445543e-01 7.56098092e-01 6.44699484e-02
-8.75944674e-01 8.16133440e-01 -3.31616968e-01 2.72641510e-01
1.10430002e+00 1.84173808e-01 -1.04161814e-01 -7.16729343e-01
-3.13492954e-01 9.50313330e-01 2.33282387e-01 3.46783698e-01
5.78075111e-01 -1.20370996e+00 -2.85065860e-01 1.30437151e-01
-2.55034387e-01 -2.69156396e-01 -5.31578735e-02 8.47890198e-01
-2.42038697e-01 3.20676416e-01 -1.35241941e-01 -3.14619988e-01
-6.73536479e-01 4.49679732e-01 3.71972442e-01 -6.43912733e-01
-4.45681542e-01 4.87560689e-01 -4.30535287e-01 -3.94476175e-01
5.51247954e-01 -3.12346250e-01 1.97191358e-01 -5.22145152e-01
6.19692743e-01 4.66960311e-01 -2.55521625e-01 2.16299996e-01
1.19936243e-01 -6.91139251e-02 -2.07858428e-01 -1.61984339e-01
1.30140388e+00 -2.23920986e-01 2.99254507e-01 3.42325062e-01
5.93824744e-01 1.06537983e-01 -2.17618084e+00 -3.83548707e-01
2.37593036e-02 -6.56141222e-01 1.61570027e-01 -8.23928177e-01
-9.48361814e-01 3.73083979e-01 5.29072881e-01 1.91720217e-01
1.27796340e+00 -4.83773887e-01 1.17569983e+00 7.43741751e-01
9.09797013e-01 -1.72075057e+00 2.15112194e-01 6.47501171e-01
4.81763661e-01 -1.23250258e+00 7.70953670e-02 3.71683598e-01
-7.32579410e-01 1.07338870e+00 7.53328800e-01 -2.81714946e-01
3.97792399e-01 4.22335625e-01 -1.35332301e-01 4.42710340e-01
-1.31914210e+00 -3.57069016e-01 -1.37882814e-01 4.80413407e-01
-1.30656198e-01 3.86622161e-01 -1.65887237e-01 4.96380448e-01
2.83005595e-01 1.21502675e-01 9.37656239e-02 1.09419513e+00
-6.61210537e-01 -1.51717949e+00 -1.55726805e-01 2.37853706e-01
-2.26936042e-01 3.60230774e-01 -2.69412175e-02 6.89079165e-01
-3.81784707e-01 9.11266267e-01 -1.73213825e-01 -2.71112561e-01
2.27268532e-01 -3.51091087e-01 7.68668771e-01 -1.54416859e-01
-6.22953057e-01 1.48337007e-01 2.04340573e-02 -1.16186547e+00
6.54005706e-02 -5.79582751e-01 -1.52076614e+00 -3.53636801e-01
-1.43337876e-01 2.12266013e-01 6.24115765e-01 1.06714880e+00
6.78975403e-01 4.13315952e-01 1.51671422e+00 -7.95507073e-01
-1.65411782e+00 -5.30575037e-01 -4.14764702e-01 5.11305481e-02
4.95686233e-01 -7.45296299e-01 -3.24383341e-02 -5.30156255e-01] | [4.089633464813232, 2.2403101921081543] |
f2ba79db-44e1-4911-99fe-0b18ac993d33 | reducing-hallucinations-in-neural-machine | 2211.09878 | null | https://arxiv.org/abs/2211.09878v2 | https://arxiv.org/pdf/2211.09878v2.pdf | Reducing Hallucinations in Neural Machine Translation with Feature Attribution | Neural conditional language generation models achieve the state-of-the-art in Neural Machine Translation (NMT) but are highly dependent on the quality of parallel training dataset. When trained on low-quality datasets, these models are prone to various error types, including hallucinations, i.e. outputs that are fluent, but unrelated to the source sentences. These errors are particularly dangerous, because on the surface the translation can be perceived as a correct output, especially if the reader does not understand the source language. We present a case study focusing on model understanding and regularisation to reduce hallucinations in NMT. We first use feature attribution methods to study the behaviour of an NMT model that produces hallucinations. We then leverage these methods to propose a novel loss function that substantially helps reduce hallucinations and does not require retraining the model from scratch. | ['Lucia Specia', 'Marina Fomicheva', 'Joël Tang'] | 2022-11-17 | null | null | null | null | ['nmt'] | ['computer-code'] | [ 5.29011428e-01 5.19452691e-01 1.10284097e-01 -1.46311060e-01
-1.08848882e+00 -6.35406673e-01 9.09752190e-01 8.33351612e-02
-2.84906387e-01 8.37160051e-01 4.11457688e-01 -2.89069235e-01
4.85172868e-01 -6.41014874e-01 -1.01328921e+00 -3.26910645e-01
4.89214629e-01 7.58043945e-01 -4.76238549e-01 -3.94672543e-01
7.76963905e-02 1.19488008e-01 -1.02307904e+00 5.45195758e-01
1.19834781e+00 3.51878345e-01 1.44988135e-01 5.62124550e-01
1.31956771e-01 9.51522231e-01 -9.43092942e-01 -7.27324784e-01
1.38095036e-01 -9.64487135e-01 -8.93476009e-01 1.11305505e-01
3.69493872e-01 -7.25032464e-02 -1.43339187e-01 1.03722715e+00
4.20098096e-01 -3.83165628e-02 8.21391702e-01 -1.16009319e+00
-1.20018637e+00 5.11171460e-01 -2.95758545e-02 -9.80567485e-02
2.63041556e-01 3.65196913e-01 8.68299186e-01 -1.08152521e+00
7.78988540e-01 1.42830181e+00 7.40561068e-01 9.93124008e-01
-1.68682575e+00 -6.36366308e-01 -1.89460710e-01 -5.81969842e-02
-1.02896476e+00 -7.88591862e-01 2.79660583e-01 -3.70432168e-01
1.11150968e+00 1.02005258e-01 3.36966455e-01 1.70457006e+00
4.37125325e-01 5.90321720e-01 1.39421368e+00 -5.35400152e-01
1.40759796e-01 5.05838990e-01 -1.86007068e-01 5.19072950e-01
9.97146890e-02 1.98636398e-01 -8.17085385e-01 -3.39436866e-02
5.40587783e-01 -3.97909701e-01 -4.13043499e-01 -8.73158574e-02
-1.45884144e+00 9.08792734e-01 2.88644284e-01 1.47326365e-01
-5.22459865e-01 5.86809963e-02 1.81200895e-02 7.82192349e-01
4.16226864e-01 1.00641131e+00 -4.21303451e-01 -2.39229370e-02
-9.72379148e-01 1.91859961e-01 7.77033627e-01 8.60942841e-01
6.78194404e-01 2.59399414e-02 -4.88193303e-01 8.61659467e-01
2.44273972e-02 6.92590773e-01 6.78818762e-01 -8.26512516e-01
5.70307016e-01 3.57294291e-01 7.74281919e-02 -5.70624709e-01
2.87435297e-02 -7.15503693e-01 -8.30257952e-01 2.08217710e-01
2.77404159e-01 -1.90780133e-01 -1.05579138e+00 2.12472391e+00
-3.46589535e-01 -1.86736703e-01 5.38701475e-01 7.25225449e-01
5.19079626e-01 8.25562775e-01 -1.46151502e-02 -3.05667460e-01
7.21400142e-01 -1.08665884e+00 -7.53472149e-01 -6.62877619e-01
7.45293617e-01 -9.33991969e-01 1.42126894e+00 3.23694497e-01
-1.36376262e+00 -4.48282719e-01 -9.59100544e-01 -1.31312504e-01
-2.58484781e-01 -5.64493835e-02 1.60472736e-01 4.53868568e-01
-1.15592706e+00 9.14194286e-01 -7.39094555e-01 -4.12230909e-01
3.31170648e-01 2.40269199e-01 -4.76034254e-01 -1.09895453e-01
-1.30741191e+00 1.60412169e+00 1.01824060e-01 2.36365572e-03
-1.32434714e+00 -6.00501359e-01 -6.76157892e-01 -3.50999013e-02
4.94336523e-03 -1.04786634e+00 1.49322259e+00 -1.50237679e+00
-1.32043934e+00 7.45649934e-01 -4.47767854e-01 -4.95432764e-01
8.54079068e-01 -2.73391426e-01 -3.59538555e-01 -2.25706503e-01
3.04267615e-01 7.48525262e-01 9.75667059e-01 -1.20486414e+00
-4.51112203e-02 -1.47378072e-01 -2.67783105e-01 3.17707360e-01
-1.92277163e-01 -5.19154444e-02 -3.79795395e-02 -6.66038334e-01
-5.48868775e-02 -1.07442129e+00 -2.58044571e-01 -8.04320872e-02
-7.19588995e-01 1.33177921e-01 2.99540997e-01 -7.71588385e-01
7.88481712e-01 -1.74222171e+00 4.77824450e-01 -3.31787691e-02
4.29995619e-02 2.81854391e-01 -4.70337361e-01 5.46318471e-01
-9.76676717e-02 3.37155104e-01 -4.89154696e-01 -6.27905428e-01
-7.97020495e-02 1.73925564e-01 -5.66176414e-01 1.51081219e-01
7.35779822e-01 1.14372563e+00 -7.88120806e-01 -1.16299260e-02
-2.45531291e-01 4.77233797e-01 -4.79014814e-01 3.89912695e-01
-2.89360374e-01 6.69217169e-01 3.21204178e-02 1.47869021e-01
2.91616023e-01 -2.81147718e-01 -1.17379911e-01 1.39978841e-01
2.89740562e-02 6.26780450e-01 -4.82830077e-01 1.73930264e+00
-9.03521657e-01 7.39330769e-01 -2.78081119e-01 -5.90700030e-01
8.08604777e-01 4.66319263e-01 -4.03089315e-01 -7.10932732e-01
-3.79191674e-02 4.73153442e-01 2.24693701e-01 -2.81474262e-01
3.63798529e-01 -7.38652050e-01 -4.05404530e-03 7.11447418e-01
1.36592895e-01 -2.72276223e-01 1.82230566e-02 2.68667847e-01
9.71143365e-01 2.63760388e-02 1.07400551e-01 -2.47944966e-01
1.08227968e-01 2.33616740e-01 5.30947089e-01 7.55918443e-01
2.87246794e-01 9.09330666e-01 4.66175705e-01 -4.61474806e-02
-1.26819527e+00 -1.13522339e+00 1.92292005e-01 6.36270583e-01
-2.44820565e-01 -4.30957586e-01 -9.34294939e-01 -5.27200401e-01
-2.86019921e-01 1.31002092e+00 -7.68147707e-01 -7.97769189e-01
-3.70309234e-01 -6.39225245e-01 8.35218668e-01 4.35293823e-01
1.34769320e-01 -1.17311227e+00 -4.04749185e-01 1.57815829e-01
-3.94454777e-01 -9.02948797e-01 -4.18473095e-01 2.51653165e-01
-8.73720288e-01 -5.84720552e-01 -7.07174599e-01 -5.46228349e-01
8.76964748e-01 -7.00640380e-02 1.37620664e+00 2.23231271e-01
2.54017174e-01 -5.63849919e-02 -1.31536826e-01 -5.13899505e-01
-1.37377679e+00 1.03706539e-01 2.43045866e-01 -8.08925927e-02
1.45604163e-01 -4.66273099e-01 -1.29142208e-02 1.31541923e-01
-8.48917246e-01 5.59253871e-01 8.69266927e-01 1.11931336e+00
3.73793155e-01 -3.00750166e-01 6.05167627e-01 -1.06150937e+00
1.12327290e+00 -5.19588351e-01 -8.59153718e-02 3.28478187e-01
-8.99138451e-01 4.13469613e-01 9.57060695e-01 -7.30182707e-01
-9.77815151e-01 -2.72721380e-01 2.19610408e-02 -4.48250562e-01
-1.81482211e-02 5.23502111e-01 -4.40862067e-02 1.84280768e-01
1.06090724e+00 4.17011470e-01 1.55473845e-02 -5.28658867e-01
3.56189996e-01 6.04783833e-01 4.68847096e-01 -4.41521168e-01
8.74285460e-01 3.14950794e-02 -2.44637325e-01 -6.74365997e-01
-8.69071186e-01 2.88699627e-01 -5.02537191e-01 1.29954636e-01
5.95557749e-01 -1.02824211e+00 8.05096701e-03 3.44575107e-01
-1.55169821e+00 -5.07198691e-01 -2.38507107e-01 2.09182426e-01
-7.03827798e-01 -1.03637256e-01 -6.31583214e-01 -5.95498323e-01
-4.51129377e-01 -1.08527911e+00 7.94213355e-01 -8.81117508e-02
-6.17255926e-01 -9.03666496e-01 1.83928013e-01 2.81514287e-01
6.73930764e-01 1.94714628e-02 1.35565031e+00 -6.92300141e-01
-4.73625213e-01 4.99900095e-02 -3.26397717e-02 5.89988768e-01
-4.84000295e-02 -4.08132344e-01 -1.03813553e+00 -2.71412313e-01
1.73715428e-01 -6.15492523e-01 8.12228858e-01 -8.05004239e-02
4.48504418e-01 -6.66050971e-01 -9.52987990e-04 5.40142953e-01
1.11931157e+00 -2.64912069e-01 6.41639113e-01 1.11723147e-01
6.94807529e-01 7.67974436e-01 1.67370111e-01 -1.40830889e-01
1.74946979e-01 5.74704587e-01 1.07898235e-01 -3.27984601e-01
-3.87368441e-01 -7.64309049e-01 7.52542198e-01 9.02505040e-01
1.62464038e-01 -2.99552202e-01 -8.95161867e-01 6.74120247e-01
-1.75413597e+00 -7.54029751e-01 -2.20404372e-01 2.22592258e+00
1.21434677e+00 1.28365815e-01 -2.45523110e-01 -1.59120515e-01
5.91024756e-01 -2.34873444e-01 -5.23378253e-01 -6.48481369e-01
-3.55038255e-01 4.25554127e-01 2.51371264e-01 6.16293311e-01
-4.51719999e-01 1.14479220e+00 6.69700861e+00 4.46654558e-01
-1.14595699e+00 5.11704683e-01 5.02903461e-01 -1.21917814e-01
-5.99449277e-01 1.36113361e-01 -3.38551342e-01 3.89531612e-01
1.45368147e+00 -2.40912527e-01 6.50177419e-01 4.09073770e-01
2.44388118e-01 2.60096848e-01 -1.49696732e+00 7.21773088e-01
2.63247639e-01 -1.21993053e+00 3.66963834e-01 -8.31639841e-02
8.66873741e-01 1.74815953e-01 2.58931398e-01 4.02510196e-01
3.25483918e-01 -1.54887187e+00 1.01910698e+00 4.05313522e-01
7.90005922e-01 -6.79047465e-01 6.08071029e-01 6.61712050e-01
-1.49475664e-01 1.65387258e-01 -2.88677633e-01 -1.93819672e-01
1.43096507e-01 3.99322063e-01 -1.00274336e+00 9.99880955e-03
1.08321875e-01 5.06442666e-01 -7.04928517e-01 5.59892774e-01
-6.65008068e-01 6.20838344e-01 -4.23001684e-02 -1.85180642e-02
1.99100032e-01 7.32873231e-02 6.34046197e-01 1.07534671e+00
4.89298642e-01 -2.09582537e-01 -1.87325060e-01 1.50688672e+00
-3.33500057e-01 8.00814256e-02 -7.56846011e-01 -3.04992974e-01
1.77628666e-01 8.17939699e-01 -8.16383064e-02 -3.86992961e-01
-2.37267196e-01 1.54634690e+00 5.63478708e-01 4.21998739e-01
-4.08699960e-01 1.53370043e-02 5.48564374e-01 1.39349639e-01
-5.37280031e-02 -3.51316035e-02 -4.49772030e-01 -1.43624306e+00
7.65120536e-02 -1.33455062e+00 -2.49469385e-01 -9.92292881e-01
-1.24275661e+00 9.71195281e-01 -3.92958701e-01 -9.54906285e-01
-5.23926616e-01 -4.06276762e-01 -5.48278034e-01 1.25348437e+00
-1.42157745e+00 -1.10809076e+00 2.17499971e-01 4.42896843e-01
7.37161517e-01 -1.89545229e-01 1.21301150e+00 -7.68651962e-02
-4.27269727e-01 6.75233006e-01 3.92705249e-03 -4.04102504e-02
7.82045305e-01 -1.13251150e+00 7.80811489e-01 1.14210260e+00
3.68969768e-01 9.72532690e-01 9.53630924e-01 -7.81781495e-01
-1.26710260e+00 -1.25441062e+00 1.45300305e+00 -8.32295239e-01
5.47646999e-01 -5.19738495e-01 -1.01925206e+00 7.91829288e-01
3.67650151e-01 -4.23737109e-01 5.93992054e-01 1.16295777e-01
-5.83868742e-01 4.30743545e-01 -1.04495704e+00 8.35283339e-01
8.65167141e-01 -8.05657029e-01 -8.17669570e-01 5.36379039e-01
6.62967920e-01 -1.27847686e-01 -4.84311640e-01 8.48826114e-03
2.62274265e-01 -8.01342368e-01 5.78593016e-01 -9.78804052e-01
8.28629136e-01 -1.52165025e-01 8.33262410e-03 -1.94959092e+00
-2.92442650e-01 -8.78405631e-01 1.10103246e-02 9.82014716e-01
1.03791559e+00 -4.61971313e-01 3.89990866e-01 6.28710866e-01
1.03661092e-03 -4.30210799e-01 -7.24207520e-01 -8.44067454e-01
6.09213233e-01 -2.65400529e-01 4.15264428e-01 1.03075421e+00
4.07136381e-02 1.07729781e+00 -6.67177320e-01 -4.13000621e-02
5.26489973e-01 -6.71347789e-03 5.53136468e-01 -1.05098963e+00
-5.17246425e-01 -3.44101280e-01 9.91421938e-02 -7.38510728e-01
3.37154597e-01 -1.21905637e+00 3.22922319e-01 -1.45965350e+00
2.69653708e-01 -1.67168945e-01 -6.28955364e-02 6.96661055e-01
-2.19703019e-01 4.24378335e-01 4.35288966e-01 4.66479450e-01
-1.32472157e-01 5.99205852e-01 1.24570966e+00 -8.45891088e-02
-5.43268472e-02 -1.19897053e-02 -9.65772688e-01 4.92013752e-01
7.93369949e-01 -8.89662623e-01 -4.01571363e-01 -7.93491602e-01
4.72334236e-01 1.04969226e-01 5.53622663e-01 -7.38741040e-01
3.32305208e-02 -3.03507354e-02 4.27444637e-01 -1.44080340e-03
4.40856308e-01 -4.63034660e-01 3.07676673e-01 5.74782550e-01
-9.44027603e-01 3.86271119e-01 6.15379065e-02 4.47056949e-01
-1.89501839e-03 -2.22960979e-01 8.46623063e-01 -4.14508879e-01
-9.70844459e-03 -2.78107468e-02 -6.09577537e-01 1.40272632e-01
4.20019418e-01 1.59802154e-01 -2.92647630e-01 -6.90358758e-01
-5.61122537e-01 2.17685085e-02 9.36123013e-01 6.06064081e-01
6.06242120e-01 -1.33494532e+00 -1.19044614e+00 2.34065607e-01
2.29770064e-01 -3.10246587e-01 -3.20408404e-01 8.12678874e-01
-8.88942927e-02 3.75351399e-01 -2.67301857e-01 -1.76142812e-01
-8.48602295e-01 4.72566903e-01 4.97342795e-01 -1.92675158e-01
-4.70285356e-01 7.23095775e-01 2.18054816e-01 -6.73689842e-01
-1.50364026e-01 -1.63265690e-01 2.45954111e-01 -2.52827108e-01
5.17681718e-01 -9.75057541e-04 3.10621768e-01 -7.71678269e-01
-1.05757914e-01 9.66762006e-02 -2.79419243e-01 -6.85401917e-01
1.03642583e+00 7.29629993e-02 -5.68508506e-02 5.48453391e-01
1.08802271e+00 -1.40938058e-01 -9.69937921e-01 -3.47925186e-01
-1.64264619e-01 -2.80778229e-01 6.29744381e-02 -1.35927200e+00
-6.08506024e-01 1.21199155e+00 2.98741251e-01 -1.27542540e-01
7.93235302e-01 -1.78174421e-01 9.64197755e-01 5.10078669e-01
2.61520684e-01 -9.30957496e-01 1.23673320e-01 6.51846528e-01
1.12391663e+00 -1.12087142e+00 -4.43651497e-01 -1.19052581e-01
-9.98936117e-01 8.87879133e-01 4.14170533e-01 -7.34956190e-02
-5.27914725e-02 5.78102581e-02 3.97481829e-01 5.73615693e-02
-1.23698068e+00 7.77903795e-02 3.11126828e-01 7.12584674e-01
6.10525787e-01 1.82990089e-01 -1.94240630e-01 5.46642959e-01
-5.71145535e-01 -9.18893218e-02 7.81511247e-01 5.71966708e-01
-1.17331840e-01 -1.25744355e+00 -2.90340126e-01 3.03552806e-01
-4.27310735e-01 -6.49743080e-01 -8.63130271e-01 3.06905210e-01
5.62521480e-02 1.16880918e+00 -1.93441778e-01 -3.46976280e-01
3.37393820e-01 3.92952979e-01 5.80520451e-01 -9.39902782e-01
-8.15640748e-01 -2.01860026e-01 2.36985549e-01 -3.88208777e-01
-1.29290208e-01 -6.98959589e-01 -1.07928920e+00 -2.51762718e-01
-2.89465576e-01 1.45752236e-01 6.08573198e-01 1.15959477e+00
4.54725057e-01 2.79918462e-01 4.79464859e-01 -5.26814699e-01
-9.00100768e-01 -1.28060770e+00 -7.95036394e-05 7.15140820e-01
5.82063258e-01 -2.24994332e-01 -3.31806362e-01 1.61106646e-01] | [11.703813552856445, 9.913617134094238] |
ffd465c6-6f04-4a56-ba42-1ed5bd877490 | detection-of-word-adversarial-examples-in | 2203.01677 | null | https://arxiv.org/abs/2203.01677v1 | https://arxiv.org/pdf/2203.01677v1.pdf | Detection of Word Adversarial Examples in Text Classification: Benchmark and Baseline via Robust Density Estimation | Word-level adversarial attacks have shown success in NLP models, drastically decreasing the performance of transformer-based models in recent years. As a countermeasure, adversarial defense has been explored, but relatively few efforts have been made to detect adversarial examples. However, detecting adversarial examples may be crucial for automated tasks (e.g. review sentiment analysis) that wish to amass information about a certain population and additionally be a step towards a robust defense system. To this end, we release a dataset for four popular attack methods on four datasets and four models to encourage further research in this field. Along with it, we propose a competitive baseline based on density estimation that has the highest AUC on 29 out of 30 dataset-attack-model combinations. Source code is available in https://github.com/anoymous92874838/text-adv-detection. | ['Nojun Kwak', 'Jiho Jang', 'Jangho Kim', 'KiYoon Yoo'] | 2022-03-03 | null | null | null | null | ['adversarial-defense'] | ['adversarial'] | [ 9.59863290e-02 -1.18179008e-01 -2.33352810e-01 -2.09984571e-01
-1.28802145e+00 -1.09653592e+00 1.11090827e+00 3.45452368e-01
-3.26311618e-01 6.40486896e-01 2.62652874e-01 -6.30187094e-01
3.23038489e-01 -9.19306040e-01 -4.58957523e-01 -5.93882382e-01
1.26510397e-01 3.50126177e-01 5.40420189e-02 -3.42353225e-01
2.65530139e-01 4.82911587e-01 -6.00267410e-01 3.10608298e-01
7.55697370e-01 6.43083572e-01 -5.67303002e-01 7.13024855e-01
-4.39656377e-02 8.58774006e-01 -1.00194776e+00 -1.46811569e+00
2.88885802e-01 -3.59834850e-01 -6.42496049e-01 -5.37692726e-01
4.73835886e-01 -2.65484303e-01 -5.25783420e-01 1.41833580e+00
8.39620352e-01 -2.42970407e-01 8.37905288e-01 -1.20668459e+00
-6.43279672e-01 9.91627574e-01 -6.80562615e-01 4.34680283e-01
3.53355616e-01 3.90482157e-01 1.04842567e+00 -7.85427094e-01
3.80016476e-01 1.27669835e+00 5.09752929e-01 8.08507383e-01
-1.00745070e+00 -1.18613410e+00 6.29414916e-02 2.00718358e-01
-1.12217724e+00 -3.71206850e-01 9.10703719e-01 -1.04548126e-01
7.08137870e-01 4.25697207e-01 2.83073515e-01 1.80904782e+00
3.22801888e-01 1.11483967e+00 1.07094729e+00 -1.46867529e-01
1.40833706e-01 3.42774093e-01 -2.67773513e-02 3.37015361e-01
3.73988003e-01 -2.15412490e-02 -2.82320321e-01 -5.82814515e-01
1.14345714e-01 -1.22568160e-01 -2.07782969e-01 8.71252343e-02
-5.49374700e-01 1.30227005e+00 4.01614994e-01 4.03913498e-01
-1.19847529e-01 -6.83859289e-02 7.52717733e-01 3.27524245e-01
8.55619490e-01 6.90467656e-01 -3.96862507e-01 -2.39280164e-01
-6.54519141e-01 6.16498888e-01 8.53661358e-01 5.72647572e-01
1.82610497e-01 2.74031311e-01 -1.86973084e-02 8.56289387e-01
5.14469966e-02 7.48311520e-01 1.96833834e-01 -3.09841663e-01
5.85389078e-01 4.38965172e-01 -1.67073503e-01 -1.10809219e+00
-7.17341751e-02 -3.54397267e-01 -7.94508159e-01 -5.38472980e-02
4.92378622e-01 -3.52854788e-01 -7.16111481e-01 1.50578785e+00
2.67127544e-01 1.16968162e-01 -5.10451198e-02 5.83391547e-01
5.04633188e-01 5.97337127e-01 4.07331496e-01 2.19699010e-01
1.35646379e+00 -5.33711910e-01 -5.35889924e-01 -5.04341185e-01
7.07910836e-01 -1.05319655e+00 9.66677666e-01 7.12093472e-01
-9.09793973e-01 4.07261737e-02 -1.01051426e+00 2.45621234e-01
-7.27583945e-01 -3.64903927e-01 6.40569150e-01 1.30511189e+00
-3.72381747e-01 3.58631998e-01 -5.12867510e-01 -2.41460294e-01
7.50821829e-01 7.69808656e-03 -2.55601496e-01 -2.68000126e-01
-1.73929656e+00 1.17370939e+00 4.59572254e-03 -1.21045701e-01
-1.12940121e+00 -6.71107888e-01 -7.61431575e-01 -2.20387071e-01
2.06323013e-01 -2.80293703e-01 1.20221794e+00 -7.01488554e-01
-1.07568395e+00 8.60671103e-01 2.68102527e-01 -9.27505732e-01
7.66221166e-01 -5.87206900e-01 -7.49888003e-01 2.36382466e-02
-6.22125641e-02 2.79654860e-01 1.12072659e+00 -9.53417659e-01
-4.31163430e-01 -3.75238597e-01 2.37137690e-01 -4.20012288e-02
-7.92821527e-01 5.86367309e-01 -1.35087088e-01 -1.01421368e+00
-7.81358242e-01 -9.24142480e-01 -2.52954692e-01 -3.66794139e-01
-7.31136978e-01 -2.20312402e-01 8.51413727e-01 -7.02752292e-01
1.37763274e+00 -1.94204628e+00 -1.48242548e-01 9.59907249e-02
9.44090113e-02 7.67772079e-01 -1.12903133e-01 7.09513903e-01
-8.71070102e-02 5.55460751e-01 -3.98735344e-01 -3.69677931e-01
1.92258302e-02 -2.61903971e-01 -7.09505200e-01 7.48541057e-01
5.02563834e-01 9.21335936e-01 -8.62149537e-01 -2.81195253e-01
2.44133011e-01 5.53316236e-01 -4.64877307e-01 2.38276452e-01
-1.33942217e-01 1.87524095e-01 -4.73016948e-01 8.74420106e-01
7.36291349e-01 1.72448307e-01 -1.59885567e-02 9.19070616e-02
4.64936882e-01 4.27069992e-01 -7.41140723e-01 1.05098724e+00
-4.79264468e-01 6.34257436e-01 1.61074344e-02 -8.91661286e-01
7.47653067e-01 2.03582510e-01 2.88970619e-01 -5.06730795e-01
4.75366473e-01 1.40284643e-01 1.67038500e-01 -8.06524754e-02
4.84057993e-01 -2.58188725e-01 -4.67823118e-01 2.60904193e-01
-2.85838749e-02 -3.60598236e-01 1.76459759e-01 7.01558590e-01
1.36498940e+00 -3.72648805e-01 2.21909523e-01 1.21039279e-01
6.40288055e-01 4.91320826e-02 1.62237734e-01 8.19723189e-01
-4.08136755e-01 4.79149699e-01 5.19833326e-01 -2.72750139e-01
-9.80666518e-01 -1.08930099e+00 -3.61868739e-01 7.67700016e-01
-2.55105764e-01 -5.81030428e-01 -8.02881598e-01 -1.23169565e+00
6.64935932e-02 9.16452467e-01 -6.92292273e-01 -4.33099419e-01
-5.48709750e-01 -9.37217534e-01 1.13282156e+00 2.92495400e-01
4.01924521e-01 -1.02828634e+00 8.01545680e-02 5.06204404e-02
-1.91677153e-01 -1.25932205e+00 -4.82571363e-01 1.15885481e-01
-3.91296059e-01 -1.26130056e+00 -6.92417920e-01 -3.57463866e-01
3.97979438e-01 -3.66792679e-02 1.13979685e+00 -3.80797870e-02
-3.69592816e-01 3.53218943e-01 -6.40395641e-01 -9.61799026e-01
-8.02034318e-01 3.09496135e-01 1.80150896e-01 -1.35095149e-01
7.94216633e-01 -2.40769595e-01 -4.65748101e-01 2.08720058e-01
-1.05900717e+00 -6.34166420e-01 4.52234149e-01 6.10564411e-01
7.29505718e-02 -5.14868237e-02 8.15369964e-01 -1.28348398e+00
1.08683777e+00 -8.75451446e-01 -4.49891627e-01 9.52800922e-03
-3.82915556e-01 -3.75577122e-01 1.00226712e+00 -6.66540682e-01
-8.11901093e-01 -3.82585794e-01 -6.99913383e-01 -3.58195394e-01
-2.56813258e-01 3.52638274e-01 -1.73829526e-01 2.73135025e-02
1.09989321e+00 5.04278392e-02 -2.02845544e-01 -2.28602946e-01
4.24003512e-01 7.04607546e-01 1.52515352e-01 -4.63803023e-01
1.32619584e+00 2.92621106e-01 -3.90502214e-01 -7.92017519e-01
-9.94776964e-01 -3.42290998e-01 -1.54278010e-01 -7.24934563e-02
6.30147994e-01 -7.46339619e-01 -3.03395241e-01 7.12624609e-01
-8.90369356e-01 -2.27230921e-01 9.87084862e-03 1.57462537e-01
-2.66640764e-02 5.60092390e-01 -7.59747267e-01 -8.36581707e-01
-6.03474021e-01 -9.12199259e-01 7.07419336e-01 3.68827730e-02
-4.89581406e-01 -1.19301331e+00 1.80227399e-01 5.65568864e-01
5.12828529e-01 2.55924940e-01 5.64598262e-01 -1.26234078e+00
-2.48660650e-02 -9.02825177e-01 2.84141898e-02 6.65293038e-01
5.24978749e-02 1.18783355e-01 -1.12416637e+00 -4.15425241e-01
1.43069714e-01 -6.44307196e-01 8.12176406e-01 1.69498071e-01
1.20480800e+00 -4.17812407e-01 -1.91348180e-01 3.01273048e-01
1.21198297e+00 -3.36804539e-02 8.03485811e-01 2.62603432e-01
7.01689482e-01 4.65400040e-01 6.22344851e-01 4.25697207e-01
6.18919954e-02 4.49593395e-01 6.37803614e-01 9.11671817e-02
1.14888318e-01 -3.95649046e-01 7.74964690e-01 4.78231579e-01
4.59980518e-01 -7.17762291e-01 -9.78442013e-01 6.35066450e-01
-1.26113820e+00 -1.16537225e+00 -1.35396764e-01 1.97732532e+00
8.97918522e-01 6.52152479e-01 3.13363165e-01 2.09987313e-01
6.04347587e-01 6.34583950e-01 -5.04382670e-01 -9.15149450e-01
-1.43860340e-01 4.06459630e-01 4.91117418e-01 4.99481291e-01
-1.35005295e+00 1.18652570e+00 5.62707567e+00 1.18922949e+00
-1.08290005e+00 5.75952679e-02 8.36130857e-01 -4.99345809e-02
-3.78314465e-01 -2.57443547e-01 -9.92588043e-01 6.61671460e-01
1.24358308e+00 -2.50209033e-01 1.83388054e-01 8.72183442e-01
-1.54447302e-01 1.39130682e-01 -7.34415710e-01 5.84317207e-01
2.58374870e-01 -1.02236974e+00 2.01445788e-01 1.70120761e-01
5.18618822e-01 1.58988491e-01 5.01936793e-01 5.68809986e-01
7.08968222e-01 -1.08319855e+00 3.60462457e-01 -1.09904058e-01
5.91477871e-01 -1.07802999e+00 8.21142495e-01 3.75187814e-01
-8.42428327e-01 -2.29011700e-02 -2.45953485e-01 1.94160625e-01
1.84395671e-01 6.44405782e-01 -8.78187001e-01 3.61663222e-01
5.69574594e-01 6.03447020e-01 -6.55414999e-01 7.17321157e-01
-5.74403644e-01 1.27983332e+00 -2.85196424e-01 -2.39882126e-01
3.23365539e-01 2.12749839e-01 6.99937344e-01 1.30881822e+00
-8.54706094e-02 -5.80530539e-02 -5.29371016e-02 5.47656655e-01
-5.12505591e-01 2.73352146e-01 -9.53916609e-01 -3.98740500e-01
5.25650918e-01 1.43319774e+00 -4.14066046e-01 -9.64168310e-02
-5.00257313e-01 9.26635981e-01 2.69835979e-01 6.61769286e-02
-1.15778315e+00 -4.38102841e-01 7.96688378e-01 1.92825481e-01
-6.67983890e-02 -1.21180164e-02 -2.21661925e-01 -1.32853818e+00
-1.37406006e-01 -1.49791479e+00 6.37787759e-01 -1.55229375e-01
-1.81840169e+00 5.97286522e-01 -1.72878042e-01 -1.19390941e+00
-3.99126142e-01 -7.13071227e-01 -8.91499162e-01 7.99950123e-01
-1.22774267e+00 -1.23762751e+00 2.55544871e-01 6.17027521e-01
6.52153730e-01 -3.84049058e-01 9.66592252e-01 4.40518737e-01
-5.85221589e-01 1.05767620e+00 -1.40468225e-01 4.92438406e-01
9.55899954e-01 -1.25127828e+00 9.68505025e-01 1.13091278e+00
3.91383529e-01 5.79548359e-01 8.33616674e-01 -8.13416362e-01
-1.15338564e+00 -1.02451849e+00 6.73147202e-01 -1.04678166e+00
1.19976544e+00 -7.09173143e-01 -9.09849763e-01 5.56516826e-01
2.69489557e-01 -8.25493261e-02 8.78092706e-01 1.50393620e-01
-6.18455350e-01 1.57141313e-01 -1.29533541e+00 7.62092769e-01
5.65013766e-01 -6.95186615e-01 -2.90366203e-01 4.48800504e-01
3.40671420e-01 -3.16954225e-01 -8.66794348e-01 3.30045432e-01
3.52198154e-01 -6.64788723e-01 1.02439725e+00 -7.06564844e-01
5.92725933e-01 1.07254811e-01 -1.22630009e-02 -1.36483455e+00
-1.65804371e-01 -5.66470802e-01 -3.34183335e-01 1.40933013e+00
6.24255002e-01 -8.59607279e-01 8.75118077e-01 4.87977505e-01
2.51041651e-01 -9.87265408e-01 -6.85485423e-01 -6.25799775e-01
7.81077743e-01 -7.86915243e-01 2.85206020e-01 1.08682001e+00
-1.68929100e-01 2.28109539e-01 -6.63691401e-01 2.45309293e-01
6.25408173e-01 -4.87580270e-01 9.60553706e-01 -9.07146692e-01
-2.50146296e-02 -4.59017187e-01 -3.37597966e-01 -6.72597468e-01
3.59694421e-01 -8.34280193e-01 -4.71705049e-01 -1.14722216e+00
9.42050964e-02 -3.47933590e-01 -1.76019594e-01 4.41783130e-01
-5.55034101e-01 7.06299186e-01 9.44879875e-02 -1.74214821e-02
-3.31956089e-01 3.86121064e-01 8.24258626e-01 -4.45163637e-01
3.01297665e-01 2.84934819e-01 -1.12612259e+00 7.33129382e-01
1.33819675e+00 -7.29627788e-01 -3.10091376e-01 -8.19672123e-02
2.15287343e-01 -4.01794612e-01 1.66810378e-01 -7.41880476e-01
-3.38224731e-02 1.23473462e-02 5.02189994e-01 -4.14723307e-01
3.56393188e-01 -5.91049671e-01 -4.01904255e-01 4.52256739e-01
-3.41397583e-01 2.66365081e-01 4.02954191e-01 4.82487738e-01
-2.22475439e-01 -3.88962865e-01 8.94643664e-01 -1.46767870e-01
-3.01713765e-01 4.91972715e-01 -3.11563909e-01 6.26310349e-01
1.10265660e+00 2.24815458e-01 -6.47477627e-01 -6.16072297e-01
-3.90690982e-01 1.92168623e-01 4.77913558e-01 7.87114382e-01
6.53641164e-01 -1.04474676e+00 -1.13275933e+00 -5.61787337e-02
1.11379340e-01 -5.14632225e-01 3.23515773e-01 4.27424699e-01
-4.01063710e-01 2.93785274e-01 3.20582986e-02 -1.70777529e-01
-1.39509070e+00 6.79319441e-01 2.17368662e-01 -6.77394331e-01
-2.01976463e-01 1.05454254e+00 -5.89679666e-02 -3.79121095e-01
1.72005698e-01 3.21752906e-01 -2.29211032e-01 8.60529616e-02
7.72266328e-01 2.45481521e-01 3.08883134e-02 -6.39006793e-01
-4.98257756e-01 -3.46900076e-02 -7.12022781e-01 4.54266882e-03
1.09629512e+00 3.92496079e-01 1.20792285e-01 1.01367630e-01
1.24334669e+00 5.37347615e-01 -6.41586363e-01 -6.52457401e-02
-2.15437338e-01 -6.05069637e-01 -2.07584992e-01 -1.01366270e+00
-9.57898557e-01 1.01498699e+00 3.90145153e-01 6.79132581e-01
9.66455281e-01 -1.56060178e-02 9.08820748e-01 1.29432520e-02
2.22683847e-01 -6.61877453e-01 1.16624996e-01 4.90245014e-01
8.08345079e-01 -1.46048856e+00 2.00972140e-01 -2.75807232e-01
-8.49454045e-01 5.90261221e-01 7.03937471e-01 -1.93343535e-01
5.66404104e-01 4.63050038e-01 2.01753944e-01 -8.47196504e-02
-7.01416433e-01 1.24799199e-01 2.82536209e-01 6.52458787e-01
4.75241810e-01 4.47212271e-02 -3.38681310e-01 3.84262592e-01
-3.15800101e-01 -7.04118073e-01 4.70872849e-01 6.29098296e-01
-5.43859601e-02 -1.52433670e+00 -2.95154512e-01 7.05431640e-01
-1.35870290e+00 -5.08393168e-01 -8.30283225e-01 6.78707957e-01
-3.53236228e-01 1.10545993e+00 -3.97884130e-01 -4.43820357e-01
3.61239910e-01 -3.07984892e-02 2.84945160e-01 -5.75617611e-01
-1.09857512e+00 -2.59848446e-01 3.07484478e-01 -1.80343673e-01
1.38521483e-02 -5.43744266e-01 -6.91218972e-01 -7.20784843e-01
-4.27918732e-01 2.14213550e-01 5.91039956e-01 5.90037346e-01
7.02335760e-02 1.34399951e-01 7.53981948e-01 -4.36453491e-01
-7.80110657e-01 -1.11242139e+00 -3.48532170e-01 5.71758807e-01
7.79481754e-02 -3.55951041e-01 -6.84762597e-01 -5.13132989e-01] | [6.033651351928711, 8.070252418518066] |
68e63160-b0c7-4b05-b6f7-456a37b202e1 | on-the-importance-of-distinguishing-word | null | null | https://aclanthology.org/N19-1222 | https://aclanthology.org/N19-1222.pdf | On the Importance of Distinguishing Word Meaning Representations: A Case Study on Reverse Dictionary Mapping | Meaning conflation deficiency is one of the main limiting factors of word representations which, given their widespread use at the core of many NLP systems, can lead to inaccurate semantic understanding of the input text and inevitably hamper the performance. Sense representations target this problem. However, their potential impact has rarely been investigated in downstream NLP applications. Through a set of experiments on a state-of-the-art reverse dictionary system based on neural networks, we show that a simple adjustment aimed at addressing the meaning conflation deficiency can lead to substantial improvements. | ['Mohammad Taher Pilehvar'] | 2019-06-01 | null | null | null | naacl-2019-6 | ['reverse-dictionary'] | ['natural-language-processing'] | [ 6.67357981e-01 2.50524729e-01 -2.82154530e-01 -2.30470911e-01
-3.96359622e-01 -6.93470418e-01 7.53171742e-01 5.18930197e-01
-6.70061588e-01 6.72762632e-01 5.00103652e-01 -6.91925943e-01
-9.17549655e-02 -7.93146312e-01 -1.88851982e-01 -2.58078933e-01
6.30422592e-01 5.78702033e-01 -1.93466693e-02 -6.18829548e-01
4.91702318e-01 3.09590995e-01 -1.29528117e+00 2.15449125e-01
8.20766866e-01 6.16843581e-01 3.73262018e-01 3.32863163e-03
-7.61244059e-01 5.07581711e-01 -7.63508916e-01 -2.59026915e-01
9.54575092e-02 -3.08090359e-01 -7.87412345e-01 -5.56167006e-01
1.09876916e-01 1.76901311e-01 -2.22394377e-01 1.33977115e+00
3.41435075e-01 2.77251929e-01 5.63312709e-01 -6.80876195e-01
-8.26741576e-01 8.97661626e-01 -6.26417249e-02 6.07595026e-01
5.13729334e-01 -1.02187589e-01 1.35420382e+00 -9.88987148e-01
6.69720769e-01 1.31596363e+00 5.76469779e-01 3.53336036e-01
-1.53493869e+00 -5.54985583e-01 5.16181886e-02 3.86478081e-02
-1.27039707e+00 -5.34230292e-01 7.00122535e-01 -1.64013356e-01
1.58338845e+00 -1.55341238e-01 5.45561969e-01 1.03832901e+00
2.33657971e-01 5.57821095e-01 9.05470371e-01 -7.32252836e-01
2.06093416e-01 -5.78761334e-03 4.37642545e-01 2.44678348e-01
4.43583101e-01 -5.39263934e-02 -5.51574886e-01 -1.24520168e-01
5.28797686e-01 -3.49475712e-01 -4.87981170e-01 -4.20111939e-02
-1.17339087e+00 1.02772033e+00 3.87124360e-01 8.89677584e-01
-3.22392195e-01 1.79995801e-02 3.40803564e-01 4.99724865e-01
5.45093000e-01 1.06017041e+00 -5.83013952e-01 -4.08235371e-01
-1.02648842e+00 2.23824650e-01 8.45949173e-01 4.93299425e-01
5.01168430e-01 -8.60598609e-02 -7.51243159e-02 9.52941537e-01
1.64075360e-01 4.13604081e-01 8.46783936e-01 -4.90769714e-01
4.73889023e-01 9.95351017e-01 5.82104065e-02 -1.18895137e+00
-3.76278579e-01 -5.35592556e-01 -5.67562878e-01 -2.84385532e-01
4.84949052e-01 -8.62126872e-02 -1.04142499e+00 1.79239309e+00
-5.15029170e-02 -3.00521664e-02 4.32725161e-01 8.86348963e-01
7.25487947e-01 5.56384921e-01 4.19093251e-01 -1.14015199e-01
1.31892741e+00 -4.76805478e-01 -8.36792409e-01 -8.60011876e-01
8.34924221e-01 -9.49143827e-01 1.18738341e+00 3.47644895e-01
-7.61832178e-01 -3.80464494e-01 -1.10473669e+00 -3.47228527e-01
-7.54625320e-01 -6.22924343e-02 8.25242400e-01 5.28013647e-01
-7.36214221e-01 6.61531210e-01 -4.39316332e-01 -6.66217685e-01
2.12332740e-01 1.41801521e-01 -3.78887624e-01 -2.94893384e-01
-1.69981813e+00 1.53127003e+00 6.13754570e-01 1.46155149e-01
2.33645476e-02 -7.10123062e-01 -7.56893992e-01 2.29834348e-01
4.59433168e-01 -6.51547968e-01 1.17354023e+00 -8.43035042e-01
-1.08902955e+00 7.97076643e-01 -3.43604207e-01 -4.13921744e-01
-1.09071784e-01 -4.41944718e-01 -5.38050354e-01 -2.15170935e-01
2.30600923e-01 6.38920665e-01 5.93433797e-01 -7.52389312e-01
-4.92865801e-01 -3.76116872e-01 -9.37769935e-02 4.13991511e-01
-3.80311489e-01 1.92527547e-02 -1.26851171e-01 -9.25922275e-01
1.70440853e-01 -6.96811140e-01 -4.05350089e-01 -3.76679212e-01
-6.46166578e-02 -3.82263869e-01 4.86229151e-01 -5.37525296e-01
1.30389047e+00 -2.07670999e+00 2.74112761e-01 1.86343461e-01
-2.25875869e-01 6.18031144e-01 -1.80808917e-01 5.92869818e-01
-2.09474295e-01 3.55335265e-01 -2.98466086e-01 9.01745632e-02
-9.35112499e-03 5.69583833e-01 -7.23843575e-01 8.74665678e-02
3.58544886e-01 9.40886080e-01 -1.08732605e+00 -2.19919663e-02
2.18173116e-01 4.60179269e-01 -3.62799406e-01 7.65919462e-02
-4.41672564e-01 1.94849744e-02 -3.79575163e-01 2.49222860e-01
2.16103479e-01 -2.56714553e-01 3.44848424e-01 -1.98994040e-01
5.65595031e-02 1.14119554e+00 -8.47151458e-01 1.73964858e+00
-5.36690414e-01 6.96079075e-01 -4.38954741e-01 -1.13046420e+00
8.28657746e-01 3.08413595e-01 -4.44641942e-03 -8.17250013e-01
9.29512158e-02 1.82216406e-01 4.08223718e-01 -2.29335636e-01
9.19618905e-01 -6.88699782e-01 -6.59716129e-02 4.98430818e-01
3.26512828e-02 -1.23315260e-01 3.47515076e-01 1.90866172e-01
1.08901215e+00 -8.16511288e-02 6.44158721e-01 -3.65541816e-01
3.11631829e-01 4.28470463e-01 5.65076113e-01 6.08445585e-01
3.25195342e-02 5.59642076e-01 6.19348586e-01 -2.94228315e-01
-7.64737368e-01 -7.56682515e-01 -6.21085204e-02 8.60058248e-01
-7.12909270e-03 -4.32259530e-01 -4.79419559e-01 -5.80053985e-01
6.82509318e-02 1.53163469e+00 -5.22058487e-01 -4.14122194e-01
-5.70621133e-01 -7.49269068e-01 7.12945223e-01 5.86692750e-01
7.86518529e-02 -1.19046700e+00 -5.40943086e-01 5.09201467e-01
-2.72033274e-01 -1.25057018e+00 -1.57319665e-01 4.04828817e-01
-8.18923116e-01 -1.00918388e+00 -4.70844448e-01 -5.94465673e-01
5.35956383e-01 3.54605377e-01 1.29901290e+00 2.87912399e-01
4.46603373e-02 -6.70676157e-02 -4.60787237e-01 -3.95083129e-01
-5.77989042e-01 3.84642810e-01 -3.44620869e-02 -3.89666229e-01
1.00523829e+00 -5.21491051e-01 -2.46433914e-01 -7.99839273e-02
-1.11152077e+00 -1.15415186e-01 5.58734834e-01 7.29587615e-01
2.52949178e-01 -9.42937806e-02 7.12515712e-01 -1.23160064e+00
1.20534086e+00 -4.43170428e-01 -3.67670834e-01 1.32386699e-01
-9.31474864e-01 3.62224996e-01 5.01634181e-01 -4.49889779e-01
-1.10961556e+00 -2.17633843e-01 -4.46531504e-01 4.14187200e-02
-1.45252183e-01 9.69242692e-01 -4.60747555e-02 1.65955484e-01
6.98103011e-01 2.63050228e-01 -1.90330476e-01 -5.16922414e-01
6.40823781e-01 5.61524987e-01 3.47520709e-01 -3.56067061e-01
4.23578113e-01 1.29761487e-01 -2.34852210e-01 -9.12765861e-01
-1.30223441e+00 -6.32174671e-01 -5.02808630e-01 4.15921748e-01
6.03906810e-01 -7.81780064e-01 1.29850179e-01 -3.36694680e-02
-1.34343576e+00 1.61363129e-02 -2.43417248e-01 2.17850104e-01
-1.57705635e-01 2.88289011e-01 -1.88778132e-01 -2.90613741e-01
-4.40963805e-01 -8.37182701e-01 5.70872426e-01 8.45927298e-02
-9.45363045e-01 -1.19043219e+00 1.90937147e-01 4.20964777e-01
6.14822745e-01 -3.24722290e-01 1.46603119e+00 -1.08551943e+00
1.14136569e-01 -1.03293829e-01 -2.74041831e-01 2.40468651e-01
3.30347925e-01 -3.73762012e-01 -8.67173254e-01 -3.34667042e-02
-1.29653797e-01 -1.40488952e-01 9.66179609e-01 3.88595387e-02
4.85267699e-01 -1.68508723e-01 -5.06556332e-01 1.81794465e-01
1.49658394e+00 1.98884914e-03 5.04963636e-01 1.87123194e-01
3.58676612e-01 4.84895557e-01 5.97044408e-01 7.69130960e-02
-1.53848538e-02 4.30618435e-01 1.17735818e-01 9.11366567e-02
-3.53115737e-01 -3.73644531e-01 4.70138602e-02 6.82115078e-01
2.44698286e-01 -4.74811345e-01 -1.13818145e+00 1.01478374e+00
-1.63447607e+00 -6.26371682e-01 1.16565019e-01 1.86820531e+00
9.48302627e-01 2.67516971e-01 -4.54102397e-01 2.04801083e-01
5.81110954e-01 3.41010958e-01 -3.27308506e-01 -5.93121529e-01
-1.63934380e-01 3.94861102e-01 2.89211184e-01 3.94048065e-01
-5.96333385e-01 1.42672849e+00 6.97036171e+00 7.21593082e-01
-1.07635474e+00 4.20699492e-02 1.45901993e-01 1.35627881e-01
-6.04192674e-01 2.51134962e-01 -8.64914834e-01 3.12117636e-01
9.39795077e-01 -4.56726164e-01 4.27491814e-01 5.80465078e-01
1.14372611e-01 -1.72461122e-01 -1.28139210e+00 8.38041544e-01
1.12087227e-01 -1.30090666e+00 4.79767203e-01 -1.30464241e-01
4.54438806e-01 1.68235600e-01 -2.57946193e-01 2.86455542e-01
3.58972937e-01 -1.21123624e+00 4.40126091e-01 1.04867555e-01
6.34088993e-01 -7.66950667e-01 9.47333813e-01 2.66642630e-01
-7.52476454e-01 7.14161387e-03 -6.47114635e-01 -3.81350189e-01
4.97054905e-01 9.85355914e-01 -8.14013302e-01 2.40972146e-01
1.22445323e-01 5.93602777e-01 -2.15869352e-01 4.91147369e-01
-7.26454198e-01 7.83883333e-01 -4.76980716e-01 -3.10282350e-01
4.01039749e-01 1.06654108e-01 6.45059288e-01 1.37095857e+00
2.39711538e-01 3.19902569e-01 -1.42096579e-01 1.11398017e+00
-1.90553740e-01 2.13103592e-01 -7.60403872e-01 -7.61736512e-01
5.68990946e-01 9.00787771e-01 -7.29632556e-01 -5.04476488e-01
-3.39353681e-01 7.84750581e-01 7.39430547e-01 3.45940053e-01
-2.32479855e-01 -2.53540546e-01 9.82101500e-01 4.30871807e-02
4.12164837e-01 -2.44685665e-01 -5.93027592e-01 -1.22753716e+00
-1.43830657e-01 -8.62390935e-01 4.03631359e-01 -5.46401381e-01
-1.37715983e+00 2.98517942e-01 -1.38280332e-01 -5.19124985e-01
-3.88899684e-01 -7.23106325e-01 -5.15435576e-01 1.11538005e+00
-1.56099916e+00 -9.56052899e-01 3.40242535e-01 2.68045366e-01
6.67106271e-01 -8.25715959e-02 1.25494051e+00 1.32366389e-01
-2.11817801e-01 2.37480789e-01 -1.81998432e-01 1.56345695e-01
5.26240289e-01 -1.07265544e+00 6.56203151e-01 9.55030859e-01
5.43240130e-01 1.08917058e+00 9.52818215e-01 -7.53149331e-01
-1.32799494e+00 -7.91429520e-01 1.50946558e+00 -3.27545166e-01
8.97666872e-01 -9.19373855e-02 -1.17143929e+00 6.35639489e-01
1.32558867e-01 -3.09226274e-01 8.83345068e-01 4.91238952e-01
-4.50371325e-01 3.32610339e-01 -9.87266064e-01 6.81621790e-01
9.55256283e-01 -5.32817185e-01 -1.63169122e+00 1.25472210e-02
8.23175907e-01 -2.18418345e-01 -6.68003380e-01 2.56332129e-01
3.72648865e-01 -4.39077854e-01 8.51650953e-01 -8.09429526e-01
3.58497530e-01 -1.80085465e-01 -1.49005681e-01 -1.70031905e+00
-4.33502734e-01 -3.97942781e-01 4.66836654e-02 1.25394571e+00
6.74657464e-01 -6.24999583e-01 4.70992208e-01 6.19167030e-01
1.40383184e-01 -4.29941118e-01 -9.34208632e-01 -6.53710008e-01
5.00057161e-01 -6.84062839e-01 4.91073012e-01 1.07452381e+00
4.24535692e-01 9.69690144e-01 2.45735303e-01 -9.95106548e-02
2.33973771e-01 2.29889050e-01 2.62005478e-01 -1.23455536e+00
-4.74387817e-02 -5.04355967e-01 -2.09212571e-01 -1.12678337e+00
2.67491430e-01 -1.08157229e+00 7.90869892e-02 -1.81492710e+00
-5.18692732e-02 -5.15056551e-01 -4.77343857e-01 7.41285801e-01
-1.79575115e-01 1.33035451e-01 1.45128369e-01 -2.70092338e-02
-8.89941528e-02 4.84496653e-01 8.25255275e-01 -1.39046118e-01
-8.68559331e-02 -4.38181221e-01 -1.23330033e+00 8.02808285e-01
9.12227750e-01 -7.01064229e-01 -5.02383053e-01 -6.51869953e-01
7.45238423e-01 -3.06739271e-01 -8.39123428e-02 -6.73022568e-01
1.01946846e-01 -3.20696026e-01 9.45115164e-02 -4.18954343e-01
2.72674650e-01 -6.63666189e-01 -1.21980309e-01 3.88485312e-01
-5.47567964e-01 8.67958739e-02 3.66616338e-01 4.36496586e-01
-2.70526022e-01 -4.86052990e-01 6.04249120e-01 -3.41709644e-01
-7.64282584e-01 -3.55348408e-01 -5.60324132e-01 4.53334630e-01
4.63353604e-01 -4.13497277e-02 -1.20896392e-01 -1.20025001e-01
-2.90633202e-01 -2.45199818e-02 3.93454641e-01 7.82807827e-01
6.20637000e-01 -1.10438836e+00 -5.15636861e-01 3.19671482e-01
9.90398675e-02 -2.47526973e-01 -3.45466077e-01 4.28477794e-01
-1.33788317e-01 7.85477579e-01 5.40199243e-02 -1.06363587e-01
-1.07638371e+00 5.28394043e-01 1.23914361e-01 -4.82176930e-01
-7.36662865e-01 6.79938674e-01 -8.17278773e-02 -2.81243354e-01
-3.13855223e-02 -5.54829180e-01 -5.10769010e-01 1.74802139e-01
5.79200447e-01 1.47321038e-02 2.49733180e-01 -6.44212425e-01
-4.11532640e-01 3.38816226e-01 -2.31392145e-01 -1.16791047e-01
1.37813997e+00 -1.45837143e-01 -1.12425745e-01 4.24419582e-01
7.15148032e-01 -2.29100913e-01 -3.67879361e-01 -4.41669166e-01
2.59293556e-01 -3.89745206e-01 4.62175667e-01 -9.30895448e-01
-7.48222947e-01 1.04816377e+00 1.21627852e-01 1.10958889e-01
7.77672410e-01 -1.10661775e-01 1.13384330e+00 6.89043462e-01
1.70659006e-01 -1.04697359e+00 -3.46120059e-01 9.73215759e-01
6.88352406e-01 -1.26066518e+00 -8.31000358e-02 -4.55358326e-01
-3.37102056e-01 1.11958766e+00 2.07668513e-01 -2.96228714e-02
5.80951571e-01 2.02272847e-01 1.66151524e-01 -2.78919369e-01
-7.98801959e-01 -2.72577763e-01 4.96395789e-02 5.03893197e-01
7.70820260e-01 -1.04305886e-01 -8.65932584e-01 6.13000810e-01
-2.20022872e-01 4.15475555e-02 4.45531547e-01 7.46039152e-01
-5.00940979e-01 -1.28537071e+00 -2.28450626e-01 4.52395082e-01
-7.86957502e-01 -7.20954061e-01 -4.65513974e-01 5.56265354e-01
3.00325286e-02 9.88428473e-01 1.38272747e-01 1.02031622e-02
2.53417850e-01 4.51467574e-01 3.46199304e-01 -9.57216263e-01
-7.13211477e-01 -2.40577519e-01 4.87708211e-01 -3.85176510e-01
-3.66157264e-01 -4.80534256e-01 -1.58215499e+00 -1.16216376e-01
-3.66805762e-01 1.48163080e-01 5.74520886e-01 1.44552112e+00
3.95792574e-01 2.66771585e-01 -2.72599578e-01 -3.22696954e-01
-6.74639344e-01 -1.22598612e+00 -3.75567406e-01 7.26893365e-01
1.39852008e-02 -7.01735914e-01 -2.87300855e-01 -2.59287879e-02] | [10.408653259277344, 8.91614818572998] |
71bad3c0-04a1-4770-af22-53b2a8f4370a | study-of-list-based-omp-and-an-enhanced-model | 2105.03774 | null | https://arxiv.org/abs/2105.03774v1 | https://arxiv.org/pdf/2105.03774v1.pdf | Study of List-Based OMP and an Enhanced Model for Direction Finding with Non-Uniform Arrays | This paper proposes an enhanced coarray transformation model (EDCTM) and a mixed greedy maximum likelihood algorithm called List-Based Maximum Likelihood Orthogonal Matching Pursuit (LBML-OMP) for direction-of-arrival estimation with non-uniform linear arrays (NLAs). The proposed EDCTM approach obtains improved estimates when Khatri-Rao product-based models are used to generate difference coarrays under the assumption of uncorrelated sources. In the proposed LBML-OMP technique, for each iteration a set of candidates is generated based on the correlation-maximization between the dictionary and the residue vector. LBML-OMP then chooses the best candidate based on a reduced-complexity asymptotic maximum likelihood decision rule. Simulations show the improved results of EDCTM over existing approaches and that LBML-OMP outperforms existing sparse recovery algorithms as well as Spatial Smoothing Multiple Signal Classification with NLAs. | ['R. C. de Lamare', 'W. S. Leite'] | 2021-05-08 | null | null | null | null | ['direction-of-arrival-estimation'] | ['audio'] | [ 1.00279093e-01 -4.95744824e-01 8.21372047e-02 -1.17160967e-02
-1.04198086e+00 -2.67209679e-01 2.49021232e-01 1.34557709e-01
-4.16372567e-02 5.04031837e-01 5.46743453e-01 -4.02016908e-01
-6.32911623e-01 -4.68296915e-01 -4.21557248e-01 -8.29634726e-01
-8.34543943e-01 1.78305164e-01 -2.53180891e-01 1.46723613e-01
3.18253368e-01 7.05971301e-01 -8.22352767e-01 -1.24553526e-02
8.01703990e-01 8.46709669e-01 3.08111012e-01 7.43764281e-01
4.74761099e-01 5.13824403e-01 -1.96210384e-01 3.88007089e-02
7.51626432e-01 -4.98009592e-01 3.78607929e-01 -5.12096845e-03
1.42587095e-01 -4.11010757e-02 -3.79438430e-01 1.00437391e+00
9.04355645e-01 4.15629178e-01 7.41979420e-01 -9.67489958e-01
-1.81100875e-01 5.37036300e-01 -1.05069423e+00 3.35892946e-01
6.34900808e-01 -6.16422594e-01 3.11193824e-01 -1.56950581e+00
1.30311966e-01 1.24034131e+00 1.18555617e+00 -6.53156161e-01
-9.06061888e-01 -7.61015952e-01 -5.20880342e-01 1.83045492e-01
-2.07217169e+00 -7.67978072e-01 7.01179564e-01 -3.12732428e-01
6.67785048e-01 3.27552110e-01 2.10666597e-01 9.00075436e-02
3.46646428e-01 5.07533967e-01 1.16900671e+00 -9.85460997e-01
2.55215257e-01 -1.75084651e-01 1.06500857e-01 6.62528574e-01
8.60303164e-01 3.40806127e-01 -8.27569067e-01 -9.39944506e-01
9.18577373e-01 -3.12198639e-01 -4.18661475e-01 -3.36657494e-01
-1.43673062e+00 8.12910616e-01 -8.96320269e-02 6.26480699e-01
-1.05730247e+00 3.24909687e-02 -1.57754749e-01 2.43775934e-01
5.19734845e-02 4.99991596e-01 7.21356422e-02 4.26596940e-01
-1.56284511e+00 6.55045956e-02 9.01825845e-01 1.04676592e+00
5.14961779e-01 7.97662377e-01 -1.71615317e-01 7.99011827e-01
9.01207149e-01 1.23027968e+00 2.13480949e-01 -7.48791635e-01
3.11597139e-01 -3.89084607e-01 4.13094699e-01 -1.59287155e+00
-3.65481108e-01 -1.45661545e+00 -9.50000405e-01 -2.14990094e-01
1.12447612e-01 -7.48112261e-01 -3.25823873e-01 1.31532025e+00
5.11581421e-01 1.02785242e+00 5.29489577e-01 1.01821661e+00
4.00144190e-01 9.37385499e-01 -4.60838675e-01 -8.72340560e-01
8.16063941e-01 -5.50580740e-01 -8.96871567e-01 -1.72941491e-01
2.92332649e-01 -1.36946416e+00 -3.89295846e-01 6.30507827e-01
-9.49068785e-01 -4.66816872e-01 -1.21825397e+00 8.93065989e-01
5.35152912e-01 4.03517365e-01 5.23496211e-01 7.61995733e-01
-8.63234580e-01 1.01350136e-01 -4.28206652e-01 5.96924908e-02
-1.77088723e-01 1.45278662e-01 -2.87639648e-01 -3.43067050e-01
-6.75885677e-01 5.93752921e-01 -9.72213224e-02 8.18797126e-02
-5.79521716e-01 -5.21158576e-01 -9.07788038e-01 -7.14096203e-02
-6.40948862e-02 -4.26040769e-01 6.21205091e-01 -5.76269090e-01
-1.00046694e+00 1.52295485e-01 -7.45327592e-01 -6.26612246e-01
-1.98780194e-01 -1.65981948e-01 -8.92286479e-01 2.74810016e-01
3.55852246e-01 -2.46872559e-01 1.24335968e+00 -1.27586615e+00
-3.65081996e-01 -1.56461105e-01 -9.65062559e-01 3.24322790e-01
3.49650651e-01 -5.95112108e-02 -9.57905948e-02 -1.16210985e+00
1.10541868e+00 -8.86609852e-01 -6.87577486e-01 -5.59329808e-01
-9.96434242e-02 4.80767339e-01 3.82928550e-01 -1.04864132e+00
1.44373584e+00 -2.47024369e+00 -1.03180639e-01 7.59159625e-01
-2.02372372e-01 -7.55973011e-02 -2.91610837e-01 7.01618135e-01
-2.98590183e-01 -8.70702684e-01 -2.11573929e-01 -9.96615216e-02
-3.24189067e-01 -3.60608399e-01 -2.12750763e-01 1.09132874e+00
-5.66888094e-01 1.19134352e-01 -8.19245756e-01 -2.27615818e-01
1.29720420e-01 3.81676853e-01 -4.30762261e-01 4.71665524e-02
5.73738337e-01 6.08163655e-01 -3.75549793e-01 8.19417059e-01
1.06554151e+00 -5.16969189e-02 2.17652142e-01 -6.04805887e-01
-5.50331652e-01 -3.39885980e-01 -2.07635713e+00 1.46891785e+00
-3.01108062e-01 7.44796932e-01 4.70334888e-01 -1.11104643e+00
1.03799355e+00 7.64999390e-01 7.73128688e-01 -2.71354258e-01
1.73230708e-01 7.72968650e-01 9.52933505e-02 -3.25990945e-01
1.82874769e-01 -1.95281096e-02 1.51836708e-01 3.03327858e-01
-1.40083089e-01 3.24622095e-01 -6.82178289e-02 2.51407981e-01
9.62558031e-01 -2.55617261e-01 1.09173667e+00 -6.61709726e-01
7.63705194e-01 -2.68169417e-04 6.66260839e-01 1.15796769e+00
4.50571999e-02 3.86622250e-01 -6.84014559e-01 3.71904485e-02
-6.74337745e-01 -9.14541423e-01 -2.83555925e-01 4.99455899e-01
2.73341566e-01 7.78641105e-02 -3.76140364e-02 1.27254799e-01
-6.67964341e-03 9.68304873e-01 2.08548322e-01 3.70061785e-01
-6.54333115e-01 -1.03372192e+00 2.87709057e-01 -2.12789789e-01
4.88518685e-01 -1.36201650e-01 -2.98884183e-01 6.79799080e-01
-1.95149422e-01 -9.27822113e-01 -2.07448483e-01 3.08701606e-03
-9.16245520e-01 -6.74001098e-01 -8.76347125e-01 -8.00779104e-01
7.46249914e-01 1.06118023e+00 5.66622913e-01 -5.01034975e-01
1.60421077e-02 6.83054984e-01 -5.99690378e-01 -3.20981681e-01
-4.32559699e-02 -9.27055001e-01 3.89374286e-01 6.37602270e-01
1.91543832e-01 -1.01347494e+00 -4.70852852e-01 2.22123727e-01
-2.07393870e-01 -1.58232346e-01 1.04782593e+00 5.63488126e-01
5.19575536e-01 3.79876316e-01 6.56180203e-01 -1.18951179e-01
6.62750602e-01 -9.68557298e-01 -6.19495153e-01 1.21173076e-01
-5.70609987e-01 -1.97525881e-02 2.57555693e-01 -5.03580511e-01
-1.12400711e+00 1.26083672e-01 9.17145237e-02 -3.31669867e-01
9.46490988e-02 9.04719114e-01 1.63337708e-01 -6.66326106e-01
5.60407698e-01 6.19832098e-01 -4.36560452e-01 -5.22097349e-01
1.18311360e-01 6.58482611e-01 5.95621109e-01 -2.56397039e-01
1.08858001e+00 4.82424706e-01 4.57088470e-01 -1.30868065e+00
-3.62906933e-01 -1.13004112e+00 -3.53849798e-01 -7.89820179e-02
2.67376035e-01 -1.25562239e+00 4.64333716e-04 3.55099410e-01
-9.46689069e-01 2.84401119e-01 1.83691278e-01 1.50881147e+00
-1.16863839e-01 6.92716539e-01 -2.23005727e-01 -1.21518862e+00
-4.06257987e-01 -7.17836380e-01 5.89772463e-01 -4.00299728e-02
-2.00751886e-01 -8.32320452e-01 1.86002508e-01 -5.25328778e-02
6.67133987e-01 1.75661638e-01 4.32334393e-01 -6.57508194e-01
-5.54190576e-01 -4.88365173e-01 3.37277986e-02 -2.10316613e-01
-4.72371560e-03 -9.34652150e-01 -2.31373295e-01 -6.25407159e-01
4.07656282e-01 5.29295385e-01 8.47694427e-02 1.17620742e+00
4.16622572e-02 -5.75156569e-01 -6.09996021e-01 7.53917158e-01
1.91639328e+00 4.87005800e-01 4.64962840e-01 2.91345805e-01
2.06027463e-01 -4.42006849e-02 8.79485726e-01 1.12776661e+00
-4.18665037e-02 6.17060244e-01 -2.70840041e-02 7.63472989e-02
-3.27516459e-02 2.52834439e-01 1.56666413e-01 1.22124267e+00
2.02398196e-01 -2.82634139e-01 -5.64637423e-01 7.12704599e-01
-1.57043183e+00 -1.07992053e+00 -8.35480928e-01 2.35325360e+00
3.35115433e-01 -5.20436287e-01 -2.89662153e-01 1.90799758e-01
6.15184247e-01 1.24297403e-01 -1.53630823e-01 8.46649334e-03
-3.54012042e-01 2.61667639e-01 1.00401509e+00 8.00599992e-01
-8.11169624e-01 1.92164183e-01 5.52846241e+00 8.28406274e-01
-9.42100465e-01 4.26069736e-01 -3.00955564e-01 4.21823949e-01
-1.55025020e-01 2.80051351e-01 -6.64314508e-01 2.27919444e-01
7.59800673e-01 -3.21909159e-01 2.83288687e-01 5.59531569e-01
5.57485998e-01 -4.14483935e-01 -4.20836061e-01 1.68831563e+00
4.12997037e-01 -1.08947980e+00 -3.52408409e-01 -6.67621866e-02
1.00057650e+00 -6.13952428e-02 -9.71889049e-02 -5.25692962e-02
-2.13092398e-02 -5.40707648e-01 6.17545784e-01 5.20995140e-01
4.20515567e-01 -3.92214000e-01 6.24930382e-01 5.31897187e-01
-1.47896993e+00 -1.02517799e-01 -3.11545372e-01 -3.18433307e-02
7.08658278e-01 1.16581845e+00 -1.05940390e+00 9.22889352e-01
1.25856385e-01 3.44842821e-01 1.19050719e-01 1.84529388e+00
1.04505472e-01 9.85025942e-01 -6.67796671e-01 3.03268343e-01
2.33912125e-01 -4.00718182e-01 1.42823172e+00 1.41842091e+00
1.30384839e+00 6.18657649e-01 4.20952320e-01 1.70130953e-01
6.16284907e-01 4.36228931e-01 -4.82004344e-01 4.14838403e-01
1.00769317e+00 8.93851876e-01 -4.69244510e-01 -4.27211851e-01
-5.46250522e-01 6.07342005e-01 -7.03120172e-01 5.84357798e-01
-4.27917451e-01 -2.73557961e-01 -3.14420313e-02 2.22926274e-01
4.30717111e-01 -7.73838937e-01 -2.99298763e-01 -9.46047843e-01
-4.54713434e-01 -1.22622800e+00 3.30031842e-01 -8.18156838e-01
-8.80832493e-01 4.48140830e-01 1.63430840e-01 -1.76774704e+00
-2.66712457e-01 4.01387503e-03 -4.72892940e-01 9.68828201e-01
-1.28994262e+00 -8.58441591e-01 1.00053072e-01 6.13714457e-01
4.96105492e-01 -6.50468767e-01 7.44439185e-01 5.63209653e-01
-1.39833251e-02 3.25075418e-01 6.43418849e-01 -1.29982725e-01
4.17949855e-01 -4.44749713e-01 -3.90113562e-01 1.58481979e+00
2.60649592e-01 7.95928776e-01 1.25830960e+00 -8.27907622e-01
-1.60936177e+00 -8.39646697e-01 8.25112283e-01 5.39323092e-01
2.80836582e-01 3.57445627e-01 -5.95394194e-01 6.38996243e-01
3.36662799e-01 -2.29352131e-01 1.14185834e+00 -1.95201144e-01
3.14363302e-03 6.01078533e-02 -9.41866279e-01 2.57812053e-01
4.05853361e-01 3.08108218e-02 -7.08576620e-01 4.74856645e-01
-1.57859728e-01 -2.53226012e-01 -8.23548555e-01 8.01157176e-01
5.46313584e-01 -7.38960028e-01 1.19069898e+00 2.91990012e-01
-5.60170591e-01 -8.29406381e-01 -1.00818944e+00 -1.16864765e+00
-1.00387752e+00 -9.22600985e-01 -1.67133331e-01 7.06771910e-01
3.35302263e-01 -5.35755455e-01 3.76485825e-01 -3.02788228e-01
-1.81867421e-01 -2.34621003e-01 -9.22131181e-01 -8.38234723e-01
-9.02978659e-01 -6.19940817e-01 1.77076712e-01 1.27402043e+00
-2.30083123e-01 2.70323008e-01 -8.30263317e-01 1.35666955e+00
1.44713271e+00 3.75734538e-01 8.14946234e-01 -9.25022304e-01
-1.00850570e+00 2.95678914e-01 -3.56958747e-01 -1.34057283e+00
-2.76844144e-01 -8.49957407e-01 3.21302861e-02 -1.52072370e+00
-1.52052626e-01 -8.12331438e-01 -1.71487421e-01 -2.13787645e-01
4.69431141e-03 6.92335367e-02 4.98447269e-02 3.64033192e-01
-4.60379601e-01 2.42436066e-01 5.90124547e-01 1.95082277e-01
-4.12268460e-01 3.53004158e-01 -3.64415377e-01 6.45912826e-01
4.30955619e-01 -8.69984686e-01 -2.07975775e-01 -3.70000750e-01
-3.04138828e-02 9.73497570e-01 5.95695265e-02 -1.46333170e+00
5.45419157e-01 4.48070914e-02 4.62944508e-01 -9.19489920e-01
4.06500995e-01 -8.31391931e-01 7.98203647e-01 6.16694689e-01
9.66455266e-02 -9.29921791e-02 1.14834644e-02 7.46800005e-01
-2.64686048e-01 -5.30408502e-01 7.40983248e-01 7.78614581e-02
-7.64178634e-01 -1.36585653e-01 -8.10513556e-01 -5.53802490e-01
7.74510920e-01 -3.28644633e-01 3.39355081e-01 -7.33161569e-01
-7.37352729e-01 -2.45150715e-01 -2.07444668e-01 -2.90668607e-01
7.86247075e-01 -1.34852839e+00 -1.25185406e+00 4.22993749e-01
-3.40002030e-01 -6.89912796e-01 3.29700649e-01 1.37542999e+00
-3.21232826e-01 6.69163108e-01 1.35584384e-01 -5.22443354e-01
-1.25671530e+00 2.12300301e-01 3.25265564e-02 -1.88702926e-01
-2.64107883e-01 9.51832175e-01 -9.82383415e-02 -6.51889816e-02
-1.13836288e-01 4.79698867e-01 -2.17908144e-01 -4.11128491e-01
5.78974664e-01 7.10792303e-01 -7.87833706e-02 -1.11748314e+00
-3.25090736e-01 6.58086002e-01 6.67028666e-01 -5.27727664e-01
1.24309731e+00 -4.03811902e-01 -3.61204594e-01 1.01320772e-02
1.18873489e+00 1.14581239e+00 -4.78769779e-01 -4.89799410e-01
-2.30159350e-02 -9.23942626e-01 3.75327557e-01 -3.77048314e-01
-4.44401652e-01 6.60218745e-02 8.96592498e-01 -4.22816008e-01
1.18757856e+00 -4.88792419e-01 4.70219791e-01 4.18147564e-01
7.47933447e-01 -3.24119568e-01 -9.66021121e-02 2.95579582e-01
9.82186913e-01 -7.22773612e-01 6.12492979e-01 -5.44315338e-01
-1.15177698e-01 9.11509514e-01 -1.12361267e-01 -4.27953243e-01
9.61848497e-01 4.56186295e-01 -4.01334651e-02 -5.69106452e-02
-1.68950573e-01 -4.06365795e-03 4.39947218e-01 6.36319935e-01
3.22964638e-01 1.56103954e-01 -8.04322183e-01 4.89947289e-01
-4.30958718e-02 -1.82648644e-01 3.65476400e-01 1.00590146e+00
-8.81727397e-01 -8.99141312e-01 -1.55702412e+00 5.06004870e-01
-4.46017444e-01 -3.64009112e-01 6.23913646e-01 2.98770189e-01
-7.16105253e-02 1.33706152e+00 -3.43676448e-01 -8.35695192e-02
2.68834885e-02 -2.99523234e-01 4.47905660e-01 -3.94498229e-01
1.63497180e-01 9.70881402e-01 2.16999039e-01 -1.39351049e-03
-3.89149696e-01 -1.01861179e+00 -1.07738912e+00 9.05543640e-02
-7.08202004e-01 7.84011781e-01 7.26212859e-01 8.33999455e-01
4.03635949e-01 1.40856445e-01 7.93447912e-01 -8.43937278e-01
-5.75815141e-01 -7.74512112e-01 -8.64638925e-01 -1.02892727e-01
2.06781343e-01 -4.28539634e-01 -4.56969261e-01 8.10242891e-02] | [6.480166912078857, 1.354430913925171] |
870a2c10-2339-43ae-8816-f68fe7832ce8 | global-and-local-relative-position-embedding | null | null | https://www.ecva.net/papers/eccv_2020/papers_ECCV/html/4920_ECCV_2020_paper.php | https://www.ecva.net/papers/eccv_2020/papers_ECCV/papers/123700171.pdf | Global-and-Local Relative Position Embedding for Unsupervised Video Summarization | In order to summarize a content video properly, it is important to grasp the sequential structure of video as well as the long-term dependency between frames. The necessity of them is more obvious, especially for unsupervised learning. One possible solution is to utilize a well-known technique in the field of natural language processing for long-term dependency and sequential property: self-attention with relative position embedding (RPE). However, compared to natural language processing, video summarization requires capturing a much longer length of the global context. In this paper, we therefore present a novel input decomposition strategy, which samples the input both globally and locally. This provides an effective temporal window for RPE to operate and improves overall computational efficiency significantly. By combining both Global-and-Local input decomposition and RPE together, we come up with GL-RPE. Our approach allows the network to capture both local and global interdependencies between video frames effectively. Since GL-RPE can be easily integrated into the existing methods, we apply it to two different unsupervised backbones. We provide extensive ablation studies and visual analysis to verify the effectiveness of the proposals. We demonstrate our approach achieves new state-of-the-art performance using the recently proposed rank order-based metrics: Kendall's $ au$ and Spearman's $
ho$. Furthermore, despite our method is unsupervised, we show ours perform on par with the fully-supervised method. | ['Sanghyun Woo', 'Yunjae Jung', 'In So Kweon', 'Donghyeon Cho'] | null | null | null | null | eccv-2020-8 | ['unsupervised-video-summarization'] | ['computer-vision'] | [-1.79301221e-02 -1.37542263e-01 -3.15894693e-01 -3.21397424e-01
-5.33410490e-01 -5.57555020e-01 4.94925767e-01 2.70633101e-01
-6.32902682e-01 4.87867117e-01 5.44963837e-01 -8.23602974e-02
-1.30444899e-01 -5.57203889e-01 -5.51044822e-01 -5.30265808e-01
-3.86534631e-01 -7.88163394e-02 3.06358278e-01 -4.14344370e-02
3.42369288e-01 4.24736142e-01 -1.59452164e+00 3.96570921e-01
6.45184398e-01 8.28781962e-01 3.01253259e-01 8.95272076e-01
-1.65601254e-01 1.15716231e+00 -3.70770782e-01 -3.89693528e-01
9.17055607e-02 -6.60694778e-01 -8.27726245e-01 2.75866717e-01
4.36245233e-01 -3.47675294e-01 -4.02490824e-01 9.26778674e-01
4.23327655e-01 3.27263296e-01 6.35209322e-01 -9.61490035e-01
-2.55728275e-01 5.18532932e-01 -6.26883507e-01 4.98915672e-01
4.74121511e-01 1.28927529e-01 1.40332353e+00 -9.09073532e-01
7.63873041e-01 9.81172562e-01 5.42194545e-01 2.54674524e-01
-1.08065820e+00 -2.60057539e-01 4.72678930e-01 4.48978692e-01
-1.06063974e+00 -5.23893237e-01 9.83479261e-01 -4.12002504e-01
1.11902785e+00 3.10461223e-01 5.37152469e-01 1.01378906e+00
1.62016660e-01 9.08933163e-01 8.58370662e-01 -5.57234049e-01
8.62113908e-02 -2.73887627e-02 4.12769854e-01 8.64551902e-01
-5.15121929e-02 -1.41192093e-01 -7.21191406e-01 2.06974491e-01
6.59364223e-01 7.31212273e-02 -2.54719079e-01 -3.50137025e-01
-1.37033105e+00 5.16658127e-01 2.42804885e-01 5.41290641e-01
-3.38581979e-01 2.42749006e-01 6.39461339e-01 3.10499281e-01
5.90080917e-01 4.28709060e-01 -2.47057527e-01 -3.15534949e-01
-1.18573070e+00 3.35102640e-02 7.76930153e-01 8.01795125e-01
6.97077930e-01 -1.27287850e-01 -2.30519146e-01 7.44664311e-01
1.63664088e-01 -9.92008671e-03 3.61477166e-01 -1.17429543e+00
3.96676779e-01 5.13480604e-01 -2.88216565e-02 -1.47396266e+00
-3.52988511e-01 -4.72314209e-01 -9.87412274e-01 5.24115376e-02
3.09787333e-01 3.87382670e-03 -4.49212790e-01 1.88766742e+00
-4.08989601e-02 3.13922137e-01 -2.66719833e-02 7.42591560e-01
6.77773476e-01 7.51742780e-01 9.84752644e-03 -7.32137561e-01
1.29460359e+00 -1.14631140e+00 -8.96493554e-01 -1.46149322e-01
4.30283964e-01 -5.42908251e-01 1.15942061e+00 3.37725520e-01
-1.18193758e+00 -6.29409552e-01 -1.04616535e+00 -1.21388398e-01
-2.24735051e-01 1.86662167e-01 4.64840859e-01 1.62080586e-01
-1.28715813e+00 7.84456253e-01 -8.81173372e-01 -6.18338764e-01
1.89151019e-01 2.77690440e-01 -5.32216311e-01 -4.71049882e-02
-1.04030991e+00 6.88337147e-01 2.79812336e-01 5.13660200e-02
-5.69087386e-01 -3.13221633e-01 -8.98502707e-01 2.33440831e-01
6.29810810e-01 -7.24991024e-01 9.75091338e-01 -9.03801739e-01
-1.57265234e+00 6.68429196e-01 -5.14517784e-01 -4.60416496e-01
4.11314785e-01 -3.81619424e-01 -6.05303654e-03 6.88918591e-01
-4.30125091e-03 6.21605158e-01 8.95328224e-01 -1.10522783e+00
-5.19028366e-01 -1.31916493e-01 2.80811518e-01 3.39276761e-01
-7.35622287e-01 2.48779461e-01 -8.86780918e-01 -8.79967988e-01
-6.99335560e-02 -6.37629211e-01 -2.40911290e-01 -5.67333326e-02
-1.14101999e-01 -2.08526224e-01 5.88865161e-01 -7.47568130e-01
1.78064442e+00 -2.22623396e+00 4.63115573e-01 4.04301286e-02
3.87390137e-01 1.62983149e-01 -2.34041452e-01 6.59945607e-01
-2.95287162e-01 1.65793106e-01 -3.21772933e-01 -6.61003470e-01
-7.37984180e-02 2.30169222e-01 -1.60649672e-01 4.01714474e-01
2.95543015e-01 9.15569365e-01 -1.05133724e+00 -6.66096091e-01
2.75392950e-01 3.00732613e-01 -7.77004182e-01 2.36090243e-01
-4.93110530e-02 3.81287456e-01 -1.74926564e-01 3.40506554e-01
1.88159853e-01 -2.13912830e-01 3.35985959e-01 -4.75905567e-01
-1.60241112e-01 8.53771269e-02 -1.07928777e+00 1.95045197e+00
-2.67500281e-01 7.87940741e-01 -8.82947072e-02 -1.39210999e+00
5.57636142e-01 3.56228232e-01 6.16446674e-01 -5.75765193e-01
2.26189122e-02 -7.93642998e-02 -2.83334762e-01 -7.72352934e-01
5.86137474e-01 -3.48149240e-02 8.27462226e-03 5.00432968e-01
2.57370323e-01 3.27885807e-01 6.99919820e-01 5.75325489e-01
1.27733505e+00 2.67015934e-01 6.22673750e-01 -3.24128568e-01
6.38868153e-01 -3.34257573e-01 4.92864251e-01 6.49686754e-01
-2.68006444e-01 6.94563210e-01 8.16322446e-01 -2.99778461e-01
-8.07654798e-01 -8.67286980e-01 4.15361434e-01 1.19601643e+00
6.42193556e-02 -9.64589179e-01 -7.65919030e-01 -8.66493523e-01
-4.24110383e-01 4.13405001e-01 -5.13536811e-01 1.74135361e-02
-7.67246306e-01 -4.66243804e-01 3.41446310e-01 5.81747413e-01
4.12441641e-01 -1.10422397e+00 -6.94066226e-01 1.85389683e-01
-4.30824935e-01 -1.31142032e+00 -5.62981606e-01 -5.47676459e-02
-8.87938082e-01 -9.69768882e-01 -5.97577691e-01 -5.89842081e-01
5.70687950e-01 3.54858100e-01 1.18463588e+00 5.77938259e-02
-1.36868089e-01 7.33387828e-01 -6.58679128e-01 2.22082883e-01
-1.41432121e-01 1.05095930e-01 1.43858165e-01 1.06832348e-01
9.10789967e-02 -8.42444241e-01 -7.17868507e-01 9.08851624e-02
-1.06124616e+00 9.43403393e-02 6.93237960e-01 7.80403614e-01
4.85961914e-01 2.13620126e-01 4.25478607e-01 -7.79264390e-01
6.61908090e-01 -1.99461371e-01 -1.54853955e-01 2.48054162e-01
-2.78812766e-01 6.19693138e-02 7.34265327e-01 -2.05875188e-01
-8.53162587e-01 -2.70944387e-02 -5.41346185e-02 -4.97947633e-01
-1.33434132e-01 8.65640640e-01 -1.71322480e-01 3.72281611e-01
3.76901329e-01 2.24723220e-01 7.33666122e-03 -3.71597797e-01
5.67384124e-01 3.62892538e-01 6.33889079e-01 -5.13856113e-01
5.99739850e-01 5.73470950e-01 -6.68463483e-02 -1.03983188e+00
-8.29617798e-01 -7.17182338e-01 -7.56729662e-01 -3.61835241e-01
9.20354307e-01 -8.24498355e-01 -6.38073266e-01 1.70254990e-01
-1.28376472e+00 -1.52916163e-01 -4.19206113e-01 4.76822913e-01
-6.03554964e-01 8.61459672e-01 -7.17860281e-01 -6.66367292e-01
-1.42318323e-01 -9.92826462e-01 6.93419397e-01 -1.63468793e-01
-3.15866709e-01 -1.14557791e+00 9.68676209e-02 1.16578132e-01
1.53840572e-01 1.62878260e-01 6.95524335e-01 -3.89536023e-01
-4.18111533e-01 -8.59099403e-02 -2.69282997e-01 6.20508134e-01
1.85977742e-01 1.63293615e-01 -7.38646865e-01 -3.64972144e-01
1.76166192e-01 -1.48744630e-02 1.19249737e+00 4.17594701e-01
1.15327942e+00 -4.05920178e-01 -1.66307539e-02 3.56339037e-01
1.42939997e+00 -1.33256957e-01 7.18359292e-01 1.72867998e-01
6.84973300e-01 8.70142221e-01 5.79608440e-01 5.02649426e-01
4.38404649e-01 5.66797256e-01 3.16624731e-01 5.31880483e-02
-7.68976584e-02 -1.15978524e-01 7.77791142e-01 1.42522573e+00
-3.72879744e-01 -3.40969533e-01 -7.75744379e-01 5.14846265e-01
-2.19705796e+00 -1.30404997e+00 1.31698579e-01 2.00970674e+00
6.88997090e-01 1.67984411e-01 1.70483977e-01 2.90391237e-01
4.85092700e-01 5.61393738e-01 -1.23202175e-01 -3.10819834e-01
-5.88203222e-02 3.47249731e-02 1.67881250e-01 5.88988483e-01
-1.16309392e+00 8.66677523e-01 6.45886040e+00 8.75445962e-01
-9.58112359e-01 9.36102644e-02 4.97430503e-01 -2.46326551e-01
-1.82144836e-01 -2.47665159e-02 -3.86021703e-01 3.36152881e-01
9.50786531e-01 4.17045057e-02 3.46591830e-01 4.73627657e-01
4.17145431e-01 -2.57054567e-01 -1.34394729e+00 1.14778733e+00
3.55554968e-01 -1.43446016e+00 6.81661516e-02 -1.74246192e-01
5.38876057e-01 -2.37819672e-01 -1.05054684e-01 1.72583714e-01
-8.16567242e-02 -7.80098677e-01 5.86377621e-01 6.15217805e-01
7.82111585e-01 -6.66937053e-01 6.28705204e-01 4.56234574e-01
-1.44049907e+00 -4.02379408e-02 -9.31644812e-02 -3.66661280e-01
2.51907855e-01 6.99209809e-01 -2.37293303e-01 8.73203397e-01
6.88917577e-01 1.28183997e+00 -6.18821561e-01 6.25082314e-01
-3.56428921e-01 4.85200733e-01 -1.93967298e-01 1.10968120e-01
2.63676792e-01 -2.24266797e-01 6.94636822e-01 1.55843151e+00
1.32523417e-01 2.54953653e-01 1.39657766e-01 4.05421525e-01
3.00007872e-02 1.69180810e-01 -6.68043554e-01 -2.26220250e-01
3.12947854e-02 1.17960548e+00 -8.14836740e-01 -6.28850758e-01
-5.14956832e-01 1.11661971e+00 4.98387843e-01 4.98425484e-01
-8.83066714e-01 -2.77441442e-01 5.04510820e-01 -3.82633391e-03
4.01789099e-01 -5.96700251e-01 -1.04832880e-01 -1.58696151e+00
2.60126889e-01 -8.64326179e-01 4.47775602e-01 -5.18050075e-01
-1.21558774e+00 5.93180954e-01 1.16291016e-01 -1.48422134e+00
-3.31126779e-01 -4.81841624e-01 -5.37003934e-01 4.04265195e-01
-1.62021673e+00 -8.72629762e-01 -2.29659468e-01 5.75246811e-01
8.72719646e-01 4.40784842e-02 6.96406484e-01 3.64800990e-01
-7.46979058e-01 4.67271835e-01 -1.33304566e-01 1.27031222e-01
8.48671079e-01 -1.13823640e+00 -3.72571894e-03 1.24721801e+00
4.87488002e-01 7.67799795e-01 7.31700540e-01 -4.00712848e-01
-1.27359450e+00 -8.87990177e-01 9.10193622e-01 -3.84434372e-01
7.82144070e-01 -2.36363187e-01 -7.79734194e-01 6.17046893e-01
5.32168567e-01 2.98337564e-02 7.34480202e-01 1.01683162e-01
-4.45612580e-01 -2.68438756e-01 -5.97103655e-01 8.28541696e-01
1.14837480e+00 -6.88696980e-01 -8.31652880e-01 2.14539114e-02
7.37059653e-01 5.01596965e-02 -6.32204771e-01 4.28347588e-01
5.58399200e-01 -1.38996112e+00 8.63897264e-01 -5.09788632e-01
8.00659835e-01 -3.14589053e-01 -2.53082275e-01 -1.13753974e+00
-3.95223051e-01 -7.49391913e-01 -3.11003417e-01 1.31104815e+00
1.63192689e-01 -2.76426166e-01 5.76725602e-01 3.69325817e-01
-1.01559266e-01 -8.04161310e-01 -7.80854225e-01 -6.51271105e-01
-3.04234087e-01 -7.31666923e-01 -1.01577111e-01 1.01480150e+00
3.62733006e-01 5.51197767e-01 -6.74369514e-01 1.12084173e-01
5.32188475e-01 1.01346202e-01 6.74301505e-01 -8.94762039e-01
-4.71660405e-01 -5.01568019e-01 -4.35932755e-01 -1.35397148e+00
2.44244426e-01 -6.73836529e-01 1.65431779e-02 -1.62425506e+00
4.26509112e-01 1.27459317e-01 -6.38750732e-01 2.59808272e-01
-1.78682223e-01 2.59683639e-01 3.29315275e-01 2.87491024e-01
-1.04596245e+00 5.23901582e-01 9.54545915e-01 -3.62678058e-02
-2.35024139e-01 -4.41336095e-01 -5.36340177e-01 9.27477658e-01
6.35640740e-01 -2.66741097e-01 -4.55889314e-01 -5.79539955e-01
3.12347144e-01 3.40818688e-02 2.48899728e-01 -1.00449753e+00
4.77127731e-01 2.17725467e-02 4.47567031e-02 -4.81975138e-01
2.17906415e-01 -8.06810141e-01 -1.64275959e-01 2.05552116e-01
-3.69357437e-01 1.24129631e-01 -1.36468709e-02 7.14218676e-01
-6.23169184e-01 -4.52705286e-02 4.88548607e-01 -2.08211824e-01
-9.12454009e-01 2.98635930e-01 -4.03270334e-01 -1.24494351e-01
8.73855293e-01 -2.54825503e-01 -1.69536158e-01 -5.38978875e-01
-9.18395996e-01 2.14725718e-01 4.13107961e-01 2.52924681e-01
7.44975686e-01 -1.10497427e+00 -4.83931899e-01 3.56725901e-02
4.84075956e-02 -2.93129653e-01 3.31514686e-01 1.23252988e+00
-5.16095817e-01 2.62327045e-01 -1.30794317e-01 -6.14814997e-01
-1.37719107e+00 6.92464828e-01 -1.26253054e-01 -5.85525632e-01
-6.59184396e-01 6.33991838e-01 3.33512872e-01 5.65410815e-02
3.04463357e-01 -5.03307045e-01 -5.15630901e-01 5.21887124e-01
6.27035439e-01 3.07127297e-01 -3.51559639e-01 -6.33650899e-01
-3.31757635e-01 8.06316733e-01 1.53169585e-02 -1.70014068e-01
1.36706483e+00 -4.54542667e-01 -4.51378614e-01 7.54865825e-01
1.32338572e+00 2.27278322e-01 -1.23323333e+00 -2.68445760e-01
7.51783252e-02 -3.59829009e-01 -1.49809524e-01 -3.33395571e-01
-9.04716611e-01 1.02605712e+00 2.23147064e-01 3.44812065e-01
1.36870170e+00 -2.60802004e-02 5.63109577e-01 3.31761420e-01
2.29425028e-01 -1.00723243e+00 4.47392464e-01 5.38898170e-01
7.87135720e-01 -1.17811298e+00 3.08827579e-01 -4.63151842e-01
-5.87156355e-01 1.09012330e+00 2.60752887e-01 -3.26012075e-01
5.70333362e-01 1.21215887e-01 -5.20251654e-02 -9.36958119e-02
-1.01364326e+00 -2.88321972e-01 4.08967197e-01 1.38865829e-01
4.99271274e-01 -3.15478534e-01 -4.80035692e-01 5.55357218e-01
2.06403986e-01 -8.30458552e-02 4.88145113e-01 9.05423105e-01
-3.61057609e-01 -1.18475497e+00 -1.76219717e-02 2.93013155e-01
-6.28789246e-01 -1.22529849e-01 -2.16401473e-01 6.23213410e-01
-7.46646523e-02 8.98475647e-01 7.47347698e-02 -4.17315692e-01
1.94967657e-01 5.14327735e-02 3.93826097e-01 -6.99789047e-01
-4.37281340e-01 3.65392327e-01 1.07672878e-01 -9.65460956e-01
-1.01356363e+00 -5.14985144e-01 -1.06646562e+00 -1.40200049e-01
-3.01531442e-02 5.17426804e-02 3.22295189e-01 9.11096692e-01
3.69917750e-01 6.82891011e-01 5.88844717e-01 -1.02352834e+00
-2.78488278e-01 -7.56286383e-01 -4.46025789e-01 5.93138874e-01
4.13937271e-01 -4.63768661e-01 -5.48021793e-01 4.30747598e-01] | [10.153538703918457, 0.5615969300270081] |
411428cf-192a-4443-b2cc-520d4617c063 | nondiscriminatory-treatment-a-straightforward | 2101.10913 | null | https://arxiv.org/abs/2101.10913v1 | https://arxiv.org/pdf/2101.10913v1.pdf | Nondiscriminatory Treatment: a straightforward framework for multi-human parsing | Multi-human parsing aims to segment every body part of every human instance. Nearly all state-of-the-art methods follow the "detection first" or "segmentation first" pipelines. Different from them, we present an end-to-end and box-free pipeline from a new and more human-intuitive perspective. In training time, we directly do instance segmentation on humans and parts. More specifically, we introduce a notion of "indiscriminate objects with categorie" which treats humans and parts without distinction and regards them both as instances with categories. In the mask prediction, each binary mask is obtained by a combination of prototypes shared among all human and part categories. In inference time, we design a brand-new grouping post-processing method that relates each part instance with one single human instance and groups them together to obtain the final human-level parsing result. We name our method as Nondiscriminatory Treatment between Humans and Parts for Human Parsing (NTHP). Experiments show that our network performs superiorly against state-of-the-art methods by a large margin on the MHP v2.0 and PASCAL-Person-Part datasets. | ['Yueming Zhang', 'Tong Zhang', 'Guoshan Zhang', 'Min Yan'] | 2021-01-26 | null | null | null | null | ['multi-human-parsing', 'human-parsing'] | ['computer-vision', 'computer-vision'] | [ 4.49770510e-01 7.02015400e-01 -2.59270109e-02 -8.36408436e-01
-6.84431076e-01 -6.22524738e-01 4.96435970e-01 1.09914966e-01
-3.78271192e-01 3.10342729e-01 1.86296850e-02 -1.29621774e-01
4.83724058e-01 -7.27603972e-01 -8.44804883e-01 -4.38660979e-01
3.76013935e-01 9.34551299e-01 5.69201291e-01 -1.72280651e-02
-2.61779845e-01 1.03763841e-01 -1.47146404e+00 6.16042256e-01
6.77814901e-01 9.04595435e-01 1.39135793e-02 5.50618112e-01
-2.29271904e-01 3.65955651e-01 -7.86068261e-01 -1.01756322e+00
4.97960001e-01 -4.82389718e-01 -1.20497143e+00 2.47544810e-01
7.27616727e-01 -3.50473225e-02 8.21398646e-02 1.01322639e+00
4.98349488e-01 -1.40973255e-01 5.22008359e-01 -1.09386671e+00
-4.46044743e-01 9.34350252e-01 -8.70094061e-01 -2.26516709e-01
3.54025900e-01 1.89322978e-01 9.57127154e-01 -6.19131625e-01
7.23552823e-01 1.56935537e+00 8.18292797e-01 9.96956170e-01
-1.26211810e+00 -4.01796639e-01 4.52290058e-01 -1.73075646e-01
-9.98739421e-01 -1.30188569e-01 4.47421014e-01 -5.80036342e-01
6.79707944e-01 4.97339904e-01 7.25525558e-01 8.75793636e-01
1.18878782e-02 1.34422374e+00 8.71238232e-01 -3.82457495e-01
7.29522035e-02 -1.17263839e-01 6.95506036e-01 8.39167953e-01
4.14056480e-01 -1.21521018e-01 -2.09655702e-01 1.53016359e-01
3.44076127e-01 -2.30612233e-01 1.02592193e-01 -3.24734151e-01
-1.36260068e+00 6.45570755e-01 4.86350119e-01 3.80341187e-02
-4.27314788e-01 -2.47238018e-02 4.19826865e-01 -3.60270262e-01
3.75608832e-01 1.14909947e-01 -8.38914871e-01 3.74133766e-01
-1.05458510e+00 3.58577043e-01 8.48989367e-01 1.19584477e+00
6.71377599e-01 -5.97956061e-01 -4.61812258e-01 8.16663444e-01
2.65698850e-01 1.54476851e-01 4.85221267e-01 -9.10730422e-01
4.46598291e-01 9.20235276e-01 -6.64219111e-02 -4.40536082e-01
-7.24121988e-01 -3.74336571e-01 -7.95214593e-01 -6.82024136e-02
7.46329904e-01 -2.12243527e-01 -1.49993455e+00 1.62322617e+00
9.25011456e-01 -1.47096246e-01 -8.65382701e-02 9.33089316e-01
1.13852954e+00 3.28651011e-01 6.17984176e-01 1.86211303e-01
2.13161993e+00 -1.30625308e+00 -5.64628601e-01 -6.96717620e-01
2.53655136e-01 -5.83431542e-01 8.26929569e-01 3.06520790e-01
-1.07753050e+00 -1.09536123e+00 -8.90367806e-01 -4.21834618e-01
-5.98249793e-01 3.74740839e-01 7.31693447e-01 7.31345952e-01
-6.08711779e-01 8.05431962e-01 -9.05272663e-01 -4.72879142e-01
5.69902182e-01 3.42693478e-01 -4.40983802e-01 1.07345439e-01
-7.37319052e-01 7.51700044e-01 5.52914143e-01 1.45752430e-01
-4.79397595e-01 -6.67688251e-01 -1.08825517e+00 -1.51406378e-01
5.42263210e-01 -1.08901358e+00 1.43204904e+00 -9.73491907e-01
-1.03148746e+00 1.60026145e+00 -2.66190976e-01 -5.59598029e-01
8.33141446e-01 -4.01350260e-01 -1.79607943e-01 1.17217831e-01
5.29164732e-01 1.35415685e+00 7.30584562e-01 -1.42485905e+00
-1.06027484e+00 -5.13129592e-01 -1.84184089e-02 -3.18750851e-02
5.13958037e-01 1.84392720e-01 -9.09357786e-01 -4.86410886e-01
5.29610157e-01 -9.28765357e-01 -3.01580399e-01 3.18109132e-02
-1.13037956e+00 -4.78846371e-01 5.79677284e-01 -9.39607620e-01
9.70396936e-01 -2.05970120e+00 2.07391173e-01 -2.05747381e-01
5.35171092e-01 2.48741746e-01 7.69770667e-02 -1.36628985e-01
-8.50626379e-02 2.73452312e-01 -6.09180510e-01 -8.47680688e-01
2.20153481e-01 3.51796836e-01 2.03141704e-01 4.63882208e-01
3.52272034e-01 1.03922009e+00 -9.14787591e-01 -1.00233984e+00
1.99106649e-01 2.54503548e-01 -5.48548341e-01 2.88934857e-01
-1.99313521e-01 3.53384286e-01 -2.92956322e-01 1.03469419e+00
8.33263814e-01 -7.87529051e-02 2.59864032e-01 -2.68254966e-01
1.63690280e-03 1.14825509e-01 -1.04967725e+00 1.69520509e+00
1.68149263e-01 4.37898226e-02 2.52638549e-01 -7.17662275e-01
5.49045622e-01 9.21776742e-02 1.28854483e-01 -2.80853599e-01
2.40254313e-01 3.86462957e-02 1.38267711e-01 -3.75577897e-01
4.72986966e-01 -1.74462557e-01 -4.39006388e-01 -1.06623009e-01
2.65468031e-01 1.85283884e-01 4.98071790e-01 5.59889451e-02
9.51327682e-01 6.33808553e-01 6.14085495e-01 -2.55814016e-01
4.02819008e-01 1.43554807e-01 1.17690396e+00 8.68402243e-01
-6.48054302e-01 9.58425343e-01 6.95935726e-01 -4.87468988e-01
-8.48633707e-01 -1.07399309e+00 -8.62392187e-02 1.39843953e+00
2.25051522e-01 -3.02383035e-01 -1.37309742e+00 -1.17406964e+00
9.05944034e-02 5.92117310e-01 -8.61287475e-01 3.23811948e-01
-8.03087771e-01 -7.31753349e-01 5.27847111e-01 7.95190692e-01
5.72881222e-01 -1.41886556e+00 -7.80541480e-01 2.57062376e-01
-3.00947547e-01 -1.24511552e+00 -4.20449555e-01 4.35296267e-01
-5.03816307e-01 -1.29915357e+00 -7.72861660e-01 -1.08281648e+00
7.70517707e-01 -2.16617763e-01 1.44610584e+00 2.23235607e-01
-5.06016850e-01 -9.53209326e-02 -4.97531921e-01 -4.47940916e-01
-4.66561288e-01 -4.29436043e-02 -2.78375387e-01 -7.63565302e-02
5.03193915e-01 -6.50337860e-02 -6.52249575e-01 3.40484262e-01
-5.35396039e-01 4.29900229e-01 7.28588998e-01 5.08651853e-01
1.05749881e+00 -1.21997237e-01 4.48024906e-02 -1.37955904e+00
-3.64129879e-02 -3.72139663e-01 -3.21582407e-01 4.65673357e-01
1.71270985e-02 -9.09758434e-02 5.59546113e-01 -4.81447041e-01
-9.24902260e-01 6.73680961e-01 -2.36108005e-01 -2.10968658e-01
-7.47565925e-01 -9.65826586e-02 -8.43505323e-01 4.88272518e-01
4.20080304e-01 -1.98987965e-02 -2.84572124e-01 -7.74229050e-01
6.97080076e-01 5.28884292e-01 1.07925034e+00 -7.71047056e-01
5.86543739e-01 3.80805850e-01 -1.87337458e-01 -5.35263062e-01
-1.02065933e+00 -7.08046794e-01 -1.23360431e+00 -6.44339398e-02
1.62928879e+00 -7.37654507e-01 -6.08649194e-01 6.50003612e-01
-1.31665623e+00 -4.21793461e-01 -3.73374939e-01 -5.49257435e-02
-3.25233728e-01 5.91618240e-01 -8.09571624e-01 -7.35319078e-01
-4.57761168e-01 -1.10642910e+00 1.55265355e+00 4.06404376e-01
-2.05908298e-01 -3.09919029e-01 -3.03521603e-01 6.34455383e-01
-3.71868104e-01 4.18191701e-01 7.10363567e-01 -1.06458390e+00
-9.45674628e-02 -4.07009929e-01 -2.30520502e-01 4.04757470e-01
-2.96239972e-01 -2.29824092e-02 -1.06250715e+00 1.25032753e-01
-2.48032674e-01 1.07533425e-01 9.21896696e-01 2.86870718e-01
1.12565994e+00 -1.74761027e-01 -6.73455775e-01 4.67799783e-01
1.11442494e+00 1.03726991e-01 5.52930176e-01 7.27216303e-02
1.06233323e+00 1.03658676e+00 6.35853291e-01 5.46147861e-02
4.90813702e-01 4.76046920e-01 1.90872818e-01 -3.75471890e-01
-3.76754194e-01 -4.35526341e-01 -8.68104920e-02 2.09972963e-01
1.19148441e-01 -4.25220966e-01 -8.44207823e-01 5.58491945e-01
-1.96192944e+00 -8.40879142e-01 -3.89700890e-01 2.07321215e+00
6.85882807e-01 3.92087072e-01 5.52773178e-01 -6.45679012e-02
1.36760461e+00 -2.04731002e-01 -4.70301241e-01 -3.76057178e-02
1.23928554e-01 2.99302310e-01 3.76380503e-01 2.46558562e-01
-1.70188594e+00 1.32852888e+00 5.82971811e+00 6.45444691e-01
-4.08675045e-01 7.66207799e-02 9.55319643e-01 3.81374717e-01
3.10081124e-01 1.62711684e-02 -1.28956974e+00 5.36820352e-01
5.94417393e-01 5.26709616e-01 -5.16671799e-02 1.17431641e+00
-2.03191578e-01 -5.69918007e-02 -1.39635694e+00 6.70850873e-01
-1.54677153e-01 -6.93875015e-01 -2.46919990e-02 -3.03857820e-03
2.62290746e-01 -3.51522624e-01 -4.82528538e-01 6.63194418e-01
1.76029220e-01 -7.47107863e-01 1.31398952e+00 2.52028644e-01
4.10830975e-01 -4.33692694e-01 7.62247741e-01 4.97524023e-01
-1.52930641e+00 3.36626992e-02 -5.06171882e-01 1.21391125e-01
4.43528563e-01 5.26793778e-01 -4.25441384e-01 6.04498327e-01
6.39048219e-01 2.60016859e-01 -7.20735073e-01 1.07441330e+00
-6.24192655e-01 5.00005782e-01 -3.67632389e-01 2.05647767e-01
2.24006958e-02 -3.90556939e-02 2.78517634e-01 1.59709036e+00
-1.52369142e-01 3.73892516e-01 6.01671994e-01 9.68837976e-01
-6.78335503e-02 -9.87064913e-02 1.97319780e-02 2.70248771e-01
3.99796546e-01 1.60642934e+00 -1.34725773e+00 -9.66861010e-01
-2.76892573e-01 1.20607233e+00 3.43765974e-01 -1.74154624e-01
-1.09027064e+00 -2.98657775e-01 4.91282433e-01 1.37595519e-01
5.28252125e-01 2.85002649e-01 -4.47950542e-01 -9.10138667e-01
1.20620154e-01 -8.18885565e-01 6.87485695e-01 -5.32964528e-01
-1.52609813e+00 6.51923180e-01 2.12181568e-01 -9.61874068e-01
-1.17606409e-01 -9.16871846e-01 -5.23039460e-01 6.36612177e-01
-1.06936228e+00 -1.55618703e+00 -2.21085653e-01 3.11338305e-01
7.01925457e-01 3.52664381e-01 6.27757013e-01 1.66181743e-01
-9.26042080e-01 6.92536175e-01 -7.93255091e-01 7.83207953e-01
4.40446377e-01 -1.72228932e+00 9.92529988e-01 1.25970101e+00
-7.49409944e-02 7.34251320e-01 7.47514307e-01 -9.84947383e-01
-8.19036007e-01 -1.23061931e+00 1.09032559e+00 -5.73271513e-01
1.50386199e-01 -6.15426481e-01 -7.00364470e-01 9.22180176e-01
9.51997489e-02 1.07869968e-01 5.39439201e-01 2.43854269e-01
-4.07150835e-01 3.61826718e-01 -1.52897441e+00 4.35331851e-01
1.27167964e+00 1.09889517e-02 -1.01805091e+00 4.60343808e-01
9.31473911e-01 -6.83432698e-01 -6.48079932e-01 6.67532086e-01
5.02061963e-01 -1.02819717e+00 1.02532387e+00 -5.42736948e-01
5.04773676e-01 -5.52575648e-01 -1.17769130e-01 -8.22776496e-01
-4.55820739e-01 -5.14755189e-01 -1.00801244e-01 1.43691182e+00
4.62536126e-01 -3.04450274e-01 8.87956023e-01 7.00763822e-01
-4.95368004e-01 -6.32356107e-01 -7.62003660e-01 -7.51235247e-01
1.51141241e-01 -2.98716605e-01 6.17959023e-01 6.70701385e-01
-7.43892863e-02 3.59690219e-01 -1.61584973e-01 3.58251303e-01
7.36113191e-01 1.68792307e-01 8.55957747e-01 -1.27857804e+00
-6.11546040e-01 -5.10785401e-01 -4.75561649e-01 -1.08294559e+00
8.62281471e-02 -7.39785433e-01 6.43384635e-01 -1.59736454e+00
5.47382891e-01 -1.51952326e-01 3.52342584e-04 9.41135764e-01
-6.87807202e-01 3.33621442e-01 4.32262778e-01 9.67362002e-02
-7.98350990e-01 -1.57448754e-01 1.10781741e+00 -3.01365443e-02
-1.58204764e-01 1.74619436e-01 -8.87150407e-01 9.80760217e-01
5.81405818e-01 -4.68891501e-01 1.79134816e-01 -2.89879978e-01
-3.72696728e-01 -1.08760372e-01 3.56008381e-01 -1.25898874e+00
-5.32880574e-02 1.92340523e-01 6.80451274e-01 -8.74810696e-01
2.24509165e-01 -4.59630787e-01 3.52990739e-02 5.14848650e-01
-8.35321993e-02 -1.87750459e-01 -2.57962923e-02 3.81196171e-01
1.27442986e-01 -2.25041345e-01 9.85827386e-01 -5.63784301e-01
-8.38827312e-01 1.91045240e-01 9.72185731e-02 1.61359668e-01
9.89025235e-01 -2.94008970e-01 -3.82180601e-01 3.82457852e-01
-1.09614587e+00 3.20042342e-01 2.84221947e-01 3.17494124e-01
6.78933039e-02 -7.13375211e-01 -6.57703221e-01 7.67214894e-02
2.66201198e-02 3.85280311e-01 2.34932184e-01 5.66901445e-01
-5.14884889e-01 8.45718384e-02 -1.54172322e-02 -6.61093175e-01
-1.53335595e+00 8.52122068e-01 4.13138509e-01 -2.79252499e-01
-7.30265915e-01 1.17499650e+00 5.35994411e-01 -8.41174841e-01
2.59062111e-01 -4.62431639e-01 -1.33424073e-01 5.89414500e-03
5.12951612e-01 2.31946439e-01 -2.64248431e-01 -8.84806156e-01
-6.05641842e-01 4.64015156e-01 -1.26363458e-02 6.92468360e-02
9.73276138e-01 2.56428093e-01 -1.44399345e-01 1.98659658e-01
9.44507897e-01 -1.14146635e-01 -1.23736048e+00 2.31617048e-01
1.94977876e-02 -6.25819862e-02 -6.84072375e-01 -1.20651317e+00
-1.12864828e+00 7.01474965e-01 5.13992488e-01 2.77301013e-01
8.76786828e-01 5.75407028e-01 8.11033607e-01 -6.28981367e-02
2.47532308e-01 -1.02339506e+00 -4.53184843e-01 3.49965006e-01
4.67305273e-01 -1.24764919e+00 9.23307389e-02 -1.09555995e+00
-7.32380688e-01 7.68520951e-01 9.53049302e-01 -2.86107864e-02
3.20290446e-01 1.08877651e-01 1.52721450e-01 -1.71769798e-01
-1.06143728e-01 -5.83679080e-01 5.75293064e-01 6.27623498e-01
5.63831508e-01 5.09080827e-01 -6.39633298e-01 1.21601868e+00
-4.92544144e-01 -4.10909116e-01 1.29024666e-02 1.08504140e+00
-6.22225165e-01 -1.24111426e+00 -4.60143864e-01 3.81512225e-01
-6.64458454e-01 2.50604004e-02 -6.78636432e-01 1.20406699e+00
8.91873002e-01 9.61772144e-01 1.15617000e-01 -3.50956053e-01
5.22095382e-01 3.85324925e-01 4.01623636e-01 -8.72855961e-01
-9.79784310e-01 1.56373039e-01 2.19829842e-01 -6.41842008e-01
-2.65847355e-01 -6.89078987e-01 -1.62171221e+00 -1.99210607e-02
-1.70296341e-01 -1.52765259e-01 5.82810104e-01 1.06534839e+00
1.36371419e-01 5.92918336e-01 -8.59190673e-02 -1.18098664e+00
-2.64391840e-01 -9.98128295e-01 -3.96495581e-01 7.80599892e-01
-1.95567816e-01 -4.62435991e-01 -1.79924428e-01 1.77917883e-01] | [8.584108352661133, -0.02253217250108719] |
ae717c30-08cf-4a8f-b579-86005a054868 | dynamic-texture-analysis-for-detecting-fake | 2007.15271 | null | https://arxiv.org/abs/2007.15271v1 | https://arxiv.org/pdf/2007.15271v1.pdf | Dynamic texture analysis for detecting fake faces in video sequences | The creation of manipulated multimedia content involving human characters has reached in the last years unprecedented realism, calling for automated techniques to expose synthetically generated faces in images and videos. This work explores the analysis of spatio-temporal texture dynamics of the video signal, with the goal of characterizing and distinguishing real and fake sequences. We propose to build a binary decision on the joint analysis of multiple temporal segments and, in contrast to previous approaches, to exploit the textural dynamics of both the spatial and temporal dimensions. This is achieved through the use of Local Derivative Patterns on Three Orthogonal Planes (LDP-TOP), a compact feature representation known to be an important asset for the detection of face spoofing attacks. Experimental analyses on state-of-the-art datasets of manipulated videos show the discriminative power of such descriptors in separating real and fake sequences, and also identifying the creation method used. Linear Support Vector Machines (SVMs) are used which, despite the lower complexity, yield comparable performance to previously proposed deep models for fake content detection. | ['Giulia Boato', 'Mattia Bonomi', 'Cecilia Pasquini'] | 2020-07-30 | null | null | null | null | ['texture-classification'] | ['computer-vision'] | [ 3.63457680e-01 -1.64448619e-01 6.79029245e-03 -5.89218996e-02
-3.13284159e-01 -6.74925745e-01 1.05352581e+00 1.08137224e-02
-2.56270975e-01 4.11631376e-01 -6.26626089e-02 9.14219320e-02
-1.19800568e-01 -5.25808632e-01 -4.81408328e-01 -9.52215433e-01
-7.11835742e-01 2.30548456e-01 4.06895638e-01 -2.12856591e-01
4.83750343e-01 9.01724815e-01 -1.85208535e+00 6.89437568e-01
1.48925886e-01 1.25578904e+00 -3.84716809e-01 7.19211638e-01
1.39960691e-01 6.50184393e-01 -6.31040752e-01 -5.53309143e-01
2.24924728e-01 -3.13182056e-01 -5.09525657e-01 1.93740696e-01
7.07262456e-01 -5.58486223e-01 -6.23589218e-01 1.11380601e+00
1.84107274e-01 -2.16642410e-01 7.36342490e-01 -1.34952533e+00
-2.80426562e-01 -1.39770824e-02 -5.24679780e-01 6.45168066e-01
7.90446401e-01 9.47517753e-02 4.17283952e-01 -7.36068189e-01
9.86857831e-01 1.27019072e+00 7.97394156e-01 2.66947180e-01
-1.19549179e+00 -5.71714997e-01 -5.15733778e-01 4.84477371e-01
-1.26558566e+00 -6.79708302e-01 1.10952163e+00 -8.50113571e-01
4.02836859e-01 3.08839202e-01 6.87526882e-01 1.65461743e+00
3.26160461e-01 5.46798646e-01 1.42575824e+00 -4.47438329e-01
5.96138500e-02 2.89410323e-01 -2.21400812e-01 1.00245607e+00
1.55689731e-01 3.21182966e-01 -8.25566590e-01 -3.82865220e-01
7.94709980e-01 -3.18215370e-01 -4.22198862e-01 -5.10068417e-01
-1.07244647e+00 9.13250744e-01 -1.00338764e-01 6.20348394e-01
-3.80385637e-01 -3.64860520e-02 6.67486489e-01 4.20175165e-01
4.48667318e-01 3.85157198e-01 1.57058716e-01 -2.40198165e-01
-1.21764696e+00 1.90788120e-01 6.90201700e-01 3.50851953e-01
3.98479849e-01 1.22835994e-01 -3.93754728e-02 3.98163825e-01
-5.08962274e-02 4.21295315e-01 6.16845131e-01 -8.39395583e-01
2.73428321e-01 2.37759888e-01 1.01479337e-01 -1.77144599e+00
-4.44909632e-02 -1.12405298e-02 -7.13876009e-01 2.92049885e-01
6.51874959e-01 3.30623001e-01 -3.77639115e-01 1.19571030e+00
4.54842597e-01 3.77787024e-01 -2.47413993e-01 8.00695062e-01
3.49107653e-01 3.80422026e-01 -2.12629810e-01 -2.81496584e-01
1.27211916e+00 -2.56615788e-01 -7.44714499e-01 3.71931970e-01
3.73206884e-01 -6.13132954e-01 6.24981403e-01 7.25564241e-01
-7.58368790e-01 -4.10900623e-01 -1.11734784e+00 3.66239130e-01
-6.22521996e-01 3.57826427e-02 3.05556118e-01 1.15937245e+00
-7.87819624e-01 1.04630339e+00 -6.79760516e-01 -1.27600193e-01
6.59239769e-01 2.41836950e-01 -8.10994029e-01 1.41521320e-01
-1.15855539e+00 6.75229013e-01 2.44219586e-01 1.78213254e-01
-7.93808579e-01 -2.64350623e-01 -8.02591205e-01 -1.43198088e-01
1.00018613e-01 3.02563488e-01 4.88102496e-01 -1.35874581e+00
-1.49763179e+00 1.20448112e+00 8.48918706e-02 -5.03305554e-01
1.01650870e+00 6.08675741e-02 -5.93674064e-01 8.03434074e-01
-1.06828071e-01 2.55332857e-01 1.61859047e+00 -1.14686942e+00
-1.30095333e-01 -4.64201212e-01 -1.87671572e-01 -4.48137760e-01
-7.05127239e-01 2.44355634e-01 1.25134647e-01 -8.04093778e-01
3.00489292e-02 -7.89442062e-01 3.35089266e-01 1.58522889e-01
-2.76102632e-01 1.90293252e-01 1.25848877e+00 -1.15989769e+00
1.02410793e+00 -2.24254584e+00 2.66608000e-01 3.56015086e-01
2.12977365e-01 4.85739857e-01 -2.53765509e-02 4.82662141e-01
-2.37396687e-01 -1.27049685e-01 -1.06239520e-01 -1.96925625e-01
-8.04495588e-02 -9.00059715e-02 -5.78503191e-01 1.31942737e+00
3.96591783e-01 6.70927703e-01 -7.81947851e-01 -6.16432786e-01
3.43910664e-01 5.99188685e-01 -1.80537418e-01 -9.62883756e-02
7.47522116e-02 5.36080182e-01 -4.32248443e-01 6.21688068e-01
8.11076880e-01 2.88288593e-01 1.79372400e-01 -6.36905730e-02
8.87596980e-03 3.15703452e-02 -1.11685288e+00 1.01847363e+00
1.11079007e-01 1.29866004e+00 1.02969393e-01 -1.26284468e+00
7.86874413e-01 4.22196478e-01 6.54412568e-01 -8.84301305e-01
4.43660617e-01 3.30327123e-01 -3.61071646e-01 -9.64684784e-01
2.83134997e-01 1.17167711e-01 1.02660097e-01 3.07179987e-01
1.65692538e-01 1.10140204e-01 5.00244908e-02 -2.71365821e-01
9.41893280e-01 1.21225230e-01 1.10865682e-01 -2.39878625e-01
7.60744512e-01 -2.96209365e-01 1.41976491e-01 4.80297059e-01
-5.85652649e-01 4.24441099e-01 9.96100128e-01 -6.48904681e-01
-9.27581668e-01 -7.86174953e-01 -2.70550162e-01 5.85505247e-01
1.01538353e-01 -1.31529942e-01 -8.82400870e-01 -7.75831759e-01
1.53726310e-01 1.96248412e-01 -8.96302640e-01 -2.01515511e-01
-8.83938551e-01 -3.63140583e-01 9.08642113e-01 -1.14103526e-01
4.09441531e-01 -1.01020241e+00 -9.91468906e-01 5.19106984e-02
-7.81915560e-02 -1.49715602e+00 1.18480682e-01 -3.34890515e-01
-6.33030832e-01 -1.18391764e+00 -6.48189366e-01 -4.05528903e-01
1.51827201e-01 7.32686147e-02 6.85750961e-01 -1.71677202e-01
-8.05833697e-01 4.85377491e-01 -3.60798568e-01 7.03416020e-02
-6.75177872e-01 -4.67281520e-01 1.89196780e-01 7.49129951e-01
1.47838190e-01 -5.59297919e-01 -2.78512925e-01 3.71716678e-01
-1.06173825e+00 -3.33433181e-01 2.04521179e-01 6.87922180e-01
-6.00436032e-02 1.41595125e-01 1.15291759e-01 -3.78760070e-01
3.73900592e-01 -4.46625859e-01 -6.55325294e-01 1.29981771e-01
8.08587596e-02 -7.30733052e-02 5.38271189e-01 -7.13343322e-01
-5.45771480e-01 -1.99018568e-01 2.52935141e-01 -8.63074243e-01
-3.53898406e-01 3.92962852e-03 1.70258239e-01 -6.60591841e-01
5.65291822e-01 5.01875281e-01 2.74603248e-01 -3.02022487e-01
-3.09181046e-02 5.09785175e-01 3.93272400e-01 -3.24434191e-01
9.52901661e-01 1.01677656e+00 4.82727557e-01 -1.49522710e+00
-4.36768569e-02 -5.12697518e-01 -8.44213843e-01 -6.44891918e-01
8.36727083e-01 -5.07967174e-01 -9.11566317e-01 8.19589496e-01
-1.28376901e+00 4.85050827e-02 1.13205612e-02 3.46661001e-01
-6.87001407e-01 9.39119697e-01 -6.42049432e-01 -1.02166927e+00
2.62583315e-01 -1.02792704e+00 1.29001665e+00 -2.76610255e-01
-1.15223818e-01 -9.33203340e-01 9.89537910e-02 4.40357834e-01
2.37497091e-01 9.32917595e-01 6.24022067e-01 -5.67889094e-01
-5.47621429e-01 -5.35882771e-01 -9.49863791e-02 3.86166662e-01
-1.34639502e-01 1.41390339e-01 -1.18475580e+00 -3.41257006e-01
4.33277518e-01 -2.32041866e-01 7.77693927e-01 4.34396192e-02
8.70492041e-01 -4.55432177e-01 -2.65518129e-01 5.01102865e-01
1.23360598e+00 -8.51133913e-02 6.65605545e-01 3.56564075e-01
4.75527883e-01 1.21936786e+00 5.23174047e-01 5.64353049e-01
-4.59929556e-01 1.15385187e+00 5.88149488e-01 3.05425376e-01
6.77448437e-02 2.42614467e-02 6.67457998e-01 1.10665679e-01
-1.57837793e-01 -7.47678205e-02 -7.35880911e-01 3.09151381e-01
-1.45117795e+00 -1.43830860e+00 -1.82590589e-01 2.30548501e+00
1.99006945e-01 4.64785062e-02 3.34710121e-01 6.34998083e-01
7.99144924e-01 5.65543413e-01 3.65946777e-02 -4.07123834e-01
-3.20369273e-01 2.32397243e-01 6.07305706e-01 2.31101796e-01
-1.50127137e+00 6.86372638e-01 5.67404890e+00 1.04867971e+00
-1.52931881e+00 1.94257498e-01 3.29520315e-01 5.76882735e-02
2.23662719e-01 -4.37542945e-01 -3.38374108e-01 7.81545281e-01
8.33303809e-01 4.28688407e-01 3.08550507e-01 5.19128263e-01
3.26466799e-01 -1.04974948e-01 -7.47696221e-01 1.08806562e+00
5.76750457e-01 -1.35583186e+00 3.97496577e-03 3.77123147e-01
3.59832615e-01 -4.73953396e-01 3.60548109e-01 -2.87747622e-01
-5.09508848e-01 -1.04999936e+00 9.12019610e-01 4.30506378e-01
7.15952873e-01 -5.54308236e-01 5.38737059e-01 1.08480304e-01
-9.11532402e-01 -2.12345809e-01 -4.82139550e-02 2.56125420e-01
1.43961925e-02 3.67425889e-01 -7.82945037e-01 3.94507855e-01
5.63535988e-01 7.84004748e-01 -5.65316021e-01 6.69015408e-01
2.59515405e-01 3.94697577e-01 -3.16078246e-01 1.68831125e-01
2.64796048e-01 -3.16371262e-01 9.54819977e-01 1.36935771e+00
2.04776898e-01 -3.02100599e-01 -1.51075661e-01 6.69247150e-01
3.55404109e-01 -2.31494214e-02 -9.55211103e-01 -4.87781852e-01
-1.53787300e-01 9.41713750e-01 -9.02346373e-01 4.21859547e-02
-9.02487636e-02 1.25677872e+00 -1.69270281e-02 1.35318041e-01
-9.02636111e-01 -2.44708434e-01 7.01003969e-01 4.61561680e-01
5.37242413e-01 -5.34883857e-01 2.04765931e-01 -1.22895825e+00
3.87892008e-01 -8.79381895e-01 1.40047178e-01 -3.83226275e-01
-8.59329939e-01 5.72464049e-01 8.90484732e-03 -1.43678129e+00
-2.79281288e-01 -9.79087949e-01 -3.07040811e-01 5.16368091e-01
-1.38251150e+00 -1.14505994e+00 -2.47954249e-01 7.04114616e-01
3.52187485e-01 -3.04756135e-01 6.86277986e-01 2.14878380e-01
-3.64540398e-01 3.73019218e-01 1.62845314e-01 4.44732070e-01
4.61561799e-01 -7.30027258e-01 2.35622182e-01 6.85122848e-01
2.78427482e-01 1.49946645e-01 1.00693059e+00 -4.49317425e-01
-1.41923213e+00 -4.82334763e-01 6.68872118e-01 -3.26837182e-01
7.13042021e-01 -5.55191100e-01 -8.30521524e-01 1.96364418e-01
-1.12157747e-01 1.30332768e-01 3.43913555e-01 -6.14227057e-01
-7.09734261e-01 1.82485670e-01 -1.39564788e+00 2.18917787e-01
6.53044939e-01 -9.14854646e-01 -3.77453178e-01 4.00472224e-01
-9.63674486e-02 3.54407839e-02 -5.09602726e-01 2.92495430e-01
1.00606489e+00 -1.66654515e+00 1.17694807e+00 -6.31099582e-01
4.56286848e-01 -2.07811259e-02 -1.97214618e-01 -6.69938087e-01
4.91759628e-02 -7.69072890e-01 -2.39719629e-01 8.09785903e-01
-3.67459357e-01 -6.00389183e-01 9.29938495e-01 -1.83578402e-01
5.10300756e-01 -5.50115526e-01 -1.44077516e+00 -8.99171531e-01
-4.16805744e-01 -2.49097928e-01 2.86122225e-02 1.19593215e+00
-7.49576539e-02 -4.89719361e-01 -5.56375325e-01 2.38463387e-01
8.85880291e-01 -1.11173302e-01 5.63669443e-01 -1.22913349e+00
-3.67381513e-01 -6.83487356e-01 -1.42516422e+00 -4.84345764e-01
3.75494242e-01 -4.73977178e-01 -4.21233952e-01 -3.85735571e-01
-2.38566533e-01 -7.35024959e-02 2.34754365e-02 -1.73501775e-01
4.72623855e-01 7.27976143e-01 1.09005161e-02 2.88851798e-01
-2.16204658e-01 5.39634764e-01 9.40046906e-01 -1.34577602e-01
1.83941782e-01 -1.28730953e-01 3.38583231e-01 7.32112646e-01
2.50996500e-01 -4.39013302e-01 1.20728873e-01 1.35192558e-01
2.94057354e-02 2.39889756e-01 8.82781088e-01 -1.29543889e+00
-1.30496919e-01 -1.02937734e-02 4.29570436e-01 -1.43012106e-01
8.29547107e-01 -8.36284459e-01 5.70109487e-02 7.70981789e-01
-1.56689107e-01 -6.13211989e-02 1.40608788e-01 8.40592742e-01
-4.92639601e-01 -1.31807134e-01 1.17668855e+00 2.05566715e-02
-5.38933277e-01 3.46681289e-02 -4.45665210e-01 -2.61260986e-01
1.29723585e+00 -6.84900701e-01 -1.69789881e-01 -3.54429036e-01
-5.95529616e-01 -5.96687317e-01 4.73696560e-01 4.45335895e-01
6.38614476e-01 -1.07695663e+00 -5.58543563e-01 5.44242620e-01
-4.58984432e-04 -1.00629330e+00 3.85777563e-01 9.26935494e-01
-1.02918530e+00 3.13154459e-01 -7.34919429e-01 -6.72615826e-01
-1.69191754e+00 7.12467134e-01 4.34582561e-01 -1.26825929e-01
-5.18821716e-01 6.66662812e-01 -1.70282349e-01 2.58361191e-01
-6.80411682e-02 1.75257251e-01 -4.14641052e-01 5.13189137e-01
6.03687823e-01 5.81337512e-01 1.84764519e-01 -1.21111381e+00
-3.07197362e-01 6.34834826e-01 1.70216605e-01 -1.85125619e-01
1.13936400e+00 1.58190265e-01 -2.12853730e-01 3.04848999e-01
1.57818866e+00 2.17723921e-01 -1.14113975e+00 5.53727113e-02
2.66763389e-01 -1.01814783e+00 -8.94165039e-02 -1.13860518e-01
-8.25303912e-01 9.61380899e-01 7.20434427e-01 6.85745656e-01
7.56836176e-01 -2.85719216e-01 7.16040313e-01 3.09189744e-02
5.30921817e-01 -8.04002464e-01 3.74379039e-01 8.85275006e-02
8.86817157e-01 -1.13005412e+00 -8.41767341e-02 -6.48689687e-01
-2.68926799e-01 1.52404201e+00 -1.51387051e-01 -3.96158129e-01
6.66038156e-01 7.91696925e-03 -2.39608467e-01 -3.17012668e-01
-1.58679456e-01 1.21340089e-01 2.55254626e-01 6.37678027e-01
-1.50133178e-01 9.98223666e-03 -1.07289903e-01 -1.62345544e-01
-7.00354874e-02 -2.80147254e-01 5.90683281e-01 9.36874926e-01
-2.23692968e-01 -9.22894955e-01 -7.01688051e-01 1.68473020e-01
-6.62044048e-01 3.88864428e-01 -5.48576534e-01 8.23203385e-01
2.26565421e-01 8.44067514e-01 1.64964963e-02 -3.70747298e-01
1.69380195e-02 9.47617516e-02 8.18247199e-01 -3.54768895e-02
-4.92665470e-01 -2.34574378e-01 4.93000746e-02 -8.27630699e-01
-4.45751905e-01 -8.75568509e-01 -2.09543392e-01 -3.25995356e-01
-8.16256180e-02 -1.46820620e-01 9.82687354e-01 9.76114631e-01
1.34831905e-01 -1.07649848e-01 9.13353622e-01 -1.34139633e+00
-5.58094501e-01 -6.85126603e-01 -7.91169524e-01 8.09827268e-01
7.60962486e-01 -1.03219950e+00 -7.25900292e-01 2.42945284e-01] | [12.514843940734863, 1.1449313163757324] |
dce9490d-1e4f-4a29-ba05-3f32de1b1f7d | uctransnet-rethinking-the-skip-connections-in | 2109.04335 | null | https://arxiv.org/abs/2109.04335v3 | https://arxiv.org/pdf/2109.04335v3.pdf | UCTransNet: Rethinking the Skip Connections in U-Net from a Channel-wise Perspective with Transformer | Most recent semantic segmentation methods adopt a U-Net framework with an encoder-decoder architecture. It is still challenging for U-Net with a simple skip connection scheme to model the global multi-scale context: 1) Not each skip connection setting is effective due to the issue of incompatible feature sets of encoder and decoder stage, even some skip connection negatively influence the segmentation performance; 2) The original U-Net is worse than the one without any skip connection on some datasets. Based on our findings, we propose a new segmentation framework, named UCTransNet (with a proposed CTrans module in U-Net), from the channel perspective with attention mechanism. Specifically, the CTrans module is an alternate of the U-Net skip connections, which consists of a sub-module to conduct the multi-scale Channel Cross fusion with Transformer (named CCT) and a sub-module Channel-wise Cross-Attention (named CCA) to guide the fused multi-scale channel-wise information to effectively connect to the decoder features for eliminating the ambiguity. Hence, the proposed connection consisting of the CCT and CCA is able to replace the original skip connection to solve the semantic gaps for an accurate automatic medical image segmentation. The experimental results suggest that our UCTransNet produces more precise segmentation performance and achieves consistent improvements over the state-of-the-art for semantic segmentation across different datasets and conventional architectures involving transformer or U-shaped framework. Code: https://github.com/McGregorWwww/UCTransNet. | ['Osmar R. Zaiane', 'Jiaqi Wang', 'Peng Cao', 'Haonan Wang'] | 2021-09-09 | null | null | null | null | ['unet-segmentation'] | ['computer-vision'] | [ 2.90416449e-01 2.14923665e-01 -1.46658659e-01 -2.81927198e-01
-9.71231043e-01 -2.44103417e-01 6.93581030e-02 -2.07966119e-01
-2.34412283e-01 5.06055474e-01 2.16216698e-01 -3.92106295e-01
1.31899104e-01 -8.78180981e-01 -8.51092458e-01 -6.07543349e-01
3.51491362e-01 4.67067398e-02 6.30435526e-01 -3.23963106e-01
-2.68636812e-02 -2.01753974e-01 -1.03738821e+00 4.13357884e-01
1.11053443e+00 1.29073012e+00 5.50990522e-01 2.82092869e-01
-5.55360496e-01 3.82629693e-01 -2.11702347e-01 -4.48948354e-01
3.26337099e-01 -6.33485794e-01 -9.13816392e-01 -6.58652782e-02
6.54089525e-02 -4.60797995e-01 -2.45289102e-01 1.13742399e+00
6.47292852e-01 -3.08903098e-01 1.61447272e-01 -1.02304983e+00
-7.53764391e-01 9.58592117e-01 -8.63641143e-01 1.79695904e-01
1.30887404e-01 8.83646011e-02 1.06715894e+00 -3.35749835e-01
4.09958154e-01 1.24131799e+00 7.82960474e-01 5.31100571e-01
-9.70368624e-01 -8.07608783e-01 4.52547133e-01 1.83243096e-01
-1.30842268e+00 4.94731367e-02 6.35018826e-01 -1.71939075e-01
6.31418467e-01 1.02819525e-01 7.25300193e-01 8.73015463e-01
2.81897098e-01 1.12141979e+00 9.19323683e-01 3.43074538e-02
-6.69235289e-02 -1.81951627e-01 1.73193231e-01 7.35989988e-01
-5.07516675e-02 -1.96924463e-01 -6.82705939e-02 3.65784496e-01
1.05539620e+00 8.86351913e-02 -3.12660396e-01 -2.98708901e-02
-1.31649506e+00 7.78583467e-01 1.07843924e+00 5.52266181e-01
-2.27305815e-01 2.29070991e-01 5.76480746e-01 1.86054453e-01
1.50417164e-01 1.58680588e-01 -7.65200436e-01 1.59325823e-02
-7.74413645e-01 -1.10445961e-01 3.07432353e-01 1.17992783e+00
6.85910225e-01 -9.10206735e-02 -4.48838174e-01 7.59096026e-01
2.87373483e-01 2.13147640e-01 7.12887049e-01 -7.12857246e-01
5.42894721e-01 7.30585992e-01 -5.11193395e-01 -5.53251266e-01
-5.48657477e-01 -7.67083526e-01 -9.04758692e-01 -3.79836708e-01
3.19495142e-01 -4.42215800e-01 -1.12199700e+00 1.69825125e+00
2.05297709e-01 5.78466237e-01 4.12822235e-03 1.04526711e+00
1.16922891e+00 5.81403017e-01 9.73895416e-02 7.58432448e-02
1.60382748e+00 -1.30342603e+00 -8.24582696e-01 -2.04996631e-01
7.23120749e-01 -9.18751121e-01 8.83915365e-01 7.23666549e-02
-9.32048082e-01 -7.71560371e-01 -1.19856083e+00 -3.33795041e-01
-3.13090056e-01 1.65539920e-01 6.36196256e-01 4.81966347e-01
-1.07224214e+00 3.29000533e-01 -6.84774041e-01 -3.13439965e-01
5.46003938e-01 3.81938607e-01 -2.47916251e-01 -1.19014591e-01
-1.42602777e+00 5.08430958e-01 5.91737926e-01 4.16097373e-01
-5.10388732e-01 -4.96284395e-01 -9.92347836e-01 6.50329143e-02
4.79918331e-01 -7.71519244e-01 1.09844029e+00 -1.08905661e+00
-1.20000148e+00 4.06803668e-01 -1.00062124e-01 -4.08372819e-01
5.99165857e-01 -1.08877309e-01 -1.83654457e-01 1.67357549e-01
3.91721666e-01 1.10566545e+00 4.35834676e-01 -1.14629591e+00
-8.06895971e-01 -3.37366968e-01 1.05432488e-01 2.77753890e-01
-1.77829936e-01 -3.24723631e-01 -1.04604840e+00 -8.50632191e-01
4.47139502e-01 -8.17174375e-01 -3.48023772e-01 -1.49486586e-01
-6.88026607e-01 6.31071031e-02 1.06754220e+00 -8.68632138e-01
1.42161977e+00 -2.05579138e+00 -3.55866887e-02 1.01351246e-01
1.38658777e-01 2.74136066e-01 -2.59152561e-01 8.28581154e-02
-1.51560053e-01 2.71313608e-01 -5.45513391e-01 -2.43034780e-01
-3.21618140e-01 2.74911612e-01 2.31134966e-01 2.97172628e-02
1.72738701e-01 1.21384680e+00 -9.78128433e-01 -6.11933291e-01
3.34552765e-01 4.98946637e-01 -6.60810888e-01 -8.32536742e-02
-4.95800525e-02 6.78135633e-01 -8.23340654e-01 8.25527310e-01
9.01169240e-01 -3.25930208e-01 5.83589114e-02 -4.89337981e-01
-1.79186016e-02 1.06934346e-01 -1.02862525e+00 2.07999730e+00
-3.96204799e-01 1.24158710e-01 2.60315448e-01 -1.05999076e+00
9.17067707e-01 3.91896546e-01 6.31682158e-01 -1.02507150e+00
4.88758534e-01 3.46089214e-01 1.32691294e-01 -4.69207168e-01
2.31693968e-01 -3.82355675e-02 -1.57145590e-01 -3.67438272e-02
1.52273059e-01 2.61487305e-01 1.48127377e-01 2.82864958e-01
9.52248216e-01 2.22843200e-01 3.97728616e-03 -2.20968425e-01
7.49384224e-01 -2.65463382e-01 9.43146348e-01 5.99218011e-01
-3.50234479e-01 9.68444824e-01 5.97629428e-01 -7.26533160e-02
-8.10393810e-01 -9.01082695e-01 -2.76743263e-01 5.89837909e-01
6.07098222e-01 -3.60564709e-01 -1.02002037e+00 -8.52411926e-01
-1.70453176e-01 2.11356938e-01 -6.06127083e-01 -5.01003005e-02
-4.20676440e-01 -8.26636851e-01 6.73552096e-01 9.38246906e-01
1.14901936e+00 -8.49931002e-01 -2.75515497e-01 5.15516400e-01
-5.44142663e-01 -1.20218825e+00 -9.24334347e-01 1.53341621e-01
-9.32230055e-01 -1.07162309e+00 -9.18116391e-01 -9.16450500e-01
6.75399125e-01 1.57145858e-01 6.03498101e-01 1.89440399e-01
-2.86243763e-02 -1.39631152e-01 -6.92999721e-01 -1.01485960e-01
-5.83862347e-05 1.91365778e-01 -7.51166105e-01 1.21239178e-01
1.86231822e-01 -3.85531574e-01 -9.60991502e-01 4.87664551e-01
-9.48599756e-01 4.73867744e-01 7.51253307e-01 1.08671916e+00
7.56542444e-01 -1.16985962e-01 7.97768176e-01 -9.91028607e-01
4.37007219e-01 -5.41507006e-01 -2.36332089e-01 1.50699645e-01
-4.53775138e-01 -1.15552284e-01 6.36677802e-01 -7.82358795e-02
-9.99836922e-01 1.13126993e-01 -6.95001602e-01 -3.99050087e-01
5.55186532e-02 3.46866429e-01 -4.20945615e-01 3.67086120e-02
3.17558795e-02 3.53317916e-01 5.66250682e-02 -5.55978179e-01
3.09293985e-01 7.59073555e-01 3.32849920e-01 -4.98936892e-01
2.34449565e-01 4.14907336e-01 -3.74425113e-01 -3.79391104e-01
-5.89056611e-01 -4.65775043e-01 -6.87342584e-01 -3.93942967e-02
1.43917108e+00 -9.80001748e-01 -3.99559647e-01 6.54800892e-01
-1.17642772e+00 -1.54182956e-01 -5.42516215e-03 4.50650215e-01
-2.89403945e-01 4.18350279e-01 -9.03161407e-01 -2.43640706e-01
-5.21640122e-01 -1.88219535e+00 1.25851595e+00 5.50122738e-01
1.74952358e-01 -8.98006678e-01 -5.75647354e-01 7.64975190e-01
3.88283610e-01 1.42106358e-02 8.55761051e-01 -5.20708025e-01
-6.70073152e-01 9.03391093e-02 -7.83984184e-01 4.53385502e-01
2.60457933e-01 -3.51572901e-01 -9.84699190e-01 -1.41913533e-01
-2.00853571e-01 6.57496303e-02 1.04166853e+00 7.29963005e-01
1.49522662e+00 6.24283254e-02 -3.98755789e-01 8.21186841e-01
1.60438633e+00 4.88564730e-01 9.14967120e-01 3.07221562e-01
9.90840077e-01 1.44331291e-01 4.30769086e-01 6.85954988e-02
6.94576025e-01 4.93345976e-01 4.50869471e-01 -6.29990041e-01
-3.17368954e-01 -2.58572459e-01 1.21665336e-01 8.59187722e-01
1.02749303e-01 -1.00635543e-01 -6.48161054e-01 6.83088839e-01
-1.98213220e+00 -4.84622121e-01 -1.12635851e-01 1.90461791e+00
6.12266958e-01 2.13040560e-01 -9.24764127e-02 -2.12013442e-03
9.82589602e-01 4.02721576e-02 -5.70737064e-01 -4.43808317e-01
-1.15431789e-02 1.55030832e-01 6.68784797e-01 3.03921580e-01
-1.08924305e+00 1.00166845e+00 5.14181614e+00 1.27455580e+00
-1.29534113e+00 4.55177516e-01 9.72915232e-01 3.44195306e-01
-3.82493347e-01 3.77767496e-02 -6.81231022e-01 7.94257939e-01
3.47898155e-01 3.99524897e-01 2.79924590e-02 4.69546735e-01
2.24780887e-01 -8.71694610e-02 -7.85767615e-01 8.46615851e-01
-1.09382384e-01 -1.27217758e+00 2.29080617e-01 -6.77094758e-02
6.29391730e-01 1.20168649e-01 -1.44055396e-01 1.74700767e-01
-3.91333736e-03 -7.10428596e-01 7.21588790e-01 3.67994569e-02
8.51924539e-01 -6.13801718e-01 1.06661546e+00 8.08816999e-02
-1.64251626e+00 -1.40478030e-01 -1.22927107e-01 3.11817229e-01
4.88992274e-01 5.86212575e-01 -4.18597877e-01 1.07057571e+00
7.10446417e-01 9.21184123e-01 -4.36287820e-01 1.05038297e+00
-1.11812567e-02 5.21956682e-01 -1.97117254e-01 3.44962567e-01
7.37179458e-01 -2.95985341e-01 4.01016742e-01 1.18934011e+00
4.34147120e-01 9.95433331e-02 3.73958915e-01 7.98471034e-01
-1.22442715e-01 1.67847425e-01 -2.89689690e-01 4.13180023e-01
3.49074781e-01 1.24964809e+00 -1.13622665e+00 -3.78263950e-01
-7.64948368e-01 1.01642346e+00 -1.66633725e-01 3.74233633e-01
-1.08151162e+00 -3.49858642e-01 5.19799292e-01 7.60980397e-02
5.64194918e-01 2.57910937e-01 -6.69035316e-01 -1.07646906e+00
9.41886157e-02 -4.90847588e-01 5.24276555e-01 -8.19662690e-01
-1.13468003e+00 7.91635036e-01 -3.29506725e-01 -1.31148052e+00
4.98714089e-01 -2.36427113e-01 -7.59942651e-01 8.26189041e-01
-1.84306812e+00 -1.51983845e+00 -1.05590671e-01 7.09680140e-01
7.45849133e-01 1.84775367e-01 3.44234824e-01 6.67231619e-01
-8.33061934e-01 7.42093384e-01 -3.93058918e-02 3.05537969e-01
4.81787205e-01 -1.17908800e+00 2.60876417e-01 8.62837672e-01
-3.38211417e-01 4.95670378e-01 1.26254439e-01 -8.42760682e-01
-1.06038773e+00 -1.17995012e+00 3.78868461e-01 6.45262823e-02
4.56756204e-01 -1.42111197e-01 -7.60574877e-01 5.83840549e-01
2.01205224e-01 1.82464812e-02 3.39104533e-01 -2.57590890e-01
4.39878181e-02 -2.52138197e-01 -1.18401599e+00 2.70840138e-01
1.06620550e+00 -2.25770235e-01 -3.08275729e-01 1.57206517e-03
1.05228055e+00 -5.46086609e-01 -8.72017980e-01 5.64369082e-01
5.69354177e-01 -9.11469281e-01 8.48024070e-01 1.06782749e-01
7.79167354e-01 -4.73088056e-01 3.72145092e-03 -9.42111313e-01
-2.46647671e-01 -2.96774060e-01 4.45521027e-01 1.31669605e+00
6.65036857e-01 -8.33922923e-01 6.28306806e-01 2.74277985e-01
-6.13272607e-01 -1.22674417e+00 -1.06024408e+00 -3.77643943e-01
7.63713419e-02 -4.38265055e-01 6.97087109e-01 8.40794325e-01
-1.88563779e-01 2.43168816e-01 -3.05133998e-01 -3.34169678e-02
3.43726665e-01 8.27560425e-02 2.34278291e-01 -7.78953433e-01
-9.92986262e-02 -4.97904032e-01 -2.97116965e-01 -1.52686024e+00
-2.13027209e-01 -1.02205682e+00 -3.12182382e-02 -1.83967054e+00
2.00739577e-01 -7.31660068e-01 -5.15911818e-01 6.02287769e-01
-2.96210885e-01 3.98163527e-01 3.25936258e-01 3.58091220e-02
-4.85167354e-01 5.91337919e-01 2.05695319e+00 -2.22247630e-01
-1.90761685e-01 -1.23027124e-01 -1.06684637e+00 5.58048308e-01
6.03454411e-01 -1.77201018e-01 -5.89685202e-01 -7.01303184e-01
-1.96252942e-01 2.60252059e-01 2.94392288e-01 -9.39910173e-01
1.75640583e-01 3.54230344e-01 4.41140503e-01 -6.75965130e-01
6.77961037e-02 -9.17973280e-01 -1.07267581e-01 5.80380321e-01
-6.11213073e-02 4.55366001e-02 2.32501805e-01 4.12471414e-01
-5.36954880e-01 1.06649637e-01 7.16569901e-01 -3.65542263e-01
-9.39552009e-01 5.21938384e-01 -4.51414511e-02 -5.80901057e-02
1.12452900e+00 -4.85868394e-01 -2.57583827e-01 -1.00206658e-01
-8.61082077e-01 8.80464315e-01 1.58716083e-01 5.51751912e-01
5.05632401e-01 -1.16353607e+00 -4.77280945e-01 3.20749551e-01
-6.83663040e-02 3.63401890e-01 7.01199412e-01 1.18628836e+00
-6.81790948e-01 5.04743636e-01 -3.15659732e-01 -6.70546174e-01
-9.21075225e-01 1.26210004e-01 4.36782837e-01 -3.71864378e-01
-7.70271003e-01 1.09757674e+00 7.49297917e-01 -5.15073121e-01
-7.43135437e-02 -6.02958441e-01 -3.57493818e-01 9.04373303e-02
2.34078392e-01 1.11415468e-01 1.02014177e-01 -7.74054885e-01
-3.25989127e-01 7.74229944e-01 -1.73147172e-01 1.75030336e-01
1.04690289e+00 -4.50439811e-01 -6.82938844e-02 2.03889143e-03
1.21996713e+00 -5.51425934e-01 -1.41788173e+00 -1.81879207e-01
-6.23693705e-01 -2.64027566e-01 2.88209796e-01 -9.29634571e-01
-1.81683946e+00 9.20665026e-01 7.29835212e-01 -1.58196509e-01
1.33517718e+00 -2.73303520e-02 1.62108862e+00 -3.25917870e-01
2.83954233e-01 -1.09455609e+00 -2.40324944e-01 3.15908521e-01
5.12961924e-01 -1.28608668e+00 -4.35903966e-01 -7.82777786e-01
-7.14119971e-01 8.88998032e-01 9.21857536e-01 1.10573843e-01
8.04500639e-01 3.04288924e-01 1.46954790e-01 -1.06005140e-01
-3.06636214e-01 -4.19881254e-01 8.39756131e-02 3.30153525e-01
5.03866553e-01 8.07657391e-02 -7.80206144e-01 9.12270308e-01
9.19236615e-03 6.31896779e-02 2.68932760e-01 7.00087905e-01
-1.20561071e-01 -1.28639412e+00 -1.96324319e-01 7.27715433e-01
-4.49069232e-01 -2.85211146e-01 7.55144730e-02 5.95767975e-01
7.62164712e-01 8.59848738e-01 7.99340308e-02 -6.02155685e-01
3.05410504e-01 -1.38721958e-01 2.19917715e-01 -4.57644463e-01
-7.92320669e-01 5.62542617e-01 -1.56180277e-01 -7.48200953e-01
-3.68951470e-01 -5.15088379e-01 -1.62585008e+00 -7.61144906e-02
-4.50675964e-01 7.15417834e-03 4.89075065e-01 1.11083627e+00
3.06845278e-01 1.05980659e+00 3.64198089e-01 -5.35752118e-01
8.48049223e-02 -9.01540458e-01 -4.82929289e-01 4.33368981e-01
2.26422071e-01 -5.57697535e-01 1.39033929e-01 -6.78606927e-02] | [14.605147361755371, -2.6272103786468506] |
aa30057a-4448-4c3f-b23d-d570ca2e9226 | instructeval-systematic-evaluation-of | 2307.00259 | null | https://arxiv.org/abs/2307.00259v1 | https://arxiv.org/pdf/2307.00259v1.pdf | InstructEval: Systematic Evaluation of Instruction Selection Methods | In-context learning (ICL) performs tasks by prompting a large language model (LLM) using an instruction and a small set of annotated examples called demonstrations. Recent work has shown that the precise details of the inputs used in the prompt significantly impacts ICL, which has incentivized instruction selection algorithms. The effect of instruction-choice however is severely underexplored, with existing analyses being restricted to shallow subsets of models and tasks, which limits the generalizability of their insights. We develop an ICL evaluation suite to conduct a thorough assessment of these techniques. The suite includes 13 open-sourced LLMs of varying scales from 4 distinct model families and covers 9 different tasks, representing a range of task types across 3 categories. In this work, we evaluate the relative performance of 7 popular instruction selection methods using our benchmark over five desiderata relevant to ICL. We discover that using curated manually-written instructions and simple instructions without any task-specific descriptions often elicits superior ICL performance than that of automatic instruction-induction methods, pointing to a lack of generalizability among the latter. We release our evaluation suite for benchmarking instruction selection approaches, and call for more rigorous and generalizable methods in this space. | ['Karthik Narasimhan', 'Ameet Deshpande', 'Mengzhou Xia', 'Chris Pan', 'Anirudh Ajith'] | 2023-07-01 | null | null | null | null | ['benchmarking', 'benchmarking'] | ['miscellaneous', 'robots'] | [ 3.78584832e-01 -2.61492729e-01 -8.22792709e-01 -5.96440077e-01
-1.11274421e+00 -9.44089711e-01 8.49172711e-01 1.48257822e-01
-4.92759854e-01 6.51876330e-01 3.15234989e-01 -1.11029959e+00
6.31880164e-02 -8.86040106e-02 -8.44388366e-01 -1.52021110e-01
-9.71094146e-02 3.78841788e-01 3.07558537e-01 -1.66455567e-01
7.71185338e-01 3.57564539e-01 -1.97248006e+00 5.27073205e-01
8.79181683e-01 4.67981189e-01 4.10376728e-01 6.90191269e-01
-2.11896479e-01 9.09676075e-01 -8.31883371e-01 4.15039249e-02
-8.46883506e-02 -2.22728863e-01 -9.05293107e-01 -2.80680865e-01
9.43724990e-01 -3.26675653e-01 6.07248023e-02 6.22933388e-01
3.24952513e-01 2.75663912e-01 5.93370199e-01 -1.40624774e+00
-2.80025691e-01 6.52141571e-01 -1.55521214e-01 4.17895943e-01
6.73540473e-01 5.19036293e-01 1.22146726e+00 -6.62068665e-01
5.28922081e-01 1.13730669e+00 6.03429854e-01 4.97985631e-01
-1.38625550e+00 -7.73351431e-01 3.74635130e-01 -2.65501719e-02
-9.34752703e-01 -6.73756778e-01 4.52395082e-01 -5.97124577e-01
1.49990058e+00 3.53778273e-01 1.42624810e-01 1.35545123e+00
3.76241773e-01 1.06834328e+00 1.49863791e+00 -7.19505072e-01
2.13477194e-01 3.56904566e-02 7.91556299e-01 6.09108567e-01
4.82372671e-01 5.00017822e-01 -7.00455844e-01 -4.07839477e-01
3.73146892e-01 -1.65010110e-01 -2.65209407e-01 -2.02433273e-01
-1.33874857e+00 7.67253578e-01 -1.65445045e-01 3.48549217e-01
1.81620941e-01 1.43014938e-01 7.52032399e-01 5.27601898e-01
2.03926470e-02 1.03625524e+00 -7.85163164e-01 -6.55125082e-01
-8.01354408e-01 2.56566644e-01 1.16951549e+00 1.12348437e+00
7.28485346e-01 1.33688271e-01 -2.87273288e-01 5.38218439e-01
6.11300543e-02 2.38001972e-01 7.59358883e-01 -1.01720393e+00
4.50508237e-01 4.02275056e-01 5.70846461e-02 -5.88804185e-01
-4.77212489e-01 -2.48474583e-01 3.38194072e-01 3.20220828e-01
5.68690121e-01 -1.22012191e-01 -4.78590608e-01 1.92076290e+00
-1.34888276e-01 -1.08742192e-01 -2.15342730e-01 5.18974066e-01
7.26896048e-01 2.84787446e-01 4.76208895e-01 5.44799678e-03
1.42800641e+00 -9.15425003e-01 -5.27860224e-01 -6.91146672e-01
1.26943982e+00 -7.94494450e-01 1.91500068e+00 5.32513976e-01
-6.21533215e-01 -6.58639431e-01 -1.19471800e+00 -9.31725204e-02
-3.03140998e-01 1.90297666e-03 9.55498517e-01 6.61763668e-01
-9.54127669e-01 4.56895828e-01 -7.85417914e-01 -2.87264585e-01
1.57110370e-03 4.08632964e-01 -5.91068044e-02 -6.75190017e-02
-8.96814167e-01 9.29319084e-01 4.17433113e-01 -6.69623673e-01
-1.12805665e+00 -7.60793865e-01 -1.06594491e+00 -1.40070319e-01
6.42675817e-01 -2.45075390e-01 1.81249428e+00 -8.92909527e-01
-1.38747001e+00 8.02280247e-01 -1.73998073e-01 -6.95421025e-02
-5.02670445e-02 -5.03145754e-01 -3.51139009e-01 -2.53039330e-01
2.03591287e-01 5.72419941e-01 6.85710311e-01 -1.32759583e+00
-7.37284124e-01 -4.62169945e-02 4.13433343e-01 1.44717112e-01
-1.29937351e-01 3.54825526e-01 -1.75499558e-01 -5.68867743e-01
-5.16197085e-01 -9.51411247e-01 8.10509622e-02 -8.22978914e-01
-3.01495433e-01 -4.48638737e-01 6.45076275e-01 -2.60851383e-01
1.57423830e+00 -2.12354708e+00 -3.15614462e-01 -3.63226056e-01
2.25855365e-01 -1.21590216e-02 -4.15934920e-01 3.84366840e-01
-1.18862845e-01 2.62788206e-01 2.04458311e-02 -2.83812016e-01
3.86705995e-01 1.56735495e-01 -5.26236832e-01 1.26750559e-01
8.64969417e-02 9.30506468e-01 -1.17534149e+00 -6.30079329e-01
1.12515509e-01 1.93625595e-02 -7.48150468e-01 3.99399757e-01
-5.89987457e-01 1.43892929e-01 -4.60487485e-01 7.76805103e-01
-1.27888843e-01 -2.34343931e-01 -2.06687562e-02 -5.06940559e-02
-2.85195470e-01 1.12393177e+00 -6.75862253e-01 1.72502303e+00
-7.37996519e-01 8.82830977e-01 -1.23858653e-01 -5.79989374e-01
6.37591064e-01 1.51182145e-01 -3.17730308e-02 -4.68435496e-01
-9.05063897e-02 3.27318609e-01 4.07250583e-01 -6.45783186e-01
3.24052066e-01 7.04868734e-02 -3.57986182e-01 8.66906583e-01
1.00083590e-01 -3.08238924e-01 3.34569871e-01 1.60952121e-01
1.45563877e+00 5.91893375e-01 4.44039583e-01 -6.18014514e-01
1.78481162e-01 3.33547890e-01 2.94849992e-01 1.17375481e+00
-5.03773272e-01 1.29543990e-01 3.68855536e-01 -3.47795844e-01
-5.15312374e-01 -7.12193251e-01 -2.91555226e-01 2.07114077e+00
-3.19814593e-01 -6.91343486e-01 -6.74330473e-01 -1.16672885e+00
1.46169245e-01 1.45451629e+00 -4.72966969e-01 -1.65409997e-01
-7.46825457e-01 -6.24302328e-01 5.19671500e-01 6.03537083e-01
4.09635678e-02 -1.27779472e+00 -9.63049114e-01 1.99365932e-02
6.25436101e-03 -1.11150563e+00 -8.34761918e-01 8.08686435e-01
-1.01717389e+00 -1.22932720e+00 2.65043974e-01 -8.43712747e-01
4.63140905e-01 1.97584376e-01 1.88224697e+00 1.36875257e-01
-3.18721421e-02 7.92984307e-01 -3.83337259e-01 -4.02605981e-01
-6.50868714e-01 3.66205573e-01 7.02225640e-02 -6.95823312e-01
1.02890897e+00 -2.87884891e-01 -1.37795329e-01 3.71457130e-01
-6.58548236e-01 -2.46594613e-03 8.34286928e-01 7.83248544e-01
2.36936390e-01 -4.09890056e-01 6.01621687e-01 -1.22678971e+00
8.57604563e-01 -5.04073977e-01 -4.78720248e-01 1.55726299e-01
-7.82908142e-01 4.42794889e-01 7.20619559e-01 -6.85958087e-01
-9.12032783e-01 -5.65278232e-02 1.19068116e-01 5.25162034e-02
-6.98817611e-01 4.42308217e-01 -1.17028840e-01 -2.04460025e-01
8.96311581e-01 1.50087893e-01 -1.01052880e-01 -4.07243222e-01
1.50318399e-01 3.92531723e-01 4.18450832e-01 -1.37247741e+00
5.58957636e-01 -2.98115402e-01 -3.72755498e-01 -5.77473104e-01
-8.30808401e-01 -1.31782293e-01 -4.58346695e-01 1.07675374e-01
4.22341228e-01 -6.58247232e-01 -6.47980154e-01 -2.13879496e-01
-7.84999907e-01 -1.09779763e+00 -1.24482870e-01 5.31301618e-01
-6.30145073e-01 2.07569912e-01 -7.70937979e-01 -3.82046670e-01
1.95620582e-03 -1.98396182e+00 1.14549768e+00 -1.89217746e-01
-9.29887474e-01 -1.05521142e+00 9.68155116e-02 1.36243835e-01
3.81165594e-01 -1.40062094e-01 1.39784229e+00 -1.29490054e+00
-4.97222960e-01 -1.01116650e-01 9.58231166e-02 8.01100358e-02
1.67934254e-01 -2.67633684e-02 -1.06755757e+00 -3.12025249e-01
9.49843675e-02 -7.44013250e-01 5.35965681e-01 2.38326237e-01
1.24131143e+00 -5.86115345e-02 -4.55349088e-01 4.93072897e-01
1.20610380e+00 3.52025539e-01 1.06483415e-01 6.62855387e-01
6.00526094e-01 6.66463852e-01 7.09757864e-01 -6.01927517e-03
2.61414081e-01 6.47797883e-01 8.29533190e-02 3.61144394e-01
-1.75455317e-01 -1.63863838e-01 9.14459944e-01 8.59512269e-01
3.97853196e-01 1.14513105e-02 -1.07436407e+00 3.71944129e-01
-1.28478253e+00 -5.48592865e-01 7.03622922e-02 2.41270399e+00
1.34078157e+00 5.32038331e-01 7.81195089e-02 -1.66409716e-01
1.66734159e-01 2.25724071e-01 -5.15025735e-01 -7.12245345e-01
3.95032644e-01 3.52888733e-01 2.88630724e-01 8.71769845e-01
-9.98126149e-01 9.36440468e-01 7.79226112e+00 9.41669047e-01
-1.00501168e+00 -8.26021656e-03 5.81538796e-01 -8.30174312e-02
-4.54771727e-01 7.76958689e-02 -1.29714441e+00 3.79660308e-01
1.42515814e+00 -1.27839431e-01 4.76478547e-01 1.06332552e+00
1.70120709e-02 -2.76473582e-01 -1.95992708e+00 5.40015638e-01
1.39224917e-01 -8.63494277e-01 1.28765926e-02 -1.39499053e-01
5.61335564e-01 1.09134726e-01 1.30365178e-01 1.04418647e+00
6.35843933e-01 -1.10764563e+00 6.12864494e-01 -7.11259991e-03
9.09187675e-01 -4.95719403e-01 3.02908659e-01 3.17881465e-01
-9.96500015e-01 -9.65650901e-02 1.56308889e-01 -4.38222170e-01
-3.66401583e-01 6.30450845e-02 -8.50161612e-01 -1.01648375e-01
4.04124647e-01 3.53202313e-01 -9.69159365e-01 5.98265648e-01
-3.05303991e-01 1.21575546e+00 -1.14823535e-01 -1.33919537e-01
2.00631946e-01 2.29053363e-01 3.65706593e-01 1.69036603e+00
-2.87936032e-01 -4.61390689e-02 6.15543306e-01 7.51162648e-01
1.62239745e-01 1.18135937e-01 -8.72751176e-01 -1.95921913e-01
7.75860965e-01 1.12935305e+00 -5.66386104e-01 -3.55934054e-01
-8.93574178e-01 3.87568712e-01 3.31204236e-01 5.16539693e-01
-6.90666497e-01 -4.42749500e-01 7.60480821e-01 1.15973108e-01
-2.75127620e-01 -3.59935164e-01 -5.92217982e-01 -9.65827405e-01
-2.58610666e-01 -1.65661967e+00 3.13272923e-01 -5.17563701e-01
-1.07574284e+00 5.03421128e-01 5.78387141e-01 -9.11605477e-01
-5.61996818e-01 -8.58726025e-01 -6.70678318e-01 7.67206967e-01
-1.33952415e+00 -5.11955857e-01 -2.35112995e-01 1.21302865e-01
9.83934879e-01 -1.74103618e-01 7.92479515e-01 5.41012548e-02
-7.35220432e-01 8.25419247e-01 -2.56262928e-01 -2.76331991e-01
1.20944846e+00 -1.49369180e+00 5.55356979e-01 6.48297369e-01
-3.83669883e-02 1.09946823e+00 8.60644162e-01 -7.24158466e-01
-1.71993387e+00 -8.51041019e-01 8.99252772e-01 -1.21640348e+00
8.25669646e-01 -4.52823788e-01 -9.49786425e-01 1.26007724e+00
3.14876199e-01 -3.62912446e-01 8.50459576e-01 3.25194448e-01
-4.26694512e-01 2.16352299e-01 -6.61208749e-01 7.65254498e-01
1.11470830e+00 -7.30054021e-01 -8.34639609e-01 3.24695647e-01
1.04844332e+00 -2.49347180e-01 -7.86041260e-01 4.69476193e-01
3.99129629e-01 -9.06157315e-01 8.37025702e-01 -7.46448338e-01
4.54708129e-01 1.06376680e-02 -3.59587878e-01 -1.14676845e+00
-2.98483849e-01 -5.12088954e-01 -2.38796920e-01 1.05505848e+00
4.66082394e-01 -6.29818738e-01 3.76855642e-01 7.56404400e-01
-7.00771093e-01 -8.35774720e-01 -1.74226686e-01 -7.59664476e-01
2.37437546e-01 -7.79415369e-01 5.29494345e-01 8.94225180e-01
2.54830599e-01 6.26661420e-01 3.10481846e-01 -2.61975408e-01
5.10569751e-01 1.90102175e-01 8.80964756e-01 -1.10752761e+00
-5.75145841e-01 -8.63726795e-01 2.10385546e-01 -1.33290219e+00
6.94736004e-01 -1.05063045e+00 2.87420839e-01 -7.74213195e-01
3.61456245e-01 -5.76661408e-01 -1.86471745e-01 6.82890356e-01
-6.40889764e-01 -3.41138542e-01 -4.85919835e-03 2.65273362e-01
-8.90313566e-01 1.07595652e-01 9.72243488e-01 8.46370161e-02
-1.61145136e-01 -1.41122848e-01 -7.66246915e-01 7.79668450e-01
6.90673947e-01 -3.03115368e-01 -6.69596910e-01 -4.71475959e-01
-8.84504095e-02 -2.20956028e-01 1.88888777e-02 -1.02190208e+00
-7.03574270e-02 -4.46608275e-01 9.03240442e-02 -1.77790821e-01
-1.62592694e-01 -5.09433985e-01 -3.26119274e-01 3.91741902e-01
-8.85683239e-01 4.18380678e-01 6.79841757e-01 2.35861659e-01
-4.30226922e-02 -6.00400150e-01 3.26449871e-01 -1.12581864e-01
-1.17365122e+00 -7.18892738e-02 -5.69508016e-01 5.41400611e-01
7.47101068e-01 -2.41820797e-01 -4.08059210e-01 -1.99176192e-01
-2.21403763e-01 1.02004312e-01 8.18471789e-01 4.71348971e-01
2.16824144e-01 -1.14853919e+00 -4.10333276e-01 2.76551127e-01
4.98445749e-01 -3.65128070e-01 -4.54817206e-01 7.23800719e-01
-1.65007010e-01 8.30043733e-01 8.28103200e-02 -6.44581735e-01
-1.00113225e+00 7.34487534e-01 1.12808272e-01 -3.74971449e-01
-4.34991270e-01 5.87266803e-01 5.03303468e-01 -5.97321808e-01
4.85213071e-01 -6.67966843e-01 -9.51163545e-02 -2.48396441e-01
5.18277466e-01 1.22816980e-01 1.40257984e-01 -2.03390062e-01
-3.11438024e-01 8.88356715e-02 -2.20022365e-01 1.81871152e-03
7.10153341e-01 2.54537672e-01 1.30644828e-01 9.65256870e-01
1.17489541e+00 1.22177124e-01 -1.32716465e+00 -1.40046969e-01
5.34856975e-01 -2.89792657e-01 6.63070306e-02 -9.51608300e-01
-5.02232611e-01 6.63500309e-01 3.13414246e-01 -1.54470265e-01
1.01099312e+00 -1.18352763e-01 4.64811444e-01 5.54366708e-01
7.91458488e-01 -7.30854690e-01 2.54536718e-01 7.84311950e-01
6.49954736e-01 -1.50204206e+00 -5.60131446e-02 -1.65321022e-01
-4.90037233e-01 1.00380552e+00 1.13918197e+00 1.19959109e-01
3.42089742e-01 6.84590220e-01 -1.04958422e-01 -1.23133101e-01
-1.13902855e+00 -9.68741626e-03 2.46882617e-01 5.48340023e-01
1.20981419e+00 1.13705277e-01 -3.80935341e-01 6.65098131e-01
-2.44215325e-01 -1.85740560e-01 3.04985970e-01 1.51623201e+00
-4.95650619e-01 -1.28494251e+00 -6.27803683e-01 8.14410329e-01
-3.95097733e-01 -4.47210342e-01 -2.18578950e-01 1.14933598e+00
-2.24141955e-01 9.22937870e-01 4.89292853e-03 -5.28551340e-01
1.96520880e-01 6.18428111e-01 5.32927334e-01 -1.08160794e+00
-8.83407056e-01 -2.54891604e-01 3.50051820e-01 -8.29849541e-01
6.52293339e-02 -7.78951585e-01 -1.13843608e+00 -1.27272591e-01
-2.69775391e-01 1.25788122e-01 4.28621113e-01 1.01215267e+00
4.25249666e-01 5.43196201e-01 2.00750425e-01 -9.93147671e-01
-9.53545153e-01 -9.80138302e-01 -4.58996482e-02 5.42982399e-01
2.90887743e-01 -9.58573580e-01 -5.40276229e-01 1.03591569e-01] | [10.617579460144043, 8.288359642028809] |
b62ae0da-41f6-463b-8e21-d27ad8c26145 | simultaneous-face-hallucination-and | 2104.06534 | null | https://arxiv.org/abs/2104.06534v2 | https://arxiv.org/pdf/2104.06534v2.pdf | Simultaneous Face Hallucination and Translation for Thermal to Visible Face Verification using Axial-GAN | Existing thermal-to-visible face verification approaches expect the thermal and visible face images to be of similar resolution. This is unlikely in real-world long-range surveillance systems, since humans are distant from the cameras. To address this issue, we introduce the task of thermal-to-visible face verification from low-resolution thermal images. Furthermore, we propose Axial-Generative Adversarial Network (Axial-GAN) to synthesize high-resolution visible images for matching. In the proposed approach we augment the GAN framework with axial-attention layers which leverage the recent advances in transformers for modelling long-range dependencies. We demonstrate the effectiveness of the proposed method by evaluating on two different thermal-visible face datasets. When compared to related state-of-the-art works, our results show significant improvements in both image quality and face verification performance, and are also much more efficient. | ['Vishal M. Patel', 'Shuowen Hu', 'Rakhil Immidisetti'] | 2021-04-13 | null | null | null | null | ['face-hallucination'] | ['computer-vision'] | [ 3.46194506e-01 -2.69018617e-02 3.25067878e-01 -7.56037593e-01
-8.71371567e-01 -5.80931425e-01 7.29016542e-01 -1.20070684e+00
2.58834176e-02 4.83431011e-01 6.93591982e-02 -1.20366439e-01
2.09738731e-01 -4.61528540e-01 -7.78777003e-01 -8.27418566e-01
3.81967783e-01 2.45888501e-01 -2.59665340e-01 -9.67298672e-02
-1.90936193e-01 7.08922029e-01 -1.55293477e+00 4.65995580e-01
4.19362098e-01 1.20269561e+00 -4.34555650e-01 8.35278869e-01
4.72121894e-01 7.43743837e-01 -5.69198012e-01 -9.90053475e-01
7.68603504e-01 -5.70836365e-01 -4.90706921e-01 1.74742639e-01
1.17764878e+00 -8.83400261e-01 -3.19900513e-01 8.67317617e-01
5.68996131e-01 -1.39808804e-01 5.72399974e-01 -1.37411714e+00
-1.08391285e+00 1.98532902e-02 -1.11404145e+00 -4.24093567e-02
6.25667453e-01 4.41430807e-01 5.32576442e-01 -7.32667089e-01
4.75267082e-01 1.54072773e+00 8.43315542e-01 1.04497755e+00
-1.18861616e+00 -8.77631307e-01 -5.45199588e-02 -8.45283568e-02
-1.22785783e+00 -1.02870536e+00 8.23033452e-01 -2.81864375e-01
7.42930830e-01 1.76280588e-01 3.87286365e-01 1.61017561e+00
1.62676081e-01 1.10399365e-01 1.33232820e+00 -2.61611938e-01
-2.44993240e-01 -9.63631347e-02 -5.34135640e-01 6.80795968e-01
1.45605952e-01 7.71736205e-01 -5.96170306e-01 -1.84716210e-01
8.09103668e-01 4.85779084e-02 -3.61650944e-01 -7.67461583e-02
-6.73474729e-01 5.25274396e-01 3.99704933e-01 1.51283711e-01
-2.38899454e-01 3.49791527e-01 4.36413884e-02 3.73709530e-01
7.47065604e-01 1.03644848e-01 -1.59866468e-04 2.37037003e-01
-1.27618873e+00 -9.10726264e-02 3.58834147e-01 8.96984875e-01
3.97318333e-01 2.56563604e-01 -1.64044201e-01 5.93616962e-01
4.86997992e-01 9.85033989e-01 -1.28129289e-01 -9.81207252e-01
4.50232029e-01 2.12013468e-01 -1.58278216e-02 -5.92041373e-01
8.61468390e-02 -9.32652727e-02 -1.01161027e+00 6.42035961e-01
3.24056774e-01 -1.75603196e-01 -1.22443783e+00 1.78168714e+00
4.15642440e-01 3.95687968e-01 5.01037091e-02 1.10766602e+00
8.79614234e-01 4.10235435e-01 -1.99473858e-01 -2.72643536e-01
1.32865977e+00 -7.97106922e-01 -7.58801281e-01 -3.27254325e-01
-2.96237975e-01 -1.04394019e+00 7.56835043e-01 1.50797650e-01
-1.12912428e+00 -7.40481496e-01 -9.73863721e-01 -2.66786397e-01
-4.06803824e-02 5.47489934e-02 4.91607696e-01 1.15200925e+00
-1.60466099e+00 3.00725311e-01 -7.36714900e-01 -2.30668917e-01
6.43811464e-01 2.52990872e-01 -6.39669895e-01 -2.78698206e-01
-9.43919241e-01 6.40066266e-01 -3.41986924e-01 5.10415018e-01
-1.14267957e+00 -8.22941959e-01 -8.74051690e-01 -2.05059901e-01
1.47441283e-01 -1.00967216e+00 1.02940381e+00 -1.15802014e+00
-1.78208351e+00 1.12504900e+00 -3.16264898e-01 -1.88728664e-02
7.27658749e-01 -1.79455608e-01 -4.27576125e-01 2.13111609e-01
-2.43315428e-01 7.09765553e-01 1.38916385e+00 -1.37971628e+00
-3.00623104e-02 -7.30980873e-01 -7.23622157e-04 -1.47638053e-01
-2.04778001e-01 5.37751257e-01 -4.30153072e-01 -4.87155318e-01
-3.39113563e-01 -1.03562689e+00 1.39125645e-01 3.65311503e-01
-3.74942839e-01 2.29636312e-01 1.20578110e+00 -7.83146203e-01
3.47077221e-01 -2.03883886e+00 -5.39317541e-02 -1.48006484e-01
1.49225459e-01 2.88537562e-01 -2.41984412e-01 1.59154534e-02
-3.07558805e-01 -8.31895322e-02 -1.35543361e-01 -9.11696076e-01
-7.90576860e-02 -7.20435455e-02 -4.97740835e-01 6.66181624e-01
2.41440296e-01 1.20317960e+00 -3.68193805e-01 -4.00155455e-01
2.92576522e-01 1.17861485e+00 -2.39280969e-01 6.52908802e-01
6.07458390e-02 9.20833409e-01 -7.61798769e-02 7.28790224e-01
1.12709224e+00 -7.95346722e-02 2.27120463e-02 -2.49576300e-01
2.90366828e-01 -1.28135175e-01 -4.91757631e-01 1.44170797e+00
-4.91678417e-01 8.56391609e-01 2.90294230e-01 -3.97103280e-01
8.81761968e-01 6.48965657e-01 3.13314855e-01 -8.37591052e-01
2.13946924e-01 -1.63953349e-01 -3.67143601e-01 -2.18627051e-01
3.19587469e-01 -3.76124620e-01 1.86684743e-01 6.20424986e-01
-1.61145791e-01 -2.19389513e-01 -4.28030998e-01 -1.47090167e-01
7.36290634e-01 4.55106735e-01 -1.32179528e-01 1.21307254e-01
5.91903090e-01 -7.06685424e-01 4.05446947e-01 3.89115542e-01
-1.31299660e-01 1.02466393e+00 2.90559828e-01 -4.22982007e-01
-1.15938532e+00 -1.00906026e+00 -1.61019508e-02 8.12779188e-01
-1.40398756e-01 -9.96724293e-02 -1.13678730e+00 -7.04585910e-01
-1.01241305e-01 2.33134061e-01 -8.69047761e-01 6.83839396e-02
-7.45315313e-01 -4.58044440e-01 7.86092639e-01 6.24945045e-01
6.68160856e-01 -9.28623199e-01 -6.43530607e-01 -5.18938303e-01
-5.21819964e-02 -1.57707870e+00 -6.56551957e-01 -5.31935632e-01
-4.74894553e-01 -9.63946640e-01 -8.84941339e-01 -4.12515044e-01
7.96940744e-01 2.87111521e-01 1.25716949e+00 3.10381114e-01
-4.40345079e-01 4.38892215e-01 -3.92892020e-04 -3.89737338e-02
-4.73652005e-01 -3.26581150e-01 6.91246763e-02 3.65754336e-01
2.13254482e-01 -4.16141927e-01 -7.69796252e-01 3.52542371e-01
-8.12604964e-01 -1.65949538e-01 3.93750340e-01 8.70503128e-01
2.25555897e-01 -1.01935372e-01 1.47708014e-01 -7.63955235e-01
1.71384320e-01 8.58818367e-02 -1.05608833e+00 4.45718735e-01
-4.95644152e-01 3.22314836e-02 8.93916130e-01 -2.60258794e-01
-1.43622804e+00 1.88304141e-01 -2.09273025e-01 -8.54704916e-01
-2.34140798e-01 -3.59349161e-01 -3.11686724e-01 -5.07453740e-01
4.24730152e-01 1.66186113e-02 -8.52819309e-02 -1.05605304e-01
3.54829788e-01 6.25245214e-01 7.64630318e-01 -4.62915897e-01
1.40366960e+00 7.72981763e-01 2.06718728e-01 -6.14285648e-01
-6.78082287e-01 6.42035753e-02 -6.92879498e-01 -2.40091816e-01
1.04143763e+00 -9.20752823e-01 -1.12401628e+00 6.46943271e-01
-1.13125861e+00 -3.47727448e-01 1.21968500e-01 1.88579991e-01
-4.84105229e-01 3.75356257e-01 -7.38104403e-01 -1.11763740e+00
-6.45703912e-01 -1.31512964e+00 1.81546950e+00 3.57265711e-01
2.83351541e-01 -8.15283716e-01 -8.47522169e-02 7.78858483e-01
6.96690261e-01 3.65912735e-01 3.24711531e-01 1.87541008e-01
-8.48578930e-01 6.18776157e-02 -3.81692767e-01 2.51733303e-01
1.42070636e-01 1.88820153e-01 -1.71541941e+00 -7.26714253e-01
8.91015381e-02 -5.04873931e-01 8.31331432e-01 3.79822075e-01
1.04172254e+00 -1.59507036e-01 -2.36001015e-01 1.13432300e+00
1.18152535e+00 7.06847617e-03 9.08604920e-01 -3.38868320e-01
8.30472231e-01 7.09331691e-01 4.73436147e-01 1.65844008e-01
4.05047834e-02 1.08679187e+00 3.40950608e-01 -4.75111008e-01
-4.35967147e-01 -2.49659330e-01 5.74245393e-01 1.26388490e-01
-1.59011275e-01 -2.65618801e-01 -6.46952689e-01 2.16841385e-01
-1.44691122e+00 -1.19086289e+00 3.11040789e-01 2.15754509e+00
4.98778760e-01 -3.76818359e-01 5.28097502e-04 -1.52303517e-01
6.78253353e-01 3.65060240e-01 -6.27124608e-01 -3.14285755e-01
-7.07854703e-02 3.22746992e-01 3.26605767e-01 5.70758820e-01
-9.13180113e-01 8.98082435e-01 6.80706072e+00 3.43626231e-01
-1.37649000e+00 3.07095379e-01 1.02369666e+00 -2.09702611e-01
-1.97998345e-01 -2.53004223e-01 -7.21076965e-01 2.56363392e-01
1.02375913e+00 4.02699232e-01 6.37697756e-01 5.68602145e-01
-1.14424594e-01 2.36935735e-01 -1.37952638e+00 1.20512962e+00
5.62893391e-01 -8.65528762e-01 -1.35258839e-01 2.99935251e-01
8.25307369e-01 -2.98094094e-01 7.81472325e-01 -1.28331587e-01
3.50253850e-01 -1.62826872e+00 5.41678011e-01 4.94805157e-01
1.51295936e+00 -7.58369923e-01 6.80964351e-01 -1.49802208e-01
-1.39429235e+00 8.10603648e-02 -4.46638584e-01 4.12917256e-01
1.49660498e-01 3.18548054e-01 -5.80737650e-01 7.26283848e-01
9.08296585e-01 5.78518450e-01 -5.97520769e-01 2.20617518e-01
-2.73369133e-01 2.59410858e-01 -1.97373837e-01 5.68595111e-01
-2.11385295e-01 -1.78983375e-01 1.40657529e-01 7.74350047e-01
4.17302936e-01 -1.71470419e-02 -2.40588993e-01 1.19220161e+00
-3.55689228e-01 -6.29449785e-01 -9.07715201e-01 2.15287626e-01
3.15951586e-01 1.34990776e+00 -4.18088943e-01 -5.95145300e-02
-5.02323687e-01 1.35301435e+00 1.37155950e-01 4.43697155e-01
-1.17972958e+00 9.31401476e-02 8.76764417e-01 8.67408328e-03
3.49429160e-01 4.30074930e-02 5.73282801e-02 -1.10912907e+00
3.71522665e-01 -9.13770914e-01 2.75554031e-01 -1.15153968e+00
-1.47009420e+00 1.05468702e+00 -1.53818846e-01 -1.06066847e+00
-4.94651824e-01 -5.26863754e-01 -3.85910898e-01 1.08102989e+00
-1.72334301e+00 -1.83078265e+00 -5.26898324e-01 8.86839271e-01
3.17008972e-01 -1.63986027e-01 8.48510921e-01 2.31662855e-01
-3.48505914e-01 1.21137369e+00 -2.33388841e-01 2.74469942e-01
9.39338446e-01 -7.66537726e-01 9.76325274e-01 1.20987284e+00
1.18548363e-01 7.64182806e-01 4.04861093e-01 -3.82751375e-01
-1.78786302e+00 -1.13208127e+00 3.74949217e-01 -8.35473001e-01
1.01505026e-01 -7.75038540e-01 -6.90583885e-01 8.61792862e-01
5.61077476e-01 6.12588406e-01 4.11556214e-01 -1.45739228e-01
-1.03622293e+00 -3.75419706e-01 -1.64026368e+00 3.84232998e-01
1.25291038e+00 -9.45724726e-01 -3.33752841e-01 1.29828528e-01
4.92751718e-01 -4.53540564e-01 -8.19057107e-01 3.25764954e-01
9.32406187e-01 -1.38484585e+00 1.24936259e+00 -1.52426958e-01
3.70729715e-01 -3.29536408e-01 2.03114599e-02 -1.15101922e+00
-7.72480741e-02 -9.56315041e-01 1.50265887e-01 1.35417747e+00
1.77165031e-01 -8.37976396e-01 8.51744652e-01 7.08131254e-01
2.04562113e-01 -3.06328982e-01 -1.12046432e+00 -6.22779787e-01
-5.98591790e-02 -3.91766503e-02 9.44524288e-01 7.66022921e-01
-5.41230381e-01 2.30098218e-01 -9.30596948e-01 4.96420652e-01
1.04741704e+00 3.25082660e-01 9.18032944e-01 -1.00946271e+00
-4.98954207e-01 -1.28637940e-01 -4.32717204e-01 -8.88331115e-01
5.11066020e-01 -3.16423953e-01 4.07581031e-02 -9.16501641e-01
5.11299968e-01 -1.74148902e-02 1.68711439e-01 4.28027332e-01
-2.20991150e-01 9.19205606e-01 1.71813294e-01 -1.02090985e-01
-3.42926025e-01 4.66128439e-01 1.22674179e+00 -4.12691012e-02
3.58575344e-01 -1.57483429e-01 -5.66446960e-01 3.16405028e-01
4.18591142e-01 -1.93827599e-01 -5.07961214e-01 -6.63973510e-01
-8.47876817e-02 4.16191012e-01 4.65890795e-01 -9.51380610e-01
1.14163183e-01 -8.33323151e-02 9.67940569e-01 -2.86811888e-01
9.51827109e-01 -1.17772925e+00 5.97490728e-01 1.78004369e-01
-2.55966067e-01 1.98412210e-01 1.23874269e-01 2.98037946e-01
-5.51417470e-02 5.32811642e-01 1.16417861e+00 1.13635451e-01
2.47921608e-02 6.18554771e-01 2.02275604e-01 -2.12296158e-01
1.03501916e+00 -2.74206251e-01 -8.50029409e-01 -7.17415452e-01
-1.16335578e-01 -2.27662399e-01 9.24528718e-01 5.92795908e-01
8.94293547e-01 -1.27202857e+00 -8.99741292e-01 7.84777105e-01
9.69151929e-02 -2.18116581e-01 3.47373694e-01 6.24912739e-01
-3.74896199e-01 4.47515607e-01 -2.43813455e-01 -7.14360893e-01
-1.80438125e+00 7.84707904e-01 6.30059659e-01 3.19516361e-02
-6.10794663e-01 7.97501802e-01 5.10109365e-01 -4.37283546e-01
9.46805775e-02 -2.96389386e-02 2.23073334e-01 -5.73862433e-01
8.09468448e-01 -1.39937609e-01 3.99367139e-02 -1.14045429e+00
-4.92605954e-01 1.21414006e+00 -1.29914703e-02 -2.75383472e-01
1.28497529e+00 -1.01707995e-01 -1.20568000e-01 -1.75250620e-01
1.24544787e+00 7.52622634e-02 -1.54761899e+00 -2.97041759e-02
-5.63572228e-01 -8.85426223e-01 -1.30194843e-01 -6.66273952e-01
-1.80358899e+00 1.07558143e+00 8.99893105e-01 -1.57942817e-01
1.41138875e+00 -6.19989336e-02 7.89318860e-01 -7.00622350e-02
3.82656246e-01 -4.23932701e-01 1.37916028e-01 1.04733430e-01
7.91990519e-01 -1.32777381e+00 5.29907234e-02 -5.24716914e-01
-4.77227271e-01 1.02201569e+00 7.89180815e-01 1.84116423e-01
3.84078801e-01 5.78539848e-01 4.29600686e-01 -2.03149825e-01
-8.64215195e-01 6.77758381e-02 4.21669543e-01 9.21359420e-01
6.51081145e-01 -2.94926941e-01 6.42448545e-01 -1.37685061e-01
-2.73335844e-01 -1.39868453e-01 2.40656123e-01 5.26292443e-01
4.01440740e-01 -9.85951543e-01 -8.17425728e-01 -9.03951228e-02
-6.56206906e-01 -8.28779787e-02 -7.13119447e-01 3.87596726e-01
-1.95015699e-01 1.26401925e+00 1.21108234e-01 -3.93184900e-01
-5.67477196e-02 -6.49453923e-02 9.05037344e-01 -1.50200233e-01
-7.09267914e-01 -7.81846419e-03 -2.47912884e-01 -7.97777891e-01
-6.17300808e-01 -3.19505364e-01 -5.84125817e-01 -7.87293851e-01
-1.82148546e-01 -2.29445100e-01 6.87051415e-01 5.87864637e-01
5.12887478e-01 4.40950751e-01 9.15837765e-01 -1.00410092e+00
-3.72907192e-01 -9.15726304e-01 -1.90035358e-01 5.79489946e-01
7.67719626e-01 -3.95718902e-01 -4.65160877e-01 5.78904599e-02] | [12.891942024230957, 0.029022470116615295] |
5e57157d-f2ed-4e35-a6a6-6ef8905568ef | nus-hlt-report-for-activitynet-challenge-2021 | null | null | http://research.google.com/ava/2021/S3_NUS_Report_AVA_ActiveSpeaker_2021.pdf | http://research.google.com/ava/2021/S3_NUS_Report_AVA_ActiveSpeaker_2021.pdf | NUS-HLT Report for ActivityNet Challenge 2021 AVA (Speaker) | Active speaker detection (ASD) seeks to detect who is speaking in a visual scene of one or more speakers. The successful ASD depends on accurate interpretation of short-term and long-term audio and visual information, as well as audiovisual interaction. Unlike the prior work where systems makedecision instantaneously using short-term features, we propose a novel framework, named TalkNet, that makes decision by taking both short-term and long-term features into consideration. TalkNet consists of audio and visual temporal encoders for feature representation, audio-visual cross-attention mechanism for
inter-modality interaction, and a self-attention mechanism to capture long-term speaking evidence. The experiments demonstrate that TalkNet achieves 3.5% and 3.0% improvement over the state-of-the-art systems on the AVA-ActiveSpeaker validation and test dataset, respectively. We will release the codes, the models and data logs. | ['Haizhou Li', 'Mike Zheng Shou', 'Xinyuan Qian', 'Rohan Kumar Das', 'Zexu Pan', 'Ruijie Tao'] | 2021-06-01 | null | null | null | the-activitynet-large-scale-activity | ['audio-visual-active-speaker-detection'] | ['computer-vision'] | [-1.84456989e-01 -9.84339491e-02 2.44308598e-02 -6.74266279e-01
-1.30503857e+00 -6.82567537e-01 8.49098265e-01 -3.14946100e-02
-2.32956007e-01 2.99404353e-01 6.36900485e-01 -1.91121072e-01
1.77457586e-01 -1.09308943e-01 -5.90162039e-01 -7.21370876e-01
-2.36573666e-01 9.69610587e-02 2.92709619e-01 9.63659808e-02
1.92987785e-01 3.97612631e-01 -1.75826025e+00 6.61200345e-01
4.23419178e-01 1.18237448e+00 2.05750391e-01 1.23546147e+00
-4.50862527e-01 1.06481755e+00 -7.95775712e-01 -2.18553439e-01
-3.50615770e-01 -3.71119916e-01 -6.39086127e-01 2.19559744e-01
5.23532212e-01 -4.13605690e-01 -3.18077534e-01 6.52820289e-01
1.06617498e+00 1.53758019e-01 1.96334317e-01 -1.50829720e+00
-7.33389914e-01 6.81796670e-01 -3.85289729e-01 7.01226652e-01
7.26342320e-01 5.55252373e-01 9.80743349e-01 -1.45123422e+00
9.34011564e-02 1.56799400e+00 2.65127748e-01 4.86087441e-01
-9.09221947e-01 -9.19281960e-01 4.55231905e-01 7.56870806e-01
-1.29457235e+00 -1.52706122e+00 8.58488083e-01 -5.33857465e-01
1.08453333e+00 4.24578935e-01 7.20512211e-01 1.30111301e+00
5.55149140e-03 1.18822050e+00 7.23570228e-01 -5.84896028e-01
5.00034615e-02 2.35537112e-01 4.28701699e-01 3.38744968e-01
-6.76343203e-01 2.87624449e-01 -1.33181679e+00 -7.02519566e-02
3.65914911e-01 -3.58079523e-01 -2.96563625e-01 3.55394669e-02
-1.00579500e+00 5.40068686e-01 1.90115720e-01 2.36399785e-01
-4.74804997e-01 2.57763285e-02 4.17733639e-01 2.65425056e-01
4.73616898e-01 -9.05334651e-02 1.21085256e-01 -4.67471272e-01
-9.43910480e-01 -3.31755817e-01 6.40753508e-01 6.23211980e-01
3.30273658e-01 3.51733148e-01 -5.05330801e-01 8.74538243e-01
7.98712492e-01 6.28040493e-01 4.50129449e-01 -7.85136342e-01
3.52259994e-01 2.32799083e-01 3.68184522e-02 -5.32831311e-01
4.54248525e-02 -2.57872939e-01 -2.16314092e-01 3.69125366e-01
-6.98719323e-02 8.10693800e-02 -9.08088863e-01 1.60272861e+00
3.71481657e-01 3.95762354e-01 1.00158811e-01 8.44074130e-01
1.42354798e+00 7.73111105e-01 -2.10261419e-02 -5.84092796e-01
1.28495383e+00 -1.10741436e+00 -1.25786519e+00 -2.54176050e-01
-2.99052835e-01 -9.91510332e-01 9.81219828e-01 2.70331264e-01
-1.38985837e+00 -8.99338961e-01 -9.55791593e-01 -1.41923413e-01
-2.39732135e-02 -1.34187430e-01 3.32989395e-01 3.28516155e-01
-1.37112141e+00 -2.79881358e-01 -7.75852740e-01 -2.06282005e-01
2.13572696e-01 2.29168937e-01 -2.94441134e-01 4.61986750e-01
-1.16693830e+00 3.98688316e-01 -3.77717823e-01 3.88689667e-01
-1.69960809e+00 -3.94174427e-01 -7.73776352e-01 1.12807482e-01
2.35402256e-01 -3.69505525e-01 1.87175536e+00 -1.22651410e+00
-1.74986506e+00 7.78486848e-01 -7.77077794e-01 -2.67503202e-01
3.94800931e-01 -2.15436950e-01 -8.35330069e-01 2.80866891e-01
-1.71052486e-01 4.67491478e-01 1.09963429e+00 -1.18031299e+00
-5.24867594e-01 -5.03666103e-01 1.79921582e-01 5.11274338e-01
-2.37537310e-01 5.95682561e-01 -8.50393057e-01 -3.90967935e-01
1.22189850e-01 -4.78883088e-01 5.73675275e-01 2.42480829e-01
-5.25752485e-01 -3.93470168e-01 1.12784934e+00 -6.81034267e-01
1.15049660e+00 -2.73377037e+00 -7.06150830e-02 -2.52092868e-01
2.90025622e-01 2.36776859e-01 -1.01464204e-01 3.20682883e-01
-2.25361556e-01 -1.97537169e-01 4.69352335e-01 -8.66634488e-01
-1.10260746e-03 -1.10034339e-01 -4.22701597e-01 3.92577857e-01
-2.27709441e-03 7.01653898e-01 -6.50837839e-01 -6.87087059e-01
2.60679692e-01 7.60293961e-01 -2.74391882e-02 6.68294907e-01
2.63240844e-01 3.63617629e-01 -5.72141930e-02 7.32812166e-01
5.28331697e-01 -3.31737585e-02 -3.35303932e-01 1.48448706e-01
-4.62544948e-01 5.90716541e-01 -1.08904219e+00 1.56927133e+00
-4.76727366e-01 1.20722163e+00 6.39336526e-01 -5.11422098e-01
6.95713639e-01 1.07638717e+00 1.39523953e-01 -8.39160144e-01
-4.94911149e-02 -3.30819041e-02 -8.00032765e-02 -8.80226672e-01
1.62456676e-01 1.03939839e-01 2.76325196e-01 2.49948516e-01
1.41011342e-01 4.82433766e-01 -1.76780477e-01 4.27073121e-01
7.76791155e-01 -4.22503948e-01 2.77292430e-02 9.39501226e-02
8.48214388e-01 -6.05757535e-01 3.31741661e-01 6.90814853e-01
-6.96622491e-01 5.46374500e-01 1.62458941e-01 3.60578783e-02
-3.13241303e-01 -1.27096462e+00 2.38468349e-02 1.61358678e+00
7.68282861e-02 -4.38243777e-01 -4.88940746e-01 -4.14134175e-01
-1.57582089e-01 6.83680296e-01 -6.67472780e-01 -1.21901378e-01
-3.16044688e-01 7.39984959e-02 3.68521690e-01 7.66153336e-01
5.65866470e-01 -1.37340939e+00 -1.56802088e-01 1.25353098e-01
-3.35818350e-01 -9.53287005e-01 -8.94681036e-01 -7.36046359e-02
-2.56863624e-01 -6.93383455e-01 -6.22996032e-01 -8.16639543e-01
1.36364296e-01 4.41813350e-01 9.69656527e-01 -3.76980662e-01
-1.99561805e-01 6.51095927e-01 -1.29205093e-01 -6.14382267e-01
-3.35159034e-01 -6.18712485e-01 -2.78114015e-03 4.13186967e-01
3.90012652e-01 -4.54752624e-01 -7.61059582e-01 3.63056213e-01
-7.45656565e-02 -6.53364277e-03 4.22184318e-01 7.22207487e-01
3.38552654e-01 -4.54603344e-01 3.91749561e-01 -1.94684163e-01
4.12250638e-01 -7.03630522e-02 -3.36420715e-01 3.23402375e-01
-2.75029182e-01 -3.93889487e-01 1.66599471e-02 -5.92507482e-01
-1.31031799e+00 4.57115173e-02 -2.16058314e-01 -6.84900165e-01
-9.55871716e-02 2.02043250e-01 -5.16838968e-01 2.56757289e-01
5.27523935e-01 4.05732572e-01 -7.55602568e-02 -4.69470978e-01
2.37647161e-01 1.27643991e+00 7.08608985e-01 1.59875993e-02
1.88130870e-01 5.16535819e-01 -8.07826877e-01 -1.03090858e+00
-6.67449653e-01 -9.74877596e-01 -4.43241328e-01 -7.00576603e-01
8.63406122e-01 -1.29289138e+00 -1.08635771e+00 5.95518351e-01
-1.32361257e+00 -1.21816263e-01 -1.87141538e-01 7.23993957e-01
-4.57409799e-01 1.01853535e-01 -5.17212689e-01 -1.53588259e+00
-5.08024216e-01 -1.26367426e+00 1.10208631e+00 6.37167543e-02
-1.87719062e-01 -7.36284912e-01 2.20590010e-01 5.36697328e-01
3.40614468e-01 -3.86864543e-01 1.43664852e-01 -5.51938295e-01
-6.54583573e-01 8.78371596e-02 6.43576086e-02 3.67035300e-01
3.43507752e-02 2.59835064e-01 -1.91254914e+00 -3.60258490e-01
-4.25206823e-03 -1.94804803e-01 8.62369001e-01 7.02870846e-01
1.02803922e+00 -3.32710415e-01 -1.90348670e-01 3.96536171e-01
7.48299718e-01 5.85283220e-01 3.46652031e-01 -2.79925048e-01
5.34610808e-01 5.95006466e-01 4.11016196e-01 4.66430992e-01
4.63723928e-01 9.22139883e-01 5.15611053e-01 -2.46452928e-01
-5.73270917e-01 -1.10572733e-01 8.76129627e-01 9.32429433e-01
1.84509113e-01 -5.23115516e-01 -9.41324830e-01 7.24910676e-01
-1.76954985e+00 -1.27024436e+00 -1.56807993e-02 2.26648855e+00
5.60187101e-01 2.21701398e-01 3.79533052e-01 4.02496994e-01
1.04414916e+00 1.56780258e-01 -6.91915393e-01 -4.71177846e-01
-1.47595078e-01 -3.28225464e-01 -3.25512588e-01 9.86618817e-01
-1.00436509e+00 6.13306880e-01 6.96660662e+00 4.74584073e-01
-1.49557686e+00 3.18922430e-01 4.34959441e-01 -4.51978654e-01
-1.13016494e-01 -2.70258784e-01 -7.30566859e-01 4.32628512e-01
1.27445757e+00 -2.50200540e-01 2.69709587e-01 7.20543087e-01
5.86767256e-01 1.00508921e-01 -1.29537904e+00 1.43447983e+00
4.88956600e-01 -8.94415140e-01 -3.18932772e-01 -5.28713614e-02
-3.74060012e-02 9.77654085e-02 2.47105986e-01 3.51859957e-01
1.27103284e-01 -8.97969604e-01 1.35653496e+00 5.13178229e-01
8.00389528e-01 -6.43718541e-01 3.72822016e-01 1.51131824e-01
-1.48478389e+00 -2.40373507e-01 4.20493037e-01 2.15141624e-01
2.35251784e-01 1.07786983e-01 -1.03270149e+00 1.03324808e-01
1.03603339e+00 6.12802029e-01 -3.39766979e-01 1.06816733e+00
-3.03020209e-01 1.00316703e+00 -3.40680219e-02 2.17630953e-01
-1.19928524e-01 5.88315487e-01 9.17015016e-01 1.14837301e+00
1.80893287e-01 -1.53276682e-01 9.32435989e-02 5.35403073e-01
8.71468410e-02 -1.50950635e-02 -4.11178201e-01 -1.96052969e-01
7.36862481e-01 1.04373622e+00 -1.26255780e-01 -3.93235952e-01
-5.38030624e-01 1.06693947e+00 7.00154975e-02 6.35151565e-01
-8.69467080e-01 -3.54573071e-01 5.66261232e-01 1.02529518e-01
3.11356902e-01 -3.97422649e-02 -4.31491286e-02 -7.89087296e-01
1.67159840e-01 -5.86522222e-01 5.92307270e-01 -1.15538871e+00
-1.01791334e+00 7.41717219e-01 -2.75896370e-01 -1.13858640e+00
-5.26140332e-01 -2.45991394e-01 -9.20950711e-01 9.27324176e-01
-1.46459448e+00 -1.23895538e+00 -4.88519609e-01 7.54802525e-01
1.06805611e+00 -3.11098367e-01 8.84523094e-01 2.52527386e-01
-6.11663997e-01 7.79935658e-01 1.76942460e-02 2.63761759e-01
1.00006568e+00 -1.19160187e+00 3.66088897e-01 6.78787887e-01
3.17461014e-01 4.36391175e-01 8.53236973e-01 -6.76461235e-02
-1.40492105e+00 -5.56685925e-01 1.18072593e+00 -4.02601302e-01
5.53600967e-01 -8.72372448e-01 -7.86366999e-01 8.11634004e-01
7.43708253e-01 1.94178730e-01 9.03432250e-01 4.42560792e-01
-4.67938006e-01 -3.62282962e-01 -8.62490773e-01 1.43008426e-01
8.34147394e-01 -1.07978940e+00 -6.45322919e-01 -6.21365244e-03
6.53918624e-01 -2.41636693e-01 -2.74674952e-01 -2.15708222e-02
8.35237026e-01 -1.13530171e+00 1.01176274e+00 -3.50771785e-01
-3.41605581e-02 -1.06394053e-01 -1.69245452e-01 -1.03854871e+00
-3.85517895e-01 -8.30988824e-01 -5.26500523e-01 1.61379349e+00
4.18621600e-01 -3.60611558e-01 2.67441124e-01 1.78900674e-01
-4.99217749e-01 -2.84298152e-01 -1.23867536e+00 -6.67218268e-01
-4.62777913e-01 -7.12557912e-01 5.20898581e-01 7.86649704e-01
2.31120899e-01 7.49391913e-01 -2.98119098e-01 3.11699003e-01
3.55733693e-01 1.12518474e-01 6.83854938e-01 -1.04685354e+00
-4.88259643e-01 -3.32484722e-01 -5.90593576e-01 -1.17957640e+00
2.08370313e-01 -3.12053204e-01 3.41870874e-01 -1.53378308e+00
3.45962375e-01 2.54073530e-01 -6.83148026e-01 5.14116645e-01
-6.90604970e-02 1.23679124e-01 -9.40133557e-02 2.09133670e-01
-8.97592187e-01 7.38621771e-01 1.15610886e+00 -3.61122042e-01
-3.66999179e-01 1.52788602e-03 -4.34812248e-01 4.59347725e-01
2.53119349e-01 -1.40094534e-02 -6.32371366e-01 -5.01990497e-01
-2.86464840e-01 5.82041502e-01 5.58423996e-01 -7.77396202e-01
6.14442110e-01 -3.44533101e-02 4.04299676e-01 -9.78193879e-01
1.05402970e+00 -4.98067826e-01 -2.20744565e-01 -2.02610251e-02
-6.45984530e-01 3.83057706e-02 3.17921370e-01 5.19559085e-01
-6.60553575e-01 3.91217351e-01 6.46943867e-01 1.34519979e-01
-7.54250348e-01 2.62042522e-01 -6.21086180e-01 -1.88789040e-01
9.34425294e-01 -2.20830381e-01 -4.17741358e-01 -9.33774948e-01
-1.25183666e+00 3.30477506e-01 -2.67470300e-01 7.35923231e-01
8.10225725e-01 -1.38074088e+00 -8.46222401e-01 4.42200035e-01
3.11110616e-01 -4.08629805e-01 7.44762242e-01 8.95969748e-01
2.32437067e-02 4.99393672e-01 1.59813732e-01 -1.11962938e+00
-2.05797505e+00 3.83849174e-01 4.25043285e-01 6.65057242e-01
-4.93325979e-01 1.41258574e+00 6.12238407e-01 2.63875812e-01
8.94317210e-01 2.07548384e-02 -4.45990324e-01 2.65416831e-01
1.03345752e+00 1.79125056e-01 1.26622245e-01 -1.12099123e+00
-7.44263113e-01 1.38879031e-01 -1.83728576e-01 -6.38767779e-01
1.00871205e+00 -7.05514133e-01 3.79332125e-01 1.14897180e+00
1.08999217e+00 2.49410048e-01 -1.39049363e+00 -4.47717935e-01
-4.93961811e-01 -5.45182347e-01 5.90564311e-01 -9.94626164e-01
-9.23188210e-01 1.43848777e+00 1.07383716e+00 3.15939665e-01
1.02232361e+00 4.50154185e-01 2.09967598e-01 -4.86611836e-02
-2.05032051e-01 -8.66562843e-01 5.27122617e-01 2.86295921e-01
1.36584687e+00 -1.44210660e+00 -4.49411452e-01 -3.03564399e-01
-7.72402465e-01 8.91147017e-01 5.96020937e-01 6.15406871e-01
7.56243110e-01 3.04787636e-01 7.10732162e-01 -1.77512020e-01
-1.34096444e+00 -2.28161484e-01 5.19933283e-01 4.18946713e-01
6.65865123e-01 1.91565068e-03 5.77297390e-01 6.50360227e-01
-2.52862036e-01 -4.50097889e-01 -1.17702581e-01 8.95539641e-01
-4.37681794e-01 -5.93527973e-01 -5.32459855e-01 -2.90203750e-01
-2.66737372e-01 -6.78179786e-03 -6.73647106e-01 3.87495339e-01
-7.67716095e-02 1.63158286e+00 5.56526661e-01 -3.87101501e-01
3.04621577e-01 2.79929191e-01 2.79283315e-01 -4.81656075e-01
-5.51067233e-01 6.55283988e-01 2.04611734e-01 -5.31874180e-01
-4.11869884e-01 -8.11958551e-01 -1.10544956e+00 -1.55398160e-01
-5.28217435e-01 2.42998868e-01 8.12051892e-01 7.50023782e-01
4.39599842e-01 7.96130061e-01 9.78843629e-01 -5.59613824e-01
-2.96282649e-01 -1.27662075e+00 -4.78411555e-01 2.97379196e-01
1.02532125e+00 -5.93517840e-01 -6.92269266e-01 1.49087742e-01] | [14.415695190429688, 5.14167594909668] |
9add3d0d-0a37-4c33-9454-6d3685f21b4f | adversarial-networks-and-machine-learning-for | 2301.11964 | null | https://arxiv.org/abs/2301.11964v2 | https://arxiv.org/pdf/2301.11964v2.pdf | Adversarial Networks and Machine Learning for File Classification | Correctly identifying the type of file under examination is a critical part of a forensic investigation. The file type alone suggests the embedded content, such as a picture, video, manuscript, spreadsheet, etc. In cases where a system owner might desire to keep their files inaccessible or file type concealed, we propose using an adversarially-trained machine learning neural network to determine a file's true type even if the extension or file header is obfuscated to complicate its discovery. Our semi-supervised generative adversarial network (SGAN) achieved 97.6% accuracy in classifying files across 11 different types. We also compared our network against a traditional standalone neural network and three other machine learning algorithms. The adversarially-trained network proved to be the most precise file classifier especially in scenarios with few supervised samples available. Our implementation of a file classifier using an SGAN is implemented on GitHub (https://ksaintg.github.io/SGAN-File-Classier). | ['Josh Angichiodo', 'Ken St. Germain'] | 2023-01-27 | null | null | null | null | ['type'] | ['speech'] | [ 2.20379844e-01 -5.45271486e-02 1.90202724e-02 -2.96248317e-01
-9.60043609e-01 -9.73264575e-01 3.77514124e-01 1.90860838e-01
-8.67312253e-02 9.13555026e-01 -3.14799458e-01 -7.65409648e-01
1.54368550e-01 -1.01244819e+00 -1.04788899e+00 -6.03066862e-01
-2.54038945e-02 5.40311515e-01 4.91957255e-02 2.18061998e-01
4.61908400e-01 7.59277701e-01 -1.12802482e+00 6.09471440e-01
4.08601582e-01 1.06187928e+00 -1.56148029e-02 1.06227601e+00
-5.94479330e-02 1.18669069e+00 -1.17442060e+00 -6.34246409e-01
3.35196197e-01 -5.51663376e-02 -1.16538310e+00 -2.86322385e-01
4.32675958e-01 -6.51434839e-01 -7.14895487e-01 1.04112172e+00
3.00382614e-01 -4.25559670e-01 6.99041665e-01 -1.59848237e+00
-9.44666326e-01 8.83115053e-01 -1.54762819e-01 5.13940454e-01
3.82498831e-01 3.79769474e-01 4.82820213e-01 -4.64016557e-01
7.40052700e-01 9.64600265e-01 8.60353053e-01 2.71344960e-01
-9.52369034e-01 -8.75517428e-01 -7.75533855e-01 3.05342942e-01
-1.48011744e+00 -5.86599290e-01 7.37967491e-01 -6.39040291e-01
5.02431333e-01 3.25204581e-01 4.28183749e-02 1.30631649e+00
5.62501431e-01 5.29593647e-01 1.19629097e+00 -4.32058275e-01
1.34120211e-01 3.57467979e-01 4.14600968e-01 5.15559256e-01
4.39726412e-01 -1.05457038e-01 -1.74573869e-01 -6.21078551e-01
5.15287399e-01 1.43243894e-02 -3.20535302e-01 2.89350122e-01
-5.92735767e-01 7.29327917e-01 4.68453281e-02 3.20531100e-01
-3.70732844e-01 2.78240163e-02 5.77922046e-01 5.51079035e-01
4.14131492e-01 5.36439717e-01 -9.37117711e-02 -2.30360150e-01
-1.29205322e+00 8.47928748e-02 8.26693118e-01 6.76918328e-01
4.62628275e-01 2.56662846e-01 9.03930068e-02 6.15977526e-01
3.00697098e-03 4.84546602e-01 3.21871847e-01 -7.63756335e-01
3.66508007e-01 2.62939751e-01 -1.27843305e-01 -8.39630961e-01
1.87605500e-01 -9.51559842e-02 -6.12198651e-01 4.10764009e-01
7.64946878e-01 -1.37950748e-01 -6.02477193e-01 9.27826762e-01
-2.39590444e-02 -2.21417382e-01 -2.76334554e-01 4.20302212e-01
6.60420001e-01 5.41516364e-01 -1.42492250e-01 2.21247450e-01
1.07930434e+00 -3.28529239e-01 -5.56672812e-01 2.27095231e-01
4.58328843e-01 -8.28440368e-01 4.88296717e-01 6.83016181e-01
-8.94469619e-01 -5.29178500e-01 -8.72936547e-01 8.54868665e-02
-6.76930726e-01 1.44230668e-02 3.57924104e-01 1.06516504e+00
-8.09626043e-01 9.52933729e-01 -5.98346710e-01 1.50933424e-02
8.68130386e-01 2.62646943e-01 -7.08175182e-01 -5.26150316e-02
-1.01598692e+00 3.55731875e-01 6.52532279e-01 -1.32453173e-01
-9.07429159e-01 -6.07537329e-01 -6.43107355e-01 1.82634056e-01
1.35269150e-01 -5.52381873e-02 1.02287698e+00 -9.03692245e-01
-9.70184445e-01 1.06606889e+00 3.51140206e-03 -6.85308695e-01
6.64683998e-01 -1.17135011e-01 -6.29556298e-01 4.22160059e-01
4.86652628e-02 1.05035633e-01 1.39254963e+00 -1.23048699e+00
-4.90127690e-02 -2.65993088e-01 -2.53581088e-02 -8.87301803e-01
-1.72701702e-01 3.26894909e-01 2.56090373e-01 -9.32586074e-01
-5.17388821e-01 -7.09532678e-01 6.61077678e-01 -8.48067459e-04
-9.56302166e-01 -2.05688044e-01 1.35644794e+00 -1.22250414e+00
1.31071651e+00 -2.01619077e+00 -5.75084209e-01 2.74320871e-01
2.55342305e-01 5.71242869e-01 2.69215852e-01 4.93387997e-01
-4.83665317e-01 4.71532315e-01 -3.07277620e-01 -1.54965654e-01
-1.17046140e-01 -9.34262276e-02 -5.82404077e-01 7.59999573e-01
2.85899550e-01 8.36081624e-01 -7.00854361e-01 -4.29933131e-01
-9.10795704e-02 2.75139183e-01 -2.68539667e-01 5.06547511e-01
-5.27532548e-02 1.44325599e-01 -3.30207944e-01 1.23344052e+00
1.02823925e+00 -3.37075055e-01 -1.13247156e-01 -1.75478086e-02
3.52989405e-01 2.44724289e-01 -1.15110171e+00 8.18835199e-01
-1.33260131e-01 1.13660908e+00 1.52219206e-01 -9.99384344e-01
9.43948209e-01 1.89406484e-01 5.97105548e-02 -1.09260648e-01
2.80167282e-01 7.71597922e-02 -2.43389130e-01 -6.29853308e-01
7.46921718e-01 1.78594097e-01 5.58781922e-02 1.10888124e+00
7.89797977e-02 3.26047838e-01 -1.39313251e-01 3.79040599e-01
1.52563989e+00 -2.46107075e-02 1.80975825e-01 -3.61855105e-02
3.58845890e-01 -1.50796056e-01 1.89480454e-01 9.65720952e-01
-3.08390945e-01 8.04323554e-01 6.08750045e-01 -3.09378892e-01
-1.20943773e+00 -9.85683680e-01 -3.66857439e-01 7.95053542e-01
-4.45344836e-01 -9.86309648e-02 -8.97410095e-01 -9.95862067e-01
1.31752089e-01 7.22217560e-01 -5.05832613e-01 -5.51996194e-02
-5.60946703e-01 -2.66574144e-01 1.36699653e+00 3.88449788e-01
2.62263179e-01 -1.36474490e+00 -7.90125430e-02 3.52992639e-02
-1.32860020e-01 -9.28591013e-01 -2.98793077e-01 2.13764638e-01
-3.73186916e-01 -1.40884101e+00 -4.96650040e-01 -1.82595223e-01
6.03196323e-01 -2.24000022e-01 7.81790316e-01 2.72201896e-01
-2.32808292e-01 1.84186116e-01 -5.75809598e-01 -5.32592833e-01
-1.11099732e+00 6.24583326e-02 4.12589237e-02 9.77454782e-02
2.90583253e-01 -5.08726716e-01 -1.76825866e-01 -7.86905810e-02
-1.17274785e+00 -7.45481968e-01 -1.19310543e-01 8.00630391e-01
4.13098037e-02 4.21889216e-01 5.23735940e-01 -1.31387663e+00
7.32516408e-01 -1.06789148e+00 -2.14730740e-01 8.18853229e-02
-4.49567467e-01 -1.86974779e-01 9.68693197e-01 -4.41615283e-01
-7.47222781e-01 -3.58849108e-01 -3.68771255e-01 -7.77433872e-01
-6.76661551e-01 1.12293229e-01 -2.34367907e-01 -1.60688922e-01
7.02303052e-01 4.50101227e-01 1.90102190e-01 -7.21338809e-01
-2.46319890e-01 1.32967210e+00 7.97998250e-01 -4.83204037e-01
1.05060709e+00 4.30876821e-01 -2.83229321e-01 -7.18270719e-01
-2.43358910e-01 -1.99626803e-01 -8.77004802e-01 -3.01663548e-01
4.04263228e-01 -5.18965960e-01 -6.06300473e-01 1.08225453e+00
-1.28605413e+00 -9.39667821e-02 6.63068295e-02 2.72526429e-03
-3.24493825e-01 9.11908209e-01 -6.42057240e-01 -8.98841739e-01
-4.37263489e-01 -9.87809360e-01 8.43588889e-01 -2.84870365e-03
-3.37592334e-01 -9.39130068e-01 -1.73334897e-01 7.09059894e-01
3.89139324e-01 4.96849447e-01 8.27011883e-01 -1.43209612e+00
-3.69266659e-01 -7.79678345e-01 -2.44682640e-01 6.50006235e-01
1.55345410e-01 4.87149656e-01 -1.39827716e+00 -2.16574624e-01
-6.80613816e-02 -4.14926231e-01 5.40002227e-01 8.32696110e-02
1.78514838e+00 -7.06568241e-01 -2.22669259e-01 7.67067790e-01
1.19436455e+00 3.56957227e-01 1.06005275e+00 5.77781260e-01
7.11349010e-01 2.45528281e-01 2.07592189e-01 5.84732354e-01
-1.12523519e-01 3.83720070e-01 4.96221691e-01 3.45621526e-01
1.15967050e-01 -2.13637173e-01 4.12228107e-01 1.39110535e-01
1.21464074e-01 -8.99164915e-01 -9.42184687e-01 2.64539778e-01
-1.03895164e+00 -1.44525230e+00 -2.58909166e-01 2.12925577e+00
6.50052011e-01 3.56323540e-01 2.38556162e-01 5.90466499e-01
1.11009669e+00 9.77144316e-02 -4.85538334e-01 -5.97999096e-01
-1.77664176e-01 4.44043487e-01 6.88640058e-01 1.04314327e-01
-1.21128976e+00 5.45861840e-01 6.00439167e+00 1.28820574e+00
-1.33394659e+00 2.25724414e-01 7.91718662e-01 1.60992980e-01
-2.02710461e-02 -2.79229462e-01 -7.69512117e-01 1.25444472e+00
1.34352124e+00 1.33885359e-02 3.94267559e-01 9.78484511e-01
1.32884011e-01 -1.08041525e-01 -9.94697630e-01 6.52837753e-01
6.44118935e-02 -1.54657471e+00 -7.39818662e-02 3.50629628e-01
-8.19732249e-02 -2.35561714e-01 -6.56457469e-02 1.20612346e-01
1.11857086e-01 -1.15049636e+00 9.33918178e-01 4.44230705e-01
1.17549205e+00 -7.28757739e-01 8.75160038e-01 3.69116366e-01
-6.34693980e-01 3.58477682e-02 -1.91246644e-01 2.14571223e-01
1.12445064e-01 5.60914636e-01 -9.98319268e-01 3.39400470e-01
7.74242938e-01 4.52988774e-01 -7.33214021e-01 7.87028730e-01
6.42560124e-02 1.29314709e+00 -2.39019081e-01 3.86748135e-01
-3.05920132e-02 3.03403914e-01 8.25551927e-01 1.44648767e+00
2.90530294e-01 5.35447299e-02 -4.62770909e-01 1.05852616e+00
-3.18590492e-01 -3.39572400e-01 -8.85199249e-01 -5.37524462e-01
8.35415781e-01 9.71694291e-01 -8.07637453e-01 -2.93695927e-01
1.49665266e-01 1.12641501e+00 1.85024202e-01 -4.98667397e-02
-7.52861857e-01 -7.62094259e-01 3.86326820e-01 5.80131829e-01
4.28556561e-01 1.29183933e-01 -3.29847753e-01 -1.04301023e+00
2.48756438e-01 -9.57485318e-01 4.91843700e-01 -9.72978592e-01
-1.36812973e+00 7.45025933e-01 -1.09047435e-01 -1.32476723e+00
-4.50104535e-01 -6.45600736e-01 -1.01008117e+00 8.23203444e-01
-9.20956492e-01 -1.02686691e+00 -7.56009445e-02 4.56142157e-01
2.73835152e-01 -6.07133090e-01 6.30012214e-01 3.08780581e-01
-6.53974354e-01 1.07373583e+00 4.25893575e-01 8.17380250e-01
7.65732229e-01 -1.12965524e+00 3.53441924e-01 9.10350323e-01
-1.76494792e-01 7.93122053e-01 4.69156027e-01 -8.79066229e-01
-1.27114666e+00 -1.14237249e+00 7.51528382e-01 -6.02257907e-01
8.02908957e-01 -3.61118078e-01 -1.52756441e+00 6.89775586e-01
1.90768316e-01 -1.16712879e-02 9.31198478e-01 -5.02932727e-01
-7.09736586e-01 1.52668104e-01 -1.68676555e+00 -2.94384688e-01
4.87498850e-01 -9.32748020e-01 -5.56395054e-01 5.30128539e-01
2.96955049e-01 -3.69886875e-01 -1.13530183e+00 -1.38934389e-01
5.96589088e-01 -1.04344141e+00 1.00682616e+00 -7.61693656e-01
9.30761576e-01 4.15575057e-02 -1.29184395e-01 -8.57576609e-01
-6.52548373e-02 -6.28170490e-01 -4.33455884e-01 1.70096898e+00
7.35982955e-02 -7.95651078e-01 7.97538638e-01 5.17712891e-01
2.28638262e-01 -4.11880195e-01 -8.93665314e-01 -9.50850785e-01
1.37949407e-01 -5.80154121e-01 8.24185073e-01 1.32217741e+00
-2.21834421e-01 -3.44659179e-01 -5.17454743e-01 2.50365317e-01
9.00588691e-01 -1.37883723e-01 7.08339632e-01 -1.02036738e+00
-5.09569824e-01 -3.06928694e-01 -6.64361775e-01 -2.83491969e-01
3.72182578e-01 -1.08896184e+00 -4.02086794e-01 -7.15829492e-01
2.39650995e-01 -6.99561596e-01 -6.41727969e-02 6.59961998e-01
4.91572469e-02 3.00490677e-01 2.21009091e-01 5.25001645e-01
9.58985370e-03 -2.11586177e-01 5.90002775e-01 -3.14278334e-01
3.25296670e-01 3.22778732e-01 -6.37885213e-01 4.78453159e-01
9.52511847e-01 -9.66958344e-01 2.14335591e-01 -4.63892939e-03
-1.80177003e-01 1.83360636e-01 7.61488497e-01 -8.77098918e-01
1.18431270e-01 1.47441924e-01 6.04819953e-01 -7.00085580e-01
5.23201451e-02 -6.57255232e-01 3.89314830e-01 3.88249576e-01
-2.68892646e-01 -2.13741422e-01 2.02433228e-01 3.77198100e-01
-1.34956762e-02 -7.59064198e-01 7.55396307e-01 -6.32878542e-02
-3.74663651e-01 8.89568478e-02 -6.18630588e-01 -1.55583158e-01
9.36272204e-01 -6.65082276e-01 -6.87264621e-01 -4.03171957e-01
-6.30112112e-01 -3.94851863e-01 7.07700074e-01 2.89128482e-01
5.62562764e-01 -9.10085022e-01 -6.08029366e-01 2.57350266e-01
-1.49326399e-01 -2.81837523e-01 2.24834561e-01 1.46856412e-01
-8.44367921e-01 3.20946962e-01 -4.84179944e-01 -3.90364885e-01
-1.51442277e+00 5.82416654e-01 1.70323685e-01 -1.99271083e-01
-3.81425172e-01 6.35125875e-01 -5.68386316e-01 -1.88103795e-01
7.15440288e-02 9.57128108e-02 5.24215326e-02 -7.01921657e-02
7.56288350e-01 8.06644499e-01 2.68977970e-01 -7.01183498e-01
-2.30522588e-01 -1.58624858e-01 -2.15280026e-01 5.48765421e-01
1.53671002e+00 3.19234639e-01 -1.13813944e-01 1.34432107e-01
1.61612952e+00 4.01771814e-02 -6.42273545e-01 -1.21478364e-01
-4.91845071e-01 -6.13625050e-01 6.67675734e-02 -5.82245886e-01
-1.09723353e+00 7.75655270e-01 4.19673622e-01 7.27271199e-01
8.93650830e-01 3.52969132e-02 7.90470004e-01 4.13380861e-01
4.87671077e-01 -5.95382392e-01 -3.27188849e-01 2.86544323e-01
7.40233064e-01 -1.29414928e+00 1.56958461e-01 -1.63219541e-01
-3.67734492e-01 1.55813956e+00 2.91846126e-01 -9.32149217e-02
7.29823172e-01 5.96818149e-01 -2.30022877e-01 -2.51256347e-01
-3.20834965e-01 4.80072439e-01 1.97458286e-02 8.88261974e-01
3.06192249e-01 8.23446140e-02 2.20609248e-01 5.01895964e-01
-4.77542371e-01 8.89171883e-02 9.98989046e-01 1.22986126e+00
-1.70011058e-01 -1.11888909e+00 -8.90896857e-01 8.83716702e-01
-1.11821616e+00 -1.27465278e-01 -5.90241253e-01 6.40282989e-01
1.41095564e-01 8.76411140e-01 2.78341174e-01 -4.64624107e-01
-3.56193572e-01 1.91883713e-01 1.65824175e-01 -3.80189538e-01
-8.34979117e-01 -5.12919605e-01 3.85477836e-03 -4.41351116e-01
-1.19734950e-01 -9.72670913e-01 -7.97630072e-01 -8.93060982e-01
-2.19842002e-01 3.56040485e-02 5.63813329e-01 9.46082890e-01
3.53435189e-01 5.96436374e-02 7.39361405e-01 -6.36757493e-01
-4.09485906e-01 -1.07714975e+00 -6.38806999e-01 3.47962916e-01
4.53048766e-01 -5.04163623e-01 -4.63493109e-01 6.90332770e-01] | [12.368803977966309, 1.0194900035858154] |
8248cdc9-3210-43dd-906a-93384acd0b9c | open-source-large-language-models-outperform | 2307.02179 | null | https://arxiv.org/abs/2307.02179v1 | https://arxiv.org/pdf/2307.02179v1.pdf | Open-Source Large Language Models Outperform Crowd Workers and Approach ChatGPT in Text-Annotation Tasks | This study examines the performance of open-source Large Language Models (LLMs) in text annotation tasks and compares it with proprietary models like ChatGPT and human-based services such as MTurk. While prior research demonstrated the high performance of ChatGPT across numerous NLP tasks, open-source LLMs like HugginChat and FLAN are gaining attention for their cost-effectiveness, transparency, reproducibility, and superior data protection. We assess these models using both zero-shot and few-shot approaches and different temperature parameters across a range of text annotation tasks. Our findings show that while ChatGPT achieves the best performance in most tasks, open-source LLMs not only outperform MTurk but also demonstrate competitive potential against ChatGPT in specific tasks. | ['Fabrizio Gilardi', 'Maria Korobeynikova', 'Juan Diego Bermeo', 'Shirin Dehghani', 'Zeynab Samei', 'Maël Kubli', 'Meysam Alizadeh'] | 2023-07-05 | null | null | null | null | ['text-annotation'] | ['natural-language-processing'] | [-4.41906869e-01 1.38063207e-01 -1.31476775e-01 -2.73190737e-01
-1.20261395e+00 -6.55926406e-01 8.93004715e-01 1.95821926e-01
-6.80470645e-01 6.66903019e-01 4.88474220e-01 -1.85324535e-01
2.09664211e-01 4.62871753e-02 -7.08024427e-02 -2.43061632e-01
1.29170775e-01 8.99307430e-01 6.16808712e-01 -1.18389107e-01
2.34207422e-01 -6.86903819e-02 -9.21343386e-01 4.17712510e-01
8.20310771e-01 3.53296399e-01 -5.20145148e-02 7.85247505e-01
-6.35909379e-01 1.03799021e+00 -6.94028020e-01 -8.22663546e-01
1.33367583e-01 5.23598418e-02 -1.46831131e+00 -6.30648732e-01
4.75790411e-01 -7.29623213e-02 -2.64870942e-01 5.73038161e-01
9.64991868e-01 3.49610716e-01 2.69620657e-01 -1.43313003e+00
-7.55461752e-01 9.65832353e-01 -5.60629368e-01 1.29287422e-01
2.71446943e-01 3.95364642e-01 1.26941812e+00 -5.44989407e-01
8.65403295e-01 1.35990715e+00 1.06104493e+00 6.96993589e-01
-1.24551415e+00 -6.24160111e-01 -2.23368093e-01 -2.29058862e-01
-1.27636933e+00 -5.73805690e-01 -1.72173053e-01 -3.33876908e-01
1.69322026e+00 2.78593183e-01 6.41404763e-02 1.27385664e+00
5.16565382e-01 9.61952627e-01 9.33078170e-01 -3.27480704e-01
1.83357835e-01 8.97304863e-02 5.84153354e-01 5.58027625e-01
-2.73130685e-02 -8.14774811e-01 -8.19143295e-01 -8.78952205e-01
7.92809650e-02 -2.91740119e-01 -2.62633245e-03 2.42023975e-01
-9.65845942e-01 7.14879930e-01 -2.73634791e-01 6.00390732e-01
-4.09835763e-02 4.19093043e-01 1.01423383e+00 1.63543478e-01
1.18000650e+00 5.98815143e-01 -7.28102744e-01 -8.23679745e-01
-9.09076691e-01 2.83581704e-01 1.28979969e+00 1.18152773e+00
4.35187846e-01 -2.17744842e-01 -7.90167153e-01 1.05668092e+00
3.58938761e-02 8.99480134e-02 7.41893768e-01 -8.67529988e-01
8.21928978e-01 3.78694564e-01 1.40298203e-01 -3.36475968e-01
-4.88649011e-01 2.42591619e-01 -2.85576642e-01 -3.14578801e-01
4.56484079e-01 -3.62210631e-01 -6.20319903e-01 1.36867619e+00
7.44343624e-02 8.23121965e-02 1.62507877e-01 5.36536455e-01
8.07514668e-01 7.39005864e-01 7.63558686e-01 1.19960746e-02
1.33900511e+00 -1.16557550e+00 -9.46864307e-01 -5.77022195e-01
1.47328758e+00 -8.15414131e-01 1.40984297e+00 -3.22536714e-02
-8.10892582e-01 -5.90406097e-02 -3.31294417e-01 -6.48884296e-01
-4.03303981e-01 -9.74694043e-02 6.14796638e-01 8.19818199e-01
-1.11145675e+00 3.55627328e-01 -9.38833356e-01 -1.00373554e+00
3.60326976e-01 -8.00110102e-02 -3.56070101e-01 1.26846045e-01
-1.23923922e+00 1.05036771e+00 3.50281596e-01 -4.82393503e-01
-8.31322134e-01 -8.17878962e-01 -4.48507607e-01 2.44476944e-01
6.50238514e-01 -1.15383096e-01 1.94577837e+00 -3.83863777e-01
-1.34030259e+00 8.96066368e-01 -1.59957379e-01 -6.90633178e-01
7.96741426e-01 -6.01980150e-01 -9.19481665e-02 -1.96637716e-02
9.24603865e-02 6.51373029e-01 3.16300035e-01 -5.46505988e-01
-4.98228908e-01 1.00432821e-01 -8.21309760e-02 1.40521720e-01
-5.78896999e-01 7.73577929e-01 -5.42729974e-01 -1.69297367e-01
-6.43361449e-01 -9.36462104e-01 -2.23166034e-01 -1.47208154e-01
-6.36405826e-01 -6.57865465e-01 1.24660051e+00 -7.24617422e-01
1.34700549e+00 -1.88454986e+00 -4.20185536e-01 -4.85231817e-01
3.01389813e-01 5.04460573e-01 -2.34662190e-01 9.10037994e-01
4.83344346e-01 8.80095363e-01 5.21952771e-02 -6.86216593e-01
1.87770143e-01 3.37299854e-01 -8.54301080e-02 2.84927130e-01
-1.61509201e-01 1.33517110e+00 -8.07426512e-01 -8.53610873e-01
-8.52788147e-03 3.04895043e-01 -3.23100984e-01 1.02385441e-02
-6.39033556e-01 3.83040197e-02 -6.12420797e-01 3.96451294e-01
1.29807055e-01 -5.67122817e-01 2.20282361e-01 4.13480043e-01
-2.89128870e-01 5.66198349e-01 -5.09903431e-01 1.76939559e+00
-4.65451002e-01 7.86740720e-01 2.61461169e-01 -1.81382105e-01
6.47337914e-01 9.23804402e-01 3.51776123e-01 -4.34262455e-01
1.09475300e-01 4.76984447e-03 -2.25662857e-01 -6.20237827e-01
9.08021748e-01 1.74032256e-01 -3.95217180e-01 1.29443944e+00
3.24245900e-01 -1.96479991e-01 1.28614873e-01 9.19427574e-01
1.31766331e+00 1.56723052e-01 2.39418074e-01 -4.38036621e-01
-1.00522302e-01 7.30761737e-02 3.13940316e-01 1.14985299e+00
-4.51102972e-01 2.32839525e-01 6.62559330e-01 -2.89168090e-01
-1.13905835e+00 -2.71673501e-01 3.15760672e-02 1.92080128e+00
-4.65512604e-01 -9.46132958e-01 -1.05359876e+00 -8.23059678e-01
-1.47914454e-01 1.01162755e+00 -2.61201024e-01 2.06554294e-01
-1.63332269e-01 -8.68374944e-01 1.44548559e+00 3.82959247e-01
5.52868307e-01 -1.09489977e+00 -5.15312433e-01 2.22578734e-01
-5.52186847e-01 -1.29422057e+00 -7.66269922e-01 1.84489474e-01
-5.57754159e-01 -7.38162518e-01 -5.02450943e-01 -3.84309679e-01
2.65345965e-02 2.07718164e-01 1.21961486e+00 2.75393259e-02
-3.98783922e-01 5.99923015e-01 -5.30380249e-01 -5.92508972e-01
-8.62151682e-01 7.09393740e-01 -1.28295422e-01 -4.12717432e-01
6.43555880e-01 -1.54535174e-01 5.85914105e-02 2.92459875e-01
-6.74920917e-01 8.07755738e-02 1.40598729e-01 4.57596034e-01
3.72657627e-02 -3.58363032e-01 6.25802219e-01 -1.10967219e+00
1.09262991e+00 -4.76437807e-01 -3.50781351e-01 5.47261834e-01
-7.04572916e-01 -1.74114466e-01 3.70836139e-01 -3.78685981e-01
-1.56173396e+00 -2.76887566e-01 -1.05010353e-01 -2.76936173e-01
9.14973617e-02 5.05343318e-01 2.04075366e-01 5.89176528e-02
9.64244485e-01 -5.28882258e-02 -4.64655012e-02 -7.39862621e-01
4.80784923e-01 9.02628422e-01 9.56621692e-02 -7.86592424e-01
3.49177063e-01 1.83199093e-01 -7.97166407e-01 -8.40316176e-01
-9.09205139e-01 -6.73865438e-01 -5.81254303e-01 1.56223461e-01
1.00433743e+00 -9.31575298e-01 -7.13395536e-01 4.59242612e-01
-1.13288128e+00 -7.39473522e-01 -1.84560210e-01 -8.55930075e-02
-2.44516864e-01 5.48964083e-01 -1.31214011e+00 -1.20074725e+00
-9.43686903e-01 -8.55498075e-01 1.08860195e+00 4.43930067e-02
-6.66710556e-01 -1.26717341e+00 1.60025448e-01 7.21018076e-01
5.24583638e-01 -3.01570296e-01 8.52486789e-01 -1.73052943e+00
-2.41932765e-01 -5.26522994e-01 -2.06306919e-01 2.00853813e-02
-1.24651745e-01 1.01504773e-01 -1.28400767e+00 -2.74738669e-01
-4.21965688e-01 -9.55978990e-01 4.33617085e-01 -1.34577723e-02
8.47428739e-01 -5.08567929e-01 -5.56564152e-01 4.78062816e-02
1.08611250e+00 -1.68738738e-01 4.06916589e-01 3.23321968e-01
7.54824042e-01 4.53882307e-01 4.78468925e-01 3.84308010e-01
3.23153079e-01 4.72146779e-01 -4.37754914e-02 2.82782584e-01
2.49746427e-01 -1.74494892e-01 3.13035607e-01 8.84270370e-01
2.69027464e-02 -8.50473762e-01 -1.40880728e+00 4.05655324e-01
-2.46591139e+00 -8.69343877e-01 -4.30396110e-01 1.83541346e+00
8.71171594e-01 1.65820420e-01 -1.04943849e-01 -4.67883110e-01
4.79970813e-01 4.25278008e-01 -4.66691822e-01 -7.87936389e-01
-1.97253257e-01 -1.78890228e-02 4.61338073e-01 1.76219210e-01
-7.99609959e-01 1.31824291e+00 7.75812054e+00 1.26098943e+00
-7.22904384e-01 7.99175560e-01 6.95238650e-01 -3.77243847e-01
1.45406693e-01 1.74728975e-01 -1.07397544e+00 3.78901303e-01
1.59671199e+00 -6.08966708e-01 1.34179935e-01 1.01474369e+00
2.89959520e-01 -1.13553174e-01 -9.30823267e-01 3.45307112e-01
1.14390990e-02 -1.38012564e+00 -5.52148260e-02 7.90384039e-02
6.11894965e-01 7.79306233e-01 -6.25753343e-01 6.64350092e-01
1.01205266e+00 -9.20142293e-01 5.67320704e-01 1.25838518e-01
6.72325790e-01 -3.86136800e-01 8.36337447e-01 5.98532796e-01
-9.07905936e-01 1.26765659e-02 -4.56150919e-01 -2.61714160e-01
3.39506030e-01 2.58180618e-01 -1.09978807e+00 4.55253989e-01
7.57367849e-01 4.74106133e-01 -6.81995451e-01 7.91424215e-01
-1.04098402e-01 1.08679450e+00 -3.00457597e-01 -1.52100965e-01
4.94283497e-01 1.35402113e-01 3.86099130e-01 1.72796977e+00
-9.45326835e-02 -6.59252563e-03 6.09953523e-01 5.53788900e-01
-3.85984391e-01 4.22792226e-01 -4.15554643e-01 -6.07886195e-01
7.44858742e-01 1.52775061e+00 -6.50768816e-01 -6.76889122e-01
-6.32147610e-01 8.75492096e-01 4.98052597e-01 1.73014551e-01
-6.86501384e-01 -1.98095649e-01 7.09510267e-01 4.27048542e-02
-1.33228377e-01 -1.14301234e-01 -1.27993971e-01 -1.01471925e+00
-3.07702422e-01 -8.48769486e-01 8.05072784e-01 -1.11578155e+00
-1.40846860e+00 6.12263501e-01 1.57381464e-02 -5.98209023e-01
-1.79852024e-01 -2.06368417e-01 -8.34520400e-01 8.47350478e-01
-9.24115241e-01 -1.40003991e+00 1.15807150e-02 3.88525575e-01
8.30128253e-01 -1.71408802e-02 1.01768816e+00 1.04513593e-01
-8.11575949e-01 6.62102163e-01 4.25464422e-01 4.60239291e-01
1.07153368e+00 -1.11576641e+00 1.11302483e+00 5.41136444e-01
6.19019382e-02 5.32014847e-01 6.66913569e-01 -8.96600306e-01
-1.05954325e+00 -1.08713531e+00 1.10348082e+00 -9.97245848e-01
1.01998556e+00 -9.02861655e-01 -1.34613681e+00 1.22598410e+00
4.98080164e-01 -3.35601084e-02 6.42558813e-01 4.18842494e-01
-6.33865654e-01 5.36564648e-01 -1.08003199e+00 5.34985483e-01
9.12343383e-01 -9.03719783e-01 -6.58657074e-01 7.51273990e-01
9.83510375e-01 -3.64624470e-01 -8.90588045e-01 -3.85863334e-01
5.38783967e-01 -5.96623063e-01 5.14437854e-01 -7.99937546e-01
2.37564787e-01 4.21879262e-01 2.68058926e-01 -1.21148133e+00
-1.65249497e-01 -1.15776300e+00 2.48308465e-01 1.61925495e+00
5.35150588e-01 -8.30793858e-01 6.15130603e-01 1.40152264e+00
-1.76510289e-01 -1.80353329e-01 -1.12701070e+00 -8.92107546e-01
3.38022172e-01 -4.79547381e-01 1.76365584e-01 1.27327442e+00
4.40966666e-01 6.37905121e-01 -6.66705430e-01 -3.37914765e-01
1.90636188e-01 -4.52208996e-01 8.40199769e-01 -1.30262542e+00
-9.42316779e-04 -5.16432635e-02 1.66026447e-02 -4.19033170e-01
3.13654184e-01 -9.63512540e-01 1.16645835e-01 -1.64531243e+00
5.55679142e-01 -3.79176497e-01 1.19443253e-01 1.23962152e+00
-1.72688961e-01 9.87939164e-02 3.25070173e-01 5.22155404e-01
-1.15560091e+00 1.60553023e-01 7.32298970e-01 -1.48132471e-02
-2.95837998e-01 -5.54195225e-01 -5.76130867e-01 6.45516098e-01
6.88656092e-01 -7.32421279e-01 -2.06210002e-01 -5.92468560e-01
8.03360045e-02 -1.14300564e-01 -6.81531131e-02 -9.04576182e-01
2.82739252e-01 -1.63562298e-01 -1.71241432e-01 -9.98742729e-02
1.86276764e-01 -2.70275861e-01 -9.80122574e-03 2.14061692e-01
-6.82059824e-01 -9.80663598e-02 4.48822320e-01 4.16293293e-01
2.58952379e-01 -3.73576522e-01 5.48404455e-01 -3.07885915e-01
-7.26718783e-01 2.97782540e-01 -8.13651621e-01 4.94181067e-01
8.14666152e-01 -2.18855068e-02 -9.01819885e-01 -4.81728762e-01
-3.68104726e-01 5.57863712e-01 3.45808893e-01 7.78816640e-01
-1.70340031e-01 -6.26719117e-01 -7.84808040e-01 -4.40658182e-01
3.25862169e-01 -2.31289133e-01 1.67362809e-01 8.06302071e-01
-2.75728524e-01 6.84291124e-01 1.32751495e-01 -4.18693751e-01
-1.51455605e+00 1.59850568e-01 1.62516877e-01 -7.13041067e-01
-9.43291903e-01 7.67678499e-01 1.26119122e-01 -6.89812243e-01
2.24686101e-01 -3.99077451e-03 2.73266405e-01 2.67344303e-02
7.82088816e-01 5.09217083e-01 1.37722820e-01 -3.33552450e-01
-2.84802526e-01 -2.66218185e-01 -4.26667869e-01 -1.82184696e-01
1.11020374e+00 -2.70788044e-01 -7.57730827e-02 6.99474573e-01
8.80203366e-01 -9.22938585e-02 -1.08766782e+00 -4.24857706e-01
5.23072183e-01 -1.68881521e-01 -2.65379611e-04 -1.12715745e+00
-4.48791116e-01 6.32885575e-01 1.95389673e-01 2.01475114e-01
3.58691990e-01 1.39893681e-01 9.40546691e-01 8.33278179e-01
5.51042080e-01 -1.32893682e+00 1.06910169e-01 1.00850952e+00
4.53206539e-01 -1.14918649e+00 1.00900546e-01 -1.11824058e-01
-9.53666747e-01 6.18767977e-01 7.81055510e-01 4.90220726e-01
2.71023214e-01 5.91093361e-01 4.41575646e-01 -3.74058217e-01
-1.48927069e+00 5.61553128e-02 -2.15021849e-01 2.45485634e-01
8.20457518e-01 -2.68602222e-01 -3.40914547e-01 5.97338617e-01
2.59057730e-02 1.77564546e-01 6.33766234e-01 1.32654154e+00
-4.47785795e-01 -7.90319622e-01 -2.03080907e-01 6.66503489e-01
-9.73502696e-01 -4.39591914e-01 -7.67563045e-01 9.46652472e-01
-5.44285715e-01 9.29746509e-01 7.09752738e-02 -8.91484320e-02
-1.85422767e-02 9.97508943e-01 -2.65724421e-01 -1.01646006e+00
-1.37889421e+00 1.10387206e-01 7.71981299e-01 -4.53563809e-01
-2.15642646e-01 -4.17479873e-01 -1.28689206e+00 -5.73443353e-01
-5.73538423e-01 7.46518731e-01 6.75551474e-01 1.01346910e+00
6.97343111e-01 6.18561730e-02 -2.96187103e-01 -5.69560051e-01
-4.23967987e-01 -1.45364618e+00 -4.22173858e-01 2.77077615e-01
-3.00257206e-01 -4.12949100e-02 -2.77966112e-01 8.09622854e-02] | [9.913901329040527, 8.779406547546387] |
fe97889d-d30a-4d73-8e6a-4dac5e778e86 | a-graph-to-graphs-framework-for | 2003.12725 | null | https://arxiv.org/abs/2003.12725v3 | https://arxiv.org/pdf/2003.12725v3.pdf | A Graph to Graphs Framework for Retrosynthesis Prediction | A fundamental problem in computational chemistry is to find a set of reactants to synthesize a target molecule, a.k.a. retrosynthesis prediction. Existing state-of-the-art methods rely on matching the target molecule with a large set of reaction templates, which are very computationally expensive and also suffer from the problem of coverage. In this paper, we propose a novel template-free approach called G2Gs by transforming a target molecular graph into a set of reactant molecular graphs. G2Gs first splits the target molecular graph into a set of synthons by identifying the reaction centers, and then translates the synthons to the final reactant graphs via a variational graph translation framework. Experimental results show that G2Gs significantly outperforms existing template-free approaches by up to 63% in terms of the top-1 accuracy and achieves a performance close to that of state-of-the-art template based approaches, but does not require domain knowledge and is much more scalable. | ['Ming Zhang', 'Jian Tang', 'Hongyu Guo', 'Chence Shi', 'Minkai Xu'] | 2020-03-28 | null | https://proceedings.icml.cc/static/paper_files/icml/2020/4152-Paper.pdf | https://proceedings.icml.cc/static/paper_files/icml/2020/4152-Paper.pdf | icml-2020-1 | ['retrosynthesis'] | ['medical'] | [ 6.00448370e-01 1.13925777e-01 -5.02378166e-01 1.83192283e-01
-8.13804865e-01 -9.79265928e-01 8.05767417e-01 3.17192435e-01
1.54287875e-01 1.00953233e+00 -6.94327876e-02 -5.56072891e-01
1.96127400e-01 -1.02747631e+00 -8.70508611e-01 -9.93105710e-01
3.23357642e-01 8.45024288e-01 4.58807021e-01 -4.01268035e-01
4.23590600e-01 6.71182215e-01 -1.13576496e+00 9.99613702e-02
1.14455032e+00 7.03458905e-01 1.63200885e-01 2.87028849e-01
-2.31036216e-01 5.28401017e-01 -2.69944847e-01 -5.69011867e-01
3.10852736e-01 -8.84662986e-01 -8.14434469e-01 -2.28468493e-01
4.46126834e-02 4.97019321e-01 -2.11415529e-01 1.05809689e+00
3.78836632e-01 2.97650874e-01 9.21034217e-01 -9.26159501e-01
-3.46836895e-01 5.59598625e-01 -2.55795091e-01 -3.07987899e-01
3.64860296e-01 -9.99422744e-02 1.03951252e+00 -1.13305593e+00
1.04091752e+00 1.14833844e+00 2.57317632e-01 6.08899772e-01
-1.64622915e+00 -8.27617764e-01 1.32787108e-01 2.41344329e-02
-1.44217753e+00 -4.99528199e-01 7.59512484e-01 -5.19468009e-01
1.24768555e+00 3.32628191e-01 6.19769812e-01 7.93621778e-01
4.59757328e-01 -8.43160823e-02 9.38392997e-01 -3.56208682e-01
5.67680359e-01 -2.03549594e-01 -2.93448180e-01 8.81904423e-01
3.62973601e-01 1.15969054e-01 -5.83505452e-01 -3.03662717e-01
4.97383744e-01 7.17196390e-02 -2.39777729e-01 -6.26153409e-01
-1.29499805e+00 1.01710737e+00 6.38099611e-01 2.45879620e-01
-4.36163098e-01 2.93406527e-02 1.53516173e-01 -4.87484746e-02
4.13166583e-01 9.27005470e-01 -4.05044168e-01 3.82348031e-01
-9.17565405e-01 3.21859717e-01 9.81823862e-01 9.97553766e-01
9.89695549e-01 -8.20725188e-02 5.53167239e-02 3.22065949e-01
2.87356317e-01 2.08591506e-01 -1.04723968e-01 -5.58545113e-01
6.24886692e-01 7.17193782e-01 2.13913307e-01 -5.22466958e-01
-4.26734447e-01 -1.71294823e-01 -5.89745700e-01 -1.10030837e-01
3.78598809e-01 -2.23121904e-02 -1.09753644e+00 1.48491740e+00
7.16230750e-01 -1.53000519e-01 2.54390687e-01 7.61596382e-01
7.07095981e-01 1.30440021e+00 2.31830195e-01 -8.19266737e-01
8.95973921e-01 -1.25551546e+00 -3.73558581e-01 -8.83737877e-02
6.83795512e-01 -9.67350602e-01 3.84779721e-01 4.96838510e-01
-1.06008589e+00 -2.69203037e-01 -1.15326810e+00 -1.14672147e-01
-6.37134552e-01 -8.54092687e-02 8.45877886e-01 7.48766363e-01
-5.94460309e-01 1.03139651e+00 -5.67866325e-01 -2.47156352e-01
-1.03673473e-01 6.78799152e-01 -4.98195797e-01 -2.15785742e-01
-9.70558524e-01 9.77160692e-01 8.50281239e-01 -1.99807584e-01
-1.15484786e+00 -1.08683574e+00 -8.68743122e-01 -1.34520248e-01
8.35696280e-01 -6.64878786e-01 1.11386549e+00 -7.62313664e-01
-1.76700759e+00 5.51343143e-01 -4.33627039e-01 -1.99766308e-02
2.72750497e-01 4.50602263e-01 -5.03492236e-01 9.80062224e-03
-3.89650762e-02 4.50017869e-01 5.26966155e-01 -1.16336739e+00
-2.90142149e-01 -1.90873101e-01 -5.84859997e-02 1.97381496e-01
3.29194963e-01 6.46298453e-02 -3.02489549e-01 -5.01267076e-01
2.25361601e-01 -1.17221236e+00 -5.99314451e-01 -4.29680258e-01
-6.41260326e-01 -3.21636051e-01 4.42720503e-01 -2.34554440e-01
1.10966480e+00 -1.53351629e+00 7.81699598e-01 6.23929620e-01
4.49783176e-01 2.60839045e-01 -1.42410412e-01 1.08633339e+00
-6.61984861e-01 1.89745888e-01 -1.98459744e-01 9.68028978e-02
-1.23276748e-01 -2.86247462e-01 -2.84601063e-01 5.08754849e-01
9.92057770e-02 7.01760173e-01 -1.16600311e+00 -2.01159775e-01
2.58173823e-01 3.30285817e-01 -5.27344108e-01 2.64383443e-02
-8.10304105e-01 4.84563649e-01 -6.60486639e-01 5.83183467e-01
5.15442371e-01 -3.55671346e-01 9.23744738e-01 -3.89697611e-01
-5.12854755e-01 8.14914107e-01 -8.99068654e-01 1.70503020e+00
2.81485710e-02 -1.91093445e-01 -4.26398844e-01 -1.01627123e+00
1.12588656e+00 4.52995926e-01 6.87010646e-01 -3.56612623e-01
1.36327192e-01 6.14930272e-01 1.44075230e-01 1.74184710e-01
4.30443764e-01 -5.70554733e-01 -1.91337261e-02 2.87201017e-01
1.16028465e-01 -3.61410618e-01 6.08914137e-01 2.52989382e-01
7.69977093e-01 3.51819158e-01 7.39996374e-01 -4.40072864e-01
6.26763463e-01 2.13100314e-01 6.52342975e-01 2.75874764e-01
3.01252425e-01 3.34401459e-01 6.44090772e-01 -5.30184984e-01
-1.33389401e+00 -8.47415447e-01 4.72210824e-01 8.33774984e-01
4.65762876e-02 -9.55752432e-01 -9.59734857e-01 -5.36355734e-01
-1.70323804e-01 6.42043889e-01 -4.98625934e-01 -2.95371383e-01
-2.93678105e-01 -7.02051342e-01 9.32758767e-03 1.22712374e-01
6.26310706e-02 -6.12500608e-01 3.28548014e-01 6.90801978e-01
9.32721719e-02 -7.88021505e-01 -7.91393220e-01 2.61250973e-01
-7.24330544e-01 -1.24915743e+00 -4.81450856e-01 -5.67015588e-01
7.07131505e-01 4.19514447e-01 7.87719190e-01 -2.26194993e-01
-1.49210125e-01 -5.18483818e-01 -1.19187362e-01 -5.40351987e-01
-8.58932436e-01 2.57045805e-01 3.93227227e-02 -2.04148926e-02
-3.76707092e-02 -4.75733519e-01 -5.32952130e-01 1.94293395e-01
-6.42357826e-01 3.36816579e-01 5.33372045e-01 5.30887127e-01
1.07621503e+00 2.57598199e-02 4.63248253e-01 -1.08892655e+00
1.57111838e-01 -3.81736606e-01 -1.06276608e+00 6.62850916e-01
-9.55115438e-01 4.62994814e-01 9.83292103e-01 -3.52711260e-01
-9.48403776e-01 5.64955652e-01 2.01053470e-01 -3.03545833e-01
3.31412584e-01 6.27804458e-01 -5.64437032e-01 -3.69212300e-01
7.09909558e-01 1.74218029e-01 -1.99972332e-01 -3.41824204e-01
4.92876649e-01 1.29428580e-01 1.55902609e-01 -7.31523275e-01
7.67828882e-01 -1.14046102e-02 7.47632265e-01 -6.28273368e-01
-7.95541227e-01 -5.60644448e-01 -5.06327450e-01 -3.40631939e-02
8.39053512e-01 -7.16947138e-01 -7.77737617e-01 -2.21270379e-02
-1.16900730e+00 -1.61396369e-01 7.39333779e-02 3.35566431e-01
-6.40379727e-01 4.28712785e-01 -3.29885632e-01 -3.78609866e-01
-5.08423150e-01 -1.57047033e+00 8.45179677e-01 5.02331089e-03
-7.74640441e-02 -7.85920441e-01 2.66565919e-01 4.26739365e-01
-1.35793939e-01 6.11119568e-01 1.25878203e+00 -6.67145252e-01
-9.29150879e-01 -1.45934805e-01 6.06761500e-02 -4.38145578e-01
3.16022277e-01 1.39451116e-01 -5.43871224e-01 -2.49811307e-01
-5.78940570e-01 1.68770365e-02 8.33530903e-01 2.29577035e-01
6.68507099e-01 -3.51826787e-01 -7.96778679e-01 4.57442999e-01
1.47691882e+00 6.97212458e-01 7.99477339e-01 1.25833545e-02
8.90711904e-01 3.94125581e-01 6.77629352e-01 1.67660594e-01
-4.97132204e-02 7.61863649e-01 3.66601437e-01 -3.50773633e-02
6.11451045e-02 -7.28957832e-01 3.60471785e-01 5.14665365e-01
-4.71220344e-01 -5.47952294e-01 -8.13087642e-01 1.16660580e-01
-1.90458083e+00 -9.22994614e-01 -1.90330461e-01 2.25095272e+00
1.08707500e+00 -1.17857926e-01 4.69559789e-01 2.06717134e-01
7.60995090e-01 1.02466255e-01 -6.65693343e-01 -3.99048090e-01
2.61648268e-01 7.35737324e-01 8.03858876e-01 5.96690476e-01
-9.13637280e-01 1.50771570e+00 6.97409010e+00 9.65664625e-01
-1.29264927e+00 -2.05017790e-01 3.33967119e-01 1.11495927e-01
-3.50429714e-01 4.72604573e-01 -8.45511675e-01 2.22207189e-01
1.17729795e+00 -5.34922183e-01 7.38905609e-01 6.50196373e-01
3.34963351e-01 2.95794547e-01 -1.35881460e+00 7.35631049e-01
-2.05525562e-01 -2.10739565e+00 2.84406275e-01 1.38366789e-01
9.68279898e-01 -3.32165569e-01 -3.63561213e-01 -7.79480562e-02
2.77948737e-01 -1.09169960e+00 9.44584548e-01 2.09025621e-01
8.65652502e-01 -9.91745293e-01 -2.09797751e-02 4.12821397e-02
-1.44018519e+00 3.63987386e-01 -2.86484689e-01 1.80061758e-01
-6.19428381e-02 3.51439685e-01 -1.02331161e+00 1.00039876e+00
-2.25641634e-02 7.21229196e-01 -3.37598696e-02 7.94562817e-01
-2.52091795e-01 4.36989993e-01 -1.43016055e-01 -3.80780458e-01
3.87444168e-01 -8.51626456e-01 2.58428544e-01 9.12989855e-01
3.95434827e-01 4.13811624e-01 4.44335788e-01 1.10084319e+00
-5.04640222e-01 3.92602861e-01 -7.23988056e-01 -5.63024521e-01
5.61258852e-01 1.12505829e+00 -9.59919274e-01 -4.59753543e-01
-2.34290957e-01 7.95968235e-01 1.26795024e-01 3.65816593e-01
-7.73133039e-01 -4.49871242e-01 6.02893293e-01 9.59721580e-02
3.80708486e-01 -2.42381766e-01 2.38479137e-01 -9.79552984e-01
-3.36333334e-01 -1.04771364e+00 2.89203674e-01 -5.86457789e-01
-7.78327584e-01 2.64005989e-01 -1.99259400e-01 -9.54938889e-01
-1.11713158e-02 -7.47966707e-01 -4.23408747e-01 8.68827581e-01
-1.28566670e+00 -9.34000015e-01 8.84294584e-02 4.13509786e-01
4.43358451e-01 3.44256163e-02 9.57233131e-01 1.69450771e-02
-6.86635315e-01 1.87134057e-01 2.86740035e-01 -5.84029675e-01
6.95737720e-01 -1.13609791e+00 5.91865897e-01 7.81174183e-01
-3.17672156e-02 7.59587944e-01 8.86730671e-01 -7.96569765e-01
-1.74518013e+00 -1.31851351e+00 9.30322886e-01 -1.65276513e-01
6.72968566e-01 -5.71007073e-01 -4.22636449e-01 5.18664360e-01
1.38025954e-02 -2.38802731e-01 8.08178067e-01 -2.05123290e-01
-6.19249642e-01 4.23527369e-03 -1.05624592e+00 8.36754203e-01
1.08498931e+00 -3.59303296e-01 -1.80578366e-01 9.01357472e-01
8.62480462e-01 -5.28840721e-01 -1.20216072e+00 4.60235067e-02
2.59281844e-01 -5.51487386e-01 1.11699760e+00 -8.97123575e-01
1.83355927e-01 -6.81128979e-01 -8.05475656e-03 -1.26584923e+00
-4.24118429e-01 -1.31431007e+00 1.37249514e-01 7.09966958e-01
9.24901724e-01 -7.11896479e-01 9.15223956e-01 6.50376737e-01
-3.64418298e-01 -4.79258448e-01 -6.19615912e-01 -9.80802119e-01
1.38440773e-01 1.36689931e-01 7.52076149e-01 1.05012548e+00
3.01544279e-01 8.46666574e-01 -3.25653732e-01 6.72456622e-02
6.32920444e-01 5.10425746e-01 5.35426080e-01 -1.22109914e+00
-4.13952917e-02 -4.37916964e-01 -1.78754821e-01 -6.82363570e-01
1.41695827e-01 -1.30339122e+00 -1.61474541e-01 -1.60647249e+00
1.85715362e-01 -2.70166636e-01 -6.67952299e-02 5.68670571e-01
3.75752710e-02 3.50174755e-02 1.00330211e-01 1.02367103e-01
-4.18326318e-01 5.57069302e-01 1.28722930e+00 -2.64456213e-01
-4.25026864e-01 -3.23183149e-01 -7.77081251e-01 1.75947055e-01
8.58758569e-01 -6.65731728e-01 -4.22784835e-01 5.62454164e-01
3.42725039e-01 2.21057743e-01 -2.91795254e-01 -6.56096816e-01
3.06207258e-02 -8.20873082e-01 2.46970393e-02 -5.63233614e-01
3.98102432e-01 -5.23288131e-01 9.44271445e-01 7.55990446e-01
-8.30259696e-02 -2.06935108e-01 6.15243129e-02 5.94847798e-01
1.98489338e-01 -7.57420212e-02 7.44713128e-01 -1.84238583e-01
-1.91876903e-01 6.27886653e-01 -3.06622684e-01 -4.02486384e-01
1.28304183e+00 -1.15571395e-01 -3.93869489e-01 6.52103871e-02
-5.48015416e-01 -1.88987002e-01 6.85617149e-01 1.53382599e-01
3.24692845e-01 -1.18143427e+00 -2.38913789e-01 -2.63752222e-01
2.23645106e-01 -1.01770824e-02 -3.14198822e-01 7.45683014e-01
-8.22569251e-01 1.01841247e+00 1.49274841e-01 -1.35169283e-01
-1.24859345e+00 1.06233573e+00 5.28372049e-01 -3.01967889e-01
-2.61981279e-01 4.39512193e-01 3.66259366e-01 -2.50920802e-01
-2.35042155e-01 -3.82481962e-01 3.07219446e-01 -1.37759686e-01
2.13229194e-01 2.39540026e-01 3.68226409e-01 -6.47663057e-01
-5.12490988e-01 6.74159408e-01 -2.35239759e-01 3.15784186e-01
1.23144901e+00 6.69335008e-01 -1.10181980e-01 -1.28133491e-01
1.05674148e+00 -9.97033063e-03 -8.72807264e-01 -5.83385080e-02
6.89067021e-02 -8.59033540e-02 -1.64797902e-02 -5.90561628e-01
-9.24966395e-01 5.39088845e-01 -7.61451423e-02 -1.43057585e-01
8.13602924e-01 9.59429294e-02 5.93190849e-01 6.43699467e-01
3.14033091e-01 -8.67983878e-01 3.43662016e-02 2.61649996e-01
7.41040945e-01 -8.34413469e-01 2.93498039e-01 -1.10241961e+00
-3.94717366e-01 1.33142555e+00 1.13008723e-01 6.42704070e-02
2.79408008e-01 -2.29592174e-01 -3.94768298e-01 -3.81443679e-01
-6.25398397e-01 -1.73462957e-01 3.88060123e-01 4.26031083e-01
5.75361073e-01 1.35207325e-01 -6.19421482e-01 2.26524159e-01
-1.61837742e-01 -1.85704276e-01 3.25271368e-01 9.28094447e-01
-4.90608841e-01 -1.77799201e+00 -2.69610733e-01 -7.85098523e-02
-3.37268323e-01 -3.03350806e-01 -1.06035435e+00 7.77319908e-01
6.40123710e-02 9.85015690e-01 -6.30396724e-01 -2.01272920e-01
4.52743828e-01 2.43604138e-01 9.80387986e-01 -8.05546641e-01
-5.17994881e-01 2.06233963e-01 3.39503676e-01 -6.08777761e-01
-5.47311902e-01 -5.44011652e-01 -1.44676244e+00 -4.84692186e-01
-6.32895470e-01 6.72335982e-01 6.98769152e-01 8.09815228e-01
4.61635768e-01 3.10303032e-01 8.04772556e-01 -9.35692549e-01
-2.06040144e-01 -4.35011506e-01 -3.70556712e-01 -2.71475781e-02
-4.13401388e-02 -7.12147951e-01 -2.56074369e-02 1.44692704e-01] | [4.53211784362793, 6.104106903076172] |
3921771a-58fd-43d4-82d3-8204fd5ba1ad | ddm-net-end-to-end-learning-of-keypoint | 2212.04575 | null | https://arxiv.org/abs/2212.04575v2 | https://arxiv.org/pdf/2212.04575v2.pdf | DDM-NET: End-to-end learning of keypoint feature Detection, Description and Matching for 3D localization | In this paper, we propose an end-to-end framework that jointly learns keypoint detection, descriptor representation and cross-frame matching for the task of image-based 3D localization. Prior art has tackled each of these components individually, purportedly aiming to alleviate difficulties in effectively train a holistic network. We design a self-supervised image warping correspondence loss for both feature detection and matching, a weakly-supervised epipolar constraints loss on relative camera pose learning, and a directional matching scheme that detects key-point features in a source image and performs coarse-to-fine correspondence search on the target image. We leverage this framework to enforce cycle consistency in our matching module. In addition, we propose a new loss to robustly handle both definite inlier/outlier matches and less-certain matches. The integration of these learning mechanisms enables end-to-end training of a single network performing all three localization components. Bench-marking our approach on public data-sets, exemplifies how such an end-to-end framework is able to yield more accurate localization that out-performs both traditional methods as well as state-of-the-art weakly supervised methods. | ['Gang Hua', 'Haoxiang Li', 'Enrique Dunn', 'Li Guan', 'Xiangyu Xu'] | 2022-12-08 | null | null | null | null | ['keypoint-detection'] | ['computer-vision'] | [-6.20516092e-02 -1.42509431e-01 -2.94690430e-01 -3.21940809e-01
-1.37816846e+00 -7.04272032e-01 8.39089692e-01 1.98498085e-01
-6.30663276e-01 1.14459090e-01 1.25786997e-02 8.13611299e-02
-4.78365123e-02 -3.49421322e-01 -1.01390827e+00 -3.54917079e-01
-1.39196321e-01 5.04257083e-01 3.06251198e-01 -2.68060248e-02
2.99567252e-01 8.45970690e-01 -1.48438215e+00 -1.18235201e-02
4.42560732e-01 1.14684844e+00 3.17411090e-04 4.41872507e-01
5.03117144e-01 4.90071803e-01 5.71110919e-02 -3.20951730e-01
7.15458989e-01 1.15904666e-01 -4.91634786e-01 1.93028182e-01
1.23897254e+00 -2.72087067e-01 -4.02173519e-01 8.46311510e-01
6.49624825e-01 7.02899545e-02 4.64335799e-01 -1.05683279e+00
-3.10598373e-01 -1.22572191e-01 -6.86040521e-01 3.88414972e-02
7.14882612e-01 2.95424700e-01 1.11306620e+00 -1.38046598e+00
7.58378923e-01 1.10416949e+00 1.21110582e+00 7.51979798e-02
-1.38856864e+00 -3.47281903e-01 -5.77760525e-02 -1.72322288e-01
-1.48029172e+00 -6.52330220e-01 9.17787194e-01 -3.56972635e-01
1.12728167e+00 -1.77113339e-01 5.80190122e-01 8.95500481e-01
8.40845481e-02 7.55261183e-01 7.77400613e-01 -6.07658565e-01
6.98873475e-02 -2.57908329e-02 -3.30913246e-01 8.46840620e-01
-5.51358471e-03 4.60523129e-01 -7.74663270e-01 -1.34166241e-01
8.46440673e-01 1.95659071e-01 -9.15669203e-02 -1.34104800e+00
-1.45675802e+00 6.66623890e-01 7.63398528e-01 1.36716858e-01
-2.05561563e-01 3.23675871e-01 2.25389436e-01 2.69248664e-01
5.25397837e-01 3.38434458e-01 -4.61643457e-01 5.01986481e-02
-1.06064260e+00 3.77005666e-01 6.75756872e-01 9.90961492e-01
1.04382646e+00 -4.84629899e-01 -4.73452434e-02 6.34026706e-01
2.76799500e-01 5.45761824e-01 8.71613100e-02 -7.84343839e-01
4.80875850e-01 5.61712205e-01 1.60341099e-01 -1.09933567e+00
-4.78709996e-01 -6.26620471e-01 -4.45449173e-01 2.35280663e-01
4.09196854e-01 3.13034922e-01 -8.49297106e-01 1.80055797e+00
5.62403142e-01 4.66119170e-01 -2.56437331e-01 9.77354407e-01
3.62459809e-01 1.12371214e-01 -3.15723985e-01 3.53462905e-01
1.23243487e+00 -9.64894533e-01 -5.55648096e-02 -3.35209548e-01
6.64795816e-01 -9.88276899e-01 1.02431560e+00 -6.10312223e-02
-1.29739952e+00 -5.64266801e-01 -1.04231131e+00 -3.69383097e-01
-3.62299800e-01 3.99368286e-01 4.51448709e-01 1.68479025e-01
-1.16274107e+00 4.68226314e-01 -9.31561291e-01 -5.77835679e-01
2.96000361e-01 4.56274837e-01 -8.26986134e-01 -2.19984904e-01
-7.33942449e-01 1.05781174e+00 1.61891863e-01 -1.68027412e-02
-8.16022396e-01 -1.02863073e+00 -1.17250025e+00 -1.74039081e-01
4.39309210e-01 -1.18236375e+00 1.03252172e+00 -5.52688301e-01
-1.17188418e+00 1.42683399e+00 1.57357473e-02 -4.69524324e-01
7.90172875e-01 -4.63207275e-01 8.03334117e-02 1.93326831e-01
4.53556091e-01 8.10841382e-01 9.03178036e-01 -1.28303659e+00
-6.97807908e-01 -5.74456096e-01 -1.16591923e-01 2.12711468e-01
9.93448123e-03 -1.11583993e-01 -9.43032801e-01 -6.52471483e-01
4.41787839e-01 -9.47551727e-01 -1.89794868e-01 4.27562267e-01
-1.70695528e-01 -8.85763690e-02 7.11055815e-01 -2.03743726e-01
6.10838890e-01 -2.12798977e+00 1.31146997e-01 3.34889382e-01
-7.74346432e-03 -9.36860144e-02 -4.04438615e-01 4.84146953e-01
-1.08716577e-01 -5.08963585e-01 -1.44589230e-01 -9.21078146e-01
1.95102319e-01 -1.20955743e-01 -2.03002736e-01 1.02425039e+00
4.17877614e-01 8.82346988e-01 -1.06837809e+00 -1.65149003e-01
5.68721473e-01 6.32498503e-01 -7.41154134e-01 3.37877780e-01
-4.72224802e-02 3.83042157e-01 -1.29646480e-01 9.69697654e-01
6.97824478e-01 -1.08870052e-01 -4.90527183e-01 -5.48606992e-01
-1.97624832e-01 1.21254742e-01 -1.31220126e+00 2.59487343e+00
-6.40551388e-01 3.88460219e-01 4.81589586e-02 -7.25130260e-01
8.33133459e-01 -1.06443539e-01 4.89500403e-01 -6.03041112e-01
3.64072695e-02 5.41154385e-01 -6.11874402e-01 -1.57739222e-01
3.57569605e-01 1.81497693e-01 -1.44376442e-01 1.74121067e-01
4.37289327e-01 -4.03592467e-01 2.22418867e-02 5.29248081e-02
1.05851781e+00 5.79377592e-01 3.14765245e-01 -1.05314650e-01
6.44995689e-01 -2.70483103e-02 2.58593082e-01 7.86155403e-01
-3.58646587e-02 1.00827515e+00 1.40705302e-01 -5.12423515e-01
-1.23105705e+00 -1.03982031e+00 -6.86604530e-02 9.62246299e-01
4.10278648e-01 -3.80266398e-01 -3.90976995e-01 -7.82709420e-01
4.08321410e-01 1.28720030e-01 -6.28627360e-01 -2.84252942e-01
-5.05465806e-01 -1.10138774e-01 4.53806162e-01 4.27415729e-01
4.15807873e-01 -4.44718063e-01 -6.02392614e-01 -4.13656905e-02
1.95651546e-01 -1.32738304e+00 -7.87806988e-01 3.78461480e-01
-5.30173779e-01 -1.15028739e+00 -6.54279888e-01 -8.89845431e-01
7.36864865e-01 5.45483828e-01 1.13699019e+00 -9.36433207e-03
-3.62366170e-01 7.76233375e-01 -2.06508726e-01 1.58022791e-02
-6.65370300e-02 2.02700496e-01 1.16592340e-01 5.15805595e-02
4.16941553e-01 -6.92609549e-01 -8.89663994e-01 4.80165124e-01
-7.15071261e-01 -2.86625147e-01 7.96406507e-01 8.32555175e-01
7.39977717e-01 -6.00675285e-01 -9.95069370e-02 -2.49295264e-01
1.12573311e-01 -1.77909464e-01 -9.07169402e-01 2.79821962e-01
-4.95266110e-01 6.91141412e-02 2.64300257e-01 -1.90465540e-01
-4.16204602e-01 6.69827461e-01 -1.44697160e-01 -8.07022810e-01
-1.60244778e-01 3.56176406e-01 -3.25766988e-02 -8.25150788e-01
7.08622932e-01 1.78579167e-01 9.88342986e-02 -4.09627587e-01
5.24311006e-01 2.07181379e-01 1.04164004e+00 -5.66001594e-01
1.35380793e+00 7.50743389e-01 1.98771030e-01 -4.53434467e-01
-9.97611344e-01 -1.01512575e+00 -1.12582076e+00 1.17567852e-01
6.50119185e-01 -1.41630065e+00 -6.06353521e-01 3.34986687e-01
-1.09697998e+00 -3.32767889e-02 -3.65006417e-01 5.87007046e-01
-9.37432051e-01 4.53585207e-01 -2.47437924e-01 -3.37885827e-01
-2.14364111e-01 -1.20880544e+00 1.79284346e+00 -1.19395107e-01
1.69759057e-02 -9.53360558e-01 4.79604214e-01 1.82935685e-01
3.41137141e-01 3.86427224e-01 2.00818598e-01 -5.13936639e-01
-7.84711540e-01 -7.81027853e-01 -3.72614413e-01 2.03418046e-01
-2.07926676e-01 -2.44534001e-01 -1.04368424e+00 -6.57234788e-01
-1.51869088e-01 -5.08192658e-01 1.13286567e+00 9.81897414e-02
5.81420183e-01 1.39946342e-01 -3.07945877e-01 1.13159466e+00
1.52805436e+00 -5.19352794e-01 2.47668073e-01 6.43315673e-01
8.37799549e-01 4.03000057e-01 7.32304811e-01 3.50500733e-01
6.34253323e-01 1.14650679e+00 6.36226475e-01 -4.07119304e-01
-1.54656067e-01 -4.84528244e-01 3.40303689e-01 7.38425776e-02
2.85546988e-01 2.48100698e-01 -9.65552688e-01 6.09292209e-01
-2.00714946e+00 -6.64324284e-01 2.80161887e-01 2.39077234e+00
6.43791080e-01 1.06967591e-01 1.46749750e-01 -1.63564071e-01
3.59092444e-01 2.02062085e-01 -4.49861795e-01 3.61521930e-01
-1.45741075e-01 1.88782450e-03 7.06535935e-01 5.27942538e-01
-1.47771573e+00 9.87810850e-01 5.42131805e+00 6.88202560e-01
-1.14538562e+00 -2.56406311e-02 2.64495462e-01 1.70111638e-02
-1.58911273e-01 3.36634755e-01 -8.02007675e-01 1.45909712e-01
3.47356945e-01 3.14021289e-01 2.11373761e-01 9.39101875e-01
3.61469761e-02 -1.01911910e-01 -1.53928447e+00 1.23535204e+00
3.97534072e-01 -1.48576665e+00 -7.05035105e-02 1.01221399e-02
6.71864033e-01 3.86878729e-01 7.05886409e-02 -5.37361503e-02
-5.90461642e-02 -6.68874979e-01 1.05716693e+00 3.63018572e-01
8.16345453e-01 -7.01907814e-01 6.75762594e-01 2.79004604e-01
-1.21712220e+00 -9.97877195e-02 -2.57208586e-01 1.15561508e-01
1.28078088e-01 6.11531436e-01 -7.22123563e-01 7.24099576e-01
6.20389283e-01 9.87049580e-01 -7.27855980e-01 1.23018670e+00
-2.83890814e-01 -2.16628388e-01 -6.33072674e-01 5.20032704e-01
4.83669013e-01 -4.03454043e-02 7.11959839e-01 1.26610637e+00
3.36178988e-01 -4.00156528e-01 5.21675766e-01 8.20593476e-01
-9.22590345e-02 -1.35769278e-01 -7.07614839e-01 4.32804376e-01
5.91642737e-01 1.34928334e+00 -6.27693117e-01 5.35816886e-02
-4.25315499e-01 1.08713579e+00 5.49453616e-01 4.05637249e-02
-6.03532732e-01 -2.18073562e-01 5.88909566e-01 2.96098232e-01
5.21411061e-01 -5.10135412e-01 -5.15723154e-02 -1.41323483e+00
2.23607674e-01 -6.48769319e-01 2.78842211e-01 -5.83568215e-01
-1.27363729e+00 2.83934325e-01 -1.82263568e-01 -1.56578970e+00
-3.83211732e-01 -4.96101886e-01 -6.06560647e-01 6.79395854e-01
-1.86976123e+00 -1.77025974e+00 -5.26528001e-01 8.21054995e-01
1.53159544e-01 -9.69415680e-02 5.70013463e-01 5.14638662e-01
-2.33677670e-01 7.71367311e-01 -1.02282509e-01 1.23406023e-01
1.17309344e+00 -1.25930691e+00 4.67251807e-01 1.09793830e+00
3.57187450e-01 6.80405557e-01 5.61388493e-01 -2.06953123e-01
-1.72678959e+00 -1.02081501e+00 9.22417998e-01 -8.40284646e-01
6.47002518e-01 -6.60589397e-01 -4.68718261e-01 5.78222334e-01
-2.05535367e-01 6.45706892e-01 1.75769135e-01 7.31425807e-02
-6.87140048e-01 -2.86895096e-01 -1.03507423e+00 4.12278712e-01
1.15827906e+00 -8.74913514e-01 -3.99989933e-01 5.04746139e-01
5.81643045e-01 -8.95884693e-01 -7.58962393e-01 4.05625045e-01
4.87698227e-01 -1.26777995e+00 1.41783321e+00 -9.07612965e-02
1.71616942e-01 -6.69881225e-01 -2.55535841e-01 -1.00829732e+00
-2.21713170e-01 -8.83196712e-01 -1.02628837e-03 1.14161158e+00
4.72299308e-02 -3.40978444e-01 8.60157788e-01 1.91207692e-01
-3.47310990e-01 -7.56539464e-01 -1.09709430e+00 -8.11060131e-01
-2.73988068e-01 -4.58787143e-01 2.76231080e-01 7.50560522e-01
-3.31625193e-01 2.30318725e-01 -3.52942288e-01 3.77846152e-01
9.91965652e-01 1.62284106e-01 1.31842554e+00 -9.63479221e-01
-1.86214298e-01 -4.65913594e-01 -8.80124509e-01 -1.33305752e+00
2.09055975e-01 -9.06347096e-01 7.89788514e-02 -1.04172421e+00
1.15741789e-02 -4.48091477e-01 -5.17679118e-02 4.46287155e-01
3.33658084e-02 6.07454598e-01 1.75493494e-01 4.09497797e-01
-7.99677849e-01 4.72616345e-01 7.16323256e-01 8.79253745e-02
-1.80257231e-01 -4.91021723e-02 -4.17238235e-01 7.51061022e-01
2.12157413e-01 -4.58363652e-01 -1.36953324e-01 -4.85158533e-01
2.24337652e-01 -1.91309363e-01 1.03474915e+00 -1.12301099e+00
7.18013585e-01 2.84313112e-01 5.26425481e-01 -7.96211958e-01
3.90836954e-01 -1.08902478e+00 -1.88976750e-01 2.78788626e-01
-3.14279199e-01 1.01947799e-01 -2.28368677e-02 6.87283218e-01
-3.63824785e-01 1.91103563e-01 8.12511563e-01 -7.99479187e-02
-8.51856649e-01 5.94103098e-01 6.37607574e-01 6.84798360e-02
1.10957396e+00 -5.36548615e-01 5.29514290e-02 -2.56003350e-01
-4.10536170e-01 2.93065429e-01 9.71573472e-01 4.20558780e-01
5.32366872e-01 -1.35929084e+00 -6.68424845e-01 5.07080019e-01
6.45790517e-01 2.25588992e-01 2.29809768e-02 1.22233224e+00
-5.52378595e-01 2.53199309e-01 8.66340008e-03 -1.13333559e+00
-1.09703624e+00 4.07643676e-01 5.42888343e-01 -2.59006619e-01
-5.76058447e-01 9.84848976e-01 -9.45187211e-02 -7.06923962e-01
5.86240888e-01 -2.76190579e-01 4.03676897e-01 -2.59038918e-02
1.95957392e-01 8.63138884e-02 3.72274071e-01 -7.68825531e-01
-6.43751264e-01 1.24270129e+00 -8.80859792e-02 5.05599529e-02
1.36539388e+00 -2.30558768e-01 2.94900942e-03 8.85500535e-02
1.65470147e+00 1.33439690e-01 -1.58143532e+00 -5.46967685e-01
2.83583328e-02 -6.95945621e-01 1.08292885e-01 -6.79979205e-01
-7.11329877e-01 6.94919884e-01 7.38234997e-01 -2.31962547e-01
8.94222736e-01 1.14587776e-01 7.27074385e-01 5.16111851e-01
3.55974019e-01 -9.40200210e-01 2.29357243e-01 6.63024902e-01
7.33326674e-01 -1.62009215e+00 1.90952763e-01 -2.51602709e-01
-1.73464537e-01 1.09683681e+00 2.87609369e-01 -5.39348662e-01
5.27909815e-01 1.31395593e-01 -1.61197409e-02 -3.51856083e-01
-4.77217108e-01 -3.82977694e-01 6.87640369e-01 6.27689481e-01
1.98903888e-01 -5.40451050e-01 3.10701251e-01 -8.19912180e-02
3.38456929e-02 -1.06864320e-02 -6.63860813e-02 9.82994676e-01
-2.44366884e-01 -9.57129121e-01 -3.90152365e-01 -2.25443870e-01
-2.53680408e-01 -1.13196727e-02 -3.49661142e-01 1.05053937e+00
1.44746915e-01 4.88127053e-01 6.28079027e-02 -3.25455099e-01
7.84608066e-01 -3.13300103e-01 6.83742583e-01 -4.45502251e-01
-6.24707878e-01 2.33282551e-01 -2.38924310e-01 -1.02627504e+00
-5.35142124e-01 -6.20866716e-01 -8.58999670e-01 -3.56446505e-02
-2.42123336e-01 -1.94069549e-01 6.54269218e-01 7.79374301e-01
7.08345115e-01 -6.89046383e-02 8.15070510e-01 -1.54215729e+00
-9.21678364e-01 -4.80698198e-01 -3.20451796e-01 6.34798467e-01
7.14594066e-01 -7.21207201e-01 -5.60475111e-01 -2.32433975e-01] | [7.978281021118164, -2.1625473499298096] |
0137b61c-ea2c-4187-b31d-fa5e347f0dfa | zooming-slow-mo-fast-and-accurate-one-stage | 2002.11616 | null | https://arxiv.org/abs/2002.11616v1 | https://arxiv.org/pdf/2002.11616v1.pdf | Zooming Slow-Mo: Fast and Accurate One-Stage Space-Time Video Super-Resolution | In this paper, we explore the space-time video super-resolution task, which aims to generate a high-resolution (HR) slow-motion video from a low frame rate (LFR), low-resolution (LR) video. A simple solution is to split it into two sub-tasks: video frame interpolation (VFI) and video super-resolution (VSR). However, temporal interpolation and spatial super-resolution are intra-related in this task. Two-stage methods cannot fully take advantage of the natural property. In addition, state-of-the-art VFI or VSR networks require a large frame-synthesis or reconstruction module for predicting high-quality video frames, which makes the two-stage methods have large model sizes and thus be time-consuming. To overcome the problems, we propose a one-stage space-time video super-resolution framework, which directly synthesizes an HR slow-motion video from an LFR, LR video. Rather than synthesizing missing LR video frames as VFI networks do, we firstly temporally interpolate LR frame features in missing LR video frames capturing local temporal contexts by the proposed feature temporal interpolation network. Then, we propose a deformable ConvLSTM to align and aggregate temporal information simultaneously for better leveraging global temporal contexts. Finally, a deep reconstruction network is adopted to predict HR slow-motion video frames. Extensive experiments on benchmark datasets demonstrate that the proposed method not only achieves better quantitative and qualitative performance but also is more than three times faster than recent two-stage state-of-the-art methods, e.g., DAIN+EDVR and DAIN+RBPN. | ['Yun Fu', 'Yulun Zhang', 'Yapeng Tian', 'Jan P. Allebach', 'Chenliang Xu', 'Xiaoyu Xiang'] | 2020-02-26 | zooming-slow-mo-fast-and-accurate-one-stage-1 | http://openaccess.thecvf.com/content_CVPR_2020/html/Xiang_Zooming_Slow-Mo_Fast_and_Accurate_One-Stage_Space-Time_Video_Super-Resolution_CVPR_2020_paper.html | http://openaccess.thecvf.com/content_CVPR_2020/papers/Xiang_Zooming_Slow-Mo_Fast_and_Accurate_One-Stage_Space-Time_Video_Super-Resolution_CVPR_2020_paper.pdf | cvpr-2020-6 | ['space-time-video-super-resolution'] | ['computer-vision'] | [ 3.98106456e-01 -2.72246599e-01 -3.20668340e-01 -3.25106233e-01
-9.51960981e-01 -8.52646958e-03 4.47555035e-01 -8.99455070e-01
-2.63038963e-01 9.17339981e-01 3.34016025e-01 -7.78784137e-03
8.73491317e-02 -7.49661028e-01 -9.50817227e-01 -6.32574975e-01
1.92116320e-01 -1.70322120e-01 4.72145587e-01 -9.32520255e-02
4.97432128e-02 4.17031318e-01 -1.64614761e+00 7.18334198e-01
6.71786785e-01 9.84614611e-01 5.49595535e-01 7.46760011e-01
3.75144021e-03 1.16293788e+00 -2.05754504e-01 9.33651626e-02
2.87112892e-01 -6.71638906e-01 -6.88985765e-01 1.74754381e-01
4.81393486e-01 -9.23413515e-01 -6.65205300e-01 9.07826602e-01
3.40053409e-01 4.71184373e-01 1.09483980e-01 -8.15093338e-01
-8.71643424e-01 6.39384866e-01 -9.20760632e-01 4.56815720e-01
6.31571233e-01 2.56015718e-01 5.52643359e-01 -1.17486441e+00
8.72596979e-01 1.49769056e+00 4.25854743e-01 8.50005209e-01
-1.13755178e+00 -8.23588550e-01 2.79824048e-01 2.24391714e-01
-1.51347029e+00 -6.52720094e-01 7.47961164e-01 -2.86684602e-01
7.64976025e-01 1.48189157e-01 4.18512553e-01 1.21126413e+00
8.90381262e-02 6.53542161e-01 9.37231600e-01 7.69294575e-02
-1.04715303e-02 -4.53626096e-01 -4.54268962e-01 4.47811544e-01
-2.49917910e-01 3.41767430e-01 -6.16252124e-01 3.28432798e-01
1.57671165e+00 3.47904265e-01 -6.53861165e-01 2.93044806e-01
-1.44225383e+00 4.47323829e-01 4.14656013e-01 4.94342148e-01
-5.71227551e-01 1.00325204e-01 2.40825862e-01 1.89887553e-01
5.67481935e-01 -1.21558480e-01 -2.35659495e-01 6.08561411e-02
-1.36253321e+00 2.66766578e-01 6.50637001e-02 1.01777649e+00
7.13765144e-01 4.76401508e-01 -3.28244150e-01 7.78789461e-01
2.25741729e-01 1.28759548e-01 4.90951210e-01 -1.37546480e+00
6.53110683e-01 2.02791356e-02 4.43015724e-01 -8.94889176e-01
-6.73920885e-02 -5.45586571e-02 -1.31704605e+00 1.17631897e-01
1.93968117e-01 3.60980779e-02 -1.00980675e+00 1.57034755e+00
2.58510381e-01 8.04011583e-01 1.44398242e-01 1.45530105e+00
1.09211218e+00 1.17895007e+00 3.10223028e-02 -8.80738735e-01
1.08328271e+00 -1.06956542e+00 -8.53893042e-01 9.73760933e-02
-8.26492980e-02 -6.27571046e-01 8.50144267e-01 3.21229577e-01
-1.69672394e+00 -1.14947534e+00 -9.56480026e-01 -5.68726063e-01
3.77907246e-01 -1.58348363e-02 1.56950533e-01 -2.67784983e-01
-1.17661810e+00 7.24267840e-01 -7.90526986e-01 1.89584985e-01
3.64538193e-01 1.66805521e-01 -3.50685388e-01 -3.28864604e-01
-1.33177650e+00 4.10849482e-01 3.86799067e-01 3.11719239e-01
-1.00221050e+00 -8.20372164e-01 -9.44441378e-01 -2.85129864e-02
4.52761322e-01 -7.87093759e-01 1.09617782e+00 -1.23350751e+00
-1.70623958e+00 4.21651036e-01 -6.17895961e-01 -3.73348147e-01
6.96411908e-01 -2.02894017e-01 -6.44894779e-01 2.50185519e-01
7.62291625e-02 6.78877950e-01 1.13486040e+00 -1.33253324e+00
-1.08242822e+00 -1.93533639e-03 2.38280613e-02 2.60218740e-01
2.09495634e-01 2.56905824e-01 -7.01018095e-01 -9.25663531e-01
5.55998795e-02 -5.01148045e-01 -2.66116768e-01 -5.77630363e-02
-6.61769509e-02 -5.96837029e-02 9.61944878e-01 -1.06747544e+00
1.43843758e+00 -2.10102558e+00 4.18681353e-01 -3.87932360e-01
3.82910967e-01 4.02432412e-01 -2.64297634e-01 -2.42570475e-01
-1.89792067e-01 1.48364678e-01 -1.03832416e-01 -4.13417310e-01
-5.72697341e-01 1.14839643e-01 -5.74429870e-01 2.64460236e-01
2.95970261e-01 7.92367041e-01 -9.75013971e-01 -4.86107588e-01
4.18715090e-01 1.14343965e+00 -3.92177880e-01 3.70240778e-01
-2.02097833e-01 9.56504881e-01 -3.72338325e-01 5.36190510e-01
7.89513767e-01 -3.39153767e-01 1.64134800e-02 -6.18368089e-01
-4.80451494e-01 1.11111231e-01 -1.13472617e+00 1.90108204e+00
-5.78435302e-01 5.58327675e-01 3.96306477e-02 -5.87125957e-01
7.85468638e-01 4.30315644e-01 5.76553702e-01 -9.05466139e-01
-5.24044149e-02 2.27649480e-01 -4.61332142e-01 -4.93176252e-01
7.61502028e-01 -7.82127976e-02 3.35820884e-01 1.57544240e-02
-2.02296913e-01 5.81145883e-01 1.14515699e-01 -7.63642490e-02
7.24647403e-01 7.12926388e-01 2.65393317e-01 2.71433145e-01
8.75891030e-01 -5.50637066e-01 1.07610929e+00 2.50857055e-01
-9.77157131e-02 1.11891973e+00 6.25588465e-03 -7.01671064e-01
-1.31927598e+00 -1.07276225e+00 1.45454407e-01 8.56530726e-01
4.51023430e-01 -2.29506180e-01 -6.59018338e-01 -2.82073021e-01
-5.21382511e-01 3.67771357e-01 -4.10842955e-01 1.42442405e-01
-1.26876044e+00 -4.37872857e-01 6.07555322e-02 7.20927060e-01
7.62821734e-01 -1.13449395e+00 -4.67781126e-01 4.73410070e-01
-6.23737037e-01 -1.43744349e+00 -8.00585032e-01 -5.64057291e-01
-8.95416975e-01 -6.30282640e-01 -1.04979670e+00 -8.39424729e-01
3.50239009e-01 6.85103893e-01 1.02404642e+00 1.53010443e-01
4.81731147e-02 -3.46800804e-01 -3.78482223e-01 5.38810253e-01
-3.86759460e-01 -3.08094949e-01 1.52480409e-01 3.25117737e-01
-2.23800843e-03 -6.49533451e-01 -9.08292532e-01 4.57473934e-01
-1.10769451e+00 6.34105980e-01 5.21281123e-01 8.89353991e-01
1.30431771e+00 1.68130204e-01 5.89413285e-01 -5.31145096e-01
7.03930780e-02 -4.85882819e-01 -5.90147197e-01 1.51264280e-01
-2.24790826e-01 -1.12772070e-01 9.92917776e-01 -6.23942792e-01
-1.42667007e+00 3.25436629e-02 -7.75216147e-02 -1.19992280e+00
-3.26229855e-02 1.45088121e-01 -2.68672824e-01 1.02388717e-01
2.70519048e-01 5.17961204e-01 -3.03467214e-01 -5.08465886e-01
3.38375419e-01 5.36184430e-01 9.24672127e-01 -2.18714386e-01
9.15690958e-01 5.38682222e-01 -1.26697615e-01 -6.54753625e-01
-7.48941004e-01 -6.91647604e-02 -5.15442431e-01 -1.67049244e-01
1.30070770e+00 -1.40402579e+00 -5.52448988e-01 4.07323092e-01
-1.24886060e+00 -4.35640514e-01 -1.51992828e-01 4.40361351e-01
-7.57188439e-01 4.07936543e-01 -9.69892502e-01 -6.02018952e-01
-3.68731111e-01 -1.26276243e+00 1.14484787e+00 4.56577450e-01
2.52361149e-01 -5.20799100e-01 -3.26803535e-01 4.25334543e-01
4.94551361e-01 3.67033154e-01 2.59475350e-01 3.23293239e-01
-1.18385041e+00 4.14953649e-01 -6.02879286e-01 2.49799132e-01
2.72156298e-01 1.34891376e-01 -6.92828417e-01 -3.94089997e-01
1.02830335e-01 1.22611821e-02 9.29466128e-01 6.08618319e-01
1.48921478e+00 -4.94203687e-01 8.87711812e-03 1.17175460e+00
1.54632640e+00 3.38077307e-01 9.54723001e-01 2.11727530e-01
1.11106002e+00 2.24987462e-01 1.06764722e+00 4.13513720e-01
5.82404256e-01 9.51697469e-01 1.96737915e-01 -1.21532574e-01
-4.56602275e-01 -4.02492642e-01 5.59256852e-01 7.58937120e-01
-7.10897982e-01 -1.40739664e-01 -2.75941163e-01 2.84260392e-01
-1.97326338e+00 -1.44191742e+00 -3.26886252e-02 2.12343502e+00
9.19368505e-01 -1.35349974e-01 1.93493545e-01 -1.67188216e-02
9.38258231e-01 5.25878429e-01 -7.08682835e-01 3.15061472e-02
-2.15965927e-01 -3.77664864e-02 2.39203125e-01 6.04027987e-01
-1.03392327e+00 1.02001488e+00 4.98751831e+00 9.84425128e-01
-1.25027585e+00 1.70771778e-01 9.76261318e-01 -3.28963071e-01
-2.74252504e-01 -2.52377808e-01 -8.33903551e-01 7.67267466e-01
9.81990933e-01 -3.48246358e-02 9.47951615e-01 5.31986296e-01
6.74449623e-01 2.42624626e-01 -9.39913750e-01 1.32615340e+00
-8.03543776e-02 -1.70421350e+00 2.99821198e-01 -2.65181512e-01
8.13924015e-01 -1.50279090e-01 8.21165815e-02 2.30048314e-01
-1.45820191e-03 -1.08046794e+00 7.98133433e-01 7.45023549e-01
1.43168938e+00 -8.70287836e-01 3.44538212e-01 1.56056166e-01
-1.80233347e+00 -1.76314026e-01 -5.49134135e-01 1.76279321e-01
6.47847712e-01 3.97620469e-01 1.64377466e-01 7.37721920e-01
1.07858050e+00 1.14303458e+00 -9.57921520e-02 5.45245826e-01
-2.12985324e-03 1.36875451e-01 1.09726675e-01 9.27553356e-01
6.52468875e-02 -1.94488972e-01 4.76087302e-01 9.83832300e-01
5.52206159e-01 6.19748533e-01 1.85711026e-01 1.04356396e+00
-1.58590391e-01 -3.51653814e-01 -2.70502716e-01 2.26206318e-01
4.74226743e-01 1.31612921e+00 -3.66803914e-01 -5.39252818e-01
-7.14982688e-01 1.17686152e+00 3.00607178e-02 5.96884906e-01
-1.18772984e+00 -3.26044448e-02 7.17523932e-01 1.92569569e-01
4.64706153e-01 -1.35946482e-01 1.77476965e-02 -1.55163038e+00
1.77979678e-01 -9.76839304e-01 1.91941723e-01 -1.06142759e+00
-1.03522694e+00 9.89253998e-01 -2.21735537e-01 -1.73882294e+00
-5.43827236e-01 2.51466483e-02 -2.22527131e-01 9.96703267e-01
-1.85420549e+00 -1.17661405e+00 -6.93864584e-01 9.23180997e-01
1.25386739e+00 2.53719147e-02 1.42710716e-01 5.78289151e-01
-6.33628488e-01 4.63010609e-01 -1.08603053e-01 4.53051254e-02
6.78964257e-01 -6.79297805e-01 5.22151828e-01 1.14845967e+00
-3.32161576e-01 3.27716976e-01 5.10258496e-01 -7.36146808e-01
-1.33817220e+00 -1.60389519e+00 5.62681019e-01 -6.61453158e-02
3.70121658e-01 9.65625606e-03 -1.31820774e+00 6.23535573e-01
-5.33668920e-02 5.31444788e-01 -7.81489443e-03 -8.49910378e-01
-2.94414401e-01 -1.85035825e-01 -1.14340949e+00 6.65641129e-01
1.33589709e+00 -5.47455370e-01 -2.82451004e-01 -8.90834406e-02
1.31377828e+00 -6.82224989e-01 -1.11347628e+00 5.85998654e-01
4.46598858e-01 -1.21040535e+00 1.35379791e+00 -2.00558364e-01
1.02755177e+00 -7.53145754e-01 -2.43015736e-01 -8.17544222e-01
-5.79627335e-01 -8.56481791e-01 -5.64881146e-01 1.22881973e+00
-1.10023953e-01 -2.31816664e-01 5.30196071e-01 6.14814341e-01
-1.34908423e-01 -7.72088051e-01 -7.84054756e-01 -6.34224176e-01
-2.53863722e-01 -3.01477611e-02 6.36100113e-01 9.27246749e-01
-5.98787129e-01 1.12575486e-01 -8.74407172e-01 2.57171422e-01
7.44479597e-01 2.38044292e-01 5.39295197e-01 -7.76641488e-01
-3.30442131e-01 -2.14362040e-01 -6.59687147e-02 -1.33184326e+00
1.77173749e-01 -3.18297088e-01 1.41448855e-01 -1.38248110e+00
1.99232906e-01 -1.77003086e-01 -3.91162574e-01 1.43230230e-01
-4.03246105e-01 4.55834299e-01 2.40231752e-01 4.88262504e-01
-6.27205968e-01 4.89710152e-01 1.56443560e+00 2.03317136e-01
-4.97523665e-01 -3.16798866e-01 -4.29064691e-01 6.91085875e-01
4.42305088e-01 -2.92735000e-04 -4.53426480e-01 -5.97769916e-01
-1.85313404e-01 9.52396035e-01 4.57218826e-01 -9.24818575e-01
7.16204420e-02 -5.67401707e-01 7.98177302e-01 -6.86704218e-01
4.99976516e-01 -6.15212500e-01 5.88753581e-01 1.17129952e-01
-4.11157966e-01 1.77778408e-01 -1.16942793e-01 6.23753071e-01
-3.97854447e-01 3.96965712e-01 1.10966170e+00 -1.87814802e-01
-1.00976765e+00 8.42052400e-01 -1.45845041e-01 -1.67042121e-01
8.78652513e-01 -4.25058097e-01 -3.60022187e-01 -3.00785512e-01
-6.17342949e-01 5.59583418e-02 6.70514703e-01 7.36020982e-01
1.14837444e+00 -1.47916853e+00 -8.51306975e-01 2.44168743e-01
-4.09053922e-01 4.85437602e-01 9.52585995e-01 7.27617323e-01
-4.63296831e-01 1.33621708e-01 -4.32594389e-01 -5.29902518e-01
-1.18232524e+00 9.80853915e-01 2.31615230e-01 -1.47971347e-01
-1.01847649e+00 6.29811049e-01 5.68112195e-01 3.54512632e-01
-6.91615120e-02 -2.96869487e-01 -3.70736033e-01 -2.23430112e-01
1.09156704e+00 3.74388337e-01 -5.65797091e-01 -1.04111481e+00
-7.18335062e-02 8.27848971e-01 -1.36226729e-01 3.89946364e-02
1.38792181e+00 -6.82010174e-01 2.94554848e-02 2.89866030e-01
1.15700936e+00 -3.93033594e-01 -1.86632526e+00 -4.23980772e-01
-4.55646843e-01 -8.86543393e-01 1.53125986e-01 -3.47702444e-01
-1.52241755e+00 6.42377436e-01 5.05861461e-01 -1.90713063e-01
1.68065500e+00 -3.47448826e-01 1.32978714e+00 -2.43062988e-01
5.32147944e-01 -8.62918913e-01 -8.71724635e-03 2.18484506e-01
9.37428415e-01 -1.21568692e+00 7.93856289e-03 -5.05429685e-01
-6.13303185e-01 1.18322515e+00 7.87642896e-01 -2.50233442e-01
1.88263476e-01 1.95489079e-01 -1.88430399e-01 4.60076690e-01
-9.50941205e-01 1.22945495e-02 3.29671383e-01 4.62133646e-01
5.26680589e-01 -2.25625932e-01 -2.36897439e-01 5.84141970e-01
1.83579445e-01 6.53764367e-01 7.44889855e-01 3.96572143e-01
-2.87048846e-01 -7.17420518e-01 -2.32509255e-01 1.11424617e-01
-5.70954502e-01 -3.15605178e-02 3.63777012e-01 5.60367405e-01
7.26528838e-02 8.83063674e-01 4.21682000e-02 -6.04240537e-01
3.39378715e-02 -5.33538520e-01 3.30593050e-01 -1.17808394e-01
-2.53479451e-01 4.00486141e-01 -1.64713398e-01 -1.16478395e+00
-7.34059155e-01 -4.47620153e-01 -1.41942966e+00 -4.79758531e-01
1.09479152e-01 -3.17032754e-01 1.46114007e-01 7.88537860e-01
4.42922205e-01 8.51702631e-01 7.17593014e-01 -1.48517871e+00
-1.36102811e-01 -5.95999897e-01 -4.05081302e-01 3.79429340e-01
7.34599710e-01 -4.75988746e-01 -2.35329822e-01 4.33075666e-01] | [11.016128540039062, -1.860333800315857] |
3e2f38e9-5f46-414e-9a27-c861e1e60526 | image-shadow-removal-using-end-to-end-deep | null | null | https://www.mdpi.com/2076-3417/9/5/1009 | https://www.mdpi.com/2076-3417/9/5/1009/pdf-vor | Image Shadow Removal Using End-to-End Deep Convolutional Neural Networks | Image degradation caused by shadows is likely to cause technological issues in image segmentation and target recognition. In view of the existing shadow removal methods, there are problems such as small and trivial shadow processing, the scarcity of end-to-end automatic methods, the neglecting of light, and high-level semantic information such as materials. An end-to-end deep convolutional neural network is proposed to further improve the image shadow removal effect. The network mainly consists of two network models, an encoder–decoder network and a small refinement network. The former predicts the alpha shadow scale factor, and the latter refines to obtain sharper edge information. In addition, a new image database (remove shadow database, RSDB) is constructed; and qualitative and quantitative evaluations are made on databases such as UIUC, UCF and newly-created databases (RSDB) with various real images. Using the peak signal-to-noise ratio (PSNR) and the structural similarity (SSIM) for quantitative analysis, the algorithm has a big improvement on the PSNR and the SSIM as opposed to other methods. In terms of qualitative comparisons, the network shadow has a clearer and shadow-free image that is consistent with the original image color and texture, and the detail processing effect is much better. The experimental results show that the proposed algorithm is superior to other algorithms, and it is more robust in subjective vision and objective quantization | ['Jinjiang Li', 'Meng Han', 'Hui Fan'] | 2019-03-11 | null | null | null | null | ['shadow-removal', 'image-shadow-removal'] | ['computer-vision', 'computer-vision'] | [ 5.10713577e-01 -3.82915735e-01 1.22308500e-01 -3.00694197e-01
-2.27807716e-01 7.31867924e-02 -4.71538864e-02 -4.71959203e-01
-3.83820415e-01 7.94773400e-01 1.08180523e-01 -3.30507278e-01
4.28546727e-01 -6.43155813e-01 -4.29388195e-01 -1.05969608e+00
2.74477959e-01 -1.81678116e-01 9.69966173e-01 -2.02279106e-01
3.39976430e-01 3.74685168e-01 -1.48035061e+00 1.67227566e-01
1.17384243e+00 1.50788176e+00 6.58665955e-01 5.70675910e-01
-1.31281450e-01 8.16708624e-01 -7.20155060e-01 -3.02119464e-01
2.60033727e-01 -4.38914090e-01 -1.52090654e-01 2.11477667e-01
2.12560609e-01 -7.22080827e-01 -4.71467912e-01 1.15538514e+00
9.04111981e-01 -4.54778597e-02 2.81134844e-01 -1.15781188e+00
-7.90303171e-01 -7.18112141e-02 -8.18036437e-01 3.47881578e-02
5.08538894e-02 3.55875492e-01 4.35485899e-01 -8.35038960e-01
3.91935050e-01 1.34986901e+00 9.18929160e-01 6.30974099e-02
-7.43080080e-01 -7.10086405e-01 -2.11327463e-01 6.04215086e-01
-1.27823436e+00 -2.10406095e-01 8.31522286e-01 1.62781611e-01
2.77401745e-01 3.01724225e-01 7.05191851e-01 7.24154115e-01
6.99249327e-01 9.69107330e-01 1.46016312e+00 -3.45978111e-01
1.09467544e-01 3.60128842e-02 -1.11767955e-01 9.40114439e-01
3.03124487e-01 3.59424770e-01 -2.78786391e-01 4.71484691e-01
7.41957664e-01 6.32606670e-02 -6.53901577e-01 -2.91388154e-01
-1.04153323e+00 4.52845216e-01 7.05452442e-01 1.06266290e-02
-8.26624259e-02 5.54647110e-02 2.80207396e-01 -1.41442746e-01
3.42580676e-01 -6.09147362e-02 -4.21795130e-01 2.82828718e-01
-9.00477529e-01 -1.44140989e-01 6.62326097e-01 7.63824284e-01
6.16074741e-01 3.65560561e-01 -1.90920427e-01 8.93480659e-01
3.55866075e-01 1.11991477e+00 4.36634481e-01 -1.02633941e+00
5.81522919e-02 3.62734139e-01 1.61789715e-01 -1.22988844e+00
-2.33590856e-01 -3.60857844e-01 -1.01671016e+00 5.73100507e-01
8.94669443e-02 5.07227257e-02 -1.33140743e+00 1.22924459e+00
1.68738708e-01 1.63531825e-01 1.95922274e-02 1.22412288e+00
1.28183591e+00 9.73685920e-01 -2.66415805e-01 -4.81230229e-01
1.26075602e+00 -1.21471965e+00 -1.19417286e+00 -3.65038216e-01
-1.52084425e-01 -1.10148489e+00 1.28348839e+00 5.03614008e-01
-9.57635105e-01 -7.59892941e-01 -1.34215510e+00 -1.38298467e-01
-4.75392759e-01 3.38430882e-01 7.01401114e-01 8.05724859e-01
-8.44557762e-01 4.00196284e-01 -2.35501170e-01 -9.12236795e-02
5.86640835e-01 9.94452462e-02 2.27771655e-01 -3.93668711e-01
-1.32762885e+00 7.30878472e-01 2.75225222e-01 4.32362497e-01
-8.46367717e-01 -4.71998513e-01 -6.35515511e-01 -3.83473262e-02
5.93707681e-01 -3.13285351e-01 1.07576561e+00 -1.27253819e+00
-1.75242198e+00 4.35439616e-01 -1.06949165e-01 -1.14375688e-01
5.57083070e-01 -1.06633097e-01 -6.60475791e-01 1.68839768e-01
-4.04654033e-02 4.19862330e-01 1.08692288e+00 -1.67168653e+00
-8.28439891e-01 -1.52105257e-01 -1.32242516e-01 4.19777930e-01
-1.23344772e-01 -2.84558795e-02 -9.65715349e-01 -7.09100187e-01
1.46527598e-02 -6.77972376e-01 -6.20851107e-03 3.99774253e-01
-2.97949433e-01 3.07073712e-01 1.42774498e+00 -9.96711671e-01
1.09177923e+00 -2.27153206e+00 -4.37943846e-01 -1.57858673e-02
1.25747263e-01 6.09434903e-01 -1.08139634e-01 -3.16754207e-02
9.69144851e-02 -3.16072494e-01 -5.26423454e-01 -4.66228463e-02
-3.12078595e-01 2.74962723e-01 -1.81953639e-01 5.75681210e-01
-3.34929407e-01 7.63470709e-01 -7.43399024e-01 -8.16709936e-01
4.67880040e-01 4.70103651e-01 8.78397450e-02 2.57586092e-01
-5.58147170e-02 9.55206007e-02 -2.20844641e-01 1.20285511e+00
1.13222468e+00 -1.86186749e-02 -2.93169767e-01 -7.13247716e-01
-2.80534744e-01 -3.18989009e-01 -1.25937271e+00 1.19172370e+00
-3.73702079e-01 1.16089129e+00 3.51537794e-01 -4.07306165e-01
8.92063975e-01 -1.06452592e-01 2.85407215e-01 -1.22223747e+00
4.48678821e-01 3.18719804e-01 -1.28005654e-01 -6.17881954e-01
4.81984228e-01 5.07377535e-02 2.92653441e-01 1.14838861e-01
-5.68420649e-01 -3.28106612e-01 2.60892101e-02 2.77715679e-02
6.52941942e-01 4.22416925e-02 -1.18852518e-01 -2.15633780e-01
5.72631180e-01 -1.70543998e-01 8.76142919e-01 4.20358747e-01
-3.34865451e-01 7.75189340e-01 8.65940526e-02 -2.86215812e-01
-9.93135154e-01 -1.21100497e+00 -1.94233283e-01 7.84747124e-01
1.28206718e+00 4.82467823e-02 -7.13459492e-01 -3.71054173e-01
-1.49870440e-01 6.90524280e-01 -3.77930760e-01 -2.75887728e-01
-4.29947644e-01 -8.16688716e-01 2.96594560e-01 5.29415548e-01
1.38741481e+00 -1.26908171e+00 -6.91654265e-01 -1.32754624e-01
-3.72851312e-01 -1.03624964e+00 -6.02164209e-01 5.13539575e-02
-5.38542807e-01 -1.16352224e+00 -9.42144096e-01 -8.76220822e-01
4.37709183e-01 7.22987950e-01 8.88672650e-01 2.19189793e-01
-6.17840469e-01 -1.11242019e-01 -2.97356337e-01 -7.03437030e-01
-1.24649778e-01 -5.93682110e-01 -2.81125128e-01 6.33008406e-02
-1.38458654e-01 -1.73927933e-01 -1.09031022e+00 5.32245457e-01
-1.12481964e+00 3.84791493e-01 1.10966492e+00 9.34735119e-01
4.33395177e-01 6.52086079e-01 1.38376951e-01 -7.75461912e-01
6.09941304e-01 3.39508295e-01 -7.08533287e-01 2.30416209e-01
-1.18368769e+00 -3.98472518e-01 6.58042789e-01 -2.95530021e-01
-1.73726773e+00 -3.52884847e-04 9.39146709e-03 -3.21204007e-01
-8.43285620e-02 -1.94093585e-02 -6.37941718e-01 -4.28938985e-01
3.36574703e-01 4.44688559e-01 -1.06532283e-01 -2.11914927e-01
2.95930952e-01 8.07133973e-01 7.27067530e-01 9.36929658e-02
7.55853713e-01 5.32132387e-01 -1.84171423e-01 -1.02742147e+00
-7.05168664e-01 -2.47091636e-01 -4.16124135e-01 -5.90171576e-01
8.58989000e-01 -7.58456111e-01 -7.97945499e-01 1.11666930e+00
-1.10076725e+00 -4.93762821e-01 3.45329614e-03 2.83105552e-01
-2.54364192e-01 7.24724412e-01 -7.18572915e-01 -5.93392015e-01
-5.07070482e-01 -1.23194313e+00 9.50269759e-01 7.80359805e-01
5.39546669e-01 -6.65412545e-01 -6.11861885e-01 5.00821471e-01
5.43643892e-01 1.20669387e-01 9.28143859e-01 3.32962036e-01
-8.66687655e-01 -1.59872305e-02 -9.21481848e-01 7.97845602e-01
1.98584557e-01 -7.10315034e-02 -1.00272143e+00 -6.78159371e-02
1.49021447e-01 3.04758903e-02 1.04066503e+00 6.66572511e-01
1.41847837e+00 -6.71325698e-02 -2.89922118e-01 7.89141595e-01
1.66929519e+00 6.71305120e-01 1.25750518e+00 3.09008449e-01
8.16228867e-01 3.00752312e-01 9.39494491e-01 5.00385799e-02
1.21587470e-01 4.08154100e-01 5.19790173e-01 -9.59960282e-01
-8.13476682e-01 8.93756747e-02 4.01696831e-01 6.43179476e-01
4.64773737e-02 -4.57423747e-01 -5.38329542e-01 2.56683946e-01
-1.58385623e+00 -7.14596927e-01 -4.91107374e-01 1.98642552e+00
7.87163615e-01 2.70492882e-01 -4.12536085e-01 1.81694523e-01
7.28992045e-01 3.39077741e-01 -7.89005637e-01 -2.40910366e-01
-5.41172981e-01 3.50837819e-02 1.08876085e+00 5.04677176e-01
-8.54064405e-01 1.04958582e+00 6.22206211e+00 1.22334516e+00
-1.36192358e+00 -5.27763851e-02 8.47215652e-01 4.62480903e-01
-9.30348486e-02 -8.68540853e-02 -3.24624360e-01 7.90878356e-01
3.19574654e-01 2.98670053e-01 4.97767895e-01 6.37074709e-01
3.03653687e-01 -7.01194584e-01 -4.20518845e-01 1.17446434e+00
4.12240058e-01 -1.19879794e+00 -2.84564883e-01 -1.65666595e-01
6.23876274e-01 -1.77967310e-01 2.27362722e-01 -8.41358397e-03
1.58824287e-02 -9.36604917e-01 7.34201550e-01 6.61076427e-01
9.64885414e-01 -7.70203531e-01 1.17903566e+00 3.50978412e-02
-1.30278146e+00 -1.57788649e-01 -3.87244463e-01 3.32728058e-01
1.20269589e-01 8.11889887e-01 -6.40565991e-01 5.22270381e-01
9.72268641e-01 4.15179580e-01 -6.28667533e-01 1.18697488e+00
-4.85799700e-01 4.84912425e-01 1.27069363e-02 -1.39020741e-01
3.64293307e-02 -3.97681624e-01 2.38606662e-01 1.38839960e+00
4.61664796e-03 2.59745479e-01 1.19833104e-01 6.02477908e-01
-1.78016331e-02 -5.52063510e-02 -1.01215854e-01 2.44584858e-01
2.16920316e-01 1.44695938e+00 -1.08821487e+00 -4.98512715e-01
-3.05676609e-01 1.46499026e+00 -5.79617798e-01 6.37493253e-01
-1.15281451e+00 -8.99836540e-01 1.77856445e-01 -3.84605527e-02
1.61043674e-01 -1.17952116e-01 -6.12440825e-01 -6.23021841e-01
-8.65480155e-02 -8.51496994e-01 -1.90022871e-01 -1.51521480e+00
-9.87543821e-01 4.39830422e-01 -3.90273035e-01 -1.06947553e+00
5.68939745e-01 -7.63670266e-01 -6.89210474e-01 8.68143678e-01
-1.74661982e+00 -1.03237629e+00 -8.82793844e-01 5.90805113e-01
7.54729450e-01 -1.79014858e-02 3.46041262e-01 5.91969669e-01
-5.96011102e-01 4.86379921e-01 4.06180948e-01 -1.50072929e-02
8.26985598e-01 -1.00478220e+00 -7.50381872e-02 1.10458076e+00
-5.45770824e-01 -1.53488055e-01 7.51364112e-01 -7.35649407e-01
-1.33544064e+00 -1.11353111e+00 3.40985298e-01 1.31852075e-01
2.53494233e-01 -2.10050330e-01 -8.48801911e-01 -1.37666851e-01
2.83316314e-01 -1.86335415e-01 3.92221212e-02 -5.28379858e-01
6.00572675e-02 -6.09977126e-01 -1.12090433e+00 7.53638625e-01
7.87814200e-01 -2.01258555e-01 -2.44276762e-01 2.67470568e-01
9.04076993e-01 -6.40793502e-01 -3.44079375e-01 6.40749574e-01
6.75918877e-01 -1.34010983e+00 1.01502502e+00 5.05505800e-01
4.84922975e-01 -5.94719470e-01 -1.52029276e-01 -1.16249442e+00
-2.39466518e-01 -2.24425659e-01 1.88553154e-01 1.34217501e+00
1.60911366e-01 -6.32497668e-01 8.65269244e-01 2.27587253e-01
-4.33588564e-01 -1.00536442e+00 -5.78422487e-01 -7.81028390e-01
-6.65517628e-01 -2.11583078e-01 4.44142073e-01 5.25964320e-01
-8.26144040e-01 4.35648054e-01 -6.40742898e-01 2.95462906e-01
7.70360410e-01 5.91831565e-01 6.68020070e-01 -9.60637212e-01
-2.35236902e-02 -2.20087767e-01 -4.91096377e-02 -1.22391629e+00
-2.11063653e-01 -1.22297883e-01 5.68459392e-01 -1.99055040e+00
2.41093084e-01 -3.88063073e-01 -2.27214813e-01 2.52872586e-01
-2.33205676e-01 4.37702924e-01 -1.75166619e-03 1.84139907e-01
-5.60811281e-01 8.64048302e-01 1.83403420e+00 -3.29155594e-01
-1.06290519e-01 1.03699550e-01 -4.77900505e-01 8.72949898e-01
6.37553394e-01 -1.10528700e-01 -2.84706324e-01 -3.63073140e-01
-3.64146769e-01 -2.31555593e-03 3.22151989e-01 -9.28858101e-01
1.94828406e-01 -2.42527515e-01 9.92782891e-01 -9.39043820e-01
6.35123193e-01 -9.05467153e-01 -1.78834990e-01 8.16646278e-01
1.38018399e-01 -4.48444098e-01 3.56756002e-02 6.20693445e-01
-1.66619569e-01 -6.56229556e-02 1.21436608e+00 1.40734380e-02
-1.14163768e+00 2.27236643e-01 -3.26286972e-01 -3.10285896e-01
9.48351681e-01 -6.67859733e-01 -4.36783493e-01 -5.96806407e-01
1.62891336e-02 9.07567218e-02 4.91056442e-01 2.69873351e-01
1.08738637e+00 -1.20893812e+00 -4.75194573e-01 3.00146222e-01
-1.03750013e-01 4.35754337e-04 2.97425330e-01 9.03299272e-01
-1.03371322e+00 1.80382028e-01 -3.63130569e-01 -5.00598371e-01
-1.46450567e+00 3.18207234e-01 3.28244418e-01 1.90720469e-01
-6.82235420e-01 8.01333547e-01 6.32259607e-01 -4.56110016e-02
5.03736675e-01 -3.95803958e-01 1.78003967e-01 -4.05638337e-01
4.26457018e-01 5.22853076e-01 -1.31925698e-02 -5.58980405e-01
-1.00932531e-01 4.29248452e-01 1.43897951e-01 1.77765951e-01
1.19286871e+00 -3.85475338e-01 -2.05384940e-01 1.79615185e-01
9.61834192e-01 1.33930445e-01 -1.49341691e+00 -2.12748364e-01
-5.53516984e-01 -9.09462333e-01 3.24400157e-01 -1.15889275e+00
-1.50072980e+00 9.40452218e-01 1.36039555e+00 9.26591083e-02
1.68545496e+00 -3.99638563e-01 1.47096896e+00 9.29311290e-02
7.10272789e-02 -1.47570062e+00 3.17115337e-01 4.49473500e-01
8.47169995e-01 -1.23563516e+00 2.74453670e-01 -6.68838501e-01
-6.69321001e-01 1.17621028e+00 6.74024343e-01 1.20562315e-01
3.50011289e-01 5.02814651e-01 2.61805445e-01 8.00474659e-02
-7.58979991e-02 -2.11024925e-01 2.68126667e-01 7.82089353e-01
-2.38792337e-02 -5.10314927e-02 -2.93655455e-01 4.93445754e-01
-1.64646506e-02 -8.74874219e-02 4.80618685e-01 5.73229373e-01
-8.25911820e-01 -5.88411808e-01 -6.65699065e-01 3.20007682e-01
-2.54110187e-01 -3.23252320e-01 -4.25780565e-01 7.75610685e-01
5.29231608e-01 1.09688771e+00 -2.23546878e-01 -4.99567986e-01
2.08186790e-01 -6.92155421e-01 3.28110069e-01 2.36097304e-03
5.65909594e-02 3.05719137e-01 3.22538354e-02 -6.80494070e-01
-2.31929183e-01 -1.43979073e-01 -1.50429559e+00 -4.73115474e-01
-5.83135724e-01 -2.38430813e-01 8.88740242e-01 8.42982650e-01
-8.55721086e-02 9.50873077e-01 8.22211623e-01 -8.72288823e-01
2.21125990e-01 -7.09491074e-01 -9.06638741e-01 3.04967880e-01
2.65471399e-01 -6.26043200e-01 -2.61849344e-01 2.38024995e-01] | [10.837363243103027, -3.9728007316589355] |
5147a27e-2be6-4bc7-b00a-5197d6990f69 | single-image-deraining-via-scale-space | 2006.05049 | null | https://arxiv.org/abs/2006.05049v2 | https://arxiv.org/pdf/2006.05049v2.pdf | Single Image Deraining via Scale-space Invariant Attention Neural Network | Image enhancement from degradation of rainy artifacts plays a critical role in outdoor visual computing systems. In this paper, we tackle the notion of scale that deals with visual changes in appearance of rain steaks with respect to the camera. Specifically, we revisit multi-scale representation by scale-space theory, and propose to represent the multi-scale correlation in convolutional feature domain, which is more compact and robust than that in pixel domain. Moreover, to improve the modeling ability of the network, we do not treat the extracted multi-scale features equally, but design a novel scale-space invariant attention mechanism to help the network focus on parts of the features. In this way, we summarize the most activated presence of feature maps as the salient features. Extensive experiments results on synthetic and real rainy scenes demonstrate the superior performance of our scheme over the state-of-the-arts. | ['Xian-Ming Liu', 'Deming Zhai', 'Junjun Jiang', 'Bo Pang'] | 2020-06-09 | null | null | null | null | ['single-image-deraining'] | ['computer-vision'] | [ 2.25237355e-01 -6.00173712e-01 4.60735440e-01 -5.48690557e-01
-2.42043108e-01 -2.72611767e-01 4.02281396e-02 -2.80738622e-01
-4.22846168e-01 7.77666330e-01 4.72470783e-02 1.42540321e-01
-5.15388208e-04 -7.38300025e-01 -6.46455824e-01 -9.30665433e-01
-1.24238923e-01 -7.66009033e-01 3.91824335e-01 -5.45734406e-01
1.89417735e-01 6.52365088e-01 -1.38916028e+00 3.28615487e-01
1.06550395e+00 9.42148864e-01 4.31300461e-01 7.47480333e-01
2.00361744e-01 8.25883746e-01 -5.75252295e-01 -7.39177223e-03
3.05297285e-01 -4.26026940e-01 -1.61973059e-01 4.13936257e-01
9.14251208e-01 -4.21389252e-01 -4.88958061e-01 1.51695609e+00
4.33223397e-01 3.10895070e-02 3.42574030e-01 -9.17004406e-01
-8.43240857e-01 -1.64955780e-01 -8.80611122e-01 1.00958955e+00
-1.44695908e-01 -3.71402316e-02 8.19329798e-01 -1.08589685e+00
3.57843071e-01 1.36276913e+00 6.32492721e-01 1.41765982e-01
-7.82921076e-01 -5.20291567e-01 4.53620881e-01 3.44212502e-01
-1.18721211e+00 -3.27269047e-01 9.23315763e-01 -8.07672814e-02
4.82926220e-01 3.89025152e-01 8.83793652e-01 6.42580926e-01
5.28691053e-01 7.63751268e-01 1.62386513e+00 -3.24199200e-01
1.11742333e-01 -1.36584982e-01 2.49431461e-01 8.11256349e-01
5.49779236e-01 1.00750074e-01 -3.15021038e-01 1.95815459e-01
1.08616006e+00 3.60953391e-01 -6.63734674e-01 -3.36120576e-01
-9.62817252e-01 7.51302361e-01 7.61247337e-01 4.28626984e-01
-4.20531243e-01 3.39357525e-01 -1.55913115e-01 5.46882510e-01
7.47655988e-01 2.00036585e-01 -5.31755447e-01 3.95460308e-01
-1.12542307e+00 8.61546993e-02 2.27815107e-01 5.36946893e-01
8.31689715e-01 2.56749094e-01 -1.48340985e-01 7.19971418e-01
2.78033227e-01 9.44548547e-01 2.90602267e-01 -7.39894867e-01
3.65019977e-01 3.45295727e-01 2.14694604e-01 -1.14111674e+00
-4.98437315e-01 -7.24303246e-01 -1.11026311e+00 5.41130900e-01
1.64099827e-01 -6.59687892e-02 -1.12695312e+00 1.44457960e+00
3.34944636e-01 4.22536165e-01 1.57899335e-01 1.24463820e+00
6.63965285e-01 7.09647834e-01 -2.96894193e-01 -5.41485071e-01
1.40067923e+00 -1.04605997e+00 -9.38730717e-01 -4.97228682e-01
-3.32512587e-01 -8.23035955e-01 8.72438431e-01 2.27622777e-01
-1.10460448e+00 -9.87317562e-01 -1.17864513e+00 7.33929034e-03
-3.69107902e-01 2.55505830e-01 5.87319911e-01 4.47426140e-01
-1.08885121e+00 7.91395485e-01 -6.45630956e-01 -4.26057935e-01
3.82112592e-01 -1.33951068e-01 -1.48500249e-01 -2.92201251e-01
-1.25188601e+00 8.56221318e-01 1.58052713e-01 6.01751149e-01
-7.23865569e-01 -3.67416024e-01 -6.88241243e-01 1.63081549e-02
-1.13918064e-02 -4.96578753e-01 7.55909383e-01 -1.37281430e+00
-1.21591485e+00 4.27635193e-01 -3.89520943e-01 -2.06190810e-01
2.83630162e-01 -5.31520545e-01 -7.96353817e-01 5.77511430e-01
-1.02028213e-01 2.19471604e-01 1.59131444e+00 -1.37320185e+00
-9.29754198e-01 -1.86395124e-01 2.93481916e-01 3.29186350e-01
-4.10259753e-01 1.60169184e-01 -4.19183493e-01 -8.56646419e-01
3.40733707e-01 -6.83831692e-01 -3.96540761e-01 3.88569713e-01
1.86431289e-01 3.72568071e-01 9.41839159e-01 -9.34675932e-01
1.18065238e+00 -2.36037731e+00 5.83933778e-02 -7.20714591e-03
3.97619396e-01 4.51458246e-01 -2.32537478e-01 9.50278528e-03
-4.84779067e-02 -1.54386446e-01 -5.05877912e-01 -4.12892215e-02
-4.91794646e-01 2.88640678e-01 -3.05286050e-01 7.55389631e-01
5.88202953e-01 7.61585057e-01 -7.04869390e-01 -5.09293795e-01
2.42458820e-01 6.43147469e-01 -2.65688568e-01 2.17383817e-01
3.71020645e-01 3.34963322e-01 -6.17230892e-01 8.72603059e-01
1.22856247e+00 -2.60840178e-01 -2.73983598e-01 -5.79835355e-01
-3.44286472e-01 -4.18345004e-01 -1.24193645e+00 1.29761052e+00
-4.78692770e-01 6.45135283e-01 2.11033061e-01 -5.75216532e-01
8.29299986e-01 -7.66678751e-02 5.22254780e-02 -6.53735101e-01
-4.42929901e-02 5.09219319e-02 -2.65291017e-02 -7.55169988e-01
4.96283680e-01 -1.32436007e-01 3.43400657e-01 -2.49597669e-01
-1.82355076e-01 -9.36554372e-02 2.58935988e-02 -6.37130216e-02
8.83995771e-01 -5.54162152e-02 5.60583711e-01 -1.52192175e-01
6.64975286e-01 -5.76805651e-01 6.14089251e-01 6.21806920e-01
-5.50241888e-01 7.95422196e-01 -6.76520076e-03 -5.11899769e-01
-9.92893040e-01 -1.07455432e+00 -1.65109023e-01 9.84816968e-01
6.17967904e-01 1.44060820e-01 -4.17064130e-01 -6.24712586e-01
3.16782109e-03 1.36057541e-01 -9.40972090e-01 -1.43015292e-02
-8.36170018e-01 -1.08776379e+00 2.07509339e-01 5.10590136e-01
1.00823307e+00 -1.09067416e+00 -8.30143452e-01 2.64615148e-01
-2.44986296e-01 -1.13736415e+00 -5.23974299e-01 2.99634576e-01
-9.38111663e-01 -8.81983459e-01 -9.20487225e-01 -9.53805625e-01
6.92789257e-01 9.40467715e-01 9.90703344e-01 -1.76583312e-03
-4.00351018e-01 1.54441908e-01 -5.70217311e-01 -2.20653728e-01
1.82777792e-01 -6.06797636e-01 -2.65182525e-01 2.66491652e-01
-4.31746133e-02 -8.00737023e-01 -9.33889627e-01 4.95972931e-02
-1.06489921e+00 -1.13241382e-01 1.13464940e+00 9.03464019e-01
4.69966441e-01 2.35304520e-01 3.68554771e-01 -5.61159790e-01
4.78484750e-01 -1.34066477e-01 -4.93519574e-01 3.76460910e-01
-3.71024281e-01 -2.30620708e-02 5.79533279e-01 -3.33200574e-01
-1.10947561e+00 5.30569926e-02 2.21282572e-01 -5.02532125e-01
-7.69159347e-02 2.51343876e-01 -1.69495016e-01 -5.99117577e-01
3.14165294e-01 5.47918260e-01 -3.30198944e-01 -5.18953681e-01
5.52101672e-01 4.30649161e-01 4.05323923e-01 -1.20757744e-01
1.26246607e+00 9.38436627e-01 7.68871382e-02 -9.31729794e-01
-9.11715925e-01 -5.95804393e-01 -6.70116603e-01 -2.66880065e-01
8.49099278e-01 -1.13186848e+00 -4.19422746e-01 5.90288937e-01
-1.07612097e+00 3.54055017e-02 -1.99504048e-01 3.44997823e-01
-2.77892590e-01 6.24644458e-01 -7.11945415e-01 -8.68394256e-01
-4.48910028e-01 -8.36210251e-01 1.09413815e+00 7.56990433e-01
6.66479051e-01 -8.02200258e-01 1.25993846e-03 -9.63337868e-02
7.78633952e-01 3.30759197e-01 5.76089919e-01 2.98318323e-02
-5.49419403e-01 1.49662465e-01 -7.44034171e-01 5.83528996e-01
1.95913106e-01 -1.27438474e-02 -1.03041506e+00 -6.38120055e-01
3.10534835e-01 -2.79010516e-02 1.26602221e+00 4.93652552e-01
8.62279773e-01 -1.16130851e-01 -3.82469781e-02 9.41953838e-01
1.83701169e+00 4.60482296e-03 6.67048395e-01 3.82254571e-01
5.99662662e-01 3.36778253e-01 8.52009594e-01 4.95528877e-01
2.34621689e-01 3.35574090e-01 7.41010249e-01 -7.52501130e-01
-4.20099646e-01 1.93161681e-01 4.21180606e-01 7.12802351e-01
-1.79111809e-01 -6.86306953e-02 -1.92244679e-01 6.11448765e-01
-1.65593529e+00 -9.93682265e-01 -7.30722696e-02 1.79786110e+00
7.30160892e-01 -6.71529621e-02 -4.01888400e-01 -2.06826076e-01
7.79088795e-01 7.01507628e-01 -4.67248261e-01 -9.38219503e-02
-7.50004292e-01 3.35554361e-01 7.16254354e-01 3.54365319e-01
-1.49047732e+00 9.34350312e-01 6.07846022e+00 6.10188484e-01
-1.24114549e+00 1.28727630e-01 4.71037269e-01 7.76323304e-02
1.27163231e-01 -2.62736008e-02 -5.92164099e-01 4.51842397e-01
4.10523266e-01 5.43570071e-02 5.49810350e-01 5.90801716e-01
4.35380876e-01 -2.28262320e-02 -5.27447402e-01 8.48510563e-01
2.80281425e-01 -7.80852735e-01 -1.20141640e-01 -2.26753145e-01
8.84414911e-01 2.45866459e-02 2.56321490e-01 9.02541429e-02
7.50046968e-02 -7.44311810e-01 7.41251826e-01 8.99497569e-01
5.94343543e-01 -5.50876439e-01 6.13618672e-01 -1.75751537e-01
-1.56182313e+00 -3.00305635e-01 -8.19378197e-01 -1.88537017e-02
-3.88563834e-02 7.92025387e-01 -1.91238418e-01 6.58602178e-01
1.08293605e+00 9.76862669e-01 -9.97662544e-01 1.32841551e+00
-4.40775126e-01 5.43812573e-01 -1.44093586e-02 2.35169232e-01
1.89920992e-01 -1.72333196e-01 4.98381585e-01 1.38704741e+00
4.17881548e-01 3.04260015e-01 4.24890071e-02 4.20218438e-01
1.31039038e-01 -6.63715228e-02 -4.07850355e-01 2.84873873e-01
1.29182681e-01 1.50541055e+00 -5.35627484e-01 -3.92316312e-01
-6.25879586e-01 1.46537232e+00 1.92660555e-01 6.88164949e-01
-9.23038125e-01 -6.27027512e-01 8.11904848e-01 -2.14630142e-01
9.11271930e-01 -3.58517855e-01 -2.38128914e-03 -1.31338406e+00
3.08020473e-01 -8.72099519e-01 1.93994101e-02 -1.03044271e+00
-1.35635877e+00 6.75053120e-01 -4.65600818e-01 -1.53581190e+00
5.92462838e-01 -4.90112126e-01 -7.27414608e-01 8.79335284e-01
-2.45009756e+00 -1.38477039e+00 -6.88193262e-01 6.81044161e-01
6.57140255e-01 2.62989223e-01 3.13732505e-01 3.51926625e-01
-3.42423618e-01 2.81171560e-01 6.34484947e-01 2.73981206e-02
9.24481988e-01 -1.23361051e+00 4.18458104e-01 1.41614890e+00
-5.73532954e-02 6.48303509e-01 8.36470604e-01 -6.01699293e-01
-1.13965976e+00 -1.36779201e+00 5.91796815e-01 1.42964825e-01
6.75141931e-01 -1.95890083e-03 -1.04868507e+00 2.79785037e-01
2.42904320e-01 6.47961557e-01 1.21276498e-01 -3.04220051e-01
-3.64837229e-01 -4.36277837e-01 -1.02674949e+00 3.41610461e-01
8.48346591e-01 -3.75098318e-01 -6.59552217e-01 2.19895527e-01
8.18654358e-01 -1.97281703e-01 -6.23116136e-01 5.10748088e-01
5.17132521e-01 -1.04782617e+00 1.14568365e+00 -4.91083384e-01
3.85610372e-01 -7.40970135e-01 -5.01372576e-01 -1.42078292e+00
-7.74723053e-01 -2.72727519e-01 -1.72483206e-01 1.09448850e+00
-6.96868002e-02 -5.38660347e-01 4.14723724e-01 -2.76171505e-01
-1.52855396e-01 -5.44459820e-01 -1.02190089e+00 -5.68074524e-01
-1.82043001e-01 1.96765512e-01 3.60231638e-01 8.66724908e-01
-6.40107930e-01 3.46839279e-02 -7.88888216e-01 9.50868309e-01
7.89900243e-01 6.19011521e-01 3.36079627e-01 -1.17081428e+00
-3.22842300e-01 -1.00949392e-01 -4.49561417e-01 -1.01920283e+00
-2.89081931e-01 -3.43332201e-01 1.98088303e-01 -1.63737428e+00
2.76375771e-01 9.33044329e-02 -8.48234534e-01 3.30545247e-01
-6.95486963e-01 5.59013903e-01 4.34374869e-01 3.05648237e-01
-9.30254400e-01 7.94136763e-01 1.54812396e+00 -1.49991006e-01
2.51393076e-02 -5.93127077e-03 -6.34169579e-01 9.23800886e-01
7.46341109e-01 -2.91280538e-01 -8.60292241e-02 -4.37685251e-01
-2.92066094e-02 1.74008664e-02 4.64358062e-01 -1.30765438e+00
-9.99183860e-03 -3.91651392e-01 8.67116630e-01 -4.69841301e-01
4.90028769e-01 -1.00539303e+00 -3.24428171e-01 5.67125499e-01
-8.14812779e-02 1.06156603e-01 5.45886196e-02 8.56184542e-01
-4.66786027e-01 7.17150494e-02 1.25103176e+00 -4.39548381e-02
-9.56048548e-01 1.83186248e-01 -2.28546470e-01 -2.80853629e-01
7.79482543e-01 -9.21843573e-02 -5.89060187e-01 -1.97145879e-01
-7.16622174e-01 4.76403162e-02 3.88565958e-01 3.45469445e-01
9.38257098e-01 -1.12130344e+00 -7.68126249e-01 3.04807037e-01
1.28191769e-01 -4.89198923e-01 5.98237693e-01 8.87111068e-01
-6.44625306e-01 5.96114546e-02 -5.04133344e-01 -2.03839928e-01
-1.42378068e+00 6.20057285e-01 4.93804544e-01 -2.16341868e-01
-5.90710640e-01 7.77660966e-01 5.90949059e-01 2.47240528e-01
-1.54516399e-01 -6.75911188e-01 -5.07225156e-01 5.83978044e-03
7.54102528e-01 2.20722497e-01 -1.51278377e-01 -6.49438620e-01
-3.59372884e-01 9.59960520e-01 -1.32461339e-01 1.69188887e-01
1.55658472e+00 -5.96382499e-01 -2.30308130e-01 4.87362176e-01
1.09567153e+00 1.74707785e-01 -1.70038736e+00 -5.74016333e-01
-4.84507471e-01 -8.13380718e-01 1.67438880e-01 -8.18397641e-01
-1.39223015e+00 8.61581385e-01 1.26446271e+00 6.89127669e-02
1.61854577e+00 -4.70383465e-01 6.70146525e-01 3.97649288e-01
1.62199408e-01 -8.89951348e-01 2.52581030e-01 3.59993249e-01
9.28072453e-01 -1.41015780e+00 2.98061609e-01 -3.72362524e-01
-5.31279564e-01 1.28527999e+00 6.35146379e-01 -4.11731094e-01
6.91053867e-01 -8.49523470e-02 2.47934401e-01 -1.88282296e-01
-3.51274401e-01 -5.66977382e-01 2.28550404e-01 5.26692152e-01
2.62416303e-01 -8.90485048e-02 -5.06603241e-01 4.17315632e-01
3.71070445e-01 -5.32219522e-02 6.56071723e-01 9.52971041e-01
-1.12192392e+00 -6.17379725e-01 -6.99906945e-01 1.46500781e-01
-4.51801181e-01 -5.71460009e-01 4.51961718e-02 6.72062337e-01
3.86509210e-01 1.17527127e+00 -2.43767470e-01 -2.17909753e-01
3.54856670e-01 -3.59830558e-01 3.23828667e-01 -1.05334416e-01
-3.46911401e-01 3.39842618e-01 -3.52078587e-01 -5.08850634e-01
-6.90547466e-01 -5.27891994e-01 -8.71114969e-01 2.02817708e-01
-3.01914096e-01 3.38342898e-02 5.60875535e-01 5.72982550e-01
8.50190595e-02 8.81092310e-01 1.14106154e+00 -8.46009493e-01
-3.39173794e-01 -9.68365252e-01 -1.07879794e+00 3.80609989e-01
9.17731524e-01 -4.76227999e-01 -5.51422715e-01 8.55601281e-02] | [10.907129287719727, -3.183201551437378] |
d8727a9b-9e14-48cf-82de-1fd25264255d | toward-a-task-of-feedback-comment-generation | null | null | https://aclanthology.org/D19-1316 | https://aclanthology.org/D19-1316.pdf | Toward a Task of Feedback Comment Generation for Writing Learning | In this paper, we introduce a novel task called feedback comment generation {---} a task of automatically generating feedback comments such as a hint or an explanatory note for writing learning for non-native learners of English. There has been almost no work on this task nor corpus annotated with feedback comments. We have taken the first step by creating learner corpora consisting of approximately 1,900 essays where all preposition errors are manually annotated with feedback comments. We have tested three baseline methods on the dataset, showing that a simple neural retrieval-based method sets a baseline performance with an F-measure of 0.34 to 0.41. Finally, we have looked into the results to explore what modifications we need to make to achieve better performance. We also have explored problems unaddressed in this work | ['Ryo Nagata'] | 2019-11-01 | null | null | null | ijcnlp-2019-11 | ['comment-generation'] | ['natural-language-processing'] | [ 3.34224761e-01 6.45072699e-01 1.10932207e-02 -5.67592084e-01
-1.32300639e+00 -5.37951708e-01 7.48331130e-01 4.01592851e-01
-6.02947474e-01 1.26285911e+00 6.19940698e-01 -8.49835515e-01
3.05109173e-01 -4.20407504e-01 -6.70184612e-01 -2.04409435e-01
3.51869017e-01 2.70299911e-01 3.22930992e-01 -4.63259876e-01
9.22529697e-01 -1.57403290e-01 -1.41325045e+00 7.40934193e-01
1.13643479e+00 3.19842696e-01 2.21648127e-01 1.03096354e+00
-2.43058011e-01 8.06320190e-01 -1.20948935e+00 -8.49273205e-01
-1.27863169e-01 -7.93020785e-01 -1.20457530e+00 -1.28858939e-01
6.31416261e-01 -3.94907266e-01 2.26108193e-01 7.45717287e-01
5.23526311e-01 2.29448870e-01 1.06113255e+00 -9.17315006e-01
-1.20347559e+00 9.95675623e-01 -3.28397572e-01 2.70447910e-01
8.26408267e-01 -4.18264009e-02 1.05919755e+00 -1.33074915e+00
8.04760277e-01 8.94912124e-01 5.41726351e-01 1.15561342e+00
-9.44784820e-01 -3.62562388e-01 -4.08954844e-02 2.16703787e-01
-6.38815105e-01 -3.08721483e-01 6.34284675e-01 -4.13129449e-01
1.02831078e+00 3.44335467e-01 3.71820569e-01 1.09587049e+00
1.85665965e-01 8.95565271e-01 1.05789888e+00 -1.08144915e+00
-1.12584107e-01 7.11708307e-01 2.29070529e-01 5.98144710e-01
1.57568946e-01 -2.00297385e-01 -7.39026487e-01 -3.72770354e-02
2.02273637e-01 -4.57123607e-01 -4.24915493e-01 2.64424622e-01
-1.07528877e+00 1.06305623e+00 2.73531526e-01 2.40714818e-01
-5.57191148e-02 -6.34575859e-02 4.66488183e-01 6.23903096e-01
6.20297253e-01 9.48891401e-01 -5.63141525e-01 -2.24430650e-01
-8.54346991e-01 5.48571467e-01 1.12365425e+00 9.03712094e-01
5.50682664e-01 -1.86348651e-02 -3.71325493e-01 9.48080122e-01
3.46697181e-01 9.55947936e-02 7.53721058e-01 -8.28025281e-01
8.51789474e-01 4.96140569e-01 5.09178102e-01 -6.84942901e-01
-1.24402903e-01 -8.91220942e-02 -2.70862222e-01 1.04579590e-01
5.14584780e-01 -7.31773913e-01 -5.62198400e-01 1.39777637e+00
-1.82909474e-01 -3.91047329e-01 3.35331231e-01 5.70765257e-01
1.22603440e+00 8.43998313e-01 -2.58915816e-02 -3.24078381e-01
9.92368400e-01 -1.39184368e+00 -7.81805634e-01 2.92110536e-02
9.65899646e-01 -1.19817019e+00 1.47468555e+00 4.59013224e-01
-1.31176019e+00 -4.97597039e-01 -1.16507494e+00 -1.14903919e-01
-2.85653800e-01 3.02576244e-01 4.34696823e-01 7.16531813e-01
-1.33601177e+00 5.50155163e-01 -3.40687215e-01 -2.91229099e-01
-1.90156445e-01 3.58174443e-02 -2.02200249e-01 1.18990466e-01
-1.10656476e+00 9.65608418e-01 8.59013200e-02 -2.90582091e-01
-3.90418440e-01 -7.52435267e-01 -7.18080223e-01 -1.31923795e-01
1.10622808e-01 -5.52274346e-01 2.12841749e+00 -9.63377893e-01
-1.58121681e+00 7.89386630e-01 -3.14283341e-01 -4.80809659e-01
4.74029928e-01 -4.92359996e-01 -3.02336901e-01 4.44829911e-02
2.60098010e-01 9.44919765e-01 3.49150956e-01 -1.13135290e+00
-6.59206927e-01 -4.22712788e-02 2.10046187e-01 4.81169701e-01
-4.61777389e-01 4.91266012e-01 2.01928779e-01 -8.30440044e-01
-3.26784492e-01 -8.43579769e-01 -1.84375674e-01 -4.69077349e-01
-1.47238791e-01 -7.56124079e-01 3.86896640e-01 -7.41741955e-01
1.42973983e+00 -1.56156361e+00 -2.22293526e-01 -2.42980808e-01
-1.96130246e-01 2.70830661e-01 -2.11236462e-01 8.68519306e-01
-4.98018675e-02 6.05058074e-01 -1.69196755e-01 -2.59198159e-01
2.91440822e-02 -2.86453754e-01 -5.71121156e-01 -3.05629224e-01
4.36428100e-01 8.41494739e-01 -1.13910103e+00 -1.78687721e-01
-4.73358661e-01 1.54462293e-01 -6.65401697e-01 5.30703366e-01
-4.93117899e-01 1.00703754e-01 -2.62557119e-01 3.56843144e-01
-2.14709993e-03 -9.65465009e-02 -3.00906628e-01 7.98239589e-01
-1.87825069e-01 1.04226065e+00 -8.28818381e-01 1.59558845e+00
-5.19451618e-01 9.44372535e-01 -3.27250630e-01 -4.77394730e-01
1.27800059e+00 5.58628976e-01 -1.70104820e-02 -5.11666596e-01
-6.08149879e-02 5.91616035e-01 2.16190308e-01 -6.41760886e-01
1.20241952e+00 -9.24578980e-02 1.73162878e-01 1.13632035e+00
8.65320340e-02 -3.42933416e-01 4.97577548e-01 4.42757398e-01
1.12320852e+00 7.44255781e-02 1.98845312e-01 -2.60709912e-01
5.22000909e-01 2.86871821e-01 -7.38853887e-02 8.78287017e-01
6.35655522e-02 8.60047162e-01 3.72292966e-01 -1.96209729e-01
-9.75540340e-01 -6.12525582e-01 1.54374093e-01 1.26354468e+00
-6.02589011e-01 -9.42924917e-01 -8.69654894e-01 -9.95542169e-01
-3.23044747e-01 1.05881643e+00 -6.37196064e-01 7.69755691e-02
-5.79667866e-01 -3.68612677e-01 4.24951434e-01 6.42085493e-01
1.07630014e-01 -1.50946462e+00 -5.83676994e-01 3.06721747e-01
-3.11883569e-01 -3.85487109e-01 -5.92753410e-01 4.15487647e-01
-8.51714313e-01 -6.30492985e-01 -9.64464247e-01 -9.00966406e-01
8.07694316e-01 8.14826414e-02 1.38862145e+00 5.41047156e-01
2.15396389e-01 3.39562804e-01 -8.38604689e-01 -1.12820256e+00
-7.06640661e-01 3.35808218e-01 -3.30219239e-01 -7.99511254e-01
5.51085413e-01 -1.09935403e-02 -2.37744823e-01 -1.08570516e-01
-6.99971199e-01 3.67587507e-02 5.31702101e-01 1.04177189e+00
2.95806527e-02 -6.59985960e-01 1.01826322e+00 -1.47348726e+00
1.47864878e+00 -3.39107960e-01 1.91916451e-02 2.20650166e-01
-7.38059878e-01 2.14644343e-01 6.24689817e-01 -3.45772356e-01
-1.22368693e+00 -1.78323194e-01 -4.64959949e-01 4.26711857e-01
-9.42790732e-02 7.69692004e-01 2.29734764e-01 1.64252520e-01
1.26957393e+00 -1.38779968e-01 -1.73961893e-01 -2.23693937e-01
2.86837369e-01 9.61511254e-01 4.40581322e-01 -8.02629769e-01
4.31584388e-01 -4.11978006e-01 -6.27392232e-01 -5.08877516e-01
-1.17672062e+00 -4.96275723e-01 -6.47464156e-01 -3.16619724e-01
3.73814732e-01 -7.29741991e-01 -1.21373788e-01 1.33496702e-01
-1.32679939e+00 -5.99479377e-01 -3.96418303e-01 2.81273633e-01
-5.62185526e-01 1.45274237e-01 -7.55510628e-01 -6.92456245e-01
-5.28981745e-01 -9.26420391e-01 5.86525023e-01 6.03234351e-01
-7.96181262e-01 -1.07229817e+00 3.67250770e-01 4.12276924e-01
4.37900662e-01 -1.92402303e-01 7.09376454e-01 -9.95474517e-01
-2.25965723e-01 1.60342485e-01 3.64720225e-02 3.83561134e-01
-2.90879793e-02 2.22467408e-01 -9.20831800e-01 5.98585745e-03
-1.33232325e-01 -7.95022845e-01 8.77598763e-01 -3.00067365e-02
1.00412846e+00 -7.65056849e-01 1.76364362e-01 -1.80995986e-02
1.00060689e+00 2.91794360e-01 5.20385146e-01 4.86285567e-01
2.69530565e-01 7.78319240e-01 8.39987218e-01 1.83003843e-01
4.69871849e-01 2.73588538e-01 5.69171086e-02 3.22991937e-01
-1.61266401e-01 -5.84328592e-01 6.22864246e-01 1.12211585e+00
5.50907571e-03 -4.21183854e-01 -9.57942784e-01 8.18319857e-01
-1.84900522e+00 -9.45303619e-01 -5.27106225e-01 1.88394189e+00
1.34469593e+00 3.48428905e-01 8.80448818e-02 1.53954118e-01
3.03969115e-01 -8.34057257e-02 -7.27462545e-02 -1.14884996e+00
3.63814414e-01 5.48246562e-01 1.04772814e-01 8.25086296e-01
-6.77400351e-01 7.75606453e-01 6.57045794e+00 1.29444212e-01
-1.00375772e+00 -1.25365958e-01 4.29955721e-01 3.71638775e-01
-8.24363768e-01 -1.01943640e-02 -9.78206277e-01 1.80172518e-01
1.33338404e+00 -4.16050404e-01 -5.74400425e-02 8.24408293e-01
5.19703291e-02 -1.42745003e-01 -1.07816446e+00 4.05985802e-01
5.37965298e-01 -1.28584862e+00 8.71780366e-02 -3.12475681e-01
1.06185794e+00 -2.12039068e-01 -6.35704026e-02 6.16428614e-01
5.79444349e-01 -1.05371809e+00 7.14083195e-01 3.40720117e-01
6.25840843e-01 -6.95855200e-01 8.76224577e-01 5.43334305e-01
-4.36121494e-01 8.47503468e-02 -5.58047652e-01 -7.46251702e-01
-1.20016538e-01 1.98810220e-01 -1.59322035e+00 3.53323549e-01
4.26908940e-01 6.24565661e-01 -8.99876595e-01 1.27056563e+00
-9.78121817e-01 8.81740093e-01 1.84777066e-01 -8.02794456e-01
2.70928979e-01 1.44000322e-01 1.27191409e-01 1.44072437e+00
3.89490515e-01 1.77221566e-01 1.11514479e-01 5.50327003e-01
-4.03345138e-01 4.97609437e-01 -6.68336391e-01 8.48051533e-02
4.36805397e-01 1.18234515e+00 -6.03265643e-01 -2.97257811e-01
-4.20655996e-01 8.34874392e-01 4.63349015e-01 3.32748502e-01
-2.87483603e-01 -7.28538871e-01 3.24939266e-02 2.56757408e-01
-6.07815161e-02 7.94919059e-02 -5.78963757e-01 -1.02791679e+00
8.26729760e-02 -9.95542824e-01 2.08982289e-01 -1.09421289e+00
-1.29105747e+00 4.66010332e-01 -2.20011368e-01 -9.12197530e-01
-8.49254370e-01 -6.53260887e-01 -1.07874167e+00 1.14589560e+00
-1.24630165e+00 -5.87370276e-01 6.88508805e-03 1.25745654e-01
1.04652035e+00 -2.08996341e-01 1.04691863e+00 -3.04896440e-02
-1.61812395e-01 7.47126341e-01 -7.92461857e-02 1.22014754e-01
1.36113966e+00 -1.82746768e+00 3.41860980e-01 9.89932895e-01
2.85075068e-01 8.99867654e-01 9.08350050e-01 -6.06385350e-01
-6.13185227e-01 -7.22929060e-01 1.69498634e+00 -7.42235243e-01
7.14123487e-01 -5.66922352e-02 -1.04632533e+00 5.96393347e-01
9.36510444e-01 -6.20738745e-01 9.58687723e-01 3.73595864e-01
-2.25975364e-01 3.32067460e-01 -7.96343863e-01 6.97488904e-01
5.96834064e-01 -4.11698610e-01 -1.16893864e+00 4.95965332e-01
9.99141514e-01 -5.42978585e-01 -5.82205832e-01 1.81471869e-01
2.94382095e-01 -8.40926647e-01 5.10888934e-01 -8.18330944e-01
1.16600835e+00 -2.53461450e-02 3.30824345e-01 -1.85200095e+00
1.36625841e-02 -6.90186799e-01 8.83947089e-02 1.54963934e+00
9.80511129e-01 -1.09473847e-01 9.25675929e-01 7.20826626e-01
-6.73870027e-01 -9.29949343e-01 -1.96988106e-01 -2.98462391e-01
5.54898262e-01 -2.57021487e-01 2.47761175e-01 8.58564854e-01
6.23160303e-01 9.34308469e-01 -6.53239787e-02 -6.84206903e-01
-9.29787457e-02 -2.42386296e-01 8.36468518e-01 -1.18115735e+00
-5.76203279e-02 -5.08613944e-01 4.47819978e-01 -9.90531385e-01
2.23062932e-01 -9.34334993e-01 4.35472846e-01 -1.85141706e+00
3.60496908e-01 -3.18684965e-01 -7.43707269e-02 5.83592832e-01
-6.67332351e-01 1.05499588e-01 1.50003418e-01 3.42335179e-02
-4.47721779e-01 1.60667464e-01 1.36133850e+00 -4.33685407e-02
-2.67900735e-01 3.03305358e-01 -1.33685434e+00 6.23298049e-01
1.00267148e+00 -5.68943977e-01 -6.19764626e-01 -5.26310802e-01
5.21064162e-01 2.14050665e-01 -1.52331397e-01 -6.06729329e-01
3.63264948e-01 -1.14073709e-01 3.80427450e-01 -6.58089876e-01
-1.52286083e-01 -2.03784734e-01 -6.68897629e-01 2.43287832e-01
-1.08899927e+00 7.30862379e-01 4.36964110e-02 7.92047158e-02
-2.99913287e-01 -1.03923762e+00 3.21580917e-01 -3.01305264e-01
-4.56678212e-01 -3.88960809e-01 -6.60864115e-01 2.61591852e-01
7.14124143e-01 2.84081399e-02 -6.82641089e-01 -6.31419361e-01
-3.97731006e-01 2.48303622e-01 3.71466130e-01 5.69753230e-01
6.87785387e-01 -1.05506551e+00 -1.07998753e+00 -2.99185347e-02
2.70033270e-01 -7.07885101e-02 -3.39714229e-01 2.86544293e-01
-5.80780506e-01 7.90306807e-01 -3.82926077e-01 -1.19092003e-01
-1.45843160e+00 1.75741002e-01 -1.90586329e-01 -5.17767072e-01
-2.73997575e-01 1.16452575e+00 -4.72491145e-01 -8.19602251e-01
4.34243917e-01 -3.27620506e-01 -6.54098153e-01 2.99294680e-01
9.28829134e-01 6.69419169e-02 4.90775973e-01 -2.37204984e-01
2.27771908e-01 -1.44240439e-01 -3.43792886e-01 -5.54475725e-01
1.20916617e+00 1.06176719e-01 2.36499712e-01 7.40495503e-01
8.64034891e-01 5.20318568e-01 -8.86281073e-01 3.66601609e-02
4.21302080e-01 -2.76876241e-01 -5.41142941e-01 -1.33294415e+00
-2.47440308e-01 1.03077078e+00 -7.33902454e-02 3.60189915e-01
7.42310584e-01 -2.22452193e-01 5.44103980e-01 8.32748115e-01
8.48061666e-02 -1.09016395e+00 4.94349599e-01 1.08434260e+00
1.10425389e+00 -1.36502993e+00 -4.46711406e-02 -3.50977182e-02
-8.81512761e-01 1.32758129e+00 1.06804037e+00 3.86737213e-02
2.13211060e-01 1.00242533e-01 4.61000055e-01 4.44701426e-02
-1.27916110e+00 1.23883523e-01 5.04564643e-01 6.48999512e-01
1.39522958e+00 -4.03625444e-02 -9.73817050e-01 5.46642900e-01
-7.87950397e-01 -1.34228384e-02 1.33343768e+00 1.08771372e+00
-7.92104900e-01 -1.51592994e+00 -3.70789289e-01 6.22020006e-01
-7.67307639e-01 -3.43625158e-01 -1.12596476e+00 8.06228340e-01
-4.68086720e-01 1.10948157e+00 -2.93994546e-01 -3.31394643e-01
1.78538769e-01 5.39688528e-01 4.18959349e-01 -1.41018462e+00
-1.16845679e+00 -3.55503350e-01 3.96420449e-01 1.27757937e-01
-4.62340921e-01 -8.28660309e-01 -1.32851338e+00 -7.16246292e-02
-4.66702491e-01 7.31955230e-01 7.53030539e-01 8.14953744e-01
-1.76549152e-01 5.68348587e-01 1.78683504e-01 -5.06408334e-01
-8.81796539e-01 -1.45975053e+00 3.15883360e-03 2.22797632e-01
1.21743634e-01 -1.45019263e-01 -1.91287667e-01 2.81910568e-01] | [11.734583854675293, 9.03400707244873] |
5cc35e3a-426b-496b-94c7-ecedb34e25c0 | breaking-shortcut-exploring-fully | 2105.05838 | null | https://arxiv.org/abs/2105.05838v2 | https://arxiv.org/pdf/2105.05838v2.pdf | Breaking Shortcut: Exploring Fully Convolutional Cycle-Consistency for Video Correspondence Learning | Previous cycle-consistency correspondence learning methods usually leverage image patches for training. In this paper, we present a fully convolutional method, which is simpler and more coherent to the inference process. While directly applying fully convolutional training results in model collapse, we study the underline reason behind this collapse phenomenon, indicating that the absolute positions of pixels provide a shortcut to easily accomplish cycle-consistence, which hinders the learning of meaningful visual representations. To break this absolute position shortcut, we propose to apply different crops for forward and backward frames, and adopt feature warping to establish correspondence between two crops of a same frame. The former technique enforces the corresponding pixels at forward and back tracks to have different absolute positions, and the latter effectively blocks the shortcuts going between forward and back tracks. In three label propagation benchmarks for pose tracking, face landmark tracking and video object segmentation, our method largely improves the results of vanilla fully convolutional cycle-consistency method, achieving very competitive performance compared with the self-supervised state-of-the-art approaches. Our trained model and code are available at \url{https://github.com/Steve-Tod/STFC3}. | ['Han Hu', 'Philip H. S. Torr', 'Zheng Zhang', 'Yue Cao', 'Zhenda Xie', 'Zhenyu Jiang', 'Yansong Tang'] | 2021-05-12 | null | null | null | null | ['landmark-tracking'] | ['computer-vision'] | [ 2.92493790e-01 -3.49834189e-02 -5.73378503e-01 -5.00162005e-01
-2.97894329e-01 -6.23659313e-01 6.31220818e-01 -5.53611405e-02
-1.66427553e-01 5.35016954e-01 -3.60227078e-02 -3.02013576e-01
2.34867200e-01 -5.16267180e-01 -9.69650924e-01 -7.53571689e-01
1.61748856e-01 1.53653741e-01 3.97510231e-01 -1.39143905e-02
5.65207861e-02 4.73863810e-01 -1.40612447e+00 3.18633527e-01
3.08952987e-01 9.68906581e-01 -1.01563461e-01 4.02703524e-01
-2.36807272e-01 6.51014745e-01 -1.48666129e-01 -5.03805101e-01
2.39169687e-01 -6.86451912e-01 -8.18351448e-01 1.87895164e-01
8.97333324e-01 -4.40840065e-01 -2.58048832e-01 1.14361918e+00
2.34835684e-01 -2.75601953e-01 6.22643471e-01 -1.44539094e+00
-5.24080575e-01 4.77549106e-01 -1.05083275e+00 1.89618587e-01
1.25602126e-01 1.06729090e-01 1.11075461e+00 -7.65111864e-01
7.96437204e-01 1.13490152e+00 7.80804813e-01 7.83409894e-01
-1.26999462e+00 -9.07647252e-01 4.57060337e-01 8.57696757e-02
-1.36969948e+00 -7.11390972e-01 9.39207315e-01 -4.88656312e-01
4.72613931e-01 2.48640552e-01 7.92345524e-01 1.13052130e+00
5.28249070e-02 8.49266231e-01 1.06179714e+00 -3.40792775e-01
-3.36517324e-03 -1.00149587e-01 2.03171253e-01 9.94989693e-01
2.24036366e-01 3.70497853e-01 -5.38286150e-01 9.99771804e-02
9.35111284e-01 1.25018165e-01 -3.40988785e-01 -6.64981186e-01
-9.08999324e-01 7.26402104e-01 6.72655702e-01 4.00542349e-01
1.35923877e-01 4.42192554e-01 2.80404478e-01 1.42134994e-01
3.93276066e-01 -5.96461110e-02 -4.73782778e-01 2.29576737e-01
-1.18858242e+00 1.28693387e-01 5.32413006e-01 9.04629052e-01
8.36808741e-01 -1.98868111e-01 -4.34453934e-01 5.05865455e-01
5.15732229e-01 1.87330410e-01 -1.66831240e-02 -1.02338839e+00
1.15031324e-01 6.18475378e-01 -1.94278836e-01 -9.88416314e-01
-2.82569915e-01 -5.88715076e-01 -7.48577833e-01 2.62025207e-01
8.55983555e-01 2.05244526e-01 -1.18805790e+00 1.76245368e+00
5.16802549e-01 5.94669938e-01 -4.36171710e-01 8.03768575e-01
7.95826197e-01 3.49458069e-01 5.96914589e-02 -2.27364838e-01
1.35693252e+00 -1.21020961e+00 -5.98000884e-01 -2.03407317e-01
5.74668467e-01 -8.77131701e-01 8.50307286e-01 -7.64370849e-03
-9.69213903e-01 -6.91647351e-01 -1.04526365e+00 -1.01925977e-01
-2.63593256e-01 1.73462510e-01 7.72435188e-01 4.47430998e-01
-1.12427211e+00 8.77254903e-01 -9.85441387e-01 -2.87647069e-01
8.19464147e-01 1.99129209e-01 -2.83246517e-01 -2.67890960e-01
-7.96784818e-01 4.11161035e-01 -1.92933413e-03 3.03143650e-01
-7.71923542e-01 -7.62162387e-01 -9.12734926e-01 -1.86200932e-01
3.46754104e-01 -4.50486362e-01 1.07966089e+00 -1.07463515e+00
-1.14392793e+00 1.24696088e+00 -3.35088968e-01 -2.95966059e-01
7.07657158e-01 -1.03127614e-01 -1.27504751e-01 5.46104684e-02
1.20604850e-01 1.25148821e+00 1.12701237e+00 -1.31053269e+00
-5.50003409e-01 -3.92136157e-01 -1.52426986e-02 -1.93753302e-01
-2.19751716e-01 8.15032646e-02 -9.92526472e-01 -6.54902160e-01
2.30167195e-01 -1.01618266e+00 1.19020581e-01 5.45479119e-01
-4.66071278e-01 -2.70656079e-01 9.44489956e-01 -5.18350244e-01
1.25971210e+00 -2.05905199e+00 3.25623841e-04 6.34360388e-02
2.39402175e-01 1.87412500e-01 -2.11432725e-01 8.97591040e-02
-2.45795101e-01 1.10996291e-01 -3.59330744e-01 -6.20983601e-01
-2.31002912e-01 2.69009441e-01 -1.67229414e-01 6.97901607e-01
4.05224174e-01 1.07030821e+00 -9.34031308e-01 -6.37200236e-01
1.99388295e-01 5.68428576e-01 -6.30330026e-01 -1.34442806e-01
-3.25917929e-01 7.41384745e-01 -2.62078494e-01 8.43125403e-01
9.06096041e-01 -4.24767077e-01 2.12475732e-01 -3.62674475e-01
-1.35780215e-01 2.24904284e-01 -1.08637714e+00 1.85526049e+00
-3.72707583e-02 7.34377861e-01 -7.24500716e-02 -8.33702683e-01
7.31070817e-01 -1.96835981e-03 5.68369150e-01 -7.35946655e-01
2.09149912e-01 7.15415850e-02 -1.25109375e-01 -2.28793770e-01
1.34353697e-01 1.44884840e-01 3.13237578e-01 2.37113789e-01
2.89760297e-03 5.83073236e-02 2.28096709e-01 1.11314230e-01
7.38028467e-01 7.50482261e-01 1.16726786e-01 -3.44512194e-01
5.74205160e-01 -1.47880092e-01 8.07917714e-01 4.27097142e-01
-4.97838497e-01 8.07630241e-01 7.14068651e-01 -5.85533381e-01
-8.76583755e-01 -9.05414701e-01 -2.50235796e-01 1.07481706e+00
3.26976955e-01 -5.31044304e-01 -8.02751362e-01 -9.52026606e-01
-2.47957418e-03 2.72253662e-01 -1.03126442e+00 -8.01543891e-02
-6.66676342e-01 -4.15636718e-01 5.49801350e-01 7.73242354e-01
6.22043610e-01 -8.59239876e-01 -4.94568288e-01 -8.42674151e-02
-1.74444377e-01 -1.13278687e+00 -6.69507205e-01 1.57913744e-01
-8.52405906e-01 -1.17371321e+00 -6.72192395e-01 -8.27939153e-01
9.66721773e-01 2.23662123e-01 1.05956376e+00 5.87545276e-01
-3.36548239e-01 -1.83966253e-02 -1.52568862e-01 -1.25315525e-02
-1.99390173e-01 7.47016771e-03 -3.88130873e-01 -1.37804765e-02
2.90156543e-01 -3.88146788e-01 -7.13772476e-01 4.29477632e-01
-7.49313533e-01 1.94981322e-01 2.45687008e-01 9.06657934e-01
6.42989218e-01 -3.41745198e-01 1.09036528e-01 -8.71277869e-01
-1.52577579e-01 -1.33304760e-01 -6.36643291e-01 3.07931960e-01
-6.30073190e-01 1.40235111e-01 1.67792052e-01 -2.48559773e-01
-6.91946268e-01 3.88821632e-01 -1.42699689e-01 -6.86427891e-01
-2.29614928e-01 -1.04377463e-01 -9.64599401e-02 -8.14170018e-02
3.48910034e-01 6.20234832e-02 5.47565967e-02 -4.07920569e-01
4.62193459e-01 1.32896915e-01 5.12331784e-01 -4.75874156e-01
9.27766800e-01 8.52721274e-01 -4.01707254e-02 -5.12887836e-01
-9.10080433e-01 -4.06957000e-01 -8.94468129e-01 -3.41990292e-01
9.55888689e-01 -8.17370713e-01 -4.81731385e-01 5.63393533e-01
-1.31069028e+00 -4.73190069e-01 -7.52508864e-02 1.01237342e-01
-2.70983547e-01 3.56198967e-01 -5.85805118e-01 -5.02184868e-01
-1.40421866e-02 -1.11290169e+00 1.28663242e+00 2.94289589e-01
-1.94527462e-01 -9.94855821e-01 3.51395160e-02 2.26172149e-01
2.14691952e-01 3.40468287e-01 7.04617023e-01 -2.35553131e-01
-5.62681794e-01 1.15150534e-01 -3.41167033e-01 2.57621795e-01
2.69160479e-01 4.73558456e-01 -1.18214226e+00 -5.49096286e-01
-3.95516843e-01 -2.37676010e-01 1.24274683e+00 4.20135647e-01
1.11036193e+00 -1.24760762e-01 -6.49474978e-01 8.48561883e-01
1.30633354e+00 -1.58065230e-01 7.85629570e-01 2.64328867e-01
8.52088809e-01 6.65394247e-01 4.55630720e-01 9.05200243e-02
3.60132456e-01 8.48118007e-01 5.43628454e-01 -3.80241364e-01
-4.97758478e-01 -4.38908756e-01 2.87835240e-01 3.93949091e-01
-2.23655552e-02 -5.27946651e-03 -6.03964090e-01 5.08383512e-01
-1.81048441e+00 -8.27198267e-01 -1.07286915e-01 2.14005423e+00
1.00344408e+00 1.70223475e-01 6.92156926e-02 1.41292498e-01
8.26812804e-01 4.30307686e-01 -4.98808384e-01 -5.08374088e-02
1.05199978e-01 2.03729019e-01 4.64763492e-01 5.12181163e-01
-1.23728597e+00 1.20323694e+00 5.51034117e+00 8.76696169e-01
-1.27988636e+00 1.26712531e-01 7.94650376e-01 -9.50327609e-03
-2.72508264e-01 2.90514290e-01 -1.05387092e+00 4.86556381e-01
3.23528945e-01 5.75355828e-01 1.79461196e-01 5.92672229e-01
4.18405421e-02 5.11014163e-02 -1.34696460e+00 7.48099267e-01
-5.48193902e-02 -1.39190328e+00 -8.75418410e-02 3.75943892e-02
8.50027561e-01 -2.69621402e-01 1.53240442e-01 6.17370196e-02
6.24788292e-02 -9.98468041e-01 8.47647548e-01 3.56967241e-01
7.56463408e-01 -5.12404144e-01 4.04042423e-01 -6.50000721e-02
-1.54081190e+00 2.04903752e-01 -2.56954044e-01 1.36249000e-02
-4.93121482e-02 6.07079268e-01 -5.67757964e-01 4.05497283e-01
8.16065907e-01 9.49428082e-01 -7.80838966e-01 8.24641585e-01
-3.75561953e-01 5.81412494e-01 -1.46473140e-01 4.01948929e-01
1.96715146e-01 -1.56848893e-01 3.80719781e-01 1.14607453e+00
-1.03143685e-01 -3.11151087e-01 -4.87820618e-02 9.34983432e-01
-1.45066276e-01 -2.20206007e-01 -5.53102612e-01 5.27080707e-02
4.68845159e-01 1.31577659e+00 -1.15027452e+00 -2.11262748e-01
-4.54668581e-01 1.12003577e+00 3.45349014e-01 3.22126269e-01
-1.23800743e+00 4.28985842e-02 6.32570326e-01 2.92696327e-01
6.70810640e-01 -1.81033283e-01 -3.76502156e-01 -8.76444697e-01
3.73889320e-02 -6.50314093e-01 4.41896617e-01 -3.95199329e-01
-1.01698363e+00 4.46662307e-01 -3.32903303e-02 -1.32111633e+00
-1.07107095e-01 -4.65668529e-01 -7.74657428e-01 3.99400413e-01
-1.80055344e+00 -1.47953844e+00 -4.04775590e-01 6.37183845e-01
5.10013282e-01 2.26133108e-01 5.43077886e-01 4.27279651e-01
-7.35380888e-01 9.52300072e-01 -2.23282233e-01 3.54976833e-01
8.60434830e-01 -9.39695179e-01 4.76615995e-01 1.06452775e+00
3.91123146e-01 6.82576060e-01 4.31791633e-01 -6.20426357e-01
-1.01984525e+00 -1.15301168e+00 8.82211685e-01 -4.44312692e-01
4.06488240e-01 -4.83187199e-01 -9.17508245e-01 7.02366889e-01
2.36365452e-01 3.94411743e-01 3.31203192e-01 -1.40376404e-01
-6.57012939e-01 -1.04143724e-01 -9.07258034e-01 5.57243943e-01
1.39080894e+00 -3.11391383e-01 -2.40686849e-01 1.55149326e-01
6.88519239e-01 -3.85161906e-01 -4.09049749e-01 6.00215912e-01
6.38487518e-01 -1.07772827e+00 1.02164090e+00 -4.91830319e-01
6.44855082e-01 -5.17719686e-01 9.73049030e-02 -7.56025851e-01
-2.40195170e-01 -6.24077499e-01 -2.33546227e-01 1.42612612e+00
3.40935886e-01 -4.13054556e-01 8.08287680e-01 1.21732794e-01
8.33322182e-02 -9.44823563e-01 -8.72063398e-01 -6.78608477e-01
1.65084437e-01 -2.80241042e-01 5.41676223e-01 1.11748111e+00
-4.58455831e-01 1.43701211e-01 -2.69264430e-01 5.82518950e-02
7.46818364e-01 2.02873319e-01 6.22747123e-01 -1.04210579e+00
-1.79625705e-01 -6.42310143e-01 -2.93801218e-01 -1.34334660e+00
1.20098434e-01 -9.26265538e-01 -2.17356458e-02 -1.33923197e+00
1.80763602e-01 -6.08939052e-01 -1.93216756e-01 8.21245790e-01
-1.67159528e-01 8.24290395e-01 1.80233017e-01 2.49378502e-01
-5.54869473e-01 3.44040960e-01 1.46046126e+00 -1.95969269e-01
-1.30729452e-01 -1.57547090e-02 -6.54881716e-01 7.84043908e-01
7.53470898e-01 -5.33967972e-01 -2.57889777e-01 -4.99461651e-01
8.41512624e-03 -4.30850357e-01 6.84125304e-01 -8.20714056e-01
2.78450668e-01 -1.96746457e-02 7.51212299e-01 -6.08441770e-01
1.76853418e-01 -7.83160925e-01 9.09869224e-02 7.57042944e-01
-2.89014012e-01 1.07324077e-02 3.03655177e-01 5.55252254e-01
-4.47587855e-02 -1.56019986e-01 1.02911484e+00 1.77372657e-02
-7.40853429e-01 4.80705857e-01 7.09185079e-02 -2.78141573e-02
1.06432152e+00 -3.46906632e-01 -3.41090888e-01 -1.45683870e-01
-6.09765589e-01 2.41989985e-01 6.25415742e-01 6.18128538e-01
4.77055937e-01 -1.39852011e+00 -5.77094197e-01 4.32614297e-01
7.34381825e-02 5.10886461e-02 1.58229917e-01 1.10290873e+00
-4.13687080e-01 2.05598846e-01 -3.20215672e-01 -1.01046729e+00
-1.38196528e+00 3.21718246e-01 4.13857520e-01 -1.03979342e-01
-6.07976198e-01 1.07369959e+00 2.90547758e-01 -3.03690851e-01
4.42675769e-01 -4.22566086e-01 -1.71375901e-01 7.29345977e-02
3.95183712e-01 1.25179365e-01 -6.22358359e-03 -6.06918395e-01
-5.79000711e-01 9.51514125e-01 -2.98828334e-01 1.49726585e-01
1.05731189e+00 -6.17569350e-02 -1.28965810e-01 1.03669420e-01
1.38918328e+00 2.29171384e-02 -1.74901259e+00 -2.35046089e-01
-1.20733745e-01 -4.95605081e-01 -4.73075621e-02 -5.23806870e-01
-1.58716369e+00 9.31097329e-01 7.03715026e-01 -1.20376386e-01
1.03236878e+00 3.08523417e-01 6.35117769e-01 -1.54567331e-01
1.17966175e-01 -8.33761096e-01 1.90260246e-01 3.48719656e-01
6.46923721e-01 -1.18869150e+00 1.55581191e-01 -6.24657691e-01
-4.97191370e-01 1.08285379e+00 7.64478922e-01 -1.61425583e-02
5.62238157e-01 3.53927374e-01 4.01597582e-02 -3.40831250e-01
-5.36486447e-01 -3.26503128e-01 3.59446555e-01 4.66005594e-01
6.04383945e-01 -1.97744578e-01 -3.24467808e-01 1.22298144e-01
5.96108586e-02 -2.92820949e-02 1.17037907e-01 9.43955183e-01
-2.05845669e-01 -1.24138153e+00 -1.51434630e-01 1.31421581e-01
-3.80569011e-01 -3.42194214e-02 -4.55800951e-01 9.63167667e-01
3.64472479e-01 6.79210186e-01 3.37277681e-01 -2.38162324e-01
2.93163583e-02 5.01579382e-02 7.81193137e-01 -3.75879914e-01
-3.17079782e-01 2.81777442e-01 -1.39795840e-01 -7.02501357e-01
-6.71333075e-01 -6.70400620e-01 -1.38042033e+00 -2.58730471e-01
-4.08851296e-01 -3.16489995e-01 4.26680535e-01 8.99363220e-01
2.10138857e-01 5.63737273e-01 6.41270339e-01 -7.81565726e-01
-2.54609197e-01 -7.07117975e-01 -2.27323711e-01 3.54654133e-01
3.93874884e-01 -8.03349614e-01 -3.53566796e-01 3.08459044e-01] | [9.124653816223145, -0.1137237399816513] |
323d5879-544e-441d-9b63-4b2aa9537960 | 3d-multi-object-tracking-using-graph-neural | 2203.10926 | null | https://arxiv.org/abs/2203.10926v2 | https://arxiv.org/pdf/2203.10926v2.pdf | 3D Multi-Object Tracking Using Graph Neural Networks with Cross-Edge Modality Attention | Online 3D multi-object tracking (MOT) has witnessed significant research interest in recent years, largely driven by demand from the autonomous systems community. However, 3D offline MOT is relatively less explored. Labeling 3D trajectory scene data at a large scale while not relying on high-cost human experts is still an open research question. In this work, we propose Batch3DMOT which follows the tracking-by-detection paradigm and represents real-world scenes as directed, acyclic, and category-disjoint tracking graphs that are attributed using various modalities such as camera, LiDAR, and radar. We present a multi-modal graph neural network that uses a cross-edge attention mechanism mitigating modality intermittence, which translates into sparsity in the graph domain. Additionally, we present attention-weighted convolutions over frame-wise k-NN neighborhoods as suitable means to allow information exchange across disconnected graph components. We evaluate our approach using various sensor modalities and model configurations on the challenging nuScenes and KITTI datasets. Extensive experiments demonstrate that our proposed approach yields an overall improvement of 3.3% in the AMOTA score on nuScenes thereby setting the new state-of-the-art for 3D tracking and further enhancing false positive filtering. | ['Abhinav Valada', 'Martin Buchner'] | 2022-03-21 | null | null | null | null | ['3d-multi-object-tracking'] | ['computer-vision'] | [ 1.51523456e-01 -1.45557508e-01 -1.52518004e-01 -1.05916448e-01
-3.81999373e-01 -6.98827922e-01 6.17629647e-01 1.72084898e-01
-3.15918267e-01 3.65291953e-01 -1.51469126e-01 -2.46490732e-01
-1.97713107e-01 -7.64185607e-01 -9.91142154e-01 -4.21563804e-01
-2.75509298e-01 5.16640127e-01 6.57296479e-01 3.79259028e-02
-3.78113061e-01 8.87429774e-01 -1.59700608e+00 -1.78998932e-01
5.11614561e-01 1.30404902e+00 9.59044099e-02 7.83709884e-01
4.10350673e-02 7.39249587e-01 -2.92234749e-01 -3.79330367e-01
5.27119279e-01 3.09666265e-02 -2.33106807e-01 3.79470825e-01
1.13113046e+00 -2.46641546e-01 -8.05884421e-01 1.18257546e+00
4.58968550e-01 2.18793392e-01 2.94245839e-01 -1.60043490e+00
-6.72557294e-01 2.18120605e-01 -6.90023780e-01 2.87839681e-01
6.95158169e-02 3.58295918e-01 8.77996445e-01 -5.97246468e-01
6.85071766e-01 1.40498960e+00 8.47349644e-01 3.63352567e-01
-1.04589593e+00 -7.09543645e-01 4.85577136e-01 3.41285974e-01
-1.33618355e+00 -1.40975609e-01 7.99765468e-01 -4.29280907e-01
1.04834580e+00 7.87053481e-02 8.46280217e-01 9.74112988e-01
1.83130652e-01 6.54661834e-01 8.21777940e-01 -7.51423761e-02
-9.34204459e-02 -1.77219361e-01 9.59800333e-02 1.00810528e+00
6.50233626e-01 2.56041408e-01 -6.52668774e-01 6.38720521e-04
6.42384827e-01 2.02802241e-01 -3.86815518e-02 -6.12447023e-01
-1.21801341e+00 5.87712586e-01 8.45088363e-01 -1.54360220e-01
-4.20286566e-01 5.30294538e-01 2.58660644e-01 1.05531417e-01
5.41788638e-01 -3.25424820e-01 -1.11786798e-01 3.12774628e-01
-6.67759061e-01 1.39170542e-01 5.52421153e-01 1.35774767e+00
5.37982225e-01 2.59128183e-01 -1.24286287e-01 3.15088570e-01
4.49398190e-01 1.11131847e+00 -4.78209466e-01 -9.20709848e-01
3.27608883e-01 9.57844198e-01 1.43293783e-01 -1.14888406e+00
-6.30534589e-01 -5.52773356e-01 -8.27977955e-01 3.64231557e-01
2.67828941e-01 -1.33284256e-01 -1.03679228e+00 1.83867526e+00
9.03164923e-01 4.66879368e-01 -2.07606122e-01 1.18853879e+00
9.75075006e-01 3.78290266e-01 5.51444329e-02 2.78490130e-03
1.23591650e+00 -8.08433592e-01 -6.82543635e-01 -3.06115538e-01
2.40636602e-01 -3.76992583e-01 4.38239455e-01 -4.95589711e-03
-8.84387136e-01 -4.94158149e-01 -8.92111599e-01 6.47103190e-02
-5.17291307e-01 2.00838912e-02 5.96388400e-01 5.70850492e-01
-1.12366927e+00 1.79005116e-01 -1.08906269e+00 -9.06442642e-01
6.31818831e-01 4.08720195e-01 -3.60058755e-01 -2.59602547e-01
-8.42938244e-01 8.60006273e-01 3.67434651e-01 2.16696069e-01
-1.12702394e+00 -5.80049455e-01 -8.86741161e-01 -1.52816802e-01
8.38438451e-01 -9.79686558e-01 6.45373464e-01 -2.37749919e-01
-1.01143944e+00 7.56505489e-01 3.14893499e-02 -6.43077970e-01
4.89742488e-01 -2.43371159e-01 -6.65884376e-01 1.79193214e-01
8.76140967e-03 8.51942658e-01 9.32237744e-01 -1.22411990e+00
-8.49163890e-01 -4.77989912e-01 3.32860112e-01 3.56199443e-01
-2.20993742e-01 -7.69515336e-02 -8.60388935e-01 -3.22259575e-01
8.13606679e-02 -1.25931561e+00 -1.40907183e-01 6.14540994e-01
-3.87919098e-01 -2.51470476e-01 1.29068840e+00 -2.43399814e-01
8.12506974e-01 -1.98683381e+00 2.39028722e-01 1.15255453e-02
5.67862093e-01 9.01113898e-02 -5.83688207e-02 2.54212409e-01
4.88107473e-01 -3.29653502e-01 -1.30009368e-01 -2.95589417e-01
1.49801597e-01 2.38381997e-01 9.45221111e-02 8.24284315e-01
4.13112581e-01 1.00173604e+00 -9.60118651e-01 -4.68002200e-01
6.62150800e-01 6.34524643e-01 -3.66082549e-01 7.67955706e-02
-4.61376578e-01 3.87209475e-01 -3.67307007e-01 1.16315937e+00
7.74692416e-01 -6.37748897e-01 2.86392514e-02 -4.46273685e-01
-2.39902079e-01 -2.67602861e-01 -1.19577086e+00 1.80151296e+00
4.83379923e-02 7.15363920e-01 2.57122099e-01 -8.39032233e-01
7.47755587e-01 -1.21445388e-01 6.56581998e-01 -4.56752419e-01
3.17350268e-01 -8.74587670e-02 -2.10977763e-01 -3.71590018e-01
7.48946249e-01 9.50481594e-02 -2.15279311e-01 4.79517244e-02
1.71348423e-01 2.16896728e-01 1.66401386e-01 3.84238183e-01
1.38625956e+00 1.71140730e-01 -6.44733310e-02 -1.36324540e-01
3.23824644e-01 4.64951605e-01 4.94177908e-01 9.11960661e-01
-5.30228853e-01 1.02432758e-01 -1.21994570e-01 -3.71160299e-01
-6.65387154e-01 -1.11504853e+00 1.73839122e-01 8.71910751e-01
7.51042128e-01 -9.78332385e-02 -9.43761468e-02 -7.61447191e-01
5.84997237e-01 2.66122073e-01 -4.81655061e-01 -3.83301638e-02
-4.58741963e-01 -4.60400641e-01 5.74576735e-01 5.90458870e-01
5.49265027e-01 -6.43545508e-01 -7.43284941e-01 2.32716784e-01
-6.19489991e-04 -1.68500614e+00 -4.28199917e-01 9.03348327e-02
-8.07715774e-01 -1.24625850e+00 -5.88482261e-01 -4.84702051e-01
6.25735819e-01 8.70426953e-01 1.00395703e+00 -1.96251333e-01
-2.22200617e-01 9.70607281e-01 -2.45579019e-01 -4.24083829e-01
3.44621912e-02 -1.36488795e-01 3.48897845e-01 1.21046998e-01
4.22579110e-01 -3.36248070e-01 -4.61346596e-01 2.77119964e-01
-6.43488586e-01 -6.25542551e-02 4.51573193e-01 4.17193830e-01
7.57691562e-01 -9.02947262e-02 2.25666031e-01 -4.46833849e-01
-1.21219838e-02 -5.94286919e-01 -1.14842796e+00 1.53073654e-01
-3.17601591e-01 -2.06231847e-01 2.95938700e-01 -5.62058628e-01
-7.07073092e-01 3.13468635e-01 3.88042569e-01 -1.28109384e+00
-2.04818413e-01 3.71651411e-01 -7.60104358e-02 -6.66138828e-01
2.93867618e-01 -3.52541059e-02 -1.54804230e-01 -2.28115559e-01
6.45044148e-01 2.50852257e-01 6.55807734e-01 -3.01786900e-01
1.22402394e+00 9.87545133e-01 4.95305836e-01 -9.55807745e-01
-8.20715845e-01 -6.68230653e-01 -4.40074950e-01 -1.03948426e+00
9.85925436e-01 -1.33014882e+00 -1.13533223e+00 4.43411648e-01
-1.06393588e+00 -2.84439743e-01 -1.82859942e-01 5.86330533e-01
-1.11294575e-01 2.61009663e-01 -2.15817645e-01 -1.02462685e+00
-2.22965598e-01 -7.98231721e-01 1.35713816e+00 2.77834713e-01
3.18957388e-01 -8.63213658e-01 -1.50184408e-01 2.40668252e-01
4.26227510e-01 6.73769057e-01 4.33183610e-01 -4.08892334e-01
-1.37338603e+00 -1.81462258e-01 -6.67244792e-01 -1.20416313e-01
-6.65424988e-02 -3.33466679e-01 -8.53954375e-01 -5.20121396e-01
-5.50027966e-01 -2.44094938e-01 8.75926733e-01 5.42841017e-01
7.65897453e-01 1.18324503e-01 -6.48197651e-01 6.87310040e-01
1.55158842e+00 -9.54047441e-02 -1.53560832e-01 -6.70031682e-02
1.19581103e+00 4.20570225e-01 5.51500201e-01 3.78257841e-01
6.99142754e-01 5.52965939e-01 1.02740335e+00 1.79366712e-02
-5.42012155e-01 -1.99856490e-01 3.01786989e-01 6.17289007e-01
2.87752990e-02 -6.88633680e-01 -9.08220410e-01 8.66946340e-01
-2.00912642e+00 -8.77635598e-01 -3.70567471e-01 1.96868157e+00
-1.52623400e-01 2.64702410e-01 2.13837981e-01 -4.02692139e-01
1.03137815e+00 3.68280739e-01 -9.63616788e-01 3.57401907e-01
-3.33608717e-01 -3.02144140e-01 1.02007174e+00 2.54797101e-01
-1.36234879e+00 8.30619872e-01 4.95195055e+00 4.66272742e-01
-1.00492251e+00 1.48758441e-01 -1.40837371e-01 -2.36582175e-01
-6.70093857e-03 -4.38600518e-02 -1.06877589e+00 1.66184887e-01
7.40490794e-01 -8.23053345e-02 4.01249856e-01 6.05293632e-01
1.37365311e-02 1.13248475e-01 -8.75281751e-01 1.04860926e+00
2.21207500e-01 -1.45401597e+00 -7.91450441e-02 1.86540410e-01
6.96641207e-01 7.87743986e-01 -1.27732173e-01 2.20253631e-01
5.99651396e-01 -4.09623981e-01 9.50632632e-01 5.61640203e-01
8.08057070e-01 -3.39608938e-01 3.28302383e-01 2.89938688e-01
-2.00474691e+00 -7.55596235e-02 -1.21970057e-01 1.24575280e-01
4.35655415e-01 4.91480142e-01 -5.14434457e-01 9.48704481e-01
9.77826834e-01 1.04912841e+00 -4.48904246e-01 1.22371149e+00
2.13058680e-01 4.10698593e-01 -9.11392629e-01 -2.16158986e-01
3.40708733e-01 -3.45554538e-02 1.13508093e+00 9.88979578e-01
4.00227249e-01 1.27277628e-01 5.64524591e-01 8.56120050e-01
-4.08076882e-01 -5.85240602e-01 -9.93906081e-01 -9.88079086e-02
5.08868694e-01 1.25655735e+00 -7.77966678e-01 -1.89506292e-01
-7.49419570e-01 6.45726085e-01 2.11914256e-01 3.74128282e-01
-1.07112551e+00 6.03444874e-03 6.00790501e-01 2.23716944e-02
7.66181171e-01 -5.46034038e-01 1.30972445e-01 -1.04379153e+00
-2.61892453e-02 -3.48892629e-01 5.39481759e-01 -7.06546783e-01
-1.44427645e+00 3.96431476e-01 6.35773763e-02 -1.44022059e+00
3.96696895e-01 -6.05520844e-01 -2.04926923e-01 4.62165594e-01
-1.78274715e+00 -1.61228812e+00 -7.92048931e-01 7.37052977e-01
2.14441627e-01 1.18900053e-01 2.84148991e-01 6.38150156e-01
-5.35986781e-01 1.83704749e-01 -1.37808844e-01 6.94944663e-03
4.44596410e-01 -9.75212693e-01 4.11033601e-01 1.06305063e+00
1.30195111e-01 1.43082097e-01 4.65662450e-01 -1.05514240e+00
-2.14171171e+00 -1.72505760e+00 6.47110879e-01 -4.85644847e-01
9.08741653e-01 -4.47358340e-01 -6.26346469e-01 7.57949948e-01
1.71895161e-01 6.35549009e-01 1.96913138e-01 -1.04281060e-01
-1.54715016e-01 -1.55191869e-01 -1.06407773e+00 3.32904994e-01
1.67156386e+00 -4.22322720e-01 -1.27389103e-01 4.48905438e-01
8.54186416e-01 -6.73542321e-01 -9.15473163e-01 6.30075157e-01
3.93028408e-01 -5.27178824e-01 1.02453709e+00 -3.85641694e-01
-3.33575279e-01 -7.96434700e-01 -4.12950873e-01 -8.38633239e-01
-3.44700366e-01 -4.91277933e-01 -7.19029129e-01 1.02902269e+00
-5.46448119e-02 -4.32400614e-01 8.92005146e-01 2.81527221e-01
-4.19698149e-01 -2.07957759e-01 -1.09878469e+00 -1.04069376e+00
-5.47298670e-01 -5.68298578e-01 4.17647332e-01 8.36477995e-01
-6.92000031e-01 3.33889753e-01 -5.00819683e-01 8.58238399e-01
1.34489608e+00 3.53532374e-01 1.01948035e+00 -1.44663012e+00
4.53224890e-02 -2.29544327e-01 -7.03942537e-01 -1.31940365e+00
8.64258334e-02 -1.04848695e+00 2.79613701e-03 -1.50395286e+00
-7.56618530e-02 -4.01306033e-01 -2.82737672e-01 4.51696038e-01
-2.90569365e-02 3.65185350e-01 4.19323325e-01 2.66504027e-02
-1.20835888e+00 7.52781868e-01 9.63172376e-01 -5.53437412e-01
5.82376793e-02 -2.81605184e-01 -1.97742268e-01 4.37934905e-01
3.39702696e-01 -5.53729415e-01 -2.18385160e-01 -7.15213060e-01
1.27718104e-02 9.73541737e-02 8.99817526e-01 -1.23504210e+00
6.85728312e-01 -1.43185005e-01 1.79562956e-01 -1.18954575e+00
6.12916529e-01 -1.41119981e+00 5.30047655e-01 4.06632274e-01
1.16074651e-01 1.03275374e-01 6.15945578e-01 1.29253542e+00
1.84200734e-01 5.67381501e-01 5.86492956e-01 2.46994346e-02
-1.10345042e+00 7.86228418e-01 -2.22242419e-02 6.73281355e-03
1.12158000e+00 -2.25891575e-01 -5.70879996e-01 -1.01480126e-01
-4.96981531e-01 7.78849900e-01 4.82008189e-01 5.46242476e-01
5.21397114e-01 -1.45661342e+00 -6.29592597e-01 4.24538217e-02
4.07816619e-01 -2.87550897e-03 4.97471392e-01 8.86132538e-01
-1.71210751e-01 3.48846495e-01 -1.06250420e-01 -1.22704542e+00
-1.24070644e+00 4.67125177e-01 2.49867752e-01 6.73057558e-03
-8.93632591e-01 7.58218169e-01 -1.87145486e-01 -4.72227067e-01
4.49943870e-01 -3.46842080e-01 3.54111232e-02 -1.99681297e-02
5.84944189e-02 5.11751413e-01 -6.64464831e-02 -9.11033690e-01
-7.76170492e-01 6.74672008e-01 3.25977683e-01 3.00290704e-01
1.21095264e+00 -3.20030540e-01 3.10995996e-01 3.50144655e-01
6.99432075e-01 -4.46937561e-01 -1.58619142e+00 -4.62573498e-01
-2.23111570e-01 -4.89032120e-01 2.90160060e-01 -6.51328206e-01
-1.21700728e+00 6.16404712e-01 8.87559533e-01 4.39588159e-01
9.49857533e-01 2.47246712e-01 8.65052700e-01 5.42010486e-01
5.84446073e-01 -6.72140956e-01 -1.28037959e-01 5.48518181e-01
4.10581231e-01 -1.49077928e+00 3.45114470e-02 -3.64212066e-01
-2.71620661e-01 7.16288745e-01 5.66397250e-01 -2.26817623e-01
6.16034687e-01 1.88074231e-01 -1.38566077e-01 -6.23462737e-01
-6.71479642e-01 -8.19583654e-01 3.74496132e-01 7.71178842e-01
-3.02110374e-01 2.19758730e-02 3.46221983e-01 7.03200474e-02
5.36947906e-01 -6.11561257e-03 6.30757660e-02 9.73419309e-01
-3.54504377e-01 -5.56619048e-01 -4.14486110e-01 5.47450423e-01
1.28584038e-02 1.76266521e-01 -4.09462124e-01 1.04528618e+00
1.90887302e-01 1.02165914e+00 8.54635611e-02 -5.72547376e-01
5.67195296e-01 -2.40824252e-01 5.93544066e-01 -2.83547580e-01
-5.01012206e-01 9.94956270e-02 2.44744763e-01 -6.74680948e-01
-1.00306439e+00 -7.06090331e-01 -1.09930587e+00 -3.89627904e-01
-6.77181423e-01 -3.61186266e-01 6.45423472e-01 8.86606753e-01
6.74169958e-01 7.95267165e-01 2.54706919e-01 -1.12610185e+00
-6.44595176e-02 -6.47009254e-01 -3.28618497e-01 2.69141138e-01
5.83040953e-01 -1.17470074e+00 -2.10792631e-01 -7.88539648e-02] | [6.439594268798828, -2.129971981048584] |
40dd74ad-9f77-45b5-a3e8-8c1ad7a0f223 | using-a-supervised-method-without-supervision | 2011.07954 | null | https://arxiv.org/abs/2011.07954v4 | https://arxiv.org/pdf/2011.07954v4.pdf | Using a Supervised Method without supervision for foreground segmentation | Neural networks are a powerful framework for foreground segmentation in video acquired by static cameras, segmenting moving objects from the background in a robust way in various challenging scenarios. The premier methods are those based on supervision requiring a final training stage on a database of tens to hundreds of manually segmented images from the specific static camera. In this work, we propose a method to automatically create an "artificial" database that is sufficient for training the supervised methods so that it performs better than current unsupervised methods. It is based on combining a weak foreground segmenter, compared to the supervised method, to extract suitable objects from the training images and randomly inserting these objects back into a background image. Test results are shown on the test sequences in CDnet. | ['Michael Werman', 'Levi Kassel'] | 2020-10-26 | null | null | null | null | ['foreground-segmentation'] | ['computer-vision'] | [ 8.11542451e-01 -2.71369182e-02 -8.30376819e-02 -4.19437259e-01
-3.87981504e-01 -3.41237813e-01 4.85494018e-01 -1.81478098e-01
-8.11859965e-01 7.55706787e-01 -3.53591561e-01 -1.13330103e-01
2.14964762e-01 -6.17308319e-01 -8.16877902e-01 -9.92922306e-01
1.17722608e-01 7.59393632e-01 1.04591072e+00 1.39043361e-01
2.61489779e-01 6.02959335e-01 -1.69518435e+00 2.77408302e-01
7.18960404e-01 8.86937022e-01 5.11624396e-01 6.48664415e-01
-2.51429617e-01 1.00639296e+00 -8.77558470e-01 -6.89342394e-02
4.51275885e-01 -6.48262203e-01 -8.64597440e-01 7.20459163e-01
6.09131753e-01 -3.69397163e-01 6.94540814e-02 1.22914970e+00
9.04579535e-02 1.71298638e-01 5.24281919e-01 -9.28029597e-01
3.65066767e-01 6.97201371e-01 -4.11264807e-01 6.99955642e-01
1.64873242e-01 9.81236175e-02 4.32388335e-01 -6.96126521e-01
8.48922431e-01 8.94812107e-01 3.73062789e-01 7.44313776e-01
-9.81150329e-01 -4.04951513e-01 2.56748170e-01 2.39636064e-01
-1.16003823e+00 -4.66591239e-01 8.07042837e-01 -5.26228368e-01
6.11962140e-01 2.37688646e-01 8.04765284e-01 8.75311494e-01
-2.89812475e-01 9.11167145e-01 8.68425369e-01 -5.38987219e-01
3.89604867e-01 2.93462545e-01 4.77770865e-01 5.77619612e-01
3.93287033e-01 -5.26572354e-02 -1.76897168e-01 6.83799013e-02
8.63166988e-01 3.12305558e-02 -5.45138538e-01 -5.57289302e-01
-1.15022802e+00 5.69806993e-01 2.34111547e-01 5.97331762e-01
-4.97219056e-01 -1.34389192e-01 2.34823987e-01 -1.01369455e-01
3.25149596e-01 1.07883491e-01 -5.19967258e-01 1.62018239e-01
-1.57863545e+00 -4.68846001e-02 8.82172465e-01 9.41991270e-01
9.11827326e-01 2.88303256e-01 7.92220011e-02 4.02683347e-01
1.86112210e-01 2.96984315e-01 6.29663706e-01 -8.22870553e-01
2.91717231e-01 8.17277014e-01 8.52367729e-02 -7.95596123e-01
-1.98006228e-01 -2.53461957e-01 -6.79406703e-01 4.34903830e-01
6.38618112e-01 -2.64941514e-01 -1.44818723e+00 1.18177760e+00
4.77913886e-01 3.35368097e-01 6.42646998e-02 9.01471615e-01
7.52420843e-01 8.73544812e-01 1.20884189e-02 -5.85734367e-01
7.53807724e-01 -1.12980103e+00 -7.49184310e-01 -2.77462989e-01
1.10349022e-01 -6.85152233e-01 6.05776131e-01 8.71217906e-01
-1.13172352e+00 -8.74178708e-01 -9.06877041e-01 4.94815320e-01
-4.84851241e-01 2.71691084e-01 3.71210992e-01 6.86761916e-01
-1.05872738e+00 6.79938793e-01 -8.86162877e-01 -2.85304308e-01
5.49380064e-01 6.22314036e-01 -1.96680620e-01 8.52560252e-02
-8.33286166e-01 6.52464390e-01 1.05667734e+00 3.76251996e-01
-1.38897657e+00 -1.29286945e-01 -7.21507490e-01 -2.40169346e-01
7.15374708e-01 -2.60212034e-01 1.02102780e+00 -1.95243824e+00
-1.52352989e+00 9.97946799e-01 -3.20966877e-02 -7.93766975e-01
8.16027522e-01 -2.90730566e-01 -1.28449053e-01 5.92626274e-01
-8.17142054e-02 7.21465528e-01 1.22440243e+00 -1.45257580e+00
-8.74193370e-01 -5.24187721e-02 -4.74284105e-02 6.28564646e-03
1.88003555e-02 1.42368972e-01 -7.07732320e-01 -5.84080279e-01
1.15873300e-01 -7.69046009e-01 -5.55022478e-01 -5.15705109e-01
-3.35826248e-01 -5.79317249e-02 1.44267166e+00 -9.17816341e-01
9.01294708e-01 -1.95722723e+00 2.87518501e-01 2.97718436e-01
9.86229107e-02 8.70482266e-01 2.22021848e-01 -1.91839337e-01
-1.51582053e-02 -2.45984703e-01 -5.21020710e-01 -1.76477864e-01
-5.80342174e-01 3.68159533e-01 2.01757386e-01 4.48561698e-01
1.60149917e-01 5.09849370e-01 -7.93121874e-01 -8.84661734e-01
3.53195041e-01 1.45330623e-01 -3.27678651e-01 3.71713400e-01
-5.81228256e-01 7.21756697e-01 -1.90981165e-01 7.32029915e-01
6.43616617e-01 -1.12379566e-01 1.61046475e-01 -5.80896400e-02
-5.46544418e-03 -1.83707118e-01 -1.44051600e+00 1.39334106e+00
2.97891110e-01 8.12978625e-01 2.58328974e-01 -1.32540703e+00
7.26608992e-01 3.27431470e-01 5.02396643e-01 -1.42698111e-02
5.20432115e-01 2.76787393e-02 -3.14734504e-02 -7.96542645e-01
2.46165499e-01 -6.67111501e-02 3.27345908e-01 1.04731984e-01
3.74952793e-01 6.06699958e-02 7.84579933e-01 2.97084339e-02
8.17551255e-01 3.50594044e-01 -2.45289188e-02 -1.77342609e-01
7.89575815e-01 2.13144735e-01 8.14542472e-01 7.11075664e-01
-2.96563983e-01 6.53720856e-01 1.79724455e-01 -6.25416875e-01
-1.02707982e+00 -7.79561698e-01 2.87694540e-02 7.26888537e-01
3.22788686e-01 1.06578648e-01 -1.46653712e+00 -9.26675498e-01
-5.36427438e-01 2.82550603e-01 -2.27119491e-01 2.27082700e-01
-8.38601112e-01 -8.12611043e-01 3.16176385e-01 4.40770000e-01
8.34235907e-01 -1.32038343e+00 -8.63949299e-01 1.90455005e-01
-1.70787051e-01 -1.34403145e+00 1.22919846e-02 4.52648550e-01
-1.19697869e+00 -1.17727685e+00 -8.58703077e-01 -1.15019858e+00
1.00568974e+00 2.36348465e-01 1.11029363e+00 3.29036176e-01
-2.47546002e-01 8.05463195e-02 -4.06787336e-01 -4.33529615e-01
-5.56725562e-01 -1.56283155e-02 1.10683041e-02 4.30183709e-01
2.67527759e-01 -2.97873735e-01 -3.03046107e-01 3.40302467e-01
-1.11454761e+00 1.74671993e-01 8.33554387e-01 4.79328185e-01
5.79862475e-01 4.17513788e-01 1.67544737e-01 -1.21761322e+00
-1.39552921e-01 -1.70660779e-01 -1.04722440e+00 1.07024163e-01
-1.58775330e-01 -4.02065605e-01 5.72461784e-01 -6.52747512e-01
-1.23549330e+00 6.26227975e-01 -5.60885184e-02 -5.09553850e-01
-7.43689716e-01 1.07414030e-01 -5.54192603e-01 -1.21556185e-01
6.38688862e-01 2.73531437e-01 -3.81159753e-01 -5.38515687e-01
1.09351486e-01 4.40870643e-01 7.75290370e-01 -2.59954661e-01
8.87812734e-01 3.84642869e-01 -4.01221097e-01 -1.06627250e+00
-5.37566900e-01 -7.10775495e-01 -1.24493885e+00 -5.50090551e-01
1.10066092e+00 -6.26408696e-01 -1.33892726e-02 7.46794641e-01
-1.17256117e+00 -5.89421809e-01 -2.55163014e-01 5.76454639e-01
-3.28716189e-01 3.48432034e-01 -6.66589797e-01 -7.93426633e-01
-5.44607006e-02 -1.05822468e+00 7.90614128e-01 6.31533563e-01
1.43474475e-01 -9.02080536e-01 -1.68337628e-01 5.64032555e-01
6.22861944e-02 3.44911993e-01 4.17890787e-01 -8.15412462e-01
-1.10850573e+00 -2.94791549e-01 -3.46062258e-02 9.52127278e-01
2.06401840e-01 3.23155731e-01 -9.03795958e-01 -6.36822283e-02
3.35697174e-01 1.25608757e-01 8.89492095e-01 4.98647839e-01
9.90108788e-01 -3.28717232e-01 -6.08029425e-01 5.85783720e-01
1.49225640e+00 7.30581820e-01 7.76755631e-01 2.43310139e-01
8.55512023e-01 4.87637162e-01 6.83150530e-01 1.28532238e-02
-3.23748827e-01 2.93551087e-01 2.78172284e-01 -4.22135055e-01
1.02361999e-01 3.47526550e-01 3.41411471e-01 6.31401658e-01
-4.05780494e-01 -2.45023087e-01 -9.21795487e-01 5.87972343e-01
-1.68259704e+00 -1.06129897e+00 -2.54477173e-01 2.04174089e+00
8.77001405e-01 6.98940217e-01 3.48290801e-01 3.46404701e-01
1.07323802e+00 2.38984004e-02 -2.48681381e-01 8.61311331e-02
-1.41856924e-01 3.33797693e-01 5.06156325e-01 4.07769024e-01
-1.50563467e+00 1.22137249e+00 6.52199221e+00 6.37661755e-01
-1.15441728e+00 3.94330099e-02 7.13324428e-01 1.32381678e-01
4.21031833e-01 5.79110086e-02 -1.05908966e+00 5.93046904e-01
8.84891748e-01 2.70387292e-01 6.88257441e-02 8.78318310e-01
3.13706338e-01 -5.81326008e-01 -1.03789532e+00 6.56667829e-01
3.28093499e-01 -1.25352180e+00 1.47222012e-01 -3.14952940e-01
9.51708853e-01 -2.11662427e-01 -4.02144969e-01 -1.09232277e-01
-1.10600039e-01 -7.62472987e-01 5.91475427e-01 6.48499131e-01
1.96738672e-02 -5.57469249e-01 1.07374537e+00 7.18164444e-01
-7.70632684e-01 6.21506386e-02 -3.92101735e-01 7.49672577e-02
2.08051562e-01 5.45916378e-01 -8.66594195e-01 4.04452622e-01
7.03919351e-01 4.84563619e-01 -7.28268385e-01 1.39192939e+00
-1.35720447e-01 9.27280545e-01 -3.34695369e-01 9.53747481e-02
4.17640507e-01 -5.10525584e-01 5.19817770e-01 1.54950261e+00
-9.19901654e-02 6.70382455e-02 3.13312441e-01 7.29063809e-01
1.75083175e-01 1.27985373e-01 -6.70762300e-01 5.68282865e-02
-1.65837213e-01 1.39549375e+00 -1.50127757e+00 -9.28430796e-01
-2.49827489e-01 1.08561945e+00 -1.88982055e-01 3.69374424e-01
-9.49794531e-01 -1.92805856e-01 -2.20156074e-01 1.09910198e-01
6.74314439e-01 -2.27842629e-01 2.71697879e-01 -9.88318622e-01
4.08650748e-02 -1.08986270e+00 3.29088897e-01 -7.39675760e-01
-8.12155366e-01 8.78807127e-01 2.49740452e-01 -1.15285695e+00
-1.86358422e-01 -8.54484618e-01 -5.74249625e-01 3.78713638e-01
-1.19071043e+00 -9.20365095e-01 -5.43513179e-01 7.11720645e-01
8.41515779e-01 -3.04610640e-01 2.85076052e-01 4.78197992e-01
-7.04417765e-01 -9.24004242e-02 -1.73864409e-01 5.40644109e-01
2.32378155e-01 -1.34938550e+00 6.23001568e-02 1.36883521e+00
4.77944046e-01 3.22958648e-01 7.12257802e-01 -8.21985781e-01
-7.77985632e-01 -9.27720070e-01 4.71353531e-01 -4.96333361e-01
2.07854450e-01 -3.35292697e-01 -1.04691839e+00 6.96211219e-01
3.50772381e-01 6.08968846e-02 3.61362278e-01 -5.67771971e-01
4.87002939e-01 -8.60977992e-02 -1.04239416e+00 3.41270924e-01
7.41650403e-01 -4.89274645e-03 -8.55075777e-01 4.02162105e-01
5.07337689e-01 -5.10907412e-01 -4.00578141e-01 4.97253567e-01
4.21653455e-03 -1.08971679e+00 8.22013617e-01 -5.08503616e-01
1.46333009e-01 -7.38202572e-01 1.63894460e-01 -7.77957559e-01
2.84630179e-01 -7.10455239e-01 -1.07192714e-02 1.25763738e+00
2.55867451e-01 -2.21635103e-01 1.27003181e+00 2.95582712e-01
1.33880615e-01 -2.89564937e-01 -6.87667012e-01 -6.38348162e-01
-4.59388226e-01 -1.45989195e-01 -2.27528792e-02 8.59136641e-01
-6.37667179e-01 2.55168527e-01 -2.73166120e-01 2.86704332e-01
8.30395758e-01 -1.29601449e-01 9.39403176e-01 -1.30479252e+00
-1.65890753e-01 -2.60126710e-01 -7.26079524e-01 -9.38966691e-01
2.26981983e-01 -4.67750818e-01 4.33214933e-01 -1.32339203e+00
1.18960038e-01 -2.46815294e-01 -1.64427757e-01 8.14726129e-02
-2.93894231e-01 3.59054059e-01 9.04113129e-02 1.07391812e-01
-8.61337483e-01 -1.07144520e-01 9.69059646e-01 -2.08254650e-01
-3.78469348e-01 4.75462466e-01 -1.43252022e-03 1.32849538e+00
7.20798492e-01 -5.41413844e-01 -2.86633223e-01 -3.37860100e-02
-5.06791353e-01 4.82945843e-03 3.26954514e-01 -1.38343930e+00
3.55352551e-01 -1.49956301e-01 7.07127512e-01 -9.09995556e-01
1.35270149e-01 -1.17986679e+00 2.50964344e-01 5.09724736e-01
-7.04547465e-02 -4.34143543e-02 1.17378727e-01 5.00886858e-01
-4.04920459e-01 -7.68163800e-01 9.57425892e-01 -6.10930026e-01
-1.26847351e+00 2.56803399e-03 -6.88723207e-01 -1.53670326e-01
1.19183373e+00 -5.37772834e-01 2.11604387e-01 -1.43335879e-01
-1.05091655e+00 -1.40572593e-01 4.28203374e-01 -2.59648971e-02
4.34026271e-01 -8.83309424e-01 -3.31552804e-01 2.72524953e-01
-4.05584067e-01 3.15482378e-01 -1.51505604e-01 6.95759952e-01
-1.18170512e+00 1.09925352e-01 -4.08913940e-01 -8.93636286e-01
-1.56352544e+00 9.52222764e-01 4.89004850e-01 -6.34673834e-02
-6.68643475e-01 7.75557935e-01 8.06340352e-02 1.17046654e-01
6.05264425e-01 -5.69348574e-01 -4.24097627e-01 1.08533725e-01
6.24886572e-01 3.52833658e-01 -4.78020608e-02 -7.75903285e-01
-1.44000784e-01 5.68580270e-01 1.21071488e-01 -8.27779174e-02
1.19612288e+00 7.32737109e-02 -2.64939636e-01 6.35482967e-01
8.40815008e-01 5.03542321e-03 -1.44260490e+00 -1.66291982e-01
4.29920077e-01 -4.21052843e-01 -2.80914038e-01 -5.51649451e-01
-1.40175319e+00 6.99733377e-01 7.84210742e-01 4.86071676e-01
1.31775296e+00 -1.50546774e-01 6.32375181e-01 6.05680168e-01
3.18033129e-01 -1.29518008e+00 1.50565445e-01 1.88166216e-01
2.36453161e-01 -1.25875652e+00 3.58676165e-02 -5.43537915e-01
-6.02331221e-01 1.30036104e+00 7.45936930e-01 -2.77334929e-01
5.20916522e-01 4.27161694e-01 4.05225694e-01 -1.29887581e-01
-2.87620544e-01 -4.78408873e-01 2.61777341e-01 6.21630251e-01
1.04052119e-01 -3.11125278e-01 -2.21207127e-01 2.64240682e-01
2.33146772e-01 1.53477028e-01 5.54900944e-01 1.17542386e+00
-7.54212677e-01 -1.09463501e+00 -5.58926761e-01 2.93705434e-01
-7.65928686e-01 1.71937212e-01 -5.81252754e-01 9.74441648e-01
6.73049092e-01 8.76554549e-01 7.22003207e-02 -3.89336050e-02
1.57207951e-01 2.47557402e-01 5.71094096e-01 -6.93367183e-01
-6.54078662e-01 5.05438209e-01 8.29359889e-02 -3.77617985e-01
-1.04683471e+00 -8.05273950e-01 -1.20950007e+00 1.63743719e-01
-5.31526506e-01 2.34794721e-01 4.67134386e-01 1.19396937e+00
-3.79212201e-01 5.14450610e-01 3.51385057e-01 -1.47886646e+00
1.32867828e-01 -7.79514968e-01 -4.06415075e-01 5.37851214e-01
3.63182157e-01 -4.65498120e-01 -4.63764936e-01 9.42978859e-01] | [9.01090145111084, -0.4129634499549866] |
7347a3de-84dd-4211-8e86-1eb987e1f5c7 | self-supervised-auxiliary-learning-with-meta | 2007.08294 | null | https://arxiv.org/abs/2007.08294v5 | https://arxiv.org/pdf/2007.08294v5.pdf | Self-supervised Auxiliary Learning with Meta-paths for Heterogeneous Graphs | Graph neural networks have shown superior performance in a wide range of applications providing a powerful representation of graph-structured data. Recent works show that the representation can be further improved by auxiliary tasks. However, the auxiliary tasks for heterogeneous graphs, which contain rich semantic information with various types of nodes and edges, have less explored in the literature. In this paper, to learn graph neural networks on heterogeneous graphs we propose a novel self-supervised auxiliary learning method using meta-paths, which are composite relations of multiple edge types. Our proposed method is learning to learn a primary task by predicting meta-paths as auxiliary tasks. This can be viewed as a type of meta-learning. The proposed method can identify an effective combination of auxiliary tasks and automatically balance them to improve the primary task. Our methods can be applied to any graph neural networks in a plug-in manner without manual labeling or additional data. The experiments demonstrate that the proposed method consistently improves the performance of link prediction and node classification on heterogeneous graphs. | ['Jung-Woo Ha', 'Kyung-Min Kim', 'Jinyoung Park', 'Dasol Hwang', 'Hyunwoo J. Kim', 'Sunyoung Kwon'] | 2020-07-16 | null | http://proceedings.neurips.cc/paper/2020/hash/74de5f915765ea59816e770a8e686f38-Abstract.html | http://proceedings.neurips.cc/paper/2020/file/74de5f915765ea59816e770a8e686f38-Paper.pdf | neurips-2020-12 | ['auxiliary-learning'] | ['methodology'] | [ 3.12850177e-01 6.13891840e-01 -6.63535714e-01 -3.30515593e-01
-2.75196612e-01 -2.59541124e-01 5.55040359e-01 4.33385700e-01
-6.06115833e-02 8.05255175e-01 6.64499700e-02 -4.08520728e-01
-1.59257159e-01 -1.13747406e+00 -6.15635812e-01 -5.56512773e-01
-1.65456310e-01 6.11855745e-01 4.15496558e-01 -4.30767924e-01
-1.32496372e-01 3.43211085e-01 -1.34960544e+00 2.52713978e-01
9.73357081e-01 8.64350796e-01 2.15108529e-01 4.63313907e-01
-4.80673611e-01 1.22417378e+00 -3.74956369e-01 -5.31741858e-01
4.78264727e-02 -3.34236145e-01 -9.36411917e-01 -1.88746989e-01
3.18578899e-01 2.11732104e-01 -4.35279965e-01 1.10596097e+00
4.19489056e-01 7.37377480e-02 5.55882692e-01 -1.72842193e+00
-5.64501941e-01 9.70720530e-01 -5.13250113e-01 2.16727749e-01
1.50827259e-01 -5.05048037e-01 1.29681385e+00 -5.15514195e-01
6.99436963e-01 1.16565204e+00 8.73695195e-01 3.82822275e-01
-1.03292465e+00 -8.01448345e-01 3.73676300e-01 3.84283870e-01
-1.18828821e+00 -2.16248423e-01 1.24722743e+00 -2.29919091e-01
9.30717766e-01 7.29761869e-02 5.64766943e-01 9.58852410e-01
-2.06703064e-03 7.93253303e-01 7.90164769e-01 -3.76685053e-01
-1.68321729e-01 3.58132482e-01 5.38844705e-01 1.13815844e+00
4.90625769e-01 -2.19736665e-01 -2.55683273e-01 -1.46095753e-01
5.66012561e-01 3.28613892e-02 -2.46312067e-01 -4.01000619e-01
-9.28092182e-01 8.23995352e-01 8.09978187e-01 3.00822824e-01
-1.86611220e-01 1.17758572e-01 6.27403736e-01 5.02986193e-01
7.55029738e-01 2.44761765e-01 -6.11832440e-01 4.02981669e-01
-3.09183240e-01 -2.13289380e-01 1.08263099e+00 1.26142514e+00
9.79939282e-01 1.25036776e-01 -2.03960970e-01 1.00408411e+00
3.54475439e-01 7.28837028e-02 2.55318910e-01 -4.90437090e-01
7.94114351e-01 1.11144745e+00 -4.39926147e-01 -1.26259172e+00
-7.37389743e-01 -8.78874183e-01 -1.11068165e+00 -9.58000943e-02
5.69150411e-02 -3.56448442e-02 -9.11860585e-01 1.85562503e+00
3.94330978e-01 3.55688870e-01 -5.12381308e-02 3.25900733e-01
1.38696563e+00 5.74325502e-01 2.99415171e-01 -1.23635843e-01
1.10027051e+00 -1.55505705e+00 -8.12579036e-01 -2.31876984e-01
9.96398687e-01 -3.69843662e-01 8.09313834e-01 -8.96571949e-02
-6.88615084e-01 -6.18651628e-01 -1.06055558e+00 -6.29700199e-02
-6.86997056e-01 -2.23829634e-02 1.08339727e+00 3.88847798e-01
-1.28135085e+00 7.19222188e-01 -5.60673118e-01 -4.47910458e-01
4.88349766e-01 4.68589872e-01 -6.03324950e-01 1.13466047e-01
-1.19023824e+00 7.12061942e-01 9.26039040e-01 -3.82156931e-02
-4.70603138e-01 -5.20147681e-01 -1.11564243e+00 3.97664189e-01
7.12921619e-01 -8.99026930e-01 8.73973072e-01 -1.07450700e+00
-1.08232510e+00 8.17423105e-01 6.59965351e-02 -2.27055073e-01
1.24185637e-01 2.04925045e-01 -6.49376214e-01 1.10513017e-01
1.79676518e-01 4.34043467e-01 7.31886148e-01 -1.25661242e+00
-7.63394296e-01 -5.62696099e-01 2.16697842e-01 1.83935851e-01
-7.59566426e-01 -4.81671900e-01 -7.09692061e-01 -8.67412806e-01
1.34860083e-01 -9.62748110e-01 -3.03185642e-01 -2.81062007e-01
-5.78802288e-01 -5.21427393e-01 9.83770967e-01 -5.64338624e-01
1.39619815e+00 -1.64480376e+00 1.77518234e-01 2.41555169e-01
6.76560521e-01 3.70014340e-01 -5.22026837e-01 4.35671389e-01
-2.00316310e-01 1.43215939e-01 -2.82020085e-02 -2.65373796e-01
-1.72102317e-01 1.11873381e-01 2.26320565e-01 -7.02028582e-03
6.41906559e-02 1.19099712e+00 -1.01135194e+00 -6.85072541e-01
-1.23700626e-01 3.02695870e-01 -2.93351531e-01 3.89571577e-01
-1.90154076e-01 1.95966154e-01 -5.88989377e-01 6.98972702e-01
5.28567374e-01 -6.74302757e-01 4.90550876e-01 -3.32370728e-01
6.96389437e-01 1.58218801e-01 -9.44763541e-01 1.59417009e+00
-5.45933545e-01 4.82701182e-01 3.13941300e-01 -1.66149783e+00
1.15393293e+00 1.65219635e-01 5.30086040e-01 -4.48770672e-01
1.57211006e-01 3.99260409e-02 1.45352587e-01 -4.99979407e-01
3.62696677e-01 1.45725310e-01 1.26264662e-01 3.74920547e-01
3.68860394e-01 3.04026008e-01 4.81189132e-01 4.39775467e-01
1.32907045e+00 1.36869550e-01 5.41867256e-01 -2.05657616e-01
7.74722576e-01 -1.82898827e-02 4.97508615e-01 7.32735276e-01
8.65849182e-02 -4.96666320e-02 4.84702826e-01 -4.54694659e-01
-7.16918647e-01 -7.87788391e-01 3.23470533e-01 1.31666505e+00
1.10380389e-01 -6.67086005e-01 -3.74791294e-01 -1.17877853e+00
-3.47759090e-02 1.82419822e-01 -4.66547161e-01 -3.77161592e-01
-4.33050990e-01 -7.87425101e-01 3.63803178e-01 8.58867049e-01
5.30470967e-01 -1.11469126e+00 2.90521383e-01 2.22567677e-01
-4.60466277e-03 -1.30562794e+00 -2.21729010e-01 5.19905150e-01
-1.22252464e+00 -1.32174480e+00 -4.34059918e-01 -1.38177180e+00
9.19955909e-01 4.45164502e-01 1.42856777e+00 5.10660648e-01
8.39990973e-02 2.41272241e-01 -4.79020000e-01 -6.72013685e-02
-5.68586111e-01 6.97057366e-01 -3.79322052e-01 -1.25028089e-01
2.03534082e-01 -9.73260522e-01 -1.84990421e-01 1.67974249e-01
-6.85401320e-01 3.16768140e-01 7.21533537e-01 1.11330760e+00
3.72941375e-01 2.40660474e-01 8.62576127e-01 -1.71846974e+00
9.78713274e-01 -7.53429770e-01 -3.81637961e-01 4.44714010e-01
-9.51718271e-01 4.68384981e-01 8.98266554e-01 -2.17688709e-01
-1.00341320e+00 -8.11513737e-02 -1.24092907e-01 -3.03726315e-01
-8.94588977e-02 9.93679047e-01 -3.98994029e-01 -3.13044369e-01
4.55573082e-01 -6.54356033e-02 1.99437644e-02 -5.07261455e-01
3.82514715e-01 5.05951822e-01 2.65522242e-01 -6.29615426e-01
8.73534977e-01 1.85216308e-01 4.77980047e-01 -5.83418906e-01
-9.51619089e-01 -5.90527892e-01 -7.25115716e-01 -8.85629430e-02
5.20347297e-01 -7.47994065e-01 -6.17952943e-01 2.62168497e-01
-8.77040863e-01 -4.54605132e-01 1.79488227e-01 2.47076526e-01
-1.88099369e-01 5.12672603e-01 -7.29595661e-01 -4.41461116e-01
-5.51141024e-01 -9.48970318e-01 7.42078066e-01 1.37903824e-01
1.19880050e-01 -1.52278805e+00 4.19256203e-02 3.67777556e-01
3.24193656e-01 1.67459369e-01 1.26062715e+00 -1.10516572e+00
-5.99856198e-01 -1.64985523e-01 -4.81637716e-01 1.63813293e-01
4.64081734e-01 -3.65605712e-01 -7.02909768e-01 -3.93475533e-01
-6.41061246e-01 -3.80444646e-01 1.13304090e+00 9.11214575e-02
1.19363403e+00 -4.39412624e-01 -8.02809775e-01 7.83200979e-01
1.39131987e+00 2.28839487e-01 3.92186821e-01 3.68776917e-01
1.26262236e+00 6.28782749e-01 3.74791652e-01 -4.87205982e-02
5.80419004e-01 5.56623459e-01 4.98205990e-01 -3.67183775e-01
-3.41024786e-01 -3.68714213e-01 -2.14711651e-02 1.04797900e+00
-1.67491838e-01 -4.60598260e-01 -8.31087351e-01 3.77903908e-01
-2.15950799e+00 -8.15617979e-01 -1.81040496e-01 1.73480368e+00
4.35944140e-01 2.65420467e-01 4.62122522e-02 1.69010729e-01
1.04145110e+00 4.42865878e-01 -5.42823017e-01 -1.57767102e-01
1.75910383e-01 1.45519540e-01 2.37978265e-01 3.08898479e-01
-1.25695097e+00 9.75547552e-01 5.95905161e+00 7.91114986e-01
-8.66058052e-01 1.97279587e-01 4.70692635e-01 6.88338041e-01
-4.93772805e-01 -2.19701305e-02 -7.16705143e-01 1.13589004e-01
7.67165661e-01 -1.17573611e-01 2.46801570e-01 1.00304639e+00
-3.31393778e-01 4.76025015e-01 -1.20052814e+00 9.49539959e-01
1.50158539e-01 -1.42437685e+00 3.37415457e-01 -1.10156365e-01
7.79261947e-01 -8.90728831e-02 -2.84926504e-01 7.50253439e-01
4.15803224e-01 -8.79604876e-01 7.35632032e-02 2.09193230e-01
4.23135102e-01 -6.34499967e-01 9.17478204e-01 3.09155643e-01
-1.73713732e+00 -1.04200594e-01 -1.65234655e-01 -7.40600303e-02
-9.49341580e-02 4.58600283e-01 -8.47754121e-01 9.35998857e-01
4.20347303e-01 1.09928024e+00 -7.95032978e-01 8.03712666e-01
-2.92921990e-01 3.72394651e-01 1.14715158e-03 -2.06087202e-01
4.03368771e-02 -3.28592986e-01 3.76809359e-01 1.04207265e+00
1.91097200e-01 -1.53481260e-01 5.60092092e-01 5.44800758e-01
-5.06587505e-01 3.70630026e-01 -8.74764621e-01 -3.40920717e-01
5.01638055e-01 1.58141434e+00 -1.00719762e+00 -2.66980350e-01
-6.59205317e-01 7.00978994e-01 9.11668539e-01 4.00967568e-01
-6.44281685e-01 -5.26221633e-01 1.24846481e-01 7.29430690e-02
1.50714337e-03 1.03197806e-01 1.39681185e-02 -1.17786741e+00
8.07051361e-02 -4.99532104e-01 9.48746026e-01 -5.79467058e-01
-1.45739436e+00 9.36818302e-01 4.10557762e-02 -1.08949888e+00
-1.66716382e-01 -8.04168999e-01 -6.91768169e-01 4.02572542e-01
-1.70308197e+00 -1.58909941e+00 -7.93860018e-01 7.66257763e-01
2.36916885e-01 -6.10152006e-01 7.41709471e-01 4.50299531e-01
-5.81023812e-01 7.26627648e-01 -1.32411972e-01 4.22972709e-01
5.07050455e-01 -1.32532656e+00 2.42723450e-01 7.17371523e-01
2.59370804e-01 4.27286446e-01 3.48142534e-01 -7.56181657e-01
-1.32287323e+00 -1.51353884e+00 6.20082378e-01 5.17052934e-02
7.09636748e-01 -2.29674682e-01 -8.45206738e-01 9.05096948e-01
1.72871858e-01 3.17095548e-01 6.79757714e-01 6.73130572e-01
-4.12128329e-01 -4.22452271e-01 -8.14981699e-01 5.04038274e-01
1.45584810e+00 -6.37348890e-01 -4.13285404e-01 3.57773513e-01
9.55059111e-01 -1.69188917e-01 -9.69890296e-01 7.54303515e-01
1.67296454e-01 -6.29638433e-01 9.46264982e-01 -8.17664921e-01
4.45321590e-01 -1.06824964e-01 1.76151872e-01 -1.60738337e+00
-4.96893167e-01 -2.33465731e-01 -5.93489230e-01 1.31897366e+00
5.67357004e-01 -8.94180000e-01 1.07557857e+00 8.08283016e-02
-3.44403923e-01 -7.60143280e-01 -5.05000949e-01 -6.49176657e-01
-4.66609120e-01 1.74236298e-03 4.90611732e-01 1.24096084e+00
3.00909188e-02 9.64472830e-01 -2.20137209e-01 8.21783841e-02
6.98264599e-01 1.65858045e-01 7.92481303e-01 -1.82633543e+00
-2.75508732e-01 -4.82823879e-01 -5.58585346e-01 -7.93201208e-01
8.55033398e-01 -1.59580874e+00 -2.98470527e-01 -1.93521190e+00
3.62879336e-01 -8.54254782e-01 -5.86701751e-01 7.82059014e-01
-5.17385006e-01 2.55641434e-02 -8.76936913e-02 -4.21572775e-02
-5.95372379e-01 5.70965350e-01 1.27137923e+00 -7.07392514e-01
-2.45610416e-01 2.03334734e-01 -8.66173804e-01 6.90503359e-01
8.60946834e-01 -4.85155702e-01 -1.00026882e+00 -1.58036709e-01
4.33584303e-01 2.35915259e-01 8.71310607e-02 -8.59591842e-01
2.45118067e-01 1.27601102e-01 9.20769721e-02 -4.70421702e-01
2.49726787e-01 -1.05304027e+00 2.64800727e-01 4.23090488e-01
-3.47927332e-01 1.15469396e-01 8.67771208e-02 8.59159529e-01
-2.96118438e-01 -2.86153823e-01 3.65452200e-01 -3.51987302e-01
-9.75543380e-01 6.47478938e-01 2.02395022e-01 1.62587985e-02
9.19000447e-01 -6.99644983e-02 -6.04966402e-01 -4.35949057e-01
-9.69008505e-01 3.28513682e-01 6.34729713e-02 7.13974535e-01
4.66072381e-01 -1.45268917e+00 -5.48687041e-01 -1.47306025e-01
3.02238584e-01 -1.19570017e-01 -8.27366635e-02 6.67361736e-01
-3.78199548e-01 2.11654469e-01 -2.98103422e-01 -4.69602138e-01
-1.37793887e+00 9.00340676e-01 1.49368346e-01 -7.97788322e-01
-6.50081992e-01 5.20620584e-01 9.49762166e-02 -6.62551641e-01
4.16008234e-01 1.60756204e-02 -6.90385044e-01 5.10647185e-02
1.02190360e-01 2.59895325e-01 1.91966236e-01 -5.24435520e-01
-1.31345823e-01 2.54466653e-01 -1.33393943e-01 5.54483712e-01
1.44655395e+00 -7.27063119e-02 -4.29092258e-01 3.09960783e-01
1.44458652e+00 -2.08509937e-01 -6.16370797e-01 -3.45088124e-01
3.15007776e-01 -4.06709686e-02 -4.28225324e-02 -3.67206156e-01
-1.62426043e+00 6.49474323e-01 1.97271541e-01 3.68302494e-01
1.21879220e+00 1.49215281e-01 5.83328903e-01 7.13830471e-01
3.86919886e-01 -9.20385122e-01 2.04957917e-01 3.62047762e-01
4.77647901e-01 -1.42263556e+00 9.86785861e-04 -1.00565600e+00
-2.37717703e-01 1.16525805e+00 1.05984998e+00 6.27198592e-02
9.15933549e-01 2.19516352e-01 -1.27135560e-01 -4.29027289e-01
-7.81740665e-01 -4.24904168e-01 4.23719436e-01 8.34554613e-01
5.28275430e-01 -1.78485662e-02 -4.63692576e-01 5.54009795e-01
1.63565367e-01 -2.24994659e-01 2.38694340e-01 8.09902787e-01
-2.73012608e-01 -1.37483656e+00 2.26216033e-01 8.47861946e-01
-1.79854915e-01 -2.17747539e-01 -3.32170248e-01 8.36612403e-01
-1.15482196e-01 9.11234260e-01 -1.15344249e-01 -6.18034244e-01
1.41038418e-01 2.74063610e-02 3.39945763e-01 -7.83113420e-01
-4.83730018e-01 -3.22253495e-01 6.26893103e-01 -4.09080505e-01
-7.47807920e-01 1.98963165e-01 -9.56747949e-01 -5.73796853e-02
-4.59137768e-01 5.30428946e-01 2.51339197e-01 9.17794347e-01
3.86330903e-01 9.36974645e-01 4.94868934e-01 -5.91766417e-01
-1.16316773e-01 -1.07254744e+00 -6.45204484e-01 6.38463438e-01
2.69680880e-02 -8.27667952e-01 -1.57018706e-01 -1.68537423e-01] | [7.353494167327881, 6.358311653137207] |
14581a94-0b91-4fe1-bd51-a71a50a748f0 | local-to-global-panorama-inpainting-for | 2303.10344 | null | https://arxiv.org/abs/2303.10344v1 | https://arxiv.org/pdf/2303.10344v1.pdf | Local-to-Global Panorama Inpainting for Locale-Aware Indoor Lighting Prediction | Predicting panoramic indoor lighting from a single perspective image is a fundamental but highly ill-posed problem in computer vision and graphics. To achieve locale-aware and robust prediction, this problem can be decomposed into three sub-tasks: depth-based image warping, panorama inpainting and high-dynamic-range (HDR) reconstruction, among which the success of panorama inpainting plays a key role. Recent methods mostly rely on convolutional neural networks (CNNs) to fill the missing contents in the warped panorama. However, they usually achieve suboptimal performance since the missing contents occupy a very large portion in the panoramic space while CNNs are plagued by limited receptive fields. The spatially-varying distortion in the spherical signals further increases the difficulty for conventional CNNs. To address these issues, we propose a local-to-global strategy for large-scale panorama inpainting. In our method, a depth-guided local inpainting is first applied on the warped panorama to fill small but dense holes. Then, a transformer-based network, dubbed PanoTransformer, is designed to hallucinate reasonable global structures in the large holes. To avoid distortion, we further employ cubemap projection in our design of PanoTransformer. The high-quality panorama recovered at any locale helps us to capture spatially-varying indoor illumination with physically-plausible global structures and fine details. | ['Yanwen Guo', 'Yan Zhang', 'Zhenyu Chen', 'Jie Guo', 'Shan Yang', 'Zhen He', 'Jiayang Bai'] | 2023-03-18 | null | null | null | null | ['hdr-reconstruction'] | ['computer-vision'] | [ 4.65433031e-01 -2.87499219e-01 8.51480812e-02 -2.26615340e-01
-5.97777009e-01 -2.90823877e-01 4.21626896e-01 -6.90388381e-01
2.54403859e-01 8.15748513e-01 4.62803066e-01 5.10543250e-02
-2.50471365e-02 -1.03537393e+00 -1.09850395e+00 -7.51115203e-01
7.19459295e-01 -1.42776161e-01 1.13218427e-01 -2.90421516e-01
-3.06145642e-02 7.35149145e-01 -1.43116307e+00 2.11045027e-01
1.12228513e+00 1.03485823e+00 6.13204718e-01 4.60149378e-01
1.16172507e-01 5.95022202e-01 -2.86951482e-01 -1.85974166e-01
6.60339415e-01 -5.66905618e-01 -6.99537992e-02 3.10755163e-01
7.62154460e-01 -6.94122851e-01 -7.36278772e-01 1.08418775e+00
1.68065980e-01 1.01195984e-01 2.24943012e-01 -7.35214591e-01
-5.03015280e-01 -9.25408825e-02 -8.36891174e-01 -3.74548763e-01
2.37910777e-01 1.29927993e-01 7.95865238e-01 -1.04744840e+00
5.86859524e-01 1.22300100e+00 6.62608564e-01 2.13847369e-01
-1.09366047e+00 -7.14751840e-01 -1.37434348e-01 -5.20485565e-02
-1.19470930e+00 -3.31532091e-01 1.35815489e+00 4.17787172e-02
3.59598964e-01 6.68317199e-01 8.14718306e-01 7.85644889e-01
6.53099179e-01 4.11143720e-01 1.27278340e+00 -2.56070822e-01
1.69401336e-02 -2.10582495e-01 -7.19414949e-01 5.96816003e-01
-9.96833742e-02 3.08433622e-01 -3.58156741e-01 3.38217430e-02
1.57474792e+00 6.19356215e-01 -8.73326957e-01 -4.26593781e-01
-1.28347099e+00 6.05923831e-01 7.62201190e-01 7.66072869e-02
-5.29646039e-01 9.70216244e-02 -6.61165267e-02 1.89031363e-01
5.24538577e-01 6.22935414e-01 -4.20989364e-01 2.38975167e-01
-1.02761459e+00 3.83098245e-01 1.48179933e-01 7.77841628e-01
9.84559119e-01 1.60424113e-01 1.76661670e-01 1.13108313e+00
1.27641678e-01 4.92492676e-01 2.98015922e-01 -1.13516247e+00
6.52465463e-01 5.36276102e-01 3.59370828e-01 -1.24967337e+00
-1.17303342e-01 -3.62384379e-01 -1.19085729e+00 3.81537557e-01
2.73217767e-01 7.50374049e-02 -9.41765904e-01 1.53246343e+00
3.43049586e-01 4.08498608e-02 -2.18382090e-01 1.07819510e+00
4.15076315e-01 1.19102609e+00 -5.61416209e-01 -4.52544868e-01
1.22119415e+00 -1.14986086e+00 -7.58287549e-01 -4.09016728e-01
-1.77046388e-01 -1.16169405e+00 1.21238184e+00 3.62274289e-01
-1.19395041e+00 -5.11724174e-01 -1.06342626e+00 -4.61872548e-01
2.35482007e-01 -1.53804854e-01 4.30233359e-01 1.06579773e-01
-8.58397305e-01 3.88827115e-01 -4.66652185e-01 -8.70407969e-02
3.16231430e-01 -1.06382772e-01 -3.84664714e-01 -8.25786412e-01
-1.06608772e+00 7.31247306e-01 -1.34080052e-01 1.40835658e-01
-7.67007530e-01 -6.98690116e-01 -9.12836373e-01 4.76480350e-02
3.20777059e-01 -7.14705944e-01 8.59989166e-01 -1.00621235e+00
-1.61722910e+00 4.42443132e-01 -1.72912955e-01 -1.30283654e-01
6.29533529e-01 -2.02040926e-01 -3.64182770e-01 4.80352454e-02
-2.71046571e-02 5.62323451e-01 1.16598678e+00 -1.50784767e+00
-4.04025793e-01 -2.53478020e-01 -1.71553642e-01 6.25110924e-01
-1.10076301e-01 -2.57640004e-01 -6.10595286e-01 -1.16520858e+00
4.95017409e-01 -6.18546724e-01 -4.08201545e-01 4.72293645e-01
-3.72108847e-01 5.15706003e-01 1.01716220e+00 -9.19978976e-01
8.86230648e-01 -1.94161689e+00 1.90605924e-01 -1.32125288e-01
1.63701147e-01 5.71602210e-02 -1.57018334e-01 2.52843529e-01
-5.88119216e-02 -4.63567674e-01 -5.83507955e-01 -4.10994262e-01
-5.29958487e-01 2.48428106e-01 -6.58515573e-01 4.07110304e-01
-5.91220520e-02 8.20995152e-01 -7.05061793e-01 -2.04156622e-01
6.44651294e-01 7.74230361e-01 -5.28455198e-01 5.09286880e-01
-3.46769065e-01 6.73032999e-01 -3.53533059e-01 8.52268219e-01
1.11715937e+00 1.14933737e-01 -2.92104751e-01 -3.74638081e-01
-4.29406524e-01 -5.83358072e-02 -8.87792528e-01 2.03739047e+00
-8.92999232e-01 6.03310227e-01 3.78453255e-01 -5.89332938e-01
1.14001012e+00 1.70701947e-02 4.72949296e-01 -9.26923633e-01
-5.23027452e-03 3.82126004e-01 -4.73087907e-01 -2.23939583e-01
6.79084957e-01 -4.15268809e-01 1.64528653e-01 1.87346458e-01
-4.99627739e-01 -7.43247330e-01 -6.57523751e-01 -3.35444659e-01
7.49608517e-01 1.84341297e-01 2.93837279e-01 8.30541551e-02
4.04223740e-01 -2.20840633e-01 6.99294448e-01 2.83021212e-01
2.31356412e-01 1.51069796e+00 -1.86599791e-03 -7.24030137e-01
-1.47169209e+00 -9.74733472e-01 -2.02553511e-01 5.12118936e-01
3.65873158e-01 -1.76663548e-01 -7.19051778e-01 -2.92232484e-01
-4.47398782e-01 5.55398941e-01 -4.73565519e-01 -7.19068274e-02
-9.17258322e-01 -5.68618953e-01 -1.06886469e-01 2.49544546e-01
9.80003595e-01 -1.09693289e+00 -7.41882682e-01 2.97063589e-01
-4.48406219e-01 -9.63455856e-01 -8.05106282e-01 -9.08272192e-02
-9.38592076e-01 -9.46883798e-01 -1.36245549e+00 -7.14496851e-01
5.52390397e-01 7.89805412e-01 8.89175594e-01 -3.73941183e-01
-1.20719112e-01 -1.29583746e-01 -8.24882835e-02 -3.95196900e-02
-1.40161619e-01 -4.60926086e-01 -3.74163181e-01 2.05532864e-01
-2.25239635e-01 -1.07453394e+00 -1.05106294e+00 6.03193521e-01
-1.15925264e+00 7.61038423e-01 7.02428281e-01 9.87517774e-01
7.66021430e-01 3.14863861e-01 4.09187470e-03 -4.75531787e-01
2.73611367e-01 -2.25292623e-01 -7.39776611e-01 6.32444173e-02
-2.72683024e-01 -3.45198005e-01 9.97741401e-01 -4.11972970e-01
-1.21732688e+00 1.08237565e-01 -1.67635888e-01 -9.51055884e-01
1.96945108e-02 8.29534456e-02 -5.78754902e-01 -2.21669048e-01
4.36812997e-01 6.74909174e-01 -8.78450572e-02 -4.81773973e-01
3.10897350e-01 2.81049192e-01 6.25015438e-01 -2.94942111e-01
8.90571594e-01 7.69968987e-01 1.35614485e-01 -8.85170996e-01
-7.34835029e-01 3.83601785e-02 -1.84721783e-01 -1.26963630e-01
8.34599555e-01 -1.05495536e+00 -1.32498339e-01 8.35450351e-01
-1.25765204e+00 -4.96864200e-01 -3.53973269e-01 2.18199477e-01
-4.79745686e-01 3.63973945e-01 -5.48401654e-01 -4.38776761e-01
-2.62716502e-01 -1.20379925e+00 1.07992887e+00 3.32155496e-01
1.60412431e-01 -6.91921532e-01 8.76996815e-02 4.97021407e-01
5.38295031e-01 3.31230402e-01 7.87758350e-01 5.92455626e-01
-1.08471966e+00 -5.79800904e-02 -4.82676715e-01 4.56418008e-01
3.95925790e-01 -3.14224899e-01 -9.89846170e-01 -1.07067041e-01
5.09858608e-01 -2.33397022e-01 8.36018682e-01 6.23316884e-01
1.32520676e+00 -4.00730282e-01 -3.49860406e-04 1.26033247e+00
1.52635098e+00 2.66384423e-01 9.54838693e-01 2.92403966e-01
1.17492127e+00 6.71866655e-01 7.07136273e-01 3.40841323e-01
3.21931630e-01 8.60069573e-01 6.19043469e-01 -2.95430243e-01
-3.14203233e-01 -7.62098670e-01 1.98855937e-01 7.48447001e-01
6.60511777e-02 -3.97516429e-01 -4.83752221e-01 3.78274381e-01
-1.62892056e+00 -5.04755735e-01 1.29749000e-01 2.35844374e+00
8.85313511e-01 -1.65558860e-01 -5.30992627e-01 1.03522744e-02
6.69915080e-01 7.25995719e-01 -6.50423169e-01 -1.20816521e-01
-3.72421473e-01 1.21667981e-01 3.68180186e-01 7.88299859e-01
-6.82126522e-01 8.83382678e-01 5.07437897e+00 1.19078708e+00
-1.43065870e+00 6.90499768e-02 8.77659857e-01 1.33359909e-01
-6.95293248e-01 5.44826873e-02 -3.42061192e-01 4.99290377e-01
2.77693987e-01 3.89937788e-01 6.28630817e-01 6.79766536e-01
3.06590885e-01 -3.53142411e-01 -5.51425815e-01 1.35361874e+00
2.52215505e-01 -1.37843168e+00 1.65341258e-01 3.42293009e-02
1.15055406e+00 -1.78102851e-01 2.06095025e-01 -3.38895649e-01
-1.83242764e-02 -1.13061845e+00 5.68847537e-01 4.90245223e-01
1.24858773e+00 -8.57384205e-01 2.00242251e-01 2.96567410e-01
-1.17367029e+00 -9.46408957e-02 -6.32507503e-01 8.36062152e-03
4.36672896e-01 8.93130720e-01 -9.04834121e-02 4.08156455e-01
6.64419115e-01 7.84821451e-01 -1.09426066e-01 9.54939425e-01
-3.09509605e-01 1.18788287e-01 -3.58569741e-01 6.20941341e-01
1.11716002e-01 -5.66212475e-01 5.98381042e-01 5.47094047e-01
7.19171643e-01 2.58838296e-01 -1.41202837e-01 1.05512869e+00
-2.51519740e-01 -6.47446737e-02 -6.71523929e-01 5.57467461e-01
3.14844847e-01 1.25414586e+00 -4.50530291e-01 -1.75934091e-01
-4.21979696e-01 1.29281127e+00 -4.51917686e-02 4.93420124e-01
-7.13109255e-01 -2.40377098e-01 6.65637374e-01 4.00799811e-01
1.45870224e-01 -1.89634606e-01 -5.22628307e-01 -1.41372716e+00
2.56451014e-02 -9.96809721e-01 -1.81875065e-01 -1.26373315e+00
-9.59719360e-01 6.13728225e-01 -2.91963786e-01 -1.54226875e+00
2.33002737e-01 -1.63845703e-01 -8.63086939e-01 9.96076047e-01
-1.63000703e+00 -1.43535781e+00 -7.74268031e-01 8.25497627e-01
9.20155764e-01 2.28404343e-01 6.13513768e-01 3.05663228e-01
-3.30020875e-01 2.64979154e-01 2.69153833e-01 -3.60229969e-01
8.87107253e-01 -7.58062780e-01 2.49095306e-01 1.04578853e+00
-1.38532639e-01 3.13195497e-01 7.55765378e-01 -6.57236516e-01
-1.28843462e+00 -1.35226774e+00 6.14120066e-01 -6.62019923e-02
2.99424380e-02 -3.10282588e-01 -9.36784863e-01 4.91696388e-01
1.07599653e-01 3.17057908e-01 -1.72431260e-01 -5.67771077e-01
-3.69925410e-01 -3.37265283e-01 -1.18029308e+00 8.44747245e-01
8.07322741e-01 -4.63819921e-01 -3.44940990e-01 2.32612357e-01
8.03858280e-01 -6.55273020e-01 -4.32669014e-01 3.90834063e-01
5.92510104e-01 -1.43583846e+00 1.11912954e+00 2.75580347e-01
1.12994039e+00 -4.07657266e-01 -2.40477026e-01 -1.35574532e+00
-7.87594616e-02 -8.58044088e-01 -2.03661621e-02 9.88428414e-01
-1.41610250e-01 -6.52588069e-01 8.23712528e-01 3.53156507e-01
-3.85468960e-01 -9.02742624e-01 -7.03901231e-01 -2.11725220e-01
-5.56020327e-02 -2.42202610e-01 7.47132659e-01 1.06233597e+00
-6.75685108e-01 7.89094046e-02 -9.79892612e-01 1.00030921e-01
6.43222928e-01 5.26775837e-01 7.94261098e-01 -7.16118932e-01
-2.66151905e-01 -1.77486744e-02 9.66538712e-02 -1.55776083e+00
-3.92682970e-01 -1.91133931e-01 2.68469751e-01 -1.33916092e+00
8.31711367e-02 -5.71988702e-01 3.71029451e-02 1.74254954e-01
-2.04974599e-02 6.18086994e-01 3.69991083e-03 4.63187486e-01
6.84290901e-02 1.01376081e+00 1.88452482e+00 1.96617514e-01
-2.90062129e-01 -2.46905815e-02 -4.20468152e-01 8.62042367e-01
6.02551401e-01 -1.74560815e-01 -4.63423669e-01 -6.50039196e-01
1.86492279e-01 5.55911005e-01 3.76485556e-01 -1.12550974e+00
1.33419782e-01 -4.63208705e-01 7.25449562e-01 -7.18022585e-01
8.42641532e-01 -9.36276734e-01 4.11057919e-01 1.63980573e-01
4.55222139e-03 2.07332671e-02 -1.95211485e-01 5.35105288e-01
-5.72549641e-01 4.19117749e-01 1.14592111e+00 -3.20214689e-01
-4.11101073e-01 7.11861372e-01 1.46843595e-02 -3.05821747e-01
7.97704339e-01 -2.72286832e-01 -7.17618391e-02 -6.01244926e-01
-1.75225455e-03 -1.47689581e-01 1.06377292e+00 2.92169333e-01
9.91938531e-01 -1.37063527e+00 -3.90694112e-01 6.96394920e-01
-6.75652400e-02 5.86799860e-01 7.34294951e-01 6.06902540e-01
-1.01440394e+00 2.54665494e-01 -4.92998838e-01 -3.37248266e-01
-1.02357781e+00 3.41204673e-01 3.75953853e-01 -1.22061320e-01
-8.88313770e-01 9.01555955e-01 8.15191329e-01 -4.09370482e-01
-7.68213049e-02 -3.42625171e-01 4.16669957e-02 -2.45173290e-01
4.82970268e-01 1.49156258e-01 -1.60717323e-01 -6.74868166e-01
1.30200922e-01 9.91871953e-01 4.85925265e-02 -1.17074167e-02
1.47573161e+00 -3.96034330e-01 -3.30812633e-01 1.45576194e-01
1.16872919e+00 4.08934206e-01 -1.93131411e+00 -3.25674564e-01
-9.39206958e-01 -9.93786871e-01 2.73106337e-01 -4.58452374e-01
-1.31935477e+00 1.08014297e+00 4.24389482e-01 -3.83282363e-01
1.49143755e+00 -3.53345037e-01 1.44212794e+00 -3.74508537e-02
3.32583696e-01 -7.22330391e-01 1.67000532e-01 3.69128615e-01
1.23969018e+00 -1.08367980e+00 1.35207236e-01 -5.78959763e-01
-3.98300827e-01 1.25359130e+00 7.79256999e-01 -3.02461237e-01
3.65969509e-01 -7.71734118e-02 -2.58328323e-03 2.46320094e-04
-3.29528719e-01 5.09727597e-01 2.70112157e-01 3.95125419e-01
-8.07724148e-02 -3.71320769e-02 7.91648328e-02 2.73554444e-01
-2.64234304e-01 -2.50517458e-01 5.22665203e-01 5.19117951e-01
-4.41360950e-01 -8.04222465e-01 -7.93976784e-01 1.26864418e-01
-7.25436211e-02 -1.75208300e-01 2.75325160e-02 5.97254038e-01
4.07301426e-01 5.68741739e-01 5.98593848e-03 -2.99557775e-01
2.05453783e-01 -5.29210150e-01 3.51190805e-01 -3.54462892e-01
-3.67733128e-02 4.54172671e-01 -2.74677217e-01 -9.37837005e-01
-1.52606413e-01 -2.47145906e-01 -6.32172644e-01 -3.14625293e-01
-9.48344395e-02 -3.14294517e-01 4.04456556e-01 6.68379784e-01
7.28692114e-02 3.86758715e-01 8.60007346e-01 -1.18504667e+00
-2.87215021e-02 -7.26068735e-01 -7.03776062e-01 1.94672540e-01
6.13642097e-01 -3.71085286e-01 -4.54748333e-01 -2.42899194e-01] | [10.325384140014648, -2.20694637298584] |
5e11fdc3-f0a9-41b6-af94-d92ec2e9e585 | data-augmentation-using-random-image-cropping-1 | 2111.08270 | null | https://arxiv.org/abs/2111.08270v1 | https://arxiv.org/pdf/2111.08270v1.pdf | Data Augmentation using Random Image Cropping for High-resolution Virtual Try-On (VITON-CROP) | Image-based virtual try-on provides the capacity to transfer a clothing item onto a photo of a given person, which is usually accomplished by warping the item to a given human pose and adjusting the warped item to the person. However, the results of real-world synthetic images (e.g., selfies) from the previous method is not realistic because of the limitations which result in the neck being misrepresented and significant changes to the style of the garment. To address these challenges, we propose a novel method to solve this unique issue, called VITON-CROP. VITON-CROP synthesizes images more robustly when integrated with random crop augmentation compared to the existing state-of-the-art virtual try-on models. In the experiments, we demonstrate that VITON-CROP is superior to VITON-HD both qualitatively and quantitatively. | ['Jaegul Choo', 'Seunghwan Choi', 'Sunghyun Park', 'Taewon Kang'] | 2021-11-16 | null | null | null | null | ['image-cropping'] | ['computer-vision'] | [ 3.50223809e-01 -9.79976635e-03 1.89658418e-01 -1.89647257e-01
-1.93806469e-01 -6.58415377e-01 4.37183738e-01 -6.80613577e-01
1.11926533e-02 7.27852762e-01 7.20247701e-02 3.04611564e-01
5.89081049e-01 -6.41234636e-01 -1.00970018e+00 -3.07429105e-01
5.48558950e-01 1.66951060e-01 2.57165045e-01 -5.03176451e-01
6.39038756e-02 3.19425583e-01 -1.32631016e+00 2.42391005e-01
6.33049428e-01 7.08205462e-01 2.64748394e-01 4.41526920e-01
1.24470279e-01 1.77677810e-01 -5.49117088e-01 -7.57894337e-01
5.71856499e-01 -7.31911421e-01 -3.51942986e-01 7.32681632e-01
7.88895845e-01 -5.23706257e-01 -1.85185656e-01 1.13018298e+00
2.88173586e-01 4.93049659e-02 4.96462822e-01 -1.24464023e+00
-1.14689863e+00 1.20408185e-01 -8.90174508e-01 -4.54492539e-01
7.57334292e-01 4.85437065e-02 3.21990460e-01 -1.18427706e+00
1.07294643e+00 1.37066185e+00 7.26292849e-01 6.82725191e-01
-1.36391723e+00 -7.27044106e-01 2.13806152e-01 -1.20354779e-01
-1.35977864e+00 -2.67525584e-01 9.65182304e-01 -3.42082173e-01
1.91732064e-01 3.47968787e-01 8.77345502e-01 1.28928292e+00
2.79312402e-01 6.56760454e-01 1.38222897e+00 -6.25179529e-01
4.13314328e-02 2.09979340e-01 -2.45970398e-01 7.06702471e-01
1.82006359e-01 1.50123924e-01 -5.15925646e-01 1.61699355e-01
1.31101120e+00 5.21187447e-02 -2.90701032e-01 -6.79117739e-01
-1.32219839e+00 3.91461343e-01 5.43675482e-01 -1.98140919e-01
-4.61325645e-01 1.87252223e-01 1.06395721e-01 5.73064201e-02
4.70761329e-01 3.68488967e-01 -1.13502137e-01 3.08139324e-01
-1.02318001e+00 4.44292188e-01 3.61288100e-01 1.19709694e+00
7.16834605e-01 3.51900578e-01 -8.51316750e-02 7.33672202e-01
3.16086739e-01 7.60875881e-01 -2.72420322e-04 -8.99572551e-01
3.33005726e-01 4.34672266e-01 4.47561562e-01 -9.86073732e-01
-9.69497412e-02 -7.93213174e-02 -7.12993503e-01 5.60569227e-01
3.23672116e-01 -2.04706386e-01 -1.58221102e+00 1.58452260e+00
4.63336736e-01 1.14160232e-01 -2.72005111e-01 1.36701155e+00
9.08336818e-01 5.92276037e-01 6.05776533e-02 2.81144921e-02
1.48170805e+00 -1.13304222e+00 -1.05260658e+00 -4.25139397e-01
-1.73304468e-01 -1.19341946e+00 1.24582946e+00 2.40281805e-01
-1.21864235e+00 -9.75453436e-01 -1.27633810e+00 -4.32836488e-02
-2.31112823e-01 1.26864508e-01 5.12442946e-01 8.36932123e-01
-8.64279687e-01 3.04775983e-01 -5.57230949e-01 -5.67814171e-01
1.13354139e-01 6.31391928e-02 -6.15591943e-01 -2.59926945e-01
-1.10273361e+00 9.14862573e-01 1.40652567e-01 3.31540674e-01
-7.43609667e-01 -6.18709862e-01 -1.15295613e+00 -3.84802341e-01
4.74534214e-01 -7.83296168e-01 1.02947569e+00 -1.41446579e+00
-1.71480370e+00 9.06489193e-01 -3.88964266e-02 1.53690875e-01
7.76956379e-01 -2.45605290e-01 -4.26221371e-01 -7.33011216e-02
1.05309084e-01 7.35657871e-01 1.02823520e+00 -1.98030019e+00
5.41411527e-02 -2.77157784e-01 1.39282465e-01 2.07370043e-01
2.52789799e-02 1.71785578e-01 -9.43555892e-01 -1.15835238e+00
2.08354965e-01 -1.23366845e+00 -1.05728485e-01 4.21187401e-01
-7.39636362e-01 6.05450571e-01 9.49742734e-01 -9.99708116e-01
8.26556444e-01 -2.03636026e+00 2.64167577e-01 4.26619276e-02
-8.87151286e-02 4.95258778e-01 -1.91438809e-01 5.15121639e-01
-1.38595954e-01 -4.58281748e-02 -2.06256524e-01 -6.27296507e-01
-2.27228165e-01 3.12112659e-01 -1.95268884e-01 4.55018073e-01
4.69955578e-02 9.99867857e-01 -7.27315545e-01 -4.33291078e-01
3.99502486e-01 5.63203990e-01 -3.03370029e-01 2.58881330e-01
-1.17380694e-01 5.77665865e-01 -1.41868889e-01 8.86686146e-01
1.14356387e+00 8.73881578e-02 3.03389996e-01 -4.71307576e-01
1.04328774e-01 -3.96060854e-01 -1.35440147e+00 1.72889984e+00
-1.90811440e-01 3.97761792e-01 -1.47227654e-02 -2.40023896e-01
9.10324156e-01 2.73070633e-01 1.56799436e-01 -6.40821695e-01
7.97781721e-02 -1.78182885e-01 -4.48432803e-01 -4.48044330e-01
9.51392949e-01 -3.56120914e-01 -1.17618866e-01 3.67389977e-01
-5.21557108e-02 -2.19288811e-01 -7.58917332e-02 -7.58850574e-02
2.21397653e-01 9.03959274e-01 2.15028137e-01 -4.64417823e-02
1.83098570e-01 1.62997231e-01 5.42929471e-01 2.61291742e-01
-1.22340985e-01 1.30537307e+00 -1.03366533e-02 -2.78191477e-01
-1.50564969e+00 -1.22964096e+00 8.99644345e-02 6.37137830e-01
6.30277872e-01 -1.64136305e-01 -1.14074016e+00 -6.69812977e-01
5.65809235e-02 6.20318651e-01 -8.38680804e-01 1.91913825e-02
-4.73564833e-01 -3.61241817e-01 3.74159724e-01 7.57840157e-01
6.96576357e-01 -1.07513881e+00 -3.85897279e-01 1.26473427e-01
-3.15116525e-01 -1.34460187e+00 -9.34351981e-01 -6.14201188e-01
-4.50899392e-01 -8.48839521e-01 -1.20210409e+00 -9.09722090e-01
1.13267362e+00 5.29883742e-01 1.03300929e+00 -1.35413618e-04
-8.72731507e-02 2.77218461e-01 -4.32227910e-01 -4.93755251e-01
-4.93020236e-01 -3.18054020e-01 1.88035384e-01 4.23310965e-01
1.70843918e-02 -1.77634969e-01 -7.86792517e-01 7.30154395e-01
-1.00975120e+00 5.93706071e-01 2.60496438e-01 6.54955566e-01
7.41572320e-01 -2.29181275e-01 2.54700810e-01 -7.87261784e-01
3.76559228e-01 1.07714131e-01 -3.12160492e-01 4.31587875e-01
-2.14241937e-01 -3.46857190e-01 4.68233764e-01 -7.63497591e-01
-1.23181808e+00 2.92498380e-01 1.67151541e-01 -6.94833815e-01
5.34490235e-02 1.15591094e-01 -2.43551254e-01 -3.27184290e-01
4.13533568e-01 5.17585985e-02 1.10252857e-01 -5.03803015e-01
4.65831786e-01 4.38277930e-01 8.23706448e-01 -3.88612479e-01
1.19013715e+00 5.26421666e-01 -9.02183130e-02 -7.40293324e-01
-6.06231391e-01 2.83389576e-02 -9.31684911e-01 -5.87911129e-01
9.92021918e-01 -8.63301039e-01 -4.42874461e-01 7.60861754e-01
-1.06031227e+00 -3.79652321e-01 -2.16092810e-01 2.38176733e-01
-4.50554460e-01 4.37712073e-01 -5.27634561e-01 -5.96019983e-01
-1.26590297e-01 -1.11069453e+00 1.17436302e+00 3.71257961e-01
-3.61403972e-01 -8.41070712e-01 5.06707765e-02 5.25183618e-01
2.23522171e-01 6.86327398e-01 4.71814603e-01 2.91303247e-01
-4.58641917e-01 -2.58483648e-01 -1.85808241e-01 4.36357945e-01
2.96563625e-01 1.87249139e-01 -7.63494909e-01 -4.11063075e-01
-4.00461972e-01 -2.10365474e-01 4.11709428e-01 2.72460997e-01
6.78825200e-01 5.33491820e-02 -1.65544659e-01 6.38954103e-01
1.58070993e+00 4.89203669e-02 9.42854345e-01 3.16927522e-01
9.27092373e-01 4.07170266e-01 1.02593184e+00 8.03229734e-02
2.69430280e-01 1.08230686e+00 3.59530121e-01 -6.72267914e-01
-5.53619027e-01 -7.50815153e-01 2.62364507e-01 5.81956863e-01
-5.98489404e-01 -2.49318779e-01 -4.26878333e-01 4.16083157e-01
-1.72501099e+00 -8.10732245e-01 -1.92623511e-01 2.19785690e+00
5.75644910e-01 -2.13518232e-01 -2.71969717e-02 -9.68802944e-02
8.20485651e-01 9.93057191e-02 -3.48531216e-01 -5.12643576e-01
-1.93743244e-01 2.14425456e-02 5.78938305e-01 2.61467546e-01
-9.44622636e-01 1.04485893e+00 7.34427643e+00 3.13243568e-01
-1.21259594e+00 5.76979369e-02 3.38689119e-01 2.28945121e-01
-1.20124795e-01 -1.55597225e-01 -5.06864250e-01 4.11799937e-01
1.60185978e-01 1.14038577e-02 5.49683988e-01 7.31430769e-01
2.27875143e-01 -6.82589132e-03 -8.67205799e-01 7.91126966e-01
5.86763144e-01 -1.01049268e+00 1.61202013e-01 5.02512231e-02
1.04372025e+00 -7.10883915e-01 2.00403348e-01 6.64755628e-02
8.13950747e-02 -8.38296294e-01 1.28113949e+00 5.72503448e-01
1.22193861e+00 -4.82157499e-01 6.01122618e-01 -2.99605131e-02
-1.16610110e+00 4.62958425e-01 -3.72212529e-01 6.85959607e-02
5.22551477e-01 5.11985412e-03 -6.88338459e-01 7.74897814e-01
5.44830859e-01 3.11355084e-01 -5.49902499e-01 7.26116002e-01
-3.28302741e-01 3.06166172e-01 5.66215739e-02 3.66187483e-01
-8.92114714e-02 -4.01589602e-01 4.35897827e-01 9.77633536e-01
4.82544243e-01 7.60194287e-02 3.21397275e-01 9.98642981e-01
-9.12164897e-02 3.97859178e-02 -5.96833944e-01 1.96641207e-01
1.10534713e-01 1.16018760e+00 -6.75657511e-01 -3.64376634e-01
-4.33947444e-01 1.77174485e+00 -4.63304743e-02 6.16821051e-01
-1.02607381e+00 1.68313161e-02 6.02215528e-01 5.20677626e-01
4.01024044e-01 -3.86818975e-01 -2.22946346e-01 -1.13806641e+00
1.54831663e-01 -9.48774576e-01 -1.69375315e-01 -1.46043324e+00
-1.21232080e+00 6.40861571e-01 7.73482397e-02 -1.57249522e+00
-3.67113985e-02 -4.81117845e-01 -3.67356807e-01 1.03886068e+00
-1.18264759e+00 -1.91617894e+00 -7.04405367e-01 5.62210619e-01
6.88261092e-01 1.10238552e-01 7.86857247e-01 2.39744842e-01
-4.23213989e-01 7.28610814e-01 -2.48785451e-01 -1.71420202e-02
9.95250285e-01 -1.00345600e+00 8.01737547e-01 1.02545881e+00
-1.13827445e-01 5.12719870e-01 9.07176912e-01 -8.18249285e-01
-1.40826118e+00 -1.08764088e+00 4.73744214e-01 -5.61622679e-01
1.47235110e-01 -5.12506545e-01 -8.46516848e-01 9.93426919e-01
3.98416430e-01 9.97774825e-02 4.34486300e-01 -4.00457889e-01
-1.24050424e-01 7.76240602e-02 -1.26219797e+00 8.98783445e-01
1.10126066e+00 -3.02360088e-01 -5.13304472e-01 8.31006467e-02
4.57335025e-01 -9.05283988e-01 -8.29625189e-01 3.53871137e-01
9.38805163e-01 -8.50927353e-01 1.22944868e+00 -5.72416544e-01
6.05366290e-01 -6.09736443e-01 -2.11629882e-01 -1.62286556e+00
-5.22047997e-01 -5.81695795e-01 2.52930313e-01 1.30854905e+00
1.85063854e-01 -3.35742503e-01 7.57419527e-01 7.12758839e-01
1.15491316e-01 -3.53404790e-01 -5.30639231e-01 -7.32306838e-01
-2.82464951e-01 -7.95736089e-02 7.59816408e-01 1.14078569e+00
-4.52305049e-01 -1.72015745e-02 -1.12958717e+00 3.97911102e-01
7.90603638e-01 1.06572412e-01 1.26200211e+00 -8.06533813e-01
-2.54071891e-01 3.13604593e-01 -4.88949180e-01 -8.24480593e-01
-1.77807942e-01 -3.89486134e-01 1.05661757e-01 -1.66517150e+00
3.09447706e-01 -2.00245336e-01 -2.66258419e-02 3.60041887e-01
-5.21127045e-01 1.01882231e+00 6.06685460e-01 8.80582258e-02
-2.51124334e-02 4.57733363e-01 1.96066523e+00 -6.55268272e-03
-2.07360029e-01 -2.62921602e-01 -5.80324054e-01 7.11814463e-01
4.85456735e-01 -1.38542280e-01 -2.87119120e-01 -4.34617788e-01
-3.62347998e-02 1.93530217e-01 5.37950099e-01 -8.14564943e-01
-1.93121210e-01 -3.36934000e-01 7.70567119e-01 -4.83471513e-01
7.15829074e-01 -8.52900147e-01 7.33222425e-01 3.04842979e-01
6.78743199e-02 2.32338861e-01 2.27016985e-01 4.51839626e-01
-4.82435785e-02 -2.46031694e-02 8.22104633e-01 -1.24035187e-01
-8.62286627e-01 1.36396825e-01 -8.89531374e-02 -4.55312371e-01
1.23373270e+00 -5.44788599e-01 -2.42632642e-01 -5.99992514e-01
-6.96338058e-01 -8.89455527e-02 1.13350224e+00 6.63411677e-01
7.19420254e-01 -1.66780329e+00 -7.00046241e-01 5.77517509e-01
1.86133146e-01 -2.90535688e-01 5.00244379e-01 4.68235403e-01
-8.10198843e-01 -1.80920046e-02 -7.08634138e-01 -4.39827651e-01
-1.67807698e+00 8.15592527e-01 2.35583484e-01 4.07306105e-02
-7.53279328e-01 4.03627217e-01 4.24769431e-01 -5.45352340e-01
-2.65263587e-01 -1.05793484e-01 1.15489788e-01 -3.82970214e-01
3.28619570e-01 1.76291645e-01 -1.10306390e-01 -9.93610024e-01
-2.49331519e-01 9.97014463e-01 7.85487890e-02 -2.41311610e-01
9.31534350e-01 -2.15534598e-01 2.02264637e-01 1.96481004e-01
6.40422225e-01 1.62599698e-01 -1.52762139e+00 -2.11159378e-01
-7.80470252e-01 -9.40314353e-01 -1.79099321e-01 -8.69446158e-01
-1.11071098e+00 4.16123509e-01 6.69317424e-01 -1.64004624e-01
1.03087723e+00 -3.44408959e-01 9.81393278e-01 -2.54791796e-01
7.15671778e-01 -1.03469801e+00 1.72206536e-01 -1.82876110e-01
1.27979517e+00 -1.09206164e+00 3.35018635e-01 -1.00218856e+00
-1.14971399e+00 9.26328480e-01 7.46149063e-01 -1.94915161e-01
4.31110471e-01 2.17712104e-01 5.24072468e-01 -7.76708424e-02
-8.06623772e-02 4.09360155e-02 4.72552478e-01 8.18676174e-01
6.24980181e-02 2.15995654e-01 -2.16574058e-01 2.81317502e-01
-1.81968600e-01 1.96501732e-01 6.56392336e-01 1.01083028e+00
6.99485838e-02 -1.23853874e+00 -9.73800242e-01 -2.15382636e-01
-1.26653522e-01 2.89973885e-01 -4.80195552e-01 1.07481587e+00
2.35866770e-01 7.33535588e-01 3.12874280e-02 -5.89130342e-01
7.83311129e-01 -6.03223816e-02 8.79954576e-01 -4.78281319e-01
-5.14538586e-01 1.29729003e-01 1.22288615e-01 -4.23558354e-01
-5.33899128e-01 -4.90318954e-01 -9.01123405e-01 -4.39248264e-01
-2.12621912e-01 -5.98124921e-01 8.21023941e-01 6.41050100e-01
1.40742823e-01 5.50318480e-01 6.18562520e-01 -1.47197199e+00
-2.40107119e-01 -8.35052133e-01 -7.27866948e-01 8.26382101e-01
3.16100232e-02 -9.27938819e-01 1.01339556e-01 4.62392300e-01] | [11.899657249450684, -0.8627941608428955] |
e605abce-6aee-4517-b7a5-6602d104b2c4 | chatgpt-edss-empathetic-dialogue-speech | 2305.13724 | null | https://arxiv.org/abs/2305.13724v1 | https://arxiv.org/pdf/2305.13724v1.pdf | ChatGPT-EDSS: Empathetic Dialogue Speech Synthesis Trained from ChatGPT-derived Context Word Embeddings | We propose ChatGPT-EDSS, an empathetic dialogue speech synthesis (EDSS) method using ChatGPT for extracting dialogue context. ChatGPT is a chatbot that can deeply understand the content and purpose of an input prompt and appropriately respond to the user's request. We focus on ChatGPT's reading comprehension and introduce it to EDSS, a task of synthesizing speech that can empathize with the interlocutor's emotion. Our method first gives chat history to ChatGPT and asks it to generate three words representing the intention, emotion, and speaking style for each line in the chat. Then, it trains an EDSS model using the embeddings of ChatGPT-derived context words as the conditioning features. The experimental results demonstrate that our method performs comparably to ones using emotion labels or neural network-derived context embeddings learned from chat histories. The collected ChatGPT-derived context information is available at https://sarulab-speech.github.io/demo_ChatGPT_EDSS/. | ['Hiroshi Saruwatari', 'Kentaro Tachibana', 'Eiji Iimori', 'Shinnosuke Takamichi', 'Yuki Saito'] | 2023-05-23 | null | null | null | null | ['chatbot', 'reading-comprehension', 'chatbot', 'speech-synthesis'] | ['methodology', 'natural-language-processing', 'natural-language-processing', 'speech'] | [-1.34499997e-01 5.88646114e-01 2.02929586e-01 -5.65316379e-01
-7.96607614e-01 -5.45204341e-01 5.62231421e-01 -1.29819617e-01
-4.54102531e-02 5.26324332e-01 1.09349537e+00 -3.84994805e-01
5.49914300e-01 -4.09416407e-01 9.81769711e-02 -4.63315606e-01
5.10406673e-01 6.84539258e-01 -3.04690421e-01 -8.43728781e-01
4.18385237e-01 -2.42832333e-01 -6.28739536e-01 8.46216917e-01
6.37837470e-01 9.03473735e-01 5.62368572e-01 1.05121517e+00
-5.66485405e-01 1.60054040e+00 -1.12080097e+00 -4.41548288e-01
-3.04963291e-01 -1.07622004e+00 -1.83050895e+00 -2.86910087e-01
-5.79626501e-01 -4.20620739e-01 -5.72595060e-01 4.73969758e-01
4.94212002e-01 4.88797694e-01 7.47389376e-01 -1.53619552e+00
-1.05128038e+00 1.16516054e+00 3.70540470e-01 9.36590284e-02
9.28507447e-01 2.07177937e-01 1.22164822e+00 -7.58460999e-01
6.71616793e-01 1.53030252e+00 5.32984316e-01 1.13482141e+00
-6.62110209e-01 -2.49855354e-01 -7.96495602e-02 4.13943470e-01
-6.66368783e-01 -4.49888140e-01 9.54596996e-01 -1.03564121e-01
1.11901546e+00 4.55412924e-01 5.67668021e-01 2.07078004e+00
1.85858514e-02 1.22537088e+00 9.21650708e-01 -4.84816909e-01
7.94606954e-02 2.42729217e-01 1.91052631e-01 2.34598070e-01
-1.32178926e+00 -3.07427496e-01 -8.20808947e-01 -2.58850962e-01
4.80119795e-01 -3.67462128e-01 -3.47408116e-01 8.10933411e-01
-1.34654069e+00 1.22494233e+00 2.48529404e-01 5.03966331e-01
-2.61348277e-01 -9.78235677e-02 8.42816293e-01 7.83358514e-01
4.79171395e-01 7.09550619e-01 -3.87474895e-01 -1.08607471e+00
1.41964912e-01 7.06709996e-02 1.39956415e+00 1.24623752e+00
1.27055362e-01 -4.51088697e-02 -4.79797930e-01 1.29819739e+00
7.12501854e-02 2.64936149e-01 5.43919921e-01 -1.38319993e+00
6.22966766e-01 5.86447954e-01 7.71192461e-02 -7.96905518e-01
-3.68085951e-01 5.51824212e-01 -4.57664222e-01 -4.18332607e-01
3.48513305e-01 -6.71277523e-01 1.86114088e-01 1.60976696e+00
2.24945948e-01 -1.94899783e-01 7.15415001e-01 9.52428043e-01
1.48932648e+00 1.25579822e+00 5.53413592e-02 -2.23508939e-01
1.57875586e+00 -1.45877516e+00 -1.05609059e+00 -2.94992805e-01
9.93417084e-01 -8.81592810e-01 1.54031551e+00 1.12561449e-01
-8.69013846e-01 -2.22193599e-01 -5.90749860e-01 -5.22719860e-01
-1.87753767e-01 2.85401106e-01 3.69601101e-01 1.25710620e-02
-7.46888876e-01 2.93099761e-01 -3.62049371e-01 -4.88936961e-01
-1.22098453e-01 -2.61419654e-01 -2.53020972e-01 2.67203599e-01
-1.65434062e+00 1.18260670e+00 5.01263365e-02 1.30762875e-01
-4.94953573e-01 -4.59462106e-01 -1.09934509e+00 1.15985923e-01
2.75504500e-01 -2.51578391e-01 1.97856915e+00 -8.85050416e-01
-2.37203336e+00 7.25574970e-01 -1.52507275e-01 -1.26065254e-01
1.74141690e-01 3.92438937e-03 -2.17236504e-01 2.74697602e-01
-1.82463378e-02 4.66180563e-01 4.06461149e-01 -8.30930352e-01
-3.29046339e-01 6.72457963e-02 2.06058934e-01 4.67448324e-01
-2.26990268e-01 5.30952632e-01 9.07987803e-02 -4.72008020e-01
-2.53063828e-01 -8.36300075e-01 1.82524726e-01 -4.04720098e-01
-6.61661744e-01 -8.58225346e-01 8.66960645e-01 -7.19184279e-01
1.04028714e+00 -2.10584021e+00 3.33864033e-01 -5.73104501e-01
6.47113025e-02 1.89606324e-01 -4.57012534e-01 1.45131493e+00
2.97078848e-01 -6.92184269e-02 -5.94809428e-02 -6.40568137e-01
2.34194368e-01 4.25638497e-01 -3.92273873e-01 -7.46547356e-02
3.15936059e-01 8.82621348e-01 -1.10758793e+00 -3.85429651e-01
2.45090500e-01 3.58475208e-01 -4.42917347e-01 1.15728140e+00
-3.39897633e-01 9.04978812e-01 -5.60648620e-01 1.12091377e-01
-3.32035907e-02 -7.92706758e-02 4.23658341e-01 2.47551560e-01
-2.69042086e-02 1.06371462e+00 -2.12877870e-01 1.63252532e+00
-9.92606103e-01 8.54082406e-01 8.62354040e-02 -7.83570766e-01
1.27088344e+00 8.52644324e-01 -9.79639515e-02 -3.86563003e-01
7.21612990e-01 1.53807318e-02 -9.29973088e-03 -9.46096718e-01
6.71299696e-01 -2.49796107e-01 -7.95045495e-01 1.09739995e+00
3.54747921e-01 -3.96890402e-01 -4.14655089e-01 3.95160526e-01
1.12789154e+00 -2.11025491e-01 4.95736092e-01 3.18180658e-02
6.41607642e-01 -9.71430447e-03 3.34483057e-01 3.96548033e-01
-4.57279205e-01 2.90066212e-01 1.09270406e+00 -3.20356309e-01
-6.77467823e-01 -6.54901981e-01 5.09632468e-01 1.65014458e+00
-4.39730808e-02 -5.31131268e-01 -9.19571042e-01 -6.71777487e-01
-4.65813100e-01 1.18968570e+00 -8.43784928e-01 -2.34499112e-01
-5.33456624e-01 7.95694366e-02 8.36865306e-01 4.63660091e-01
4.79619622e-01 -1.85135400e+00 -3.40577483e-01 3.69144589e-01
-9.89045143e-01 -1.16176867e+00 -8.36775482e-01 4.63422872e-02
-2.91838765e-01 -9.32865679e-01 -4.64115292e-01 -1.17712641e+00
2.50818700e-01 2.55809147e-02 8.89132619e-01 2.96683848e-01
1.65084526e-01 3.73540699e-01 -1.17090261e+00 -1.51661471e-01
-1.27204239e+00 2.30980486e-01 -2.16551214e-01 -1.63239807e-01
4.30134058e-01 -7.85848022e-01 -2.28747383e-01 3.87476891e-01
-2.85894901e-01 4.63444591e-01 -1.29841834e-01 1.12707019e+00
-3.93994421e-01 -7.84291744e-01 9.80562329e-01 -7.81762064e-01
1.45385671e+00 -7.62926579e-01 3.93627822e-01 2.28524819e-01
1.76758230e-01 -1.70628086e-01 9.65540826e-01 -6.14556074e-01
-1.35417378e+00 -5.40980756e-01 -6.14813626e-01 -1.55440584e-01
-4.25959110e-01 4.30524111e-01 -1.62792310e-01 5.09365559e-01
6.18952572e-01 2.92393357e-01 2.85077214e-01 -4.55607951e-01
4.81400728e-01 1.55953956e+00 7.20708311e-01 -9.38434541e-01
1.14531398e-01 -2.15032369e-01 -1.04311943e+00 -9.41917300e-01
-8.63523424e-01 -4.16323632e-01 -3.00854772e-01 -5.46930969e-01
1.00636876e+00 -4.84733582e-01 -1.29922998e+00 4.98242527e-01
-1.78204429e+00 -9.53559160e-01 4.31119837e-02 2.16971859e-01
-9.75641966e-01 2.68449217e-01 -1.28044128e+00 -1.08375800e+00
-4.81001914e-01 -9.24227953e-01 8.00500333e-01 2.83662140e-01
-9.67653930e-01 -1.18746078e+00 1.24871552e-01 7.67714322e-01
4.40927505e-01 -2.56804768e-02 9.66395020e-01 -1.33442843e+00
3.50716531e-01 1.66778024e-02 1.82477515e-02 4.89604056e-01
1.72763705e-01 -1.30860686e-01 -9.33162987e-01 3.52425188e-01
4.21940744e-01 -9.75930572e-01 -1.08471461e-01 -2.23372325e-01
7.99989581e-01 -9.74730432e-01 2.89559394e-01 1.51752263e-01
5.63758731e-01 4.64149654e-01 5.01312375e-01 -2.35780384e-02
3.55268538e-01 1.02098858e+00 7.53684640e-01 6.67194307e-01
7.46192396e-01 6.20784461e-01 1.17047399e-01 4.16836679e-01
5.78315817e-02 -5.24722934e-01 8.15021992e-01 1.49991941e+00
1.23461202e-01 -5.35383046e-01 -8.13421071e-01 5.16386747e-01
-1.97886741e+00 -1.02744520e+00 -2.39495084e-01 1.07006872e+00
1.16152442e+00 -4.21942830e-01 1.53511852e-01 -1.14190057e-01
7.00365543e-01 3.91934216e-01 -2.13157088e-01 -1.37695253e+00
1.62853420e-01 -1.61493756e-02 -4.75451112e-01 8.76385927e-01
-3.45525920e-01 1.28263581e+00 4.85404539e+00 5.14800310e-01
-1.01879776e+00 3.64057571e-01 5.18699467e-01 1.95992559e-01
-1.35840178e-01 -1.01691730e-01 -2.50579923e-01 4.99992222e-01
1.18046427e+00 -4.76098835e-01 6.60067558e-01 8.65965247e-01
3.75603795e-01 1.12562284e-01 -1.19167507e+00 7.75506914e-01
2.81721056e-01 -1.24777079e+00 -3.15607578e-01 -4.28012550e-01
1.30705222e-01 -3.08731735e-01 -2.49318838e-01 8.17784309e-01
4.21375781e-01 -8.98598969e-01 5.94240785e-01 1.06629498e-01
5.87536395e-01 -5.11194944e-01 8.90203714e-01 6.48477614e-01
-7.60154426e-01 -1.19568177e-01 -9.70843732e-02 -5.81702590e-01
4.56696630e-01 -5.19440889e-01 -1.44224703e+00 3.29245746e-01
4.12642747e-01 5.87214053e-01 1.87195595e-02 6.61900118e-02
-7.52130449e-01 9.67329919e-01 -4.63826349e-03 -7.06901670e-01
4.53745812e-01 -3.69218558e-01 6.78200901e-01 1.31224084e+00
1.06838554e-01 6.76835954e-01 4.78132367e-02 9.96594906e-01
-2.19475150e-01 3.26915890e-01 -4.74592417e-01 -4.43791747e-01
1.00629640e+00 1.19790423e+00 -1.30019546e-01 -3.16589594e-01
-1.40072316e-01 1.31187248e+00 4.42827195e-01 1.74184486e-01
-6.32900536e-01 -5.48305154e-01 8.49786937e-01 -4.25606638e-01
-6.31229058e-02 2.50308346e-02 -5.85549176e-02 -8.46302032e-01
-2.37139061e-01 -1.00540531e+00 2.36058429e-01 -1.32012093e+00
-1.60405886e+00 1.09029567e+00 -1.58673897e-01 -8.98550153e-01
-5.33559203e-01 -5.32890499e-01 -1.53912497e+00 8.13691318e-01
-7.46069193e-01 -1.23317719e+00 -2.27588564e-01 5.33610821e-01
1.13762641e+00 -2.68640429e-01 1.28261244e+00 -3.35609436e-01
-5.38647115e-01 6.29728556e-01 -4.07321453e-01 7.82340407e-01
7.63812125e-01 -1.17028213e+00 4.36217278e-01 1.77088100e-02
-2.57465780e-01 4.70177621e-01 9.43267286e-01 -3.11011165e-01
-1.16813004e+00 -5.97632587e-01 1.45822752e+00 -6.28988504e-01
1.01707184e+00 -5.04548490e-01 -9.68801558e-01 8.15281093e-01
1.00235665e+00 -5.79242885e-01 1.10545719e+00 3.16312373e-01
-2.60601401e-01 5.18449843e-01 -1.06939137e+00 9.14210379e-01
9.08097208e-01 -9.00417864e-01 -1.52626157e+00 4.16216046e-01
1.07374918e+00 -6.63585305e-01 -9.57354724e-01 -5.77699721e-01
3.04127157e-01 -7.62036741e-01 3.77331167e-01 -1.02641058e+00
9.40838039e-01 3.17915410e-01 -2.07989484e-01 -1.67897677e+00
1.94153681e-01 -1.17500889e+00 3.36433351e-01 1.47027850e+00
3.64407688e-01 -6.74381971e-01 1.92249686e-01 6.59698784e-01
-6.81575000e-01 -6.71487451e-01 -1.11270320e+00 -4.93897766e-01
4.51081008e-01 -4.17204410e-01 6.07410371e-01 1.23150992e+00
1.23206365e+00 9.80512738e-01 -5.59519529e-01 -2.60503709e-01
-2.43550375e-01 3.57611454e-03 8.91205370e-01 -9.03969347e-01
-1.98347926e-01 -3.49344760e-01 1.76766068e-01 -1.21361947e+00
6.67151272e-01 -8.62118125e-01 5.38177907e-01 -1.57705331e+00
-2.78387308e-01 -2.05062509e-01 4.95163381e-01 6.95159793e-01
-5.63291647e-02 -3.33594620e-01 4.14301813e-01 -2.16523581e-03
-5.36888361e-01 9.82364178e-01 1.35789526e+00 5.94046956e-04
-4.03461456e-01 -1.96869373e-02 -8.04258049e-01 4.58498597e-01
1.07967317e+00 -4.35658604e-01 -3.84632051e-01 -1.81418315e-01
-6.01748973e-02 9.52275753e-01 1.09021515e-01 -4.86367106e-01
2.53051400e-01 -5.39502084e-01 -5.05740345e-01 -9.82313156e-02
7.20630050e-01 -3.52268428e-01 -4.11171615e-01 -2.62872241e-02
-1.03740430e+00 -5.22928461e-02 -3.97445448e-02 1.74404696e-01
-4.15824980e-01 -3.66986245e-01 4.66726214e-01 -4.23377044e-02
-3.07521194e-01 -1.88020512e-01 -1.15559840e+00 4.00403261e-01
6.54317796e-01 4.68394794e-02 -5.70710599e-01 -1.19249308e+00
-7.48575330e-01 4.69693929e-01 1.21174879e-01 8.84004474e-01
7.22155333e-01 -1.41806734e+00 -7.84874856e-01 -2.29330719e-01
2.49915987e-01 -1.10246345e-01 5.60842872e-01 8.27334583e-01
-2.24500835e-01 1.70939356e-01 -2.08941415e-01 -1.39170647e-01
-1.43127859e+00 1.03395514e-01 2.59365648e-01 5.25172763e-02
-8.56581509e-01 9.90796566e-01 -7.08483160e-02 -8.87983441e-01
1.41841903e-01 -4.09255415e-01 -3.13574135e-01 8.97744447e-02
6.87716126e-01 8.65558088e-02 -3.93287241e-01 -7.34725416e-01
-1.07035093e-01 -3.29016596e-01 -7.55881192e-03 -4.09929305e-01
1.33553076e+00 -3.72933894e-01 -5.32931574e-02 8.44403327e-01
1.46256745e+00 -1.26731664e-01 -9.40304101e-01 -1.32224888e-01
-9.33957621e-02 -2.10405231e-01 -4.63459074e-01 -9.73157167e-01
-5.45666218e-01 1.03857827e+00 -4.50609446e-01 4.52308863e-01
5.55697858e-01 4.17951286e-01 1.38631845e+00 6.70109928e-01
1.25729859e-01 -1.34662640e+00 4.93428737e-01 1.27929413e+00
1.60905445e+00 -1.07563233e+00 -8.31974208e-01 -2.33079419e-01
-1.66009557e+00 1.43042099e+00 7.91269898e-01 -4.90094237e-02
3.13301772e-01 2.33572777e-02 8.54718864e-01 -2.64877737e-01
-1.35203743e+00 3.10519397e-01 -3.96170855e-01 8.11003029e-01
4.86320078e-01 2.03197554e-01 -9.20753404e-02 1.21701062e+00
-9.60857451e-01 -3.86475861e-01 9.11927521e-01 7.92542875e-01
-2.92732030e-01 -1.14242411e+00 -1.63561285e-01 -1.91106703e-02
9.60699767e-02 -7.94689879e-02 -1.09421945e+00 5.97269535e-01
-5.92409015e-01 1.46087849e+00 1.15704991e-01 -7.65602529e-01
4.74131435e-01 6.70127749e-01 1.48346489e-02 -1.01198018e+00
-1.16509342e+00 -4.38312680e-01 7.98841774e-01 -4.18489099e-01
-1.50995716e-01 -5.77548683e-01 -1.51321268e+00 -6.08848274e-01
1.91921219e-02 8.39566827e-01 3.85141671e-01 1.22747040e+00
1.38613164e-01 2.13528231e-01 1.13951004e+00 -7.93952107e-01
-4.79001969e-01 -1.67454779e+00 -2.36737818e-01 3.93107474e-01
1.19424351e-01 -3.12779099e-01 -6.77258313e-01 -2.53994763e-01] | [12.89903736114502, 7.789802551269531] |
d64c9848-e958-404e-88ad-dfc0f13d9158 | document-intelligence-metrics-for-visually | 2205.11215 | null | https://arxiv.org/abs/2205.11215v1 | https://arxiv.org/pdf/2205.11215v1.pdf | Document Intelligence Metrics for Visually Rich Document Evaluation | The processing of Visually-Rich Documents (VRDs) is highly important in information extraction tasks associated with Document Intelligence. We introduce DI-Metrics, a Python library devoted to VRD model evaluation comprising text-based, geometric-based and hierarchical metrics for information extraction tasks. We apply DI-Metrics to evaluate information extraction performance using publicly available CORD dataset, comparing performance of three SOTA models and one industry model. The open-source library is available on GitHub. | ['Adam Karwan', 'Krzysztof Wilkosz', 'Zhuoyu Han', 'Swapnil Gupta', 'Jonathan Degange'] | 2022-05-23 | null | null | null | null | ['document-ai'] | ['natural-language-processing'] | [-6.48483559e-02 5.97570688e-02 9.91874561e-02 -1.47184670e-01
-7.82927096e-01 -1.09382355e+00 1.15678799e+00 8.50159883e-01
-1.07315667e-01 3.01342398e-01 4.12509590e-01 -4.20443207e-01
-4.20229822e-01 -1.05007780e+00 -1.56602219e-01 8.66526067e-02
-1.94963321e-01 3.76037717e-01 2.82852560e-01 -4.16132249e-02
9.88634050e-01 1.19902003e+00 -1.62547529e+00 6.96222723e-01
6.77516460e-01 1.20513046e+00 2.01082394e-01 1.31539309e+00
-7.11115539e-01 9.37549412e-01 -9.98060167e-01 -4.21574742e-01
1.36455357e-01 2.32103705e-01 -1.17133725e+00 -1.80404261e-01
5.36456823e-01 -1.39899537e-01 -2.80119628e-01 6.85518563e-01
5.84111154e-01 -3.21180195e-01 1.13104856e+00 -1.08790624e+00
-6.62623227e-01 5.66014349e-01 -3.72445196e-01 3.91998291e-01
1.02600002e+00 -3.36923271e-01 1.03966761e+00 -1.08573949e+00
1.40101397e+00 1.22843885e+00 3.53369266e-01 -1.38227999e-01
-8.40100527e-01 -2.79495925e-01 -2.97761280e-02 2.95144111e-01
-1.44598722e+00 -1.72133967e-01 7.08894908e-01 -7.41155624e-01
1.47104299e+00 6.64710760e-01 5.22105455e-01 8.14598858e-01
2.40419909e-01 1.27288234e+00 7.97449112e-01 -5.09519517e-01
1.33036435e-01 3.22106004e-01 6.99975312e-01 7.06718862e-01
5.28026104e-01 -4.38801914e-01 -5.31849623e-01 8.31599906e-02
4.87575799e-01 -6.38475120e-01 -1.20453686e-02 -3.29450876e-01
-8.99046123e-01 6.10917866e-01 3.39740545e-01 5.23056805e-01
-1.76370978e-01 -3.50739598e-01 5.21785259e-01 2.28264853e-01
4.17483658e-01 6.80559874e-01 -4.85822052e-01 -2.76146799e-01
-9.07629550e-01 5.51817238e-01 7.54674017e-01 1.49458790e+00
2.56706625e-01 -1.53798014e-01 -8.21112156e-01 8.47446620e-01
3.28463584e-01 3.41655016e-01 5.49721234e-02 -4.24280882e-01
9.33459938e-01 1.18712389e+00 -7.62746856e-02 -1.41340673e+00
-8.62099469e-01 -1.72328532e-01 -6.04762316e-01 2.23351642e-01
1.00219466e-01 1.62125081e-01 -9.10221756e-01 3.88927847e-01
3.14137310e-01 -7.74044096e-01 -6.08977675e-02 4.88448709e-01
1.87358510e+00 4.46146220e-01 -1.02324642e-01 7.07618371e-02
1.49539196e+00 -4.80604410e-01 -9.01699066e-01 1.36902317e-01
6.34787261e-01 -9.36562061e-01 8.70886564e-01 9.00183201e-01
-1.08496606e+00 -3.15083236e-01 -1.45769274e+00 -6.73826993e-01
-1.25044477e+00 2.73961931e-01 4.34959978e-01 5.28531373e-01
-7.86099434e-01 5.23181379e-01 -3.50536555e-01 -2.62255251e-01
9.45057392e-01 1.95560977e-04 -5.06828487e-01 -1.56783371e-03
-5.94678819e-01 8.71498168e-01 3.78147751e-01 -2.67823249e-01
-5.36726892e-01 -6.31892622e-01 -9.97276306e-01 -3.69256169e-01
2.44974449e-01 -3.65146637e-01 1.21914113e+00 3.70144695e-01
-1.06463861e+00 1.06197906e+00 1.18679158e-01 -1.56718507e-01
6.20075285e-01 -3.61425042e-01 -2.43744388e-01 4.98415917e-01
7.71165416e-02 1.49445042e-01 5.14270306e-01 -1.55681312e+00
-5.61219037e-01 -7.88463771e-01 -4.69513126e-02 5.27885519e-02
-1.96464166e-01 2.55333930e-01 -7.66424477e-01 -6.70633614e-01
1.82954501e-02 -1.56947553e-01 8.80641118e-02 -4.87855822e-02
-1.12872088e+00 -4.17413801e-01 1.09988403e+00 -1.08978426e+00
1.43809879e+00 -1.71452522e+00 4.07182612e-02 3.25083852e-01
4.83496517e-01 3.54637176e-01 -3.36528830e-02 3.61722559e-01
2.37609938e-01 6.44677579e-01 -7.94117600e-02 -2.58411020e-01
2.04735041e-01 -5.77545226e-01 9.42241624e-02 1.11315168e-01
4.25955921e-01 8.80326092e-01 -6.77819073e-01 -1.04685795e+00
5.60021341e-01 7.45663881e-01 1.18682601e-01 2.43406817e-01
-2.90971428e-01 -1.18042789e-01 -8.28440011e-01 8.87269795e-01
1.09108043e+00 1.84660852e-02 -1.80597946e-01 -3.95971060e-01
-5.78589082e-01 4.48000282e-01 -1.33879185e+00 1.74476767e+00
-3.38930935e-01 1.08231795e+00 -2.91069359e-01 -2.80201346e-01
1.17086470e+00 2.36892849e-02 3.25108081e-01 -1.14941645e+00
5.85877180e-01 -3.84280771e-01 -9.05124724e-01 -5.59516430e-01
1.16021621e+00 1.05205452e+00 -1.68217406e-01 3.90619993e-01
9.31383371e-02 -8.61458778e-01 7.41630256e-01 7.52783298e-01
1.53431785e+00 3.69267702e-01 4.99681473e-01 -1.46587580e-01
5.77949941e-01 4.54548091e-01 -5.81812523e-02 6.10261023e-01
2.98338085e-01 7.97961235e-01 7.42667973e-01 -1.47293851e-01
-1.15034974e+00 -1.03772902e+00 -5.91968358e-01 6.34587586e-01
-1.54703498e-01 -1.12453568e+00 -7.17720687e-01 -8.79290760e-01
1.39047220e-01 1.10981405e+00 -5.91121376e-01 5.24601698e-01
-3.31041068e-01 -5.05838394e-01 6.26324236e-01 6.31867588e-01
3.02648067e-01 -1.09753609e+00 -7.65109360e-01 -1.00881435e-01
3.42838377e-01 -1.34642220e+00 2.07987383e-01 3.09275717e-01
-4.88033056e-01 -1.24894726e+00 -4.18454766e-01 -5.38563311e-01
4.39115763e-01 1.72444031e-01 1.57056367e+00 1.94966912e-01
-1.00378597e+00 6.37102604e-01 -6.90791667e-01 -9.66809571e-01
-1.15867049e-01 3.04660559e-01 -8.12358558e-01 -8.08683574e-01
5.02787113e-01 3.07636261e-02 -4.70672518e-01 -2.78190523e-02
-1.03104246e+00 1.84607103e-01 2.78538913e-01 -1.70535609e-01
7.40287781e-01 -8.17563385e-04 -2.27491725e-02 -8.82213891e-01
9.83334780e-01 -3.10461968e-01 -7.79859245e-01 3.46421152e-01
-5.14292359e-01 2.33878568e-01 4.45269555e-01 2.95989394e-01
-9.52916682e-01 -2.18905509e-02 -7.19357282e-02 2.58695334e-01
-3.93235296e-01 4.10079688e-01 -4.33686733e-01 2.48416200e-01
4.86987680e-01 -1.98713496e-01 -6.55946136e-01 -9.05228376e-01
5.95294535e-01 1.09253776e+00 5.16279876e-01 -3.02329391e-01
8.28267276e-01 3.88541788e-01 -1.89847946e-02 -1.20182979e+00
-5.83445668e-01 -6.53651357e-01 -1.17781758e+00 -2.81067282e-01
8.58431816e-01 -3.51779372e-01 -4.37376499e-01 3.36909920e-01
-1.48680925e+00 2.07786467e-02 -1.80236936e-01 -4.78101261e-02
-2.93169677e-01 1.43679693e-01 -4.08202082e-01 -8.36687505e-01
-8.44157934e-01 -7.91128337e-01 1.30481565e+00 2.42846176e-01
-3.09021235e-01 -8.26979399e-01 3.88636023e-01 2.26712897e-01
9.77823883e-03 8.26218605e-01 9.92045701e-01 -6.63242877e-01
-4.34642702e-01 -5.34638584e-01 -9.99045551e-01 -4.57336195e-02
-1.35136619e-01 7.25297511e-01 -1.06663859e+00 2.71901846e-01
-6.90968215e-01 -7.98741952e-02 6.78174078e-01 1.54522985e-01
1.30661511e+00 -1.37525663e-01 -7.56904066e-01 4.13233042e-01
1.66392481e+00 3.45272869e-01 8.68404388e-01 8.06669235e-01
9.82872486e-01 6.46510065e-01 8.59115839e-01 8.08035314e-01
6.21423900e-01 6.26935065e-01 4.85108763e-01 1.31961741e-02
-3.60627621e-01 -4.97456305e-02 -1.54011115e-01 7.26896584e-01
-1.24223091e-01 -6.88806236e-01 -1.39681113e+00 4.35588866e-01
-1.42306960e+00 -6.09604120e-01 -8.62409890e-01 1.85321558e+00
6.21641219e-01 2.49241292e-01 1.24818541e-01 8.12800586e-01
2.08810583e-01 -6.77530095e-02 -1.66557699e-01 -8.10020745e-01
-2.21795872e-01 4.32823777e-01 3.90712261e-01 1.49237916e-01
-1.27577150e+00 7.75078773e-01 6.87407160e+00 7.08707511e-01
-4.85847563e-01 -2.56074160e-01 1.18650533e-01 7.87155330e-03
-3.15352023e-01 -2.78075010e-01 -1.16139936e+00 -8.76624584e-02
9.84215736e-01 -3.66712898e-01 -1.87089920e-01 8.46288681e-01
2.49785781e-01 -3.48011404e-01 -1.19137812e+00 7.74209678e-01
3.09717208e-01 -1.55118454e+00 3.10232103e-01 1.25164330e-01
2.24646211e-01 -1.20991178e-01 -1.66700661e-01 -1.70918867e-01
3.10442537e-01 -1.19688785e+00 9.78343546e-01 7.18966782e-01
9.39855158e-01 -9.99729931e-01 6.08240068e-01 -2.27944091e-01
-1.52127230e+00 3.25490922e-01 -2.24307105e-01 5.65615892e-01
-5.27267493e-02 7.54258454e-01 -1.13864982e+00 9.62227643e-01
9.78052080e-01 9.51075256e-01 -1.50871313e+00 1.33270657e+00
-1.64622486e-01 -1.10401504e-01 -1.93201657e-02 -3.12499017e-01
-1.40847430e-01 -3.23729403e-02 5.65421462e-01 2.03269577e+00
-5.41693419e-02 -2.39824131e-01 -4.68267202e-01 7.44657099e-01
2.35414449e-02 6.01575255e-01 -8.51871550e-01 -2.43091866e-01
3.69050860e-01 1.63055313e+00 -1.09206808e+00 -1.49435639e-01
-2.87139088e-01 7.44574070e-01 1.58618093e-01 -1.76018313e-01
-3.92872185e-01 -1.37598670e+00 2.45944187e-01 1.91100731e-01
4.22417015e-01 -5.33585966e-01 -8.12786281e-01 -5.92519283e-01
1.88101992e-01 -8.63717735e-01 5.19617736e-01 -1.12647593e+00
-8.51063728e-01 9.99616861e-01 3.76283079e-01 -1.43630457e+00
-1.55292690e-01 -1.05781698e+00 -2.29407132e-01 3.88440520e-01
-1.53899610e+00 -1.21892488e+00 -8.96892250e-01 4.05747324e-01
4.83480692e-01 -1.91102669e-01 9.57992136e-01 1.67004868e-01
-7.64528215e-01 3.91311854e-01 4.51257341e-02 5.77391148e-01
2.29502603e-01 -1.77875674e+00 7.70908058e-01 7.57738769e-01
2.56253600e-01 2.46335909e-01 5.46503782e-01 -5.66615283e-01
-1.59081399e+00 -1.05838263e+00 7.19032347e-01 -1.19053793e+00
4.75823313e-01 -6.04698896e-01 -8.76144707e-01 3.02191496e-01
2.49099851e-01 -3.61443192e-01 6.33177102e-01 -3.80138010e-01
-5.94643176e-01 3.24145287e-01 -1.41146171e+00 2.43648589e-01
1.01229584e+00 -4.63258028e-01 -5.44551313e-01 4.53480631e-01
6.85197055e-01 -3.77453476e-01 -1.38860285e+00 2.61992484e-01
5.21989524e-01 -8.66422176e-01 1.15343976e+00 -2.77741641e-01
6.67391956e-01 -2.37008154e-01 -9.92554706e-03 -9.66525555e-01
-1.43420743e-02 -3.01323861e-01 -5.79421639e-01 1.57662642e+00
7.50123739e-01 8.25892910e-02 5.92624426e-01 1.79299772e-01
6.75351499e-03 -5.34457743e-01 -4.77166057e-01 -6.82220995e-01
1.42609864e-01 -7.87086487e-01 8.17588151e-01 5.50211132e-01
1.49514899e-01 5.63969791e-01 4.84529763e-01 -2.66447663e-03
5.55261850e-01 -1.08231060e-01 1.06138694e+00 -1.71386945e+00
4.76225525e-01 -8.22758138e-01 -6.66857779e-01 -4.68720526e-01
-3.63667101e-01 -1.09723616e+00 -2.01715514e-01 -2.57942033e+00
6.41409382e-02 -1.66980878e-01 2.15785861e-01 3.92145246e-01
4.95874405e-01 3.25166613e-01 2.10891888e-01 -6.23011552e-02
-8.23282719e-01 1.44065186e-01 1.05359399e+00 -4.97091860e-01
-4.29288536e-01 -3.99915844e-01 -6.24020159e-01 5.91880977e-01
6.18344367e-01 -4.72239584e-01 -1.57949284e-01 -3.89519185e-01
4.43215340e-01 -4.75348830e-01 -1.73244125e-03 -9.86599624e-01
-8.47289618e-03 9.34300497e-02 8.41645718e-01 -1.56029594e+00
1.27777472e-01 -6.52324140e-01 -3.78102690e-01 2.53779087e-02
-1.84816241e-01 2.50463277e-01 5.65438569e-01 -2.52658385e-03
2.63501052e-02 -5.09399593e-01 3.07160169e-01 5.88702448e-02
-4.38871264e-01 1.18101716e-01 -3.69449019e-01 3.33260559e-02
9.40963745e-01 -4.19635147e-01 -8.08550179e-01 7.86959101e-03
-2.65028715e-01 1.95183799e-01 3.35263491e-01 7.69499481e-01
9.85724628e-01 -1.13372910e+00 -8.29029262e-01 -1.23999499e-01
5.35269320e-01 2.77774990e-01 -4.50876385e-01 2.65424043e-01
-1.22932732e+00 8.55897903e-01 -2.99054891e-01 -3.65049273e-01
-1.60635793e+00 5.58908820e-01 -2.04481766e-01 -4.65623617e-01
-7.92755663e-01 5.33226550e-01 -5.24402678e-01 -3.80572528e-01
4.17699248e-01 -5.64827204e-01 -1.18760371e+00 5.14589727e-01
1.00671399e+00 9.87711132e-01 7.98116207e-01 -5.42938948e-01
-6.77236855e-01 8.21656525e-01 -2.57234663e-01 -3.68762314e-02
1.53362083e+00 1.63651392e-01 4.98093665e-02 3.17969888e-01
1.33527565e+00 2.13965122e-02 -6.32584631e-01 -6.08564168e-02
7.14741230e-01 -5.22178233e-01 5.89992106e-01 -8.42179000e-01
-8.05431068e-01 7.80756891e-01 5.57317078e-01 7.43665993e-01
1.04701412e+00 -5.80830574e-02 2.46999681e-01 4.12587643e-01
2.31146395e-01 -1.49013579e+00 1.08598314e-01 4.11128521e-01
1.55250800e+00 -9.96256888e-01 7.82721579e-01 -7.13166058e-01
-7.04817772e-01 1.33566535e+00 2.88193256e-01 -1.43812289e-02
9.30752039e-01 7.33440399e-01 -1.95575044e-01 -7.45908558e-01
-4.14989024e-01 -3.92070502e-01 8.19679916e-01 1.08109331e+00
6.58140957e-01 -2.30004564e-01 -1.48495421e-01 6.52530134e-01
-6.84026897e-01 -1.57228455e-01 6.66741073e-01 1.41798639e+00
-9.63600874e-02 -9.34065521e-01 -4.62151974e-01 7.13448584e-01
-3.53512168e-01 1.45679906e-01 -1.19436848e+00 9.70881939e-01
-3.33004087e-01 1.11719096e+00 3.01973224e-01 -5.20922840e-01
8.32793057e-01 -2.96716988e-01 5.60212970e-01 -8.00073564e-01
-6.90113246e-01 -3.63902599e-01 3.96872669e-01 -6.80604875e-01
-2.23950073e-01 -8.17621171e-01 -1.40224695e+00 -1.93079412e-01
-2.29743406e-01 -2.30715588e-01 1.18031681e+00 6.39556885e-01
3.83863211e-01 7.08823979e-01 4.30631787e-01 -7.94069231e-01
3.68474036e-01 -1.02197981e+00 -3.57931137e-01 3.79491448e-02
2.24267378e-01 -4.80397075e-01 -1.84389904e-01 -6.00254424e-02] | [11.672042846679688, 2.6363253593444824] |
8b8203aa-16e3-434d-966a-902d51269b5d | lenia-biology-of-artificial-life | 1812.05433 | null | https://arxiv.org/abs/1812.05433v3 | https://arxiv.org/pdf/1812.05433v3.pdf | Lenia - Biology of Artificial Life | We report a new system of artificial life called Lenia (from Latin lenis "smooth"), a two-dimensional cellular automaton with continuous space-time-state and generalized local rule. Computer simulations show that Lenia supports a great diversity of complex autonomous patterns or "lifeforms" bearing resemblance to real-world microscopic organisms. More than 400 species in 18 families have been identified, many discovered via interactive evolutionary computation. They differ from other cellular automata patterns in being geometric, metameric, fuzzy, resilient, adaptive, and rule-generic. We present basic observations of the system regarding the properties of space-time and basic settings. We provide a broad survey of the lifeforms, categorize them into a hierarchical taxonomy, and map their distribution in the parameter hyperspace. We describe their morphological structures and behavioral dynamics, propose possible mechanisms of their self-propulsion, self-organization and plasticity. Finally, we discuss how the study of Lenia would be related to biology, artificial life, and artificial intelligence. | ['Bert Wang-Chak Chan'] | 2018-12-13 | null | null | null | null | ['artificial-life'] | ['miscellaneous'] | [-4.81954902e-01 -2.03445539e-01 2.19414547e-01 4.50942546e-01
9.22320604e-01 -1.04476011e+00 9.86150324e-01 -2.06121847e-01
-1.34892121e-01 1.05892015e+00 -1.43466353e-01 -1.74514666e-01
-2.26268992e-01 -1.07298672e+00 -3.29546750e-01 -1.16283429e+00
-6.43761218e-01 6.09478891e-01 5.43226004e-01 -7.94965148e-01
2.82014668e-01 7.19417393e-01 -1.73980856e+00 -4.07882750e-01
9.54095304e-01 2.83235252e-01 1.05998203e-01 9.36809182e-01
1.66611448e-01 2.37945810e-01 -2.60203242e-01 -7.03345537e-02
1.60704762e-01 -5.53899109e-01 -4.93850827e-01 1.34972304e-01
-3.60772908e-01 4.91359264e-01 -3.96554708e-01 8.01322281e-01
1.10879093e-01 2.44121268e-01 1.29078650e+00 -9.38713789e-01
-9.97690737e-01 2.13263586e-01 -2.00892001e-01 1.60514116e-01
7.84286857e-02 3.63923460e-01 2.15783402e-01 -6.96982801e-01
8.60491455e-01 1.21864295e+00 9.22466397e-01 8.12414169e-01
-9.68097091e-01 -8.18790942e-02 -2.45417610e-01 -2.75026232e-01
-1.77621210e+00 -3.03134412e-01 1.78986222e-01 -7.13994861e-01
9.52532649e-01 4.25183952e-01 1.22489452e+00 6.51833951e-01
1.06677055e+00 -1.33742616e-02 1.27501082e+00 -5.26803374e-01
7.74274349e-01 -2.53407974e-02 -3.35198380e-02 1.15336800e+00
1.04569423e+00 4.09514531e-02 -4.96751890e-02 -4.05602276e-01
1.17009223e+00 -2.11133763e-01 7.70582706e-02 -4.65387911e-01
-1.18099260e+00 2.88410902e-01 9.93466377e-03 7.80526400e-01
-2.11385205e-01 2.84868717e-01 -1.87193796e-01 9.00744647e-02
-1.04684897e-01 7.03749657e-01 -3.48969191e-01 -1.19120575e-01
1.02257393e-02 2.64015168e-01 1.17702258e+00 7.25113451e-01
6.75118923e-01 2.96830535e-01 5.73092043e-01 8.92009914e-01
1.42632991e-01 6.86795652e-01 7.17487216e-01 -1.09379649e+00
-5.17953873e-01 7.35914886e-01 6.10041767e-02 -1.05822074e+00
-6.85370386e-01 -3.41078371e-01 -1.09289861e+00 3.03916007e-01
3.63943577e-01 -1.48077145e-01 -3.80164117e-01 1.47966492e+00
3.00103188e-01 2.57527590e-01 1.77486837e-01 4.14401680e-01
2.36968443e-01 7.46205032e-01 -2.64355242e-02 -5.11171043e-01
1.22509801e+00 -6.81750596e-01 -5.21773577e-01 4.96303827e-01
4.26075161e-01 -5.54390140e-02 8.94837916e-01 2.23149151e-01
-1.10733795e+00 -1.08082585e-01 -1.17872441e+00 5.74739635e-01
-9.80473936e-01 -2.94589639e-01 6.18600011e-01 9.57031548e-01
-1.18228078e+00 5.92747211e-01 -1.16801047e+00 -1.08183491e+00
-1.51320547e-01 2.79310942e-01 -1.86625943e-01 9.17897224e-01
-8.54685068e-01 8.11893344e-01 7.93026239e-02 -5.96327066e-01
-5.86827219e-01 5.24491351e-03 -3.02811503e-01 -2.97580063e-01
-1.80488363e-01 -1.17702544e+00 7.96172202e-01 -4.86083478e-01
-1.81217420e+00 9.62583601e-01 -6.53574094e-02 -3.07493567e-01
-1.27413813e-02 6.34594321e-01 -6.56831026e-01 -6.84349611e-02
-6.67414814e-02 4.15446192e-01 2.02634141e-01 -1.48436356e+00
-4.68766451e-01 -2.82845914e-01 -2.10384682e-01 6.15219586e-02
-4.75137681e-01 -3.72889012e-01 -1.27650291e-01 -8.83956671e-01
2.31695008e-02 -1.23108327e+00 -5.99189222e-01 -1.03284664e-01
8.16025212e-03 -4.37644005e-01 6.08086288e-01 1.78623199e-01
1.31655049e+00 -1.98962665e+00 3.25567633e-01 2.33292893e-01
1.50572345e-01 -6.15420677e-02 1.04199395e-01 7.66548634e-01
6.86345339e-01 5.44592917e-01 -4.33721453e-01 3.25831205e-01
-1.89449087e-01 6.97039127e-01 -2.57979929e-02 6.84793174e-01
-2.12410927e-01 7.47275591e-01 -9.04741466e-01 -5.13530493e-01
-1.31518781e-01 2.69096792e-01 -6.28947616e-01 -3.01539660e-01
-2.16300506e-02 5.29978216e-01 -6.22523367e-01 8.45253170e-01
3.41240495e-01 -1.99019983e-01 -5.10057844e-02 6.01561189e-01
-7.20116735e-01 -4.56930608e-01 -8.09613645e-01 8.42535138e-01
-1.93647400e-01 4.22285646e-01 1.76441863e-01 -6.14235401e-01
9.32883382e-01 1.93227157e-02 4.47416365e-01 -2.13515699e-01
2.64001787e-01 4.52494442e-01 2.67195970e-01 -5.98813474e-01
4.24264550e-01 3.37945260e-02 -2.35995486e-01 5.87550819e-01
-3.60743292e-02 -3.77976120e-01 3.03109258e-01 -6.67729368e-03
1.22781181e+00 -1.62150890e-01 7.85356283e-01 -1.24611068e+00
7.34639049e-01 -9.26664285e-03 4.21458036e-01 7.91347504e-01
-4.28249151e-01 1.50123999e-01 8.11170712e-02 -6.75857425e-01
-1.14944839e+00 -1.24602890e+00 -3.72756630e-01 9.25890625e-01
9.07167017e-01 -1.97033629e-01 -1.15112841e+00 3.97124648e-01
6.68852702e-02 2.67571419e-01 -1.06167459e+00 -2.11667299e-01
-7.10069597e-01 -8.25729609e-01 6.95155025e-01 7.20111057e-02
6.37039125e-01 -1.26160431e+00 -9.66178060e-01 1.97823420e-01
3.61485273e-01 -7.04263151e-01 -1.69745028e-01 -2.50457413e-02
-9.43036020e-01 -6.67979002e-01 -2.35628307e-01 -7.78145373e-01
5.69265962e-01 1.19072966e-01 1.05360436e+00 4.42707151e-01
-4.12076712e-01 5.67555010e-01 -2.18185633e-01 -2.90653408e-01
-6.92765236e-01 -5.00915758e-02 1.01630628e+00 -2.25605220e-01
-1.57198742e-01 -8.95415366e-01 -4.10314083e-01 8.36558640e-01
-7.07847118e-01 -1.81090504e-01 2.19896361e-01 5.74672163e-01
6.97924435e-01 2.12086514e-01 6.07641459e-01 -3.98604721e-01
6.29283369e-01 -7.60583341e-01 -1.81809425e-01 2.74411976e-01
-3.49894285e-01 7.83166010e-03 9.18447077e-01 -4.98083353e-01
-9.09064054e-01 -5.59330545e-02 1.83510005e-01 3.25153232e-01
-2.84938484e-01 -1.30751431e-01 -1.92164611e-02 -4.53492969e-01
8.39481831e-01 7.00291514e-01 1.39846936e-01 -2.46546000e-01
2.04710618e-01 6.09143913e-01 7.83447266e-01 -9.78027642e-01
6.81739688e-01 9.90571022e-01 3.83494198e-01 -1.56055295e+00
1.75182834e-01 1.17526725e-01 -7.24143624e-01 -4.74602669e-01
8.00506711e-01 -1.83140561e-01 -7.58123875e-01 7.63491571e-01
-6.63723111e-01 -4.60894227e-01 -4.44295406e-01 -8.39831308e-03
-1.13471353e+00 1.97608650e-01 -7.15061069e-01 -1.06421256e+00
-1.03087708e-01 -5.35156786e-01 5.48744917e-01 5.48680723e-01
-2.95765102e-01 -1.40886354e+00 7.79080570e-01 -4.68961477e-01
6.91747308e-01 4.60102618e-01 9.58888173e-01 -2.92604685e-01
-2.72679657e-01 1.81830332e-01 5.58610618e-01 -6.66014135e-01
2.81317145e-01 6.97193086e-01 -3.66776586e-01 -2.11264342e-01
-1.56522706e-01 3.18989486e-01 5.56770802e-01 3.43609869e-01
7.50193894e-01 -5.98863304e-01 -7.94782639e-01 6.91564083e-01
1.34293890e+00 9.00970757e-01 2.37200230e-01 4.04369473e-01
7.51307309e-02 6.68684959e-01 9.39738303e-02 4.90621150e-01
3.06255102e-01 3.76237720e-01 2.79768467e-01 3.42104375e-01
4.00397554e-02 6.77080452e-02 4.56346780e-01 1.49741173e+00
-8.65172327e-01 -2.85808295e-01 -1.07029200e+00 5.34411073e-01
-1.80289710e+00 -1.06846201e+00 -7.11047277e-02 1.83028698e+00
4.63479608e-01 -2.50344425e-01 5.82399547e-01 -1.44679978e-01
1.01402688e+00 -7.02013612e-01 -6.06462717e-01 -6.98206663e-01
-8.37201118e-01 -1.76122069e-01 5.62595069e-01 4.23819363e-01
-8.51530075e-01 1.03389609e+00 8.38904190e+00 6.49671018e-01
-9.82489705e-01 -2.77536362e-02 5.18350780e-01 3.61176223e-01
-1.62844226e-01 -9.58112478e-02 -4.88880485e-01 5.60902536e-01
8.98495317e-01 -5.79882324e-01 5.62659860e-01 4.84241813e-01
1.94159120e-01 -7.98798352e-02 -4.97177571e-01 6.97612762e-01
-2.47667909e-01 -1.47050714e+00 2.19839469e-01 4.89214778e-01
1.05947328e+00 -2.45957673e-02 -5.60422987e-02 -2.50497945e-02
6.93423450e-01 -8.63388002e-01 6.78369641e-01 9.51755404e-01
7.13379323e-01 -6.32014453e-01 1.33407369e-01 5.74115753e-01
-1.43586397e+00 -3.21621925e-01 -6.29564643e-01 -5.97964823e-01
3.89797762e-02 1.40115721e-02 -2.91182846e-02 4.37414832e-02
7.57887423e-01 6.75823867e-01 -7.38274038e-01 1.19415641e+00
6.99417651e-01 4.94743556e-01 -5.56659341e-01 -1.03598785e+00
1.43626228e-01 -8.44166994e-01 9.50871885e-01 1.32061529e+00
6.41919434e-01 8.52632523e-01 -5.55744953e-02 7.09348023e-01
4.64907736e-01 2.33423397e-01 -9.81142223e-01 -1.07035294e-01
6.50339901e-01 1.20917761e+00 -1.50680590e+00 -4.22473013e-01
1.01889677e-01 3.90406817e-01 -2.62322128e-01 4.16147038e-02
-5.41523695e-01 -5.26294053e-01 8.54851484e-01 3.33514750e-01
4.30159084e-02 -7.55112827e-01 -6.40774906e-01 -1.19049263e+00
-7.35198677e-01 -4.16846246e-01 -1.25355730e-02 -5.09461939e-01
-1.40175724e+00 7.23385334e-01 -6.59228787e-02 -1.16428697e+00
-2.20534131e-01 -8.27668726e-01 -7.32456684e-01 -3.03294379e-02
-1.07150719e-01 -8.27743173e-01 -1.62290737e-01 6.76506877e-01
2.77890772e-01 -7.25959659e-01 8.47257853e-01 -4.24409211e-01
-4.71774638e-01 3.08594078e-01 7.71797836e-01 -3.97857726e-01
-7.58987069e-02 -1.11986578e+00 5.38980365e-01 4.08262193e-01
-2.52202213e-01 7.86266088e-01 1.07368457e+00 -5.50502181e-01
-1.61729968e+00 -9.68487859e-01 3.65526468e-01 -5.82103908e-01
9.27475929e-01 -4.23899859e-01 -7.43206322e-01 1.44107118e-01
4.29107368e-01 -3.58348668e-01 4.89739865e-01 -4.00369287e-01
1.89209044e-01 2.57317722e-01 -1.30351245e+00 1.31480753e+00
1.70627224e+00 4.13630195e-02 -4.47245657e-01 1.06212229e-01
4.86745656e-01 3.80781412e-01 -7.41067469e-01 3.20701271e-01
1.00178301e+00 -9.94766176e-01 6.44653320e-01 -4.02761608e-01
-1.99052796e-01 -5.75852394e-01 7.95750543e-02 -1.32954240e+00
-1.04517484e+00 -9.08303320e-01 1.10678631e-03 7.31449246e-01
1.66665345e-01 -1.27627337e+00 3.67927372e-01 -1.41585320e-01
-1.76214188e-01 -7.99153924e-01 -9.29820418e-01 -1.26014042e+00
7.28346825e-01 4.20516163e-01 7.18041480e-01 1.08814681e+00
7.00388610e-01 7.74530768e-02 1.26158595e-01 -2.08264083e-01
5.09563506e-01 -2.08369102e-02 5.94308615e-01 -1.43659985e+00
-3.37088078e-01 -9.80909586e-01 -6.58400357e-01 -6.86478376e-01
2.89914459e-02 -5.51309943e-01 -3.98504920e-02 -1.15882552e+00
1.47839576e-01 -6.76169038e-01 4.55033183e-02 2.48491868e-01
4.74542171e-01 3.91236097e-01 -3.00524652e-01 6.55055702e-01
-6.07436717e-01 3.07621449e-01 1.36392140e+00 2.38060981e-01
-3.93616825e-01 -3.11323017e-01 -3.51177335e-01 9.07831013e-01
1.08466959e+00 8.90402421e-02 -1.71482652e-01 1.03010997e-01
3.73422384e-01 -9.49687287e-02 -1.14215702e-01 -1.47174382e+00
3.91057253e-01 -7.37529814e-01 -1.06009163e-01 -2.35126212e-01
1.83892161e-01 -5.29783905e-01 5.96969366e-01 1.13360989e+00
1.93328336e-01 4.43220228e-01 -1.14960365e-01 6.38582826e-01
2.72458553e-01 1.29222246e-02 1.22228003e+00 -1.69364080e-01
-3.72903883e-01 1.39796570e-01 -1.69700646e+00 -1.76650435e-01
1.61725593e+00 -8.11662257e-01 -6.07107222e-01 -1.09113134e-01
-8.26694071e-01 -1.46871090e-01 1.32828748e+00 1.38648957e-01
3.03732641e-02 -1.31205046e+00 -4.20989931e-01 2.36109734e-01
1.35510817e-01 -6.84654772e-01 -1.28261700e-01 4.47912216e-01
-1.29017365e+00 4.44353044e-01 -9.69195962e-01 -5.20738661e-01
-6.74817204e-01 5.14471471e-01 7.60706484e-01 3.20275933e-01
-2.38352001e-01 3.77525449e-01 4.88561660e-01 -4.49251175e-01
-3.92164379e-01 -1.40977517e-01 -3.77401680e-01 -5.44975340e-01
1.38580725e-01 7.23511636e-01 -4.78817195e-01 -1.00769532e+00
-4.28554982e-01 1.11443603e+00 7.89208591e-01 -1.24483727e-01
1.01857197e+00 -3.54925394e-01 -9.30732310e-01 8.52910280e-01
3.98140073e-01 1.21119998e-01 -7.37748206e-01 3.33866835e-01
-5.15351072e-02 -6.96744695e-02 -9.56195474e-01 -3.30839872e-01
-4.72815871e-01 4.14291382e-01 2.27719039e-01 9.67489839e-01
8.70709717e-01 2.04612032e-01 4.47755933e-01 8.98387253e-01
9.57241833e-01 -1.06343126e+00 8.20126012e-02 1.16034102e+00
9.43435013e-01 -2.17650443e-01 -4.41142648e-01 -3.28398317e-01
-2.05306455e-01 1.00546432e+00 7.22538650e-01 -8.51836145e-01
1.11074686e+00 7.84363210e-01 -3.42541635e-01 8.45025629e-02
-1.35993290e+00 6.30093962e-02 -2.76871055e-01 9.55780327e-01
2.30646238e-01 4.54662055e-01 -8.55098188e-01 6.85617447e-01
-6.01731062e-01 -4.83299822e-01 9.49668169e-01 8.15888464e-01
-1.13168502e+00 -7.47409523e-01 -5.03728032e-01 8.02622810e-02
-1.26052797e-01 4.94492263e-01 -8.27342331e-01 7.87076294e-01
3.87036204e-01 6.90696537e-01 4.61358398e-01 -6.23341441e-01
-1.06132872e-01 -1.90520838e-01 5.44296801e-01 -2.17726901e-01
-4.23772395e-01 -2.90580630e-01 -1.97989896e-01 7.33646825e-02
-2.56986558e-01 -8.20292592e-01 -1.40394557e+00 -8.15899968e-01
-1.18860863e-01 5.58249772e-01 4.66167063e-01 7.31959939e-01
4.66858566e-01 -7.96554144e-04 5.43254852e-01 -7.72465706e-01
8.20766017e-02 -5.83891273e-01 -8.79258215e-01 1.10989973e-01
2.40342692e-02 -8.72518420e-01 -2.84851164e-01 4.52835739e-01] | [5.580287933349609, 4.140581130981445] |
220e99e8-85a3-4951-b934-81016e560877 | context-aware-relative-object-queries-to | null | null | http://openaccess.thecvf.com//content/CVPR2023/html/Choudhuri_Context-Aware_Relative_Object_Queries_To_Unify_Video_Instance_and_Panoptic_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Choudhuri_Context-Aware_Relative_Object_Queries_To_Unify_Video_Instance_and_Panoptic_CVPR_2023_paper.pdf | Context-Aware Relative Object Queries To Unify Video Instance and Panoptic Segmentation | Object queries have emerged as a powerful abstraction to generically represent object proposals. However, their use for temporal tasks like video segmentation poses two questions: 1) How to process frames sequentially and propagate object queries seamlessly across frames. Using independent object queries per frame doesn't permit tracking, and requires post-processing. 2) How to produce temporally consistent, yet expressive object queries that model both appearance and position changes. Using the entire video at once doesn't capture position changes and doesn't scale to long videos. As one answer to both questions we propose 'context-aware relative object queries', which are continuously propagated frame-by-frame. They seamlessly track objects and deal with occlusion and re-appearance of objects, without post-processing. Further, we find context-aware relative object queries better capture position changes of objects in motion. We evaluate the proposed approach across three challenging tasks: video instance segmentation, multi-object tracking and segmentation, and video panoptic segmentation. Using the same approach and architecture, we match or surpass state-of-the art results on the diverse and challenging OVIS, Youtube-VIS, Cityscapes-VPS, MOTS 2020 and KITTI-MOTS data. | ['Alexander G. Schwing', 'Girish Chowdhary', 'Anwesa Choudhuri'] | 2023-01-01 | null | null | null | cvpr-2023-1 | ['panoptic-segmentation', 'video-instance-segmentation', 'video-semantic-segmentation', 'multi-object-tracking', 'multi-object-tracking-and-segmentation'] | ['computer-vision', 'computer-vision', 'computer-vision', 'computer-vision', 'computer-vision'] | [-2.39627697e-02 -6.66263878e-01 -3.26269448e-01 -2.65148491e-01
-6.30178869e-01 -8.49763751e-01 4.76163149e-01 1.82404265e-01
-6.77376866e-01 4.51517671e-01 -2.28641063e-01 1.78853527e-01
3.80508602e-02 -4.45553720e-01 -6.16846800e-01 -5.33667922e-01
-1.04133829e-01 5.18177927e-01 1.61559486e+00 -2.48802692e-01
9.56592485e-02 6.60565376e-01 -1.86014128e+00 3.94641548e-01
4.55757529e-01 1.03547132e+00 3.97489071e-01 1.01338351e+00
-2.66104907e-01 9.12236691e-01 -4.94664162e-01 -1.35805309e-01
5.06464958e-01 -1.26815706e-01 -1.08997452e+00 3.64387542e-01
1.04444873e+00 -4.79053766e-01 -4.20536876e-01 8.55114520e-01
1.32808998e-01 3.98774445e-01 2.98708707e-01 -1.45245194e+00
-3.22407216e-01 1.47767991e-01 -7.84380257e-01 9.54109907e-01
5.86695135e-01 3.55484009e-01 7.95649946e-01 -5.38762212e-01
1.15636075e+00 1.38776290e+00 6.42791271e-01 6.14731610e-01
-1.02286124e+00 -4.49006408e-01 6.64076447e-01 3.49285364e-01
-1.38111758e+00 -4.95149314e-01 3.32036734e-01 -5.66000164e-01
7.45684743e-01 7.44493067e-01 9.22588468e-01 7.63904870e-01
2.93691549e-03 1.22500062e+00 5.80319464e-01 2.01237455e-01
-7.30571849e-03 -1.38060555e-01 1.16124853e-01 4.50027615e-01
-1.10157855e-01 -2.93757357e-02 -5.23669600e-01 8.85382965e-02
6.77836895e-01 2.94616908e-01 -4.00032789e-01 -3.11866045e-01
-1.41696596e+00 4.52470750e-01 2.23471835e-01 4.10177916e-01
-3.57178301e-01 5.64694464e-01 4.42417502e-01 2.15251818e-01
4.92428422e-01 -3.01252007e-02 -7.28648841e-01 -2.88028538e-01
-1.42004800e+00 5.44420242e-01 5.93412459e-01 1.36861479e+00
6.78622782e-01 7.28158504e-02 -4.60126072e-01 5.26324749e-01
3.98340940e-01 5.54139137e-01 2.04073578e-01 -1.35726285e+00
3.13883632e-01 3.47919971e-01 3.26128513e-01 -1.09063184e+00
-3.50308031e-01 6.57903925e-02 -4.25499409e-01 2.10869368e-02
4.53196824e-01 2.61738189e-02 -1.11672008e+00 1.46811211e+00
8.22633684e-01 6.69668853e-01 -2.88319141e-01 1.01517415e+00
1.07711196e+00 9.65127945e-01 3.76616985e-01 -4.57960188e-01
1.64392817e+00 -1.20563018e+00 -8.65671635e-01 -1.56691790e-01
3.08900923e-01 -9.45947886e-01 6.04103148e-01 1.45370722e-01
-1.44076645e+00 -9.22050714e-01 -5.92349350e-01 -1.30398795e-01
-5.40001273e-01 -1.69744357e-01 4.39697921e-01 3.92919689e-01
-1.12371290e+00 4.44514960e-01 -9.07855272e-01 -5.44511735e-01
5.41196406e-01 3.88845176e-01 -2.29582280e-01 -1.93918981e-02
-8.63751709e-01 4.04979259e-01 6.81575119e-01 -9.42331087e-03
-8.55580688e-01 -8.65411103e-01 -6.22980893e-01 -1.04946986e-01
7.90260732e-01 -7.18756080e-01 1.38548839e+00 -1.07266963e+00
-1.25698793e+00 8.61960471e-01 -3.75578374e-01 -4.10144657e-01
7.04517484e-01 -4.39018756e-01 -5.87922275e-01 4.93231058e-01
2.73243308e-01 1.28994942e+00 8.83573532e-01 -1.14135778e+00
-1.42482996e+00 -9.62222889e-02 2.37160429e-01 1.97040722e-01
1.07920781e-01 5.34210503e-01 -1.41684318e+00 -6.58775568e-01
1.42610416e-01 -9.69947696e-01 -1.95604578e-01 4.60163355e-01
-1.95887148e-01 -2.64271885e-01 1.76143324e+00 -4.41922873e-01
1.39897192e+00 -2.18289304e+00 8.94034058e-02 -3.07650626e-01
1.94993913e-01 4.55501050e-01 -1.29668757e-01 1.65477227e-02
2.92167276e-01 4.61725555e-02 -1.19708836e-01 -3.77826214e-01
-2.19535902e-01 3.99525315e-01 -1.70557693e-01 3.86095703e-01
9.10418108e-02 1.04630649e+00 -1.02053547e+00 -1.02047956e+00
5.20457506e-01 3.24301034e-01 -6.45558119e-01 -2.20699552e-02
-7.82450736e-01 3.95008296e-01 -4.12906975e-01 9.30103719e-01
7.00268388e-01 -3.23402911e-01 -2.81642199e-01 -4.91649747e-01
-3.14655602e-01 -2.68054724e-01 -1.60279560e+00 1.71171749e+00
3.07135046e-01 9.47610259e-01 1.83848277e-01 -6.39512479e-01
2.89585739e-01 3.33700180e-01 9.55128968e-01 -7.20846474e-01
-9.85218212e-02 4.65693437e-02 -3.80635947e-01 -7.95841753e-01
9.58682239e-01 4.64036703e-01 1.55606002e-01 3.83885130e-02
-1.03820041e-01 -3.00923418e-02 8.30123603e-01 3.22864562e-01
9.09365594e-01 4.85267460e-01 5.90990447e-02 -1.88039348e-01
5.61355054e-01 3.44357669e-01 7.83906877e-01 8.70804012e-01
-5.31147718e-01 9.39745903e-01 6.55897930e-02 -6.93223000e-01
-6.48247659e-01 -1.02369845e+00 9.80895162e-02 1.54953134e+00
8.57932448e-01 -4.40126270e-01 -4.71847594e-01 -6.40775144e-01
-1.14840940e-01 2.31019706e-01 -4.46993023e-01 5.85039139e-01
-9.84668076e-01 -5.58518112e-01 3.06609184e-01 4.79951650e-01
5.11348665e-01 -1.06619668e+00 -1.22348189e+00 5.63305855e-01
-4.51632977e-01 -1.58189845e+00 -1.01815832e+00 -2.04830378e-01
-7.87504792e-01 -1.23747349e+00 -8.87592554e-01 -6.03936672e-01
1.89524472e-01 6.53298855e-01 1.28025579e+00 1.11901760e-01
-5.31626046e-01 6.42906249e-01 -4.29955959e-01 -5.60069159e-02
-1.39208704e-01 -1.64841294e-01 -1.96205899e-01 9.78176817e-02
5.30603118e-02 -9.42017287e-02 -9.37784374e-01 7.54364610e-01
-1.11977887e+00 1.36516675e-01 2.25547493e-01 2.00507268e-02
9.74358857e-01 -1.87300652e-01 1.47902235e-01 -8.11108589e-01
-6.01170920e-02 -3.84170055e-01 -7.22081959e-01 4.79933649e-01
5.24372756e-02 -3.72225583e-01 1.32585928e-01 -7.62985468e-01
-8.28859806e-01 2.38244936e-01 -7.00577348e-02 -8.11107755e-01
-2.71171004e-01 -1.55530736e-01 3.23094755e-01 1.08899958e-02
3.22668731e-01 2.18985662e-01 -4.57883239e-01 -4.99075830e-01
4.65368271e-01 2.53823161e-01 9.11053538e-01 -5.23000896e-01
5.71819901e-01 8.94085050e-01 -2.06980065e-01 -8.85811865e-01
-6.28602087e-01 -1.23176503e+00 -6.14895344e-01 -3.87666106e-01
1.40694463e+00 -9.75363195e-01 -7.74274886e-01 3.63653064e-01
-1.40633881e+00 -4.03805196e-01 -4.77261901e-01 2.19316438e-01
-7.00948298e-01 4.45089936e-01 -6.12235963e-01 -4.68183011e-01
-2.83993572e-01 -1.31794453e+00 1.24553847e+00 5.53571939e-01
-1.06926151e-01 -8.48993719e-01 -5.88226058e-02 1.54304579e-01
4.64818388e-01 4.08455610e-01 1.28565937e-01 -4.56343859e-01
-1.17907906e+00 9.05416161e-02 -3.68025571e-01 -2.91600525e-02
7.93084800e-02 4.64302719e-01 -6.69353008e-01 -3.17328781e-01
-3.95398021e-01 7.27575943e-02 7.20719457e-01 5.76686502e-01
1.12222517e+00 -2.89052218e-01 -5.99810183e-01 7.44600236e-01
1.37048090e+00 4.41514075e-01 5.97775638e-01 2.47064024e-01
8.34151924e-01 3.99520278e-01 9.17389154e-01 3.96852165e-01
5.15526831e-01 1.08761728e+00 3.71733636e-01 -3.40842307e-02
-3.31734240e-01 1.80235967e-01 3.13234299e-01 4.55121219e-01
-2.37730175e-01 -5.25364757e-01 -7.01025486e-01 6.58771336e-01
-2.15096164e+00 -1.33136690e+00 -5.77240765e-01 1.72682750e+00
4.28189069e-01 9.53729544e-03 4.47883040e-01 -3.07057530e-01
6.20289981e-01 4.07143593e-01 -4.93891627e-01 1.54468164e-01
-2.18155026e-01 -1.83140710e-01 5.54852366e-01 4.07963842e-01
-1.38840115e+00 1.18313479e+00 6.08097982e+00 8.92375946e-01
-1.18317962e+00 4.17689055e-01 4.47577983e-01 -2.79604703e-01
1.56085342e-01 6.11569993e-02 -9.30839658e-01 6.56850994e-01
6.06696367e-01 1.24245256e-01 1.82965681e-01 6.55479670e-01
3.55378002e-01 -3.71364743e-01 -1.21273768e+00 1.27329993e+00
-1.24214284e-01 -1.64348173e+00 3.95637657e-03 -2.72111744e-01
8.36077094e-01 3.13477546e-01 -1.65871277e-01 1.57607004e-01
-4.36040498e-02 -5.05122602e-01 1.26469827e+00 6.54333830e-01
3.93772334e-01 -2.64401466e-01 3.28008771e-01 4.11754400e-02
-1.81880677e+00 6.55759545e-03 -2.49350369e-02 5.68419516e-01
6.02879763e-01 -6.43641800e-02 -3.44446689e-01 5.19222200e-01
1.29559946e+00 8.44478667e-01 -7.86827743e-01 1.42361581e+00
4.77299452e-01 1.64605170e-01 -7.48095870e-01 2.88261771e-01
5.60791373e-01 -3.84553410e-02 6.59722567e-01 1.60947323e+00
2.14430064e-01 3.25968802e-01 6.44996166e-01 5.53808093e-01
2.11913392e-01 -5.89062423e-02 -6.98612854e-02 1.31679401e-01
4.24320638e-01 1.12959146e+00 -1.27696514e+00 -8.22176516e-01
-4.86887187e-01 9.95552480e-01 -3.95615160e-01 6.56158328e-01
-1.14764357e+00 -1.46553302e-02 8.14024270e-01 5.32877743e-01
7.74299383e-01 -2.98848480e-01 4.37681109e-01 -1.12864280e+00
-5.63715771e-02 -7.23072529e-01 6.12176180e-01 -8.73426557e-01
-7.51728952e-01 6.38837337e-01 5.00135660e-01 -1.39390481e+00
-2.88684487e-01 -2.48980120e-01 -5.36827326e-01 2.01064885e-01
-1.52724183e+00 -1.10366690e+00 -5.16749263e-01 9.07369316e-01
1.13521302e+00 3.88174653e-01 8.68347958e-02 8.43369782e-01
-4.48340386e-01 1.14775069e-01 -6.58388510e-02 6.57748803e-02
5.57980537e-01 -1.04138041e+00 3.28981668e-01 9.01213586e-01
3.75211477e-01 1.14998192e-01 8.22547853e-01 -5.10389745e-01
-1.34241164e+00 -1.21973634e+00 4.42122787e-01 -5.92545629e-01
3.96545976e-01 -5.29630445e-02 -1.05024362e+00 6.68965399e-01
1.25273466e-01 6.77216709e-01 9.59205702e-02 -5.23731649e-01
-3.87696847e-02 -8.46749693e-02 -1.03372622e+00 4.00360018e-01
1.13200319e+00 -1.23335883e-01 -2.82687783e-01 4.86302465e-01
9.47444260e-01 -7.23555803e-01 -7.73085892e-01 4.26255494e-01
3.99450779e-01 -1.12778354e+00 1.35527563e+00 -5.58003485e-01
-9.89801809e-02 -7.70631850e-01 -2.78508246e-01 -4.40097511e-01
-3.68996859e-02 -1.01048827e+00 -4.28392738e-01 1.15986907e+00
-8.67039040e-02 -7.60581568e-02 9.28949714e-01 6.23041928e-01
-1.34478718e-01 -7.67078161e-01 -1.00420654e+00 -7.85446823e-01
-5.85410118e-01 -6.67894423e-01 3.99780542e-01 7.54727423e-01
-7.54498720e-01 -1.89902037e-01 -3.39552939e-01 1.89498633e-01
4.51617450e-01 1.95203319e-01 9.29932594e-01 -1.10860658e+00
-2.62324184e-01 -5.59817016e-01 -5.66703022e-01 -1.62942958e+00
-2.52046347e-01 -2.78420746e-01 5.26670553e-02 -1.45044184e+00
-1.99648493e-04 -5.55698395e-01 5.88405952e-02 4.54343885e-01
-6.32630214e-02 5.49276650e-01 7.32441068e-01 5.09004474e-01
-1.51620531e+00 9.89149660e-02 1.28280771e+00 -2.44355232e-01
-5.04651368e-01 1.30040914e-01 -8.68677571e-02 8.50796223e-01
3.94301265e-01 -6.11314356e-01 -4.56641823e-01 -6.06949151e-01
-1.30245164e-02 2.82972842e-01 5.56008816e-01 -1.06338394e+00
5.00613093e-01 -5.71264386e-01 1.82473093e-01 -1.07456338e+00
7.28901505e-01 -9.81314301e-01 6.21693552e-01 3.87103528e-01
2.09738910e-01 4.04948324e-01 4.23008800e-01 7.68041968e-01
-2.52496362e-01 -1.22192442e-01 8.48932743e-01 -2.39141047e-01
-1.43371737e+00 8.20971847e-01 -3.06691885e-01 3.97427917e-01
1.49917853e+00 -7.19254434e-01 -5.86899221e-02 -4.94251139e-02
-8.68451715e-01 6.54442072e-01 3.34501415e-01 8.48320365e-01
6.07902348e-01 -1.12216723e+00 -6.53093934e-01 -8.08310732e-02
2.60736141e-02 2.28184775e-01 5.91140449e-01 1.01407719e+00
-8.97619128e-01 3.51879179e-01 -2.65263245e-02 -1.33543217e+00
-1.57664001e+00 5.81663430e-01 2.58552283e-01 8.53496492e-02
-8.28933299e-01 8.48164201e-01 3.32056791e-01 1.88264251e-01
3.06419581e-01 -7.33327508e-01 -1.73851371e-01 3.99059743e-01
5.71814060e-01 4.20952380e-01 -3.28062624e-01 -9.97037768e-01
-4.22535539e-01 9.27258611e-01 -2.69776553e-01 1.29415728e-02
1.10393822e+00 -5.70093215e-01 1.97669432e-01 4.35607105e-01
1.16686153e+00 -2.65126437e-01 -1.67990804e+00 -3.28647792e-01
1.92876048e-02 -8.16590488e-01 -3.10387731e-01 -2.66939104e-01
-1.39085305e+00 5.86695194e-01 7.63924897e-01 4.77724612e-01
1.10779548e+00 2.59803206e-01 9.41127598e-01 1.98126554e-01
2.63034850e-01 -1.11468673e+00 1.23309299e-01 4.11213338e-01
6.53600335e-01 -1.30214632e+00 1.29305065e-01 -5.61099529e-01
-5.34503639e-01 1.09699667e+00 7.53170431e-01 1.74659580e-01
5.76935112e-01 1.48764208e-01 -3.14025063e-04 -2.30217740e-01
-7.11289823e-01 -4.29594427e-01 4.68287557e-01 4.18719769e-01
1.32411881e-03 -2.89217353e-01 3.09493132e-02 -2.00019658e-01
4.02416468e-01 -3.80669087e-02 3.14090997e-01 1.08236098e+00
-4.82436150e-01 -7.44679630e-01 -4.78230417e-01 3.14030558e-01
-5.91161609e-01 2.55394101e-01 1.78009048e-01 1.07148302e+00
3.35288316e-01 7.07319200e-01 3.55644733e-01 3.95655297e-02
2.44969904e-01 -1.84637547e-01 3.68929893e-01 -3.66285652e-01
-6.39148474e-01 3.03744256e-01 -2.93799698e-01 -9.02817786e-01
-1.08347309e+00 -7.95914710e-01 -1.38583291e+00 -2.68129140e-01
-3.40002328e-01 -5.21606207e-02 4.43448097e-01 7.83447504e-01
4.03121889e-01 6.09090805e-01 1.26653090e-01 -1.22699177e+00
-3.71262953e-02 -5.11678398e-01 -1.25479788e-01 7.38495588e-01
5.41919827e-01 -5.10520160e-01 2.73631811e-02 6.69207215e-01] | [9.074602127075195, -0.14852751791477203] |
f446ee99-e3be-47f2-b291-17ed4fe62860 | faster-voxelpose-real-time-3d-human-pose | 2207.10955 | null | https://arxiv.org/abs/2207.10955v1 | https://arxiv.org/pdf/2207.10955v1.pdf | Faster VoxelPose: Real-time 3D Human Pose Estimation by Orthographic Projection | While the voxel-based methods have achieved promising results for multi-person 3D pose estimation from multi-cameras, they suffer from heavy computation burdens, especially for large scenes. We present Faster VoxelPose to address the challenge by re-projecting the feature volume to the three two-dimensional coordinate planes and estimating X, Y, Z coordinates from them separately. To that end, we first localize each person by a 3D bounding box by estimating a 2D box and its height based on the volume features projected to the xy-plane and z-axis, respectively. Then for each person, we estimate partial joint coordinates from the three coordinate planes separately which are then fused to obtain the final 3D pose. The method is free from costly 3D-CNNs and improves the speed of VoxelPose by ten times and meanwhile achieves competitive accuracy as the state-of-the-art methods, proving its potential in real-time applications. | ['Yizhou Wang', 'Rujie Wu', 'Chunyu Wang', 'Wentao Zhu', 'Hang Ye'] | 2022-07-22 | null | null | null | null | ['3d-pose-estimation', '3d-multi-person-pose-estimation'] | ['computer-vision', 'computer-vision'] | [-3.88344496e-01 -3.00947487e-01 1.40366480e-01 -3.79652798e-01
-7.92632222e-01 -4.61792469e-01 4.00149703e-01 -3.02637797e-02
-5.75527430e-01 3.11162710e-01 2.62800097e-01 3.71618092e-01
3.24218273e-01 -7.24373817e-01 -6.56187534e-01 -2.29752257e-01
1.38783187e-01 1.12044728e+00 2.18042478e-01 1.69720441e-01
8.19030106e-02 5.92383206e-01 -1.40345383e+00 -3.82723302e-01
4.83209729e-01 1.02472067e+00 -4.98735905e-01 6.02170408e-01
2.03610197e-01 -3.12961429e-01 -5.28797626e-01 -5.16247928e-01
3.57457966e-01 1.83982942e-02 -3.05208772e-01 5.25013804e-01
9.94598210e-01 -6.67908549e-01 -3.85029584e-01 7.90805042e-01
4.14139777e-01 4.17762473e-02 5.44226646e-01 -1.03906083e+00
-2.90282428e-01 -2.93172449e-01 -9.77688611e-01 -3.81582558e-01
9.95631933e-01 -1.54923610e-02 6.82253778e-01 -1.12159979e+00
6.64448142e-01 1.57055795e+00 1.00154173e+00 5.53024471e-01
-1.08730590e+00 -5.90720057e-01 3.17873180e-01 -3.57171357e-01
-1.75542212e+00 -6.92141131e-02 7.06439853e-01 -5.58412254e-01
8.54771912e-01 1.37414321e-01 1.18267763e+00 7.17936873e-01
6.42238632e-02 5.61901748e-01 9.34265554e-01 -9.35208425e-02
-6.69598654e-02 -2.61505455e-01 -1.08553641e-01 1.00244451e+00
2.09176749e-01 -1.64551929e-01 -5.31671107e-01 -2.12310582e-01
1.33412516e+00 3.70700628e-01 -1.16106324e-01 -8.62040222e-01
-1.35109532e+00 6.32979035e-01 6.44260526e-01 -2.94777572e-01
-3.35879445e-01 3.63814652e-01 1.46037638e-01 -5.11424243e-01
6.50791168e-01 -3.51482742e-02 -3.97363722e-01 -4.73081693e-02
-8.55853856e-01 1.05589819e+00 4.83249158e-01 1.06201470e+00
5.94665229e-01 -4.36601877e-01 8.47944804e-03 4.00649011e-01
5.24859548e-01 8.08965623e-01 -1.61743030e-01 -1.01031554e+00
7.50181973e-01 8.56031716e-01 2.76686817e-01 -1.07215369e+00
-7.73330927e-01 -2.80298680e-01 -7.34842062e-01 4.34575826e-02
7.38703251e-01 -1.67786852e-02 -7.05761135e-01 1.32216895e+00
8.25456083e-01 2.97653414e-02 -6.67791784e-01 1.22759926e+00
6.80167854e-01 3.30022395e-01 -2.30537638e-01 3.04403126e-01
1.70334411e+00 -8.97887528e-01 -2.77179718e-01 -4.00549144e-01
1.62211880e-01 -4.44429576e-01 7.45669961e-01 4.34271365e-01
-1.16308033e+00 -4.67091799e-01 -9.51905906e-01 -3.63183737e-01
-1.94708966e-02 3.91975313e-01 6.46289051e-01 6.44922554e-01
-6.64813995e-01 4.41872448e-01 -1.11128831e+00 -1.22428469e-01
4.83035386e-01 4.90704328e-01 -6.69451416e-01 8.89603421e-02
-6.17434442e-01 6.94097936e-01 -2.38969952e-01 9.24976096e-02
-5.48479736e-01 -7.72211790e-01 -1.19117570e+00 -2.63387132e-02
4.34081882e-01 -1.07871735e+00 1.15591300e+00 1.49913235e-02
-1.45873511e+00 1.03294230e+00 -4.09851551e-01 -1.99622829e-02
1.07422018e+00 -8.43600631e-01 1.81746542e-01 1.26861483e-01
2.54699260e-01 5.91977537e-01 6.61086917e-01 -1.03251314e+00
-3.66256237e-01 -1.24771953e+00 -6.37346581e-02 4.55781728e-01
-1.94675550e-01 -6.24375828e-02 -9.89661992e-01 -2.67535150e-01
7.17309773e-01 -9.18570638e-01 -3.56740892e-01 6.73274457e-01
-7.53374994e-01 -3.03576648e-01 5.90129793e-01 -7.28710175e-01
6.70953631e-01 -1.99077535e+00 3.20484519e-01 2.03640729e-01
6.81034148e-01 -1.80156589e-01 3.73935312e-01 -5.22953905e-02
3.48446548e-01 -2.40921844e-02 1.69574976e-01 -1.00610459e+00
-2.45614331e-02 -2.94042200e-01 2.68866122e-01 1.06116998e+00
1.39371306e-02 8.42947364e-01 -7.47821927e-01 -5.80463529e-01
5.53393364e-01 9.01330054e-01 -6.92516804e-01 7.42986128e-02
1.30021170e-01 5.17273486e-01 -5.56229711e-01 7.15612531e-01
8.76470745e-01 -1.45049587e-01 -5.45168482e-02 -2.40514398e-01
-2.44458020e-02 4.45121266e-02 -1.41446531e+00 1.91176283e+00
-2.32081220e-01 3.14492822e-01 1.15395680e-01 -4.27320987e-01
8.97656620e-01 2.21614778e-01 7.00556695e-01 -2.32654348e-01
2.71125048e-01 -2.06106179e-03 -6.95542574e-01 -1.15728743e-01
3.75951469e-01 -2.19520614e-01 -4.83539969e-01 2.68185377e-01
-1.09504253e-01 -3.31489474e-01 -2.30551139e-01 -1.74410388e-01
4.82520700e-01 5.27069449e-01 3.13684165e-01 1.26095489e-01
3.92363459e-01 -2.57489353e-01 5.92216432e-01 2.03065693e-01
-2.72997260e-01 1.01389492e+00 4.39592481e-01 -7.85380423e-01
-1.17299926e+00 -1.52263176e+00 -3.04933283e-02 4.38412696e-01
1.76928505e-01 -8.20925653e-01 -9.49049056e-01 -5.78763843e-01
4.61899132e-01 1.65869836e-02 -5.18434942e-01 2.41510779e-01
-7.64677167e-01 -1.28272042e-01 2.08946109e-01 7.46199012e-01
4.73384261e-01 -1.29972965e-01 -9.01254237e-01 -1.90489069e-02
-1.87598661e-01 -1.38805664e+00 -8.55771005e-01 -4.31403190e-01
-7.07421362e-01 -1.21674132e+00 -1.01206875e+00 -4.32304412e-01
8.84717286e-01 2.70384878e-01 9.36255872e-01 -1.15517065e-01
-3.75103414e-01 2.61248887e-01 9.44922343e-02 -3.20591152e-01
6.18569613e-01 -2.36505613e-01 4.53845710e-01 -2.73529083e-01
4.05446053e-01 -4.26671416e-01 -8.94968212e-01 3.55592191e-01
-7.43680745e-02 -2.52772239e-03 -2.16263887e-02 2.90596783e-01
9.30177391e-01 1.09249703e-03 -2.26077199e-01 -4.85719621e-01
1.10077098e-01 2.13864282e-01 -8.76989067e-01 -2.22320437e-01
-6.41892180e-02 -4.12193507e-01 2.94254959e-01 -2.94697791e-01
-5.80925822e-01 5.76833725e-01 -2.44719014e-02 -5.13677955e-01
-1.48133814e-01 -2.76254490e-02 -1.90044224e-01 1.16695855e-02
3.82889032e-01 -5.21124564e-02 -4.62815985e-02 -6.32220328e-01
4.08986986e-01 4.14745688e-01 7.62551308e-01 -5.16772628e-01
1.02774632e+00 8.06242585e-01 2.26869941e-01 -6.38605595e-01
-1.08963180e+00 -5.89350164e-01 -1.37409329e+00 -3.91404480e-01
1.20978141e+00 -1.25551093e+00 -1.21614110e+00 5.67499638e-01
-1.53662217e+00 3.04509997e-01 1.04280092e-01 5.68416297e-01
-5.00969350e-01 4.22536343e-01 -4.62944061e-01 -8.26730490e-01
-4.05533046e-01 -1.20877433e+00 1.74657440e+00 2.91517824e-01
-4.04431939e-01 -6.50335252e-01 -4.69237417e-02 6.19917631e-01
-1.44986957e-01 6.25671446e-01 4.68492121e-01 -4.37719598e-02
-5.76982081e-01 -8.40276003e-01 -1.17020965e-01 -1.78967312e-01
-3.64646703e-01 -1.30295753e-01 -8.38055253e-01 -1.11156322e-01
-3.30526978e-01 -4.77284528e-02 4.87992704e-01 7.26612806e-01
1.11228764e+00 1.60671517e-01 -6.00917578e-01 9.52590048e-01
1.18718398e+00 -5.41738689e-01 1.40217990e-01 2.70659953e-01
8.90098870e-01 4.75715160e-01 2.88466483e-01 7.40470350e-01
8.85156810e-01 1.01024485e+00 4.39652532e-01 2.41768397e-02
2.02074740e-02 -5.53979635e-01 1.54152811e-01 3.90992999e-01
-3.50101262e-01 3.35693210e-01 -7.15882778e-01 2.08159223e-01
-1.58855152e+00 -4.99682903e-01 -2.64296114e-01 2.64435005e+00
4.15292740e-01 1.94864854e-01 6.49240434e-01 -3.83845419e-02
6.01235449e-01 1.57722011e-01 -5.43468416e-01 1.70729160e-01
2.30104402e-01 -2.39283647e-02 4.19544935e-01 3.49629641e-01
-1.31816339e+00 8.57263982e-01 6.61566067e+00 8.60784128e-02
-6.81112528e-01 4.47815321e-02 3.84217829e-01 -6.40750110e-01
1.15392908e-01 -1.68033436e-01 -1.20840287e+00 1.53803051e-01
4.09110427e-01 1.28043517e-01 2.26573542e-01 1.06109071e+00
3.42052765e-02 -1.63574964e-01 -1.28974056e+00 1.40960979e+00
3.66122931e-01 -1.08103335e+00 -2.09896877e-01 3.97965461e-01
5.53903580e-01 -1.83637142e-01 -6.99590519e-02 1.51776716e-01
-1.22186460e-01 -1.00894558e+00 1.13680398e+00 5.50321400e-01
9.06799376e-01 -8.83357286e-01 4.94978726e-01 5.60801208e-01
-1.53143716e+00 2.72904277e-01 -5.63817918e-01 -3.56468797e-01
5.52651525e-01 6.16084576e-01 -4.92846906e-01 2.43800387e-01
7.26482868e-01 5.27284682e-01 -4.87765729e-01 1.06297517e+00
-2.01962948e-01 -3.36889088e-01 -5.86403847e-01 -6.55852025e-03
-1.26779303e-01 -3.47490191e-01 4.50856447e-01 8.48290503e-01
4.83261198e-01 -9.00893286e-03 5.45704365e-01 1.00505877e+00
-1.07484296e-01 -1.07457682e-01 -3.40354115e-01 4.59016353e-01
5.08477449e-01 1.23968935e+00 -7.85097241e-01 -2.77933687e-01
-3.88264120e-01 1.13890386e+00 3.16723347e-01 -4.41837823e-03
-1.03004289e+00 -1.75036579e-01 8.55949581e-01 5.25879145e-01
2.88832396e-01 -7.19078124e-01 -3.45603704e-01 -1.34679258e+00
5.91320813e-01 -3.36370856e-01 3.43853310e-02 -7.50492811e-01
-1.04849577e+00 5.48385680e-01 4.49250825e-02 -1.26509261e+00
-1.04014277e-01 -8.50633502e-01 -2.62067169e-01 8.84797573e-01
-1.00128889e+00 -1.31499755e+00 -5.58226705e-01 5.66352129e-01
3.13564807e-01 2.38722488e-01 7.60594845e-01 8.26612711e-02
-4.73168015e-01 5.59982598e-01 -3.68643135e-01 4.47875082e-01
5.27943730e-01 -1.21836865e+00 8.20360780e-01 5.01969457e-01
1.77602172e-01 6.40800834e-01 4.43497509e-01 -6.88009501e-01
-1.64969289e+00 -8.42956066e-01 1.05834901e+00 -1.01175201e+00
1.73904866e-01 -9.89968598e-01 -2.77361274e-01 7.14560390e-01
-5.97243130e-01 3.95293593e-01 4.74433362e-01 4.32592362e-01
-7.37321556e-01 4.00744937e-02 -1.22220314e+00 5.42433739e-01
1.28770018e+00 -3.87871146e-01 -4.19604838e-01 5.34676194e-01
5.00609994e-01 -1.04249799e+00 -9.47166383e-01 -5.84478118e-02
8.14153790e-01 -1.06206751e+00 1.53843808e+00 -3.29778731e-01
2.85386860e-01 -3.31708968e-01 -7.38104954e-02 -1.09609580e+00
-3.47340405e-01 -2.32470036e-01 -2.20826969e-01 8.57075334e-01
-9.57216509e-03 -2.87534922e-01 1.20220792e+00 7.80061424e-01
1.90464437e-01 -8.58792841e-01 -1.32006049e+00 -5.00896096e-01
-1.27665147e-01 -6.91243947e-01 7.66867816e-01 2.49642953e-01
-1.38989925e-01 3.32715899e-01 -3.86098027e-01 2.57039547e-01
1.01365221e+00 2.42280707e-01 1.23693216e+00 -1.51753879e+00
-1.13996536e-01 -3.94115061e-01 -7.50941217e-01 -1.49493504e+00
-9.14177671e-02 -4.72711474e-01 -8.80145729e-02 -1.45801973e+00
4.10489678e-01 -3.66658047e-02 4.77990091e-01 2.62572527e-01
-2.10537121e-01 4.82620031e-01 3.57352853e-01 1.11559974e-02
-6.57685280e-01 5.87444782e-01 1.34821665e+00 2.33212505e-02
-1.07625430e-03 1.34916633e-01 -5.26595712e-01 1.08604681e+00
1.65169358e-01 -1.97991148e-01 1.77216530e-01 -5.63751161e-01
1.16467038e-02 1.15325704e-01 6.38497353e-01 -1.24653649e+00
6.39817491e-02 -3.28822657e-02 1.15263152e+00 -1.12088454e+00
9.77750123e-01 -7.42736518e-01 -5.33008464e-02 3.20401013e-01
1.01929203e-01 2.32716963e-01 -1.13926075e-01 5.02335906e-01
2.21444935e-01 4.04313117e-01 6.71437263e-01 -4.61087674e-01
-3.18269245e-02 7.84194529e-01 2.68321007e-01 -2.29417354e-01
1.12953460e+00 -4.61881280e-01 1.40133619e-01 -4.06312287e-01
-5.83832204e-01 2.49719307e-01 7.40102351e-01 3.29622090e-01
9.57481563e-01 -1.58578312e+00 -7.92807698e-01 4.98073071e-01
4.01760004e-02 5.82314551e-01 3.38294268e-01 8.85912538e-01
-7.05993831e-01 6.04577839e-01 1.85734071e-02 -9.87278283e-01
-1.48322296e+00 4.90680903e-01 4.44636643e-01 -1.36666611e-01
-1.05885291e+00 9.96183515e-01 1.27781838e-01 -7.25677729e-01
2.92013347e-01 -3.00841570e-01 1.66631997e-01 -2.82719851e-01
8.45523477e-01 4.21736389e-01 -1.41000792e-01 -1.08927572e+00
-6.84057057e-01 1.35393965e+00 8.05755183e-02 -2.43826181e-01
1.38720834e+00 -5.76859228e-02 6.06803484e-02 1.54726356e-01
1.23984504e+00 1.20676272e-01 -1.60799730e+00 -2.04709083e-01
-3.94120634e-01 -9.61813867e-01 -1.13050722e-01 -3.05267751e-01
-1.21617734e+00 9.45402920e-01 3.88596237e-01 -2.47314394e-01
5.68925619e-01 3.18941116e-01 9.47310984e-01 7.71098435e-02
5.96340299e-01 -9.08723712e-01 7.42221847e-02 4.08708453e-01
8.50823462e-01 -8.90781760e-01 6.11935079e-01 -8.37253451e-01
-3.12013328e-01 1.12589192e+00 6.53211713e-01 -5.32455564e-01
6.58350706e-01 1.16971552e-01 -3.85371521e-02 -3.54399532e-01
-8.68723765e-02 1.96032405e-01 6.68528676e-01 8.02194893e-01
4.98022258e-01 3.55802774e-01 1.62952214e-01 6.32515013e-01
-6.91927969e-01 -1.49974644e-01 7.46155903e-03 5.93816996e-01
-2.33599469e-01 -7.67715037e-01 -8.93114448e-01 3.51132870e-01
-1.95698053e-01 4.72504258e-01 -3.98759961e-01 7.84640193e-01
2.17158973e-01 5.85834086e-01 4.09856558e-01 -4.07308131e-01
7.29274273e-01 -1.24773659e-01 8.87402415e-01 -6.41534567e-01
-4.02667522e-01 4.66384292e-01 -1.02936439e-01 -9.21172202e-01
6.54094294e-03 -9.83190000e-01 -1.18084061e+00 -4.93211806e-01
-1.48022771e-01 -2.86758512e-01 8.44841957e-01 9.16704059e-01
2.21232846e-01 1.30271956e-01 2.90401965e-01 -1.57713962e+00
-5.43492734e-01 -7.08582640e-01 -5.80283403e-01 5.33033490e-01
2.15391353e-01 -1.03027785e+00 -6.62008896e-02 -2.10948274e-01] | [7.02118444442749, -1.0028704404830933] |
aad7d170-49f4-4362-a333-1cd579a0f8b4 | ogb-lsc-a-large-scale-challenge-for-machine | 2103.09430 | null | https://arxiv.org/abs/2103.09430v3 | https://arxiv.org/pdf/2103.09430v3.pdf | OGB-LSC: A Large-Scale Challenge for Machine Learning on Graphs | Enabling effective and efficient machine learning (ML) over large-scale graph data (e.g., graphs with billions of edges) can have a great impact on both industrial and scientific applications. However, existing efforts to advance large-scale graph ML have been largely limited by the lack of a suitable public benchmark. Here we present OGB Large-Scale Challenge (OGB-LSC), a collection of three real-world datasets for facilitating the advancements in large-scale graph ML. The OGB-LSC datasets are orders of magnitude larger than existing ones, covering three core graph learning tasks -- link prediction, graph regression, and node classification. Furthermore, we provide dedicated baseline experiments, scaling up expressive graph ML models to the massive datasets. We show that expressive models significantly outperform simple scalable baselines, indicating an opportunity for dedicated efforts to further improve graph ML at scale. Moreover, OGB-LSC datasets were deployed at ACM KDD Cup 2021 and attracted more than 500 team registrations globally, during which significant performance improvements were made by a variety of innovative techniques. We summarize the common techniques used by the winning solutions and highlight the current best practices in large-scale graph ML. Finally, we describe how we have updated the datasets after the KDD Cup to further facilitate research advances. The OGB-LSC datasets, baseline code, and all the information about the KDD Cup are available at https://ogb.stanford.edu/docs/lsc/ . | ['Jure Leskovec', 'Yuxiao Dong', 'Maho Nakata', 'Hongyu Ren', 'Matthias Fey', 'Weihua Hu'] | 2021-03-17 | null | null | null | null | ['graph-regression'] | ['graphs'] | [-2.42136672e-01 2.52687067e-01 -7.17691123e-01 -9.11539271e-02
-8.35978329e-01 -4.66093004e-01 4.32863504e-01 5.34166873e-01
-8.25173780e-03 8.88378203e-01 -1.02528809e-02 -5.64513564e-01
-3.06744725e-01 -9.69472885e-01 -8.36833119e-01 -2.11696699e-01
-7.77193248e-01 7.15020001e-01 2.67592400e-01 -2.71994293e-01
-8.98801759e-02 4.47521418e-01 -9.24465299e-01 1.04926966e-01
5.55201828e-01 6.94284499e-01 1.22252762e-01 7.39271700e-01
9.05518383e-02 6.61016762e-01 -3.82031351e-01 -5.98300099e-01
3.21563214e-01 9.87723544e-02 -9.65649664e-01 -1.63019508e-01
7.81403899e-01 9.70787331e-02 -7.96245515e-01 8.45775485e-01
5.51190436e-01 -1.44872919e-01 1.93377495e-01 -1.96774530e+00
-4.72733647e-01 9.53761339e-01 -6.29024863e-01 1.41919591e-02
4.52167451e-01 8.67458954e-02 1.49406826e+00 -4.78614300e-01
9.66913402e-01 1.23236442e+00 9.78634119e-01 2.43340626e-01
-1.16114092e+00 -6.21755302e-01 4.46729183e-01 2.50902414e-01
-1.40974045e+00 -1.26194477e-01 8.04752290e-01 -3.72418493e-01
1.28440070e+00 2.58743525e-01 7.29843259e-01 1.03637791e+00
4.77610016e-03 5.99936724e-01 7.90236413e-01 -3.26656640e-01
-5.33630699e-02 -2.10009038e-01 3.12719673e-01 1.17710006e+00
8.29700232e-01 -2.34800145e-01 -5.16502321e-01 -2.50668466e-01
6.17526889e-01 -2.40077779e-01 -1.74361080e-01 -6.27465069e-01
-1.24488568e+00 7.99617529e-01 8.01252961e-01 7.99150243e-02
-1.07027076e-01 5.58850050e-01 5.92030525e-01 5.79685152e-01
6.33690238e-01 5.01177132e-01 -6.95993602e-01 -4.62441333e-02
-3.25416058e-01 2.72598773e-01 1.18493581e+00 1.44068074e+00
8.93741548e-01 -2.31842980e-01 2.42811978e-01 7.64114082e-01
2.29193404e-01 2.90771842e-01 -3.13189417e-01 -7.76349723e-01
8.80248606e-01 9.38869357e-01 -4.87281591e-01 -1.45227551e+00
-6.26320899e-01 -3.69100630e-01 -1.02741230e+00 -2.65538305e-01
2.31932014e-01 7.74526075e-02 -5.56185424e-01 1.65018368e+00
2.58681566e-01 2.93429255e-01 -3.68937939e-01 4.46702451e-01
1.18983245e+00 3.60417873e-01 -5.95105141e-02 5.68711311e-02
1.06214166e+00 -1.18567741e+00 -3.75207365e-01 -3.49254608e-01
1.39427280e+00 -4.84077662e-01 1.18596101e+00 2.35076457e-01
-8.67111087e-01 -4.08641160e-01 -1.09844589e+00 -1.11236088e-01
-7.66055107e-01 -2.52876699e-01 1.34431434e+00 4.38629359e-01
-1.49365592e+00 6.63652778e-01 -6.64661169e-01 -7.67655015e-01
5.72318077e-01 4.17049259e-01 -7.59689391e-01 -4.04138863e-01
-1.10703254e+00 5.08550107e-01 3.98729742e-01 -2.36369297e-01
-7.27764368e-01 -7.14724183e-01 -8.79820764e-01 -2.52869844e-01
7.38205850e-01 -7.23802567e-01 6.82919562e-01 -9.48767439e-02
-7.09304452e-01 1.08277142e+00 2.38422513e-01 -4.39180166e-01
1.88424855e-01 1.96918044e-02 -5.71702600e-01 -7.00786859e-02
1.02920331e-01 5.51494718e-01 1.91858888e-01 -9.04726684e-01
-4.39805567e-01 -2.70715058e-01 2.87070185e-01 -3.23557220e-02
-5.51094174e-01 -2.27206334e-01 -1.07295370e+00 -4.28444803e-01
-1.63407922e-01 -1.12997973e+00 -2.95934409e-01 -3.87206674e-01
-7.79616177e-01 -6.21941090e-01 8.00483704e-01 -2.88192362e-01
1.40682292e+00 -1.85966468e+00 2.20763758e-01 2.48441771e-01
1.01257229e+00 1.46284199e-03 -5.81756890e-01 8.63781035e-01
-1.34172007e-01 4.20044929e-01 2.96977758e-01 -5.05276203e-01
1.31307185e-01 6.43042922e-02 8.50661099e-02 5.25055766e-01
-1.31343469e-01 1.31425130e+00 -9.01041210e-01 -6.00854933e-01
1.84141025e-02 3.33641954e-02 -4.24884558e-01 -4.04079743e-02
-3.63646060e-01 1.06790155e-01 -3.92028451e-01 9.36673462e-01
5.78295350e-01 -9.84229326e-01 6.54367149e-01 -3.38917673e-01
4.55353916e-01 3.62249583e-01 -9.69239175e-01 1.86825681e+00
-2.11439162e-01 6.53936505e-01 2.84280837e-01 -1.09195435e+00
8.33821476e-01 -2.94090927e-01 6.19458973e-01 -4.07116264e-01
-1.63320839e-01 2.25928992e-01 -6.26437813e-02 -1.83551274e-02
3.93611223e-01 4.95758891e-01 -2.83712208e-01 4.54698443e-01
5.07339165e-02 -2.68684089e-01 6.97127700e-01 1.09660256e+00
1.73504305e+00 -4.59449925e-02 3.34339231e-01 -3.64366591e-01
2.12911129e-01 2.47799847e-02 3.29798698e-01 7.11490631e-01
-2.19271183e-01 7.79269487e-02 8.95635605e-01 -6.79495633e-01
-7.19307303e-01 -7.49870777e-01 2.48233497e-01 1.08178616e+00
2.10704312e-01 -1.37406969e+00 -3.70372266e-01 -1.14232123e+00
4.93950486e-01 6.90245479e-02 -4.70241696e-01 -1.11389928e-01
-5.99938750e-01 -8.18246245e-01 4.53257769e-01 3.57702196e-01
2.97793537e-01 -7.41674185e-01 7.25587249e-01 2.15212796e-02
-3.65508385e-02 -1.57255054e+00 -5.38781106e-01 -1.47634923e-01
-9.89679933e-01 -1.50509083e+00 -1.89031899e-01 -9.89777207e-01
5.69362283e-01 4.59656894e-01 1.77822089e+00 5.84890604e-01
-5.50860882e-01 5.31224430e-01 -4.34067458e-01 -4.84442227e-02
-3.21467698e-01 8.59077930e-01 -4.61576832e-03 -6.97332084e-01
1.05841920e-01 -7.86993444e-01 -2.99366355e-01 3.40521991e-01
-1.77788556e-01 1.77120104e-01 5.91418862e-01 2.94808209e-01
5.65563321e-01 7.34070092e-02 8.06118667e-01 -1.49930191e+00
7.63928175e-01 -5.85353851e-01 -7.04705656e-01 2.81051129e-01
-1.06094658e+00 -1.68192327e-01 3.15543383e-01 -4.73692752e-02
-1.12446561e-01 -4.17361677e-01 2.46809609e-02 -1.88945442e-01
2.89257258e-01 7.14529932e-01 -1.82520449e-01 -5.11981726e-01
4.66988951e-01 -4.14617062e-01 -1.53389484e-01 -4.54327434e-01
4.68084961e-01 2.86346793e-01 7.68407285e-02 -8.39987338e-01
1.01621568e+00 -3.30276415e-02 5.77157497e-01 -6.83318734e-01
-7.56634712e-01 -4.95722383e-01 -5.45364201e-01 -2.39256576e-01
5.21954000e-01 -1.15631390e+00 -9.42003727e-01 5.61708152e-01
-7.92568862e-01 -9.21719015e-01 2.51203515e-02 3.10167938e-01
-1.59133315e-01 4.60644752e-01 -9.88916755e-01 -1.48344785e-01
-3.21564704e-01 -9.94394600e-01 1.22850180e+00 -1.38912424e-01
-8.55966937e-03 -1.62938344e+00 3.68147530e-02 5.38768649e-01
3.44863236e-01 5.10225058e-01 1.18387806e+00 -6.84016168e-01
-9.56589818e-01 -3.64726961e-01 -4.20112371e-01 9.26898122e-02
1.03198178e-01 -7.46250227e-02 -3.25943559e-01 -8.17041755e-01
-1.11591780e+00 -7.78586090e-01 7.68307269e-01 7.00238645e-02
1.19497144e+00 -5.08362614e-02 -8.92959297e-01 7.96293139e-01
1.44340360e+00 -5.94187856e-01 4.21248496e-01 2.27839798e-01
1.36133826e+00 2.08444566e-01 5.45760095e-01 4.47570579e-03
7.71382987e-01 7.37117708e-01 5.92480719e-01 -2.44992271e-01
-6.53309226e-01 -5.15585840e-01 2.36081511e-01 1.32586098e+00
-3.29768568e-01 -5.68325639e-01 -1.03003919e+00 3.25867653e-01
-1.86947632e+00 -3.68971378e-01 -6.90116644e-01 2.02413535e+00
5.60439944e-01 4.72846746e-01 2.86069125e-01 -2.57161975e-01
7.55532444e-01 5.01975119e-01 -4.79660988e-01 -2.07661036e-02
-2.54961252e-01 1.21495716e-01 7.21311688e-01 6.19362831e-01
-1.23365188e+00 1.28337193e+00 6.30396748e+00 1.10142744e+00
-7.05628991e-01 1.63080599e-02 5.27176976e-01 -4.29084189e-02
-2.91220725e-01 2.17431501e-01 -9.62912619e-01 1.37609929e-01
1.02279270e+00 -4.33142424e-01 6.24214590e-01 9.96631980e-01
-3.61951143e-01 1.92965001e-01 -1.15234542e+00 9.94207621e-01
-1.40213400e-01 -1.78228021e+00 -4.05107923e-02 6.77989900e-01
9.00952637e-01 6.27741158e-01 -2.04018578e-01 7.73224056e-01
7.79334426e-01 -9.61504459e-01 -9.57816616e-02 2.71779835e-01
1.07562184e+00 -3.96178842e-01 5.16506076e-01 1.73935637e-01
-1.79226887e+00 1.85454398e-01 -4.60339546e-01 -1.64753377e-01
-3.55517212e-03 8.63795280e-01 -9.34454560e-01 9.63821232e-01
6.63249552e-01 1.25477755e+00 -1.03630626e+00 7.27625549e-01
-1.77235276e-01 8.02441359e-01 -2.89534539e-01 -7.77379647e-02
1.27232755e-02 -2.24792182e-01 4.98583615e-01 1.05908036e+00
-2.27053035e-02 -3.35179716e-01 6.98103249e-01 5.15710235e-01
-9.42015231e-01 2.25348607e-01 -8.81315947e-01 -4.11351800e-01
7.53496945e-01 1.67136753e+00 -6.66625559e-01 -1.14915006e-01
-6.18557692e-01 5.38816988e-01 9.19648588e-01 3.02120894e-01
-9.18205559e-01 -3.77747089e-01 5.13462722e-01 3.09728384e-01
-1.24306418e-01 -4.40953732e-01 1.88205838e-01 -1.10943758e+00
-5.70447035e-02 -1.02433157e+00 7.15951085e-01 -5.57859719e-01
-1.61891162e+00 5.28441072e-01 -2.68241335e-02 -6.02550805e-01
2.16794401e-01 -7.90661693e-01 -5.46162665e-01 6.40510798e-01
-1.36539042e+00 -1.55349529e+00 -6.37271106e-01 5.81421733e-01
-1.37949195e-02 -3.42704684e-01 7.88812518e-01 5.52170694e-01
-6.07239127e-01 6.60784125e-01 -2.14564145e-01 2.29986966e-01
9.66889024e-01 -1.41769779e+00 1.02276695e+00 5.95327973e-01
3.49307030e-01 4.54427063e-01 2.41029859e-01 -1.02624106e+00
-1.91118670e+00 -1.38408852e+00 7.23189771e-01 -5.41228831e-01
1.04585218e+00 -8.83943439e-01 -5.61408818e-01 1.25345063e+00
-2.04416364e-01 4.84478652e-01 4.69585568e-01 8.97991836e-01
-4.08189654e-01 -4.64260638e-01 -7.15422213e-01 4.50233549e-01
1.70175254e+00 -5.18603206e-01 2.05927938e-01 1.04011703e+00
1.17522848e+00 -3.14377427e-01 -1.32785177e+00 5.03705442e-01
2.02790126e-01 -5.60975790e-01 1.03672743e+00 -8.80538523e-01
-6.23214506e-02 5.98484837e-02 -2.01222822e-01 -1.13538671e+00
-5.13915122e-01 -8.90515625e-01 -4.33649778e-01 1.20133829e+00
5.54128408e-01 -8.94798875e-01 1.10258424e+00 1.80611283e-01
-2.55225420e-01 -9.76539016e-01 -6.03071213e-01 -9.89207804e-01
3.45759206e-02 -3.11913043e-01 5.44784129e-01 1.08209050e+00
-3.32828774e-03 6.63219333e-01 -3.70732695e-01 8.06432404e-03
9.15929198e-01 2.78760880e-01 1.46638453e+00 -1.43052673e+00
-3.55969459e-01 -3.72912198e-01 -7.20068812e-01 -1.07649732e+00
3.43225986e-01 -1.63381994e+00 -6.97709262e-01 -1.93271863e+00
3.45073342e-01 -7.94995546e-01 -1.01860799e-01 7.67223418e-01
-3.17694157e-01 4.52724814e-01 2.12799251e-01 2.48049840e-01
-1.18733799e+00 1.30585015e-01 1.35686445e+00 -2.03635126e-01
4.92698187e-03 2.54235566e-02 -8.25487912e-01 4.35145795e-01
9.06974494e-01 -2.73897380e-01 -5.59061825e-01 -2.97243685e-01
7.23355353e-01 -2.54799247e-01 3.59270453e-01 -7.78368294e-01
3.59990925e-01 -4.32306528e-02 -1.71837918e-02 -3.89419466e-01
2.50567108e-01 -3.89871538e-01 1.42617136e-01 4.73754227e-01
-1.05572514e-01 1.68657154e-01 1.82400852e-01 7.60134041e-01
8.13488849e-03 2.88799644e-01 3.26437682e-01 -7.49522448e-02
-7.72151232e-01 9.07149494e-01 3.97621959e-01 4.36821640e-01
1.17986453e+00 1.32950082e-01 -8.91062379e-01 -5.86025894e-01
-5.58978021e-01 5.72269261e-01 4.23259079e-01 3.84537160e-01
3.62257421e-01 -1.36012185e+00 -8.95484805e-01 -1.75749790e-02
3.47265571e-01 -4.65928167e-02 8.58673379e-02 1.07370615e+00
-6.29778147e-01 3.70653212e-01 2.29834050e-01 -3.15871835e-01
-1.41255546e+00 7.30410218e-01 7.84299374e-02 -9.02683914e-01
-8.03920984e-01 1.04108334e+00 -1.18659690e-01 -7.57283509e-01
2.79109955e-01 -1.39098421e-01 3.21907252e-01 -4.04002339e-01
3.18785496e-02 6.05464518e-01 1.18538745e-01 -2.18433425e-01
-5.01965225e-01 4.51415420e-01 -1.39257491e-01 6.87604010e-01
1.58581603e+00 -1.18794203e-01 -3.89640003e-01 2.59557754e-01
1.31730306e+00 3.80796432e-01 -7.02181876e-01 -2.71630317e-01
1.50635764e-01 -1.70104295e-01 -1.63893402e-01 -7.17274129e-01
-1.24184215e+00 5.59829295e-01 -9.70547199e-02 5.56787372e-01
8.65165889e-01 5.44702828e-01 9.05187249e-01 5.95221162e-01
8.07254374e-01 -7.24288464e-01 1.68318212e-01 3.92033428e-01
8.27854872e-01 -1.32627606e+00 5.46741605e-01 -1.03307474e+00
-2.39221618e-01 9.33388531e-01 7.54046082e-01 -1.26512051e-01
9.33896363e-01 2.73320645e-01 -5.35411119e-01 -6.60178542e-01
-9.82967436e-01 -8.40211287e-02 3.54430795e-01 7.21609652e-01
3.48189116e-01 3.62808615e-01 -3.89997996e-02 4.43271875e-01
-2.47193232e-01 -3.54267031e-01 2.74234056e-01 5.88790178e-01
-1.95266157e-01 -1.60912883e+00 2.65809804e-01 6.88863814e-01
1.12750279e-02 -2.75069684e-01 -7.55966485e-01 1.29421246e+00
-2.18541041e-01 9.13566709e-01 -2.99204230e-01 -6.48911417e-01
3.12162578e-01 -3.68796378e-01 6.72867894e-01 -6.77074671e-01
-1.96589842e-01 -3.06285053e-01 7.55008161e-01 -9.90597785e-01
-4.95620966e-02 -2.73297846e-01 -1.07012308e+00 -9.47922826e-01
-3.20431262e-01 1.83906332e-01 6.11567438e-01 2.34897867e-01
6.57940626e-01 5.99997163e-01 4.20210093e-01 -5.75436056e-01
-2.25721359e-01 -9.53986645e-01 -8.97238910e-01 4.11292851e-01
-2.18846902e-01 -5.29465020e-01 -5.35946012e-01 -3.37452114e-01] | [7.012141227722168, 6.144567012786865] |
ebeed233-6985-46e2-972d-d87238d5fd2a | conditional-generative-data-free-knowledge | 2112.15358 | null | https://arxiv.org/abs/2112.15358v4 | https://arxiv.org/pdf/2112.15358v4.pdf | Conditional Generative Data-free Knowledge Distillation | Knowledge distillation has made remarkable achievements in model compression. However, most existing methods require the original training data, which is usually unavailable due to privacy and security issues. In this paper, we propose a conditional generative data-free knowledge distillation (CGDD) framework for training lightweight networks without any training data. This method realizes efficient knowledge distillation based on conditional image generation. Specifically, we treat the preset labels as ground truth to train a conditional generator in a semi-supervised manner. The trained generator can produce specified classes of training images. For training the student network, we force it to extract the knowledge hidden in teacher feature maps, which provide crucial cues for the learning process. Moreover, an adversarial training framework for promoting distillation performance is constructed by designing several loss functions. This framework helps the student model to explore larger data space. To demonstrate the effectiveness of the proposed method, we conduct extensive experiments on different datasets. Compared with other data-free works, our work obtains state-of-the-art results on CIFAR100, Caltech101, and different versions of ImageNet datasets. The codes will be released. | ['Libo Zhou', 'Yang Yang', 'Xinyi Yu', 'Linlin Ou', 'Ling Yan'] | 2021-12-31 | null | null | null | null | ['conditional-image-generation'] | ['computer-vision'] | [ 4.71519113e-01 2.61747122e-01 -3.18891406e-01 -4.31553423e-01
-5.18852055e-01 -4.16350543e-01 4.33599085e-01 -2.43427485e-01
-6.82703912e-01 1.06897247e+00 -1.42721951e-01 -2.79934436e-01
2.55993642e-02 -1.10279524e+00 -9.99079347e-01 -1.06463504e+00
3.06716800e-01 2.69249290e-01 1.25643713e-02 1.72810495e-01
-1.21334128e-01 1.76131293e-01 -1.19835103e+00 4.85425033e-02
1.06881523e+00 1.01183033e+00 4.50826705e-01 2.61988521e-01
2.44099528e-01 8.45951676e-01 -5.74721158e-01 -6.74020708e-01
2.96065211e-01 -4.47462052e-01 -7.10211337e-01 -9.07510966e-02
2.62243211e-01 -5.82475245e-01 -6.45592511e-01 1.35333276e+00
5.14717877e-01 1.70650557e-01 5.69354653e-01 -1.18661523e+00
-1.02333760e+00 8.73219728e-01 -2.38889456e-01 -1.09189652e-01
-2.58656204e-01 2.80021966e-01 7.59619296e-01 -8.13041151e-01
4.14212435e-01 1.06374216e+00 1.85692325e-01 8.47326398e-01
-1.10666811e+00 -1.26251042e+00 -9.47950333e-02 2.43830353e-01
-1.57066441e+00 -3.59503269e-01 9.06710982e-01 -1.16813727e-01
2.45700493e-01 -2.50180177e-02 4.86678243e-01 1.11982346e+00
-4.52441245e-01 1.04194605e+00 1.13412285e+00 -3.45889360e-01
1.26370341e-01 5.62960207e-01 -1.89161360e-01 7.33617544e-01
3.15319121e-01 3.11675519e-01 -3.61967593e-01 2.52993423e-02
7.51761734e-01 2.49616489e-01 -6.91083550e-01 -4.02804881e-01
-1.00281167e+00 8.96402419e-01 6.47320092e-01 -1.14643209e-01
-3.80755924e-02 5.52570187e-02 1.59857407e-01 3.14318568e-01
3.17144632e-01 1.21414013e-01 -4.31532145e-01 1.07559137e-01
-8.77860188e-01 2.02184096e-01 6.96007431e-01 1.12332737e+00
9.54111874e-01 1.34509414e-01 -2.81532705e-01 7.70616651e-01
4.19120044e-01 5.36072433e-01 6.83403194e-01 -6.25378311e-01
7.48333037e-01 4.00691688e-01 -3.82970244e-01 -9.07771647e-01
3.87771130e-01 -6.36667073e-01 -1.17256582e+00 -1.01933014e-02
-2.46723369e-02 -4.08003688e-01 -1.15984845e+00 1.85315752e+00
3.55355173e-01 6.90421581e-01 5.71489930e-01 6.50617123e-01
9.75669563e-01 7.65202582e-01 6.13072738e-02 -1.17337413e-01
9.67931747e-01 -1.02876842e+00 -6.40279353e-01 -9.99090970e-02
4.65350091e-01 -3.33063662e-01 9.59180415e-01 4.50135678e-01
-9.43865716e-01 -5.93578696e-01 -1.19336224e+00 -3.45041528e-02
-3.14011812e-01 3.33682686e-01 5.94470322e-01 6.15790427e-01
-6.40068829e-01 4.53033894e-01 -7.22313404e-01 3.09595764e-01
1.12240863e+00 3.39022428e-01 -3.12574208e-01 -4.09946650e-01
-1.37476635e+00 5.07110715e-01 1.06366479e+00 1.40064759e-02
-1.47056234e+00 -6.89768791e-01 -8.80475402e-01 2.49826416e-01
4.87674534e-01 -5.46703041e-01 1.09153855e+00 -8.41856420e-01
-1.56222486e+00 6.53331220e-01 4.62131113e-01 -6.89704537e-01
6.71838582e-01 -6.18090220e-02 -1.22732267e-01 9.54789668e-02
-3.13176125e-01 8.18496406e-01 1.01006472e+00 -1.29942155e+00
-4.12119627e-01 -1.17217846e-01 -7.58158276e-03 1.93859354e-01
-6.72101676e-01 -4.74170446e-01 -5.46723604e-01 -7.37970591e-01
-2.10886046e-01 -7.91931272e-01 -2.14100733e-01 -1.68210030e-01
-6.04229450e-01 -8.74593034e-02 1.06156981e+00 -5.09597123e-01
1.17444444e+00 -2.34671140e+00 -3.78730223e-02 3.39241683e-01
2.99071550e-01 7.77859330e-01 -3.07171028e-02 -8.31834301e-02
-8.74616876e-02 2.19887644e-01 -4.60361302e-01 -5.50803185e-01
-9.10626650e-02 5.37267566e-01 -6.25057280e-01 1.04688033e-01
1.50327787e-01 9.13257658e-01 -7.51860201e-01 -5.99823296e-01
3.35522331e-02 6.85073197e-01 -6.70683682e-01 4.46772605e-01
-2.97305018e-01 4.76160079e-01 -6.45107329e-01 3.71010393e-01
9.79577243e-01 -1.70319065e-01 -2.08111722e-02 -7.68568218e-02
3.85903955e-01 4.91816327e-02 -9.69630182e-01 1.77239156e+00
-3.42158854e-01 2.69709289e-01 -2.08197668e-01 -1.26837552e+00
1.04716682e+00 3.35372597e-01 -1.19913869e-01 -2.92136282e-01
3.79619658e-01 6.90619722e-02 -7.78628513e-02 -3.63449425e-01
2.19364092e-01 -4.78727631e-02 1.73150718e-01 2.95002759e-01
3.06665212e-01 -1.08263396e-01 -2.48073917e-02 3.14373136e-01
7.73612559e-01 6.81950292e-03 -1.18225083e-01 1.75517052e-01
5.72407246e-01 -3.61770034e-01 8.15105617e-01 5.12964070e-01
-1.83505397e-02 5.79983771e-01 2.84318000e-01 -1.59803331e-01
-7.34175980e-01 -1.03957355e+00 -1.34885415e-01 6.47125781e-01
3.48150097e-02 -5.13847530e-01 -8.60600173e-01 -1.04190874e+00
-1.63628951e-01 6.96900189e-01 -6.32271647e-01 -6.54391825e-01
-3.72547686e-01 -6.68851793e-01 6.82635128e-01 4.97860253e-01
1.20953131e+00 -1.05241585e+00 -2.05794737e-01 -4.25482765e-02
-2.07060784e-01 -1.05133104e+00 -2.65619159e-01 1.19755790e-01
-8.47096741e-01 -9.93380427e-01 -6.85017824e-01 -1.00390029e+00
1.13197744e+00 1.01682827e-01 6.20757401e-01 9.94781703e-02
-1.91265479e-01 -7.76624978e-02 -2.59037048e-01 -6.10480905e-01
-2.61841685e-01 1.94060564e-01 -1.21527314e-01 8.54080170e-02
3.45870584e-01 -7.25101054e-01 -6.33386970e-01 2.81258039e-02
-1.12351775e+00 2.57681131e-01 8.73345375e-01 1.06816673e+00
9.44613397e-01 4.89551038e-01 5.48388004e-01 -1.13756335e+00
6.27582848e-01 -5.69320619e-01 -7.03076839e-01 2.48128280e-01
-7.84719348e-01 2.98034370e-01 7.86822915e-01 -6.28638685e-01
-1.29168439e+00 1.87803030e-01 -4.52679992e-02 -8.41014981e-01
-1.24605149e-01 5.49974680e-01 -4.35684651e-01 -1.61774918e-01
5.48470795e-01 6.68805242e-01 -1.05499156e-01 -5.75560689e-01
4.85178411e-01 6.10208273e-01 7.82282591e-01 -7.81643569e-01
1.04285622e+00 2.69221634e-01 -1.74635679e-01 -3.08488131e-01
-9.06366706e-01 1.95012659e-01 -3.96971345e-01 3.59081924e-01
7.66364276e-01 -1.14920580e+00 -5.84483862e-01 4.86190200e-01
-9.08031166e-01 -2.60690510e-01 -4.41058248e-01 6.28936172e-01
-3.44084680e-01 9.32586193e-02 -3.26612443e-01 -4.98810202e-01
-4.75572437e-01 -1.17696011e+00 4.67636257e-01 5.01856804e-01
7.51831472e-01 -8.70189548e-01 1.00743445e-02 4.40760434e-01
3.77614230e-01 2.52129614e-01 7.46054947e-01 -8.67953420e-01
-9.30026054e-01 -2.02531248e-01 -2.28734270e-01 9.56479788e-01
-4.39024484e-03 -2.60053188e-01 -1.05008531e+00 -3.51098448e-01
1.01096854e-01 -9.55658615e-01 1.27698958e+00 -9.90273356e-02
1.93713582e+00 -7.56332397e-01 -3.81559938e-01 1.11304331e+00
1.30508327e+00 1.32353514e-01 7.15757728e-01 -1.71321221e-02
8.27498019e-01 1.70694679e-01 4.15568173e-01 4.54062790e-01
4.37042445e-01 1.05117761e-01 5.73406637e-01 1.92787610e-02
5.22317272e-03 -7.41932690e-01 1.31706163e-01 7.75152445e-01
-5.25421351e-02 -1.41849115e-01 -6.36564612e-01 5.19712687e-01
-1.63701534e+00 -7.81664670e-01 4.27804053e-01 2.17326880e+00
1.58061647e+00 4.96280491e-02 -4.84454036e-01 5.60481735e-02
5.86884320e-01 4.53229696e-02 -6.57824457e-01 1.00183316e-01
2.30230428e-02 5.52679062e-01 6.21400893e-01 2.10439876e-01
-1.15145993e+00 1.05380309e+00 4.92485476e+00 1.21598160e+00
-1.18585229e+00 1.86245859e-01 8.39182675e-01 -2.75452435e-02
-3.64650697e-01 8.20697546e-02 -9.54597354e-01 5.31552732e-01
7.00482786e-01 -3.06059957e-01 3.98728728e-01 1.16188121e+00
-2.39228591e-01 2.20237404e-01 -9.42650616e-01 1.11235702e+00
5.96428663e-02 -1.27694142e+00 2.18044877e-01 -2.61309277e-02
1.05577028e+00 -5.99186197e-02 3.54650289e-01 5.70100009e-01
6.14672601e-01 -1.24152243e+00 2.00426981e-01 4.44944948e-01
1.04284418e+00 -1.10178661e+00 6.04485631e-01 6.11077845e-01
-8.68915498e-01 -4.92051058e-02 -8.23468924e-01 2.24074095e-01
-1.88237324e-01 7.03711152e-01 -1.09863186e+00 5.84018528e-01
5.14517486e-01 6.70173764e-01 -5.57959497e-01 9.75417376e-01
-7.78441072e-01 1.02154577e+00 -2.77480781e-01 2.93344289e-01
9.51542929e-02 -1.37444004e-01 6.24060817e-03 9.10988927e-01
2.40535676e-01 2.52909303e-01 2.12982610e-01 1.02378082e+00
-7.55774558e-01 5.64789809e-02 -6.88535929e-01 -1.78084120e-01
7.46955514e-01 1.12852180e+00 -2.11156249e-01 -3.68078083e-01
-1.58361524e-01 9.07049358e-01 6.52306795e-01 4.69225943e-01
-7.76269794e-01 -7.05702364e-01 4.15756017e-01 -2.82678187e-01
4.27648246e-01 -8.57514367e-02 5.68883605e-02 -1.38805783e+00
-3.01631931e-02 -7.95812428e-01 3.74866188e-01 -5.50952673e-01
-1.14068925e+00 4.14016634e-01 1.01313524e-01 -1.08980012e+00
-8.76289606e-02 -2.11199820e-01 -8.19846332e-01 9.43999708e-01
-1.86729014e+00 -1.12176692e+00 -4.58230168e-01 1.08092058e+00
6.18494563e-02 -3.85293037e-01 8.22295427e-01 3.58591735e-01
-6.96204543e-01 1.10272777e+00 2.37279311e-01 6.38649821e-01
8.06883991e-01 -9.63517189e-01 3.20398770e-02 8.53360951e-01
1.67432651e-01 7.62452483e-01 1.95683971e-01 -4.92056310e-01
-1.10737455e+00 -1.42750633e+00 5.50839245e-01 1.43731823e-02
2.03008711e-01 -4.11068946e-01 -9.45487142e-01 9.71365392e-01
1.47028461e-01 3.25707614e-01 7.22325325e-01 -2.97063887e-01
-4.52668697e-01 -3.64221394e-01 -1.34145737e+00 3.34177673e-01
8.27639461e-01 -4.00576770e-01 -6.27013683e-01 4.16716754e-01
1.17363489e+00 -5.19237816e-01 -7.44482815e-01 4.69275713e-01
2.08472162e-01 -4.52627510e-01 9.42827582e-01 -5.74389338e-01
5.72080672e-01 -2.75919318e-01 5.37596047e-02 -1.39021385e+00
-5.18961577e-03 -6.04854763e-01 -2.54859298e-01 1.46850574e+00
3.40134859e-01 -7.39015996e-01 1.01054370e+00 4.45761889e-01
-2.60664243e-02 -8.96963477e-01 -6.37963355e-01 -6.99359119e-01
1.46011770e-01 -2.30085358e-01 8.95476401e-01 1.25808775e+00
-4.03238446e-01 1.90681070e-01 -6.16087317e-01 1.24596789e-01
7.55408347e-01 2.09871903e-02 8.49236786e-01 -9.29604471e-01
-4.76812929e-01 -1.01346411e-02 -2.91816473e-01 -1.21422553e+00
2.56953478e-01 -1.18455374e+00 -3.76013368e-02 -1.13429236e+00
1.90436289e-01 -9.13294077e-01 -5.00036538e-01 9.38509822e-01
-2.82133579e-01 8.84901658e-02 7.43290260e-02 9.62520689e-02
-4.15898055e-01 9.92730200e-01 1.66295564e+00 -1.88719273e-01
4.92420122e-02 7.06517026e-02 -8.93412828e-01 4.80652213e-01
1.03559566e+00 -6.70046270e-01 -1.11633539e+00 -6.00857496e-01
-7.79474527e-02 -2.14359999e-01 4.92888242e-01 -1.07346177e+00
3.71889114e-01 -2.67818213e-01 4.79259074e-01 -4.11643535e-01
2.46695384e-01 -9.05105054e-01 -4.52839136e-02 4.64786708e-01
-5.34122467e-01 -5.52948177e-01 -4.76381853e-02 7.14318454e-01
-4.04290199e-01 -3.52510661e-01 8.05559337e-01 -1.05837964e-01
-4.65400070e-01 8.92047644e-01 5.44634819e-01 2.61568487e-01
1.08170664e+00 2.40854397e-01 -5.04886746e-01 -2.13139802e-01
-5.70183814e-01 4.34530646e-01 9.84039977e-02 1.93206102e-01
9.20137942e-01 -1.49524248e+00 -7.43344963e-01 6.41105115e-01
4.74428535e-02 7.41316080e-01 2.39185914e-01 3.77787799e-01
-3.75997484e-01 1.60349265e-01 -1.74465224e-01 -2.94204444e-01
-1.12102818e+00 7.02121675e-01 3.47725451e-02 -2.38887057e-01
-4.00218487e-01 1.24776745e+00 3.75655115e-01 -5.56803226e-01
4.47625607e-01 -2.56436974e-01 -6.34699389e-02 -3.25132489e-01
6.81907475e-01 -5.21869399e-02 -2.68475056e-01 -1.61110356e-01
4.21988629e-02 -7.22370856e-03 -5.51547766e-01 8.06130096e-02
1.19635272e+00 1.09007567e-01 1.40633926e-01 -1.46668166e-01
1.48788035e+00 -1.62489682e-01 -1.49828601e+00 -7.02759445e-01
-7.58797765e-01 -5.60962319e-01 1.43895671e-01 -8.46718013e-01
-1.57504880e+00 1.15043950e+00 5.09006798e-01 -2.55098671e-01
1.36735213e+00 -1.61423802e-01 9.99327660e-01 5.83113134e-01
2.32698768e-01 -8.44219863e-01 6.49961606e-02 4.07003105e-01
5.35316229e-01 -1.19606590e+00 -2.61175986e-02 -3.01588267e-01
-6.54562891e-01 8.22180569e-01 8.86013210e-01 -3.23615484e-02
8.19005847e-01 5.28395437e-02 -2.21266732e-01 9.92183834e-02
-7.27123141e-01 -1.74907073e-02 3.10799666e-02 6.81544006e-01
-9.92063284e-02 9.68120471e-02 -2.38701597e-01 9.90216434e-01
-4.55536664e-01 2.51754999e-01 2.60839105e-01 9.80707824e-01
-4.11101490e-01 -1.27505672e+00 -9.62883979e-03 4.53323931e-01
-4.99805719e-01 -4.09730464e-01 -1.12106435e-01 5.89887559e-01
4.44538385e-01 6.36605382e-01 -3.39815050e-01 -5.76134026e-01
6.14349265e-03 7.06668124e-02 3.17401081e-01 -6.21697843e-01
-2.58903623e-01 -4.12891060e-01 -2.79291689e-01 -1.48193955e-01
-1.82964712e-01 -2.49465004e-01 -1.31792378e+00 -2.71271557e-01
-4.93234396e-01 4.33910340e-01 5.98368645e-01 7.96233475e-01
3.43008757e-01 5.68209052e-01 7.35514104e-01 -3.08623344e-01
-8.09776962e-01 -8.60065341e-01 -3.36756676e-01 3.67939472e-01
2.06259906e-01 -5.73346913e-01 -2.98343211e-01 3.51328880e-01] | [9.459946632385254, 3.3428030014038086] |
78d0ee74-1187-4a24-9611-d940dfbe5ade | frequency-and-spatial-domain-based-saliency | 2010.04022 | null | https://arxiv.org/abs/2010.04022v1 | https://arxiv.org/pdf/2010.04022v1.pdf | Frequency and Spatial domain based Saliency for Pigmented Skin Lesion Segmentation | Skin lesion segmentation can be rather a challenging task owing to the presence of artifacts, low contrast between lesion and boundary, color variegation, fuzzy skin lesion borders and heterogeneous background in dermoscopy images. In this paper, we propose a simple yet effective saliency-based approach derived in the frequency and spatial domain to detect pigmented skin lesion. Two color models are utilized for the construction of these maps. We suggest a different metric for each color model to design map in the spatial domain via color features. The map in the frequency domain is generated from aggregated images. We adopt a separate fusion scheme to combine salient features in their respective domains. Finally, two-phase saliency integration scheme is devised to combine these maps using pixelwise multiplication. Performance of the proposed method is assessed on PH2 and ISIC 2016 datasets. The outcome of the experiments suggests that the proposed scheme generate better segmentation result as compared to state-of-the-art methods. | ['Zanobya N. Khan'] | 2020-10-08 | null | null | null | null | ['skin-lesion-segmentation'] | ['medical'] | [ 8.26849043e-01 -2.15569302e-01 1.59306116e-02 -8.28475058e-02
-6.83648825e-01 -4.20827538e-01 4.74241585e-01 4.60765749e-01
-3.37990880e-01 7.48657048e-01 1.82103753e-01 5.41645242e-03
-3.01635891e-01 -4.83768582e-01 -2.59676367e-01 -7.28438795e-01
3.24804604e-01 -4.25944209e-01 8.89674485e-01 -1.95425954e-02
8.62006903e-01 2.17673257e-01 -1.61611581e+00 2.13243589e-01
1.45699143e+00 7.40239620e-01 4.64787215e-01 7.22518742e-01
-3.28271240e-01 4.39752609e-01 -4.39789653e-01 -1.61763668e-01
1.47102758e-01 -7.83509970e-01 -7.43164778e-01 3.16853523e-01
2.09377244e-01 4.77325246e-02 3.58082533e-01 1.45005786e+00
4.45986778e-01 3.59293446e-02 8.80939543e-01 -8.65705132e-01
-5.45627236e-01 -1.04356252e-01 -1.28415918e+00 4.78213906e-01
4.10153449e-01 -1.08957641e-01 3.61499071e-01 -5.36840558e-01
7.20565081e-01 9.09931302e-01 4.71778214e-01 1.63028255e-01
-9.69566524e-01 -1.87272608e-01 -9.51223299e-02 4.12316203e-01
-1.21547174e+00 -6.10452034e-02 8.93024147e-01 -3.58586729e-01
3.15854698e-01 7.17353106e-01 6.40269697e-01 3.22608173e-01
3.08391154e-01 4.69327599e-01 1.91211319e+00 -7.86343694e-01
2.34520286e-01 3.88878942e-01 -5.21809421e-02 7.67083406e-01
3.55007142e-01 -2.74964511e-01 -3.77862304e-01 -6.04412071e-02
7.05754519e-01 -6.35932386e-02 -1.86724752e-01 -2.09164992e-01
-7.77674317e-01 5.82120895e-01 6.43665850e-01 4.22994673e-01
-6.65404081e-01 -2.02320144e-01 1.14380144e-01 -3.43026578e-01
1.33919507e-01 1.99974701e-01 2.92394996e-01 1.68736875e-01
-1.10596740e+00 1.21099703e-01 3.13590437e-01 4.05717313e-01
4.62658763e-01 -2.74609178e-01 -3.17840368e-01 8.69224370e-01
3.38770628e-01 2.05365136e-01 5.33892691e-01 -5.73241889e-01
1.62789524e-02 6.93153918e-01 1.76695004e-01 -9.76079047e-01
-2.83341348e-01 -2.43731886e-01 -4.65062678e-01 5.19450605e-01
2.37273976e-01 -8.82297084e-02 -1.52275908e+00 1.20598614e+00
5.77284455e-01 3.53464395e-01 1.13853894e-01 1.13283658e+00
7.08836794e-01 4.14338768e-01 5.21154165e-01 -5.33369854e-02
1.47597325e+00 -9.38431978e-01 -8.62706125e-01 1.31768301e-01
5.47702946e-02 -1.18778133e+00 7.01243997e-01 3.36687654e-01
-1.25247180e+00 -3.41547668e-01 -1.33437085e+00 1.39091864e-01
-4.15736258e-01 3.50827515e-01 3.91668856e-01 6.49604321e-01
-1.02055812e+00 2.75492132e-01 -5.10022938e-01 -8.60334873e-01
3.71870995e-01 1.59292713e-01 -1.84241429e-01 6.79973885e-02
-8.99905860e-01 1.21175444e+00 2.93523639e-01 1.96125060e-02
-9.09654647e-02 -4.17328924e-01 -5.52831769e-01 -2.46484026e-01
8.41484740e-02 -3.04054826e-01 7.60560215e-01 -1.08184576e+00
-1.23923850e+00 8.96664083e-01 -3.17292094e-01 -2.89724618e-01
4.39678550e-01 2.05005646e-01 -5.02578437e-01 6.28946662e-01
1.18986510e-01 5.99590003e-01 7.78126299e-01 -1.31414664e+00
-1.22192812e+00 -3.17328453e-01 -1.24033004e-01 7.59142935e-01
5.74233755e-02 8.96206275e-02 -3.68854165e-01 -4.87375587e-01
2.72863239e-01 -8.46345901e-01 -2.67709553e-01 6.22358434e-02
-5.21892130e-01 6.07476495e-02 8.08922946e-01 -9.35944319e-01
1.45455670e+00 -1.95603895e+00 -1.75332099e-01 4.68896776e-01
9.15414914e-02 4.80100602e-01 2.51171380e-01 2.56417572e-01
5.81894219e-02 1.51989520e-01 -4.66222912e-01 3.09483707e-01
-3.91740978e-01 -3.95706028e-01 4.15796697e-01 4.76412952e-01
5.17141938e-01 4.15368438e-01 -7.85568655e-01 -9.01527464e-01
4.77370024e-01 4.74221230e-01 2.00635120e-02 -2.46556178e-01
1.43711999e-01 2.94315994e-01 -4.82866883e-01 9.73880649e-01
1.10634696e+00 1.10987779e-02 2.88889706e-02 -4.20620620e-01
-3.35802674e-01 -3.35059851e-01 -1.32400382e+00 1.59737384e+00
-1.73573136e-01 3.05232435e-01 1.16244830e-01 -6.51226461e-01
8.38886142e-01 2.20843390e-01 3.36080521e-01 -6.16932809e-01
1.99448943e-01 3.23090225e-01 1.79238096e-01 -8.41064453e-01
5.31901717e-01 -2.41759285e-01 3.25349897e-01 1.35707662e-01
-7.39955381e-02 -5.23718633e-02 3.54461759e-01 -8.69305655e-02
8.43458295e-01 1.13967091e-01 4.92029577e-01 -3.37005496e-01
8.95848334e-01 2.42487520e-01 3.21726948e-01 3.43430847e-01
-7.46599197e-01 6.19462192e-01 3.39292824e-01 1.60932645e-01
-7.52940834e-01 -1.25142729e+00 -3.15683842e-01 6.35483921e-01
7.30292737e-01 3.23775321e-01 -9.60942388e-01 -5.41421652e-01
-4.97782193e-02 3.60637218e-01 -7.32036948e-01 1.95447102e-01
-3.07743520e-01 -8.88283312e-01 1.46806762e-01 9.16671529e-02
7.89907217e-01 -9.04977500e-01 -8.57637405e-01 5.53915836e-02
8.28872770e-02 -5.63207328e-01 -3.85922164e-01 -2.19389066e-01
-6.62206888e-01 -1.18103576e+00 -1.20736015e+00 -1.24825895e+00
9.37186062e-01 4.85507131e-01 3.96168798e-01 1.02025783e-02
-8.79614234e-01 -1.06196754e-01 -4.60768759e-01 -5.28952062e-01
-4.19452004e-02 -1.45811379e-01 -5.45131624e-01 2.14004591e-01
4.38514590e-01 -2.32494339e-01 -1.13260591e+00 -1.42903894e-01
-1.06320310e+00 1.74404144e-01 9.22383785e-01 6.05169296e-01
5.34863949e-01 1.79997861e-01 7.51811683e-01 -8.88109565e-01
9.97123539e-01 -6.15607440e-01 -3.58482212e-01 5.21994352e-01
-4.11377102e-01 -8.74314085e-02 1.92791775e-01 -1.29867390e-01
-1.33825505e+00 3.10733795e-01 2.44674087e-01 1.89043641e-01
-2.21935958e-01 3.79645288e-01 4.51776892e-01 -4.56592709e-01
6.87578380e-01 3.25011492e-01 3.10119003e-01 -3.79070610e-01
1.79893851e-01 8.96687150e-01 7.54219294e-01 -8.78681466e-02
5.01654387e-01 4.54266548e-01 1.95664331e-01 -7.01583207e-01
-3.30712825e-01 -6.19260073e-01 -3.56384724e-01 -4.76179332e-01
1.03695178e+00 -5.64453185e-01 -3.85561585e-01 4.94629681e-01
-8.06401134e-01 4.36895102e-01 3.27494025e-01 3.04165959e-01
-9.85576510e-02 4.78302330e-01 -3.63669783e-01 -1.10061789e+00
-5.71965992e-01 -1.12618411e+00 7.68833339e-01 1.17451704e+00
-7.32380850e-03 -8.90756369e-01 -6.73414320e-02 2.72595912e-01
4.75055128e-01 7.66556442e-01 8.38678539e-01 -1.54727951e-01
-3.18851680e-01 -1.95867404e-01 -5.66633284e-01 7.35261440e-02
5.74701369e-01 2.80630231e-01 -7.72046924e-01 1.52403757e-01
-1.20613992e-01 6.70082122e-02 8.99703801e-01 6.07965887e-01
8.30998957e-01 4.12468659e-03 -3.51207852e-01 2.53476918e-01
2.13897681e+00 5.48081696e-01 6.38031304e-01 2.48613343e-01
4.26539242e-01 8.88336241e-01 8.41475010e-01 1.77914813e-01
1.70045033e-01 2.56962627e-01 1.91180021e-01 -6.86990499e-01
-4.55782712e-01 -9.65736620e-03 -2.19923928e-01 5.01308858e-01
-1.04353867e-01 1.18802786e-01 -6.32840931e-01 9.20291841e-01
-1.61191726e+00 -7.29571998e-01 -4.23698694e-01 2.18855786e+00
9.27204549e-01 5.00814095e-02 2.03901947e-01 2.07090676e-01
1.23690200e+00 -6.45473674e-02 -3.40455562e-01 -7.78410435e-01
-1.66560799e-01 6.02380455e-01 7.79445350e-01 6.41109288e-01
-1.16056466e+00 7.04981565e-01 6.05514860e+00 9.42530930e-01
-1.49321294e+00 5.80165200e-02 6.19207025e-01 3.81835662e-02
-3.01227480e-01 -2.37386506e-02 -1.92518860e-01 8.31518412e-01
3.82583946e-01 -2.15775132e-01 1.10954843e-01 2.07250655e-01
1.86855510e-01 -1.09346139e+00 -1.60807833e-01 7.08793819e-01
1.88438520e-01 -1.27820718e+00 -2.23727211e-01 -3.47422034e-01
9.56559896e-01 -4.81089145e-01 3.91732782e-01 -5.53185642e-01
-5.87174259e-02 -9.93659496e-01 3.12653303e-01 7.73868620e-01
6.92323327e-01 -7.21840203e-01 6.68411076e-01 -1.67048946e-01
-1.02388644e+00 1.83598042e-01 5.23867197e-02 2.38868326e-01
3.75799499e-02 3.04619759e-01 -1.00251782e+00 6.36024773e-01
3.72584373e-01 2.47353062e-01 -9.35870051e-01 1.73925722e+00
1.57537282e-01 3.47426236e-01 -2.28986040e-01 -3.28992605e-01
4.72241253e-01 -3.82137030e-01 4.98735249e-01 1.15299964e+00
3.81190568e-01 -1.45186976e-01 -2.95348525e-01 8.08028519e-01
5.46747267e-01 5.87332904e-01 -3.87437165e-01 2.20283225e-01
5.30030191e-01 1.41401517e+00 -1.06701934e+00 -2.66109258e-01
-3.88658226e-01 1.03834856e+00 -3.01005691e-01 3.70396316e-01
-7.23153830e-01 -8.85670066e-01 2.13633012e-02 1.85158893e-01
1.35722324e-01 3.40144724e-01 -6.39928401e-01 -3.87125671e-01
6.46848828e-02 -4.24047917e-01 3.29858363e-01 -6.95183098e-01
-9.26073551e-01 4.16136175e-01 -1.80834625e-02 -1.20821786e+00
1.01095743e-01 -2.78866738e-01 -9.91727173e-01 1.30884147e+00
-1.80061233e+00 -1.16062891e+00 -5.26480496e-01 5.09866297e-01
3.40946972e-01 1.29647985e-01 6.11578166e-01 1.24276303e-01
-4.17929590e-01 3.38571340e-01 1.63435370e-01 -4.13704127e-01
7.19139278e-01 -1.52858186e+00 -3.36150318e-01 9.83971059e-01
-4.63468134e-01 3.31774652e-01 7.33238935e-01 -8.23456287e-01
-8.09709430e-01 -8.04800272e-01 5.27990401e-01 2.28842869e-01
3.13316256e-01 2.48403445e-01 -6.61355138e-01 -1.91291183e-01
5.35318375e-01 -1.93162099e-01 7.22397566e-01 -4.92906272e-01
3.09412420e-01 -1.50345163e-02 -1.79797971e+00 7.24268794e-01
1.40392542e-01 -1.93449348e-01 -4.27905709e-01 1.21114708e-01
1.44888908e-01 -3.50119025e-01 -7.81727910e-01 4.29036260e-01
5.73336661e-01 -1.08470309e+00 6.65225625e-01 8.06026384e-02
3.75889897e-01 -6.44601822e-01 2.15584382e-01 -1.04814339e+00
-2.96972930e-01 -4.81462151e-01 4.45797563e-01 1.20619154e+00
3.55694830e-01 -5.14684021e-01 9.07125771e-01 2.99378037e-01
1.25928432e-01 -8.39071631e-01 -7.01464534e-01 -5.20143844e-02
-2.91341066e-01 6.00082576e-01 2.43076131e-01 6.43251896e-01
8.15447643e-02 -1.59000698e-02 -1.15827397e-02 -4.93703224e-02
7.81990469e-01 -9.15460812e-04 3.38717960e-02 -1.07615626e+00
6.98882937e-02 -3.71166676e-01 -4.55629081e-01 5.18039204e-02
-6.68693304e-01 -6.02912426e-01 1.04146026e-01 -1.72721028e+00
3.55115056e-01 -4.49431092e-01 -7.50645697e-01 1.47737801e-01
-6.41512632e-01 6.34492040e-01 2.87871182e-01 -4.81144451e-02
-2.52255887e-01 -7.87745640e-02 1.56947422e+00 1.17290532e-02
-3.14311355e-01 -2.30598226e-01 -9.43280697e-01 5.63673496e-01
9.91878927e-01 7.62571022e-03 -5.20910144e-01 7.53875226e-02
-4.73348409e-01 -7.06210807e-02 2.74858654e-01 -1.27416861e+00
2.60881722e-01 -3.35347116e-01 5.62618315e-01 -5.65825343e-01
2.71090806e-01 -4.39014167e-01 8.29632580e-02 5.23074746e-01
-2.52555311e-01 -2.50279665e-01 1.81549832e-01 5.00472724e-01
-2.50143498e-01 -4.91027027e-01 1.16064930e+00 -2.27482080e-01
-9.02883589e-01 -4.10160333e-01 -2.64314055e-01 -3.60126257e-01
1.48774898e+00 -7.09749818e-01 -4.53823537e-01 -1.05202692e-02
-5.85083544e-01 -2.27727033e-02 4.84605521e-01 1.71956599e-01
5.48482239e-01 -1.24506032e+00 -6.85379207e-01 -1.93046536e-02
1.16648845e-01 -4.28401053e-01 7.38774002e-01 1.21519184e+00
-1.11067808e+00 3.68719161e-01 -9.23441410e-01 -4.31700706e-01
-1.59474790e+00 1.31708607e-01 9.59134027e-02 -1.10745721e-01
-9.39482972e-02 7.81944335e-01 -9.17762369e-02 2.91616887e-01
-3.70907970e-02 -3.38841647e-01 -6.05915487e-01 -8.31962600e-02
3.55059803e-01 4.83563006e-01 -6.64199740e-02 -8.16822708e-01
-4.15587991e-01 8.78563821e-01 -1.70096368e-01 -2.54902929e-01
8.21007967e-01 -1.79761261e-01 -1.78782552e-01 9.40732285e-02
1.01507664e+00 6.51457161e-02 -1.05781567e+00 5.65789156e-02
-3.28128338e-02 -6.75642788e-01 2.37031758e-01 -1.32281637e+00
-7.11523175e-01 5.82508922e-01 1.23202634e+00 1.97548315e-01
1.51250327e+00 -4.83191997e-01 7.54434049e-01 -6.94617808e-01
-1.12530798e-01 -1.57767034e+00 -2.61516720e-01 -3.50292444e-01
6.93908691e-01 -1.26169527e+00 1.59758344e-01 -7.45377362e-01
-9.34373736e-01 8.58908534e-01 5.87493658e-01 -4.75324541e-01
5.91449976e-01 -6.63884506e-02 1.55370384e-01 -2.98954491e-02
-3.45847994e-01 -6.33369744e-01 5.28987825e-01 8.45317006e-01
4.11559820e-01 7.31139854e-02 -1.33118200e+00 3.19477916e-01
2.28311643e-01 2.05144361e-01 6.58184111e-01 1.27886164e+00
-7.82675028e-01 -9.75786328e-01 -6.30140603e-01 7.89748192e-01
-8.70672762e-01 9.09633636e-02 -3.29814762e-01 6.65281594e-01
3.93996954e-01 1.15238178e+00 -8.34232047e-02 -2.37297520e-01
-5.73743284e-02 -1.30112201e-01 5.67112327e-01 -3.24809253e-01
-5.71601331e-01 3.63185495e-01 6.79453984e-02 -2.05982849e-01
-6.12464488e-01 -5.07492781e-01 -1.30175769e+00 1.44677550e-01
-2.90001690e-01 6.76400363e-02 1.05321741e+00 6.46528125e-01
1.56550914e-01 5.68041205e-01 5.79604208e-01 -4.42333013e-01
-1.78748116e-01 -9.80509043e-01 -8.33804786e-01 5.63774168e-01
2.66676217e-01 -7.36455679e-01 -3.73950675e-02 1.27283692e-01] | [15.55118179321289, -2.9936695098876953] |
8df6c22b-fcdc-4915-83f8-9120b684a98e | emovie-a-mandarin-emotion-speech-dataset-with | 2106.09317 | null | https://arxiv.org/abs/2106.09317v1 | https://arxiv.org/pdf/2106.09317v1.pdf | EMOVIE: A Mandarin Emotion Speech Dataset with a Simple Emotional Text-to-Speech Model | Recently, there has been an increasing interest in neural speech synthesis. While the deep neural network achieves the state-of-the-art result in text-to-speech (TTS) tasks, how to generate a more emotional and more expressive speech is becoming a new challenge to researchers due to the scarcity of high-quality emotion speech dataset and the lack of advanced emotional TTS model. In this paper, we first briefly introduce and publicly release a Mandarin emotion speech dataset including 9,724 samples with audio files and its emotion human-labeled annotation. After that, we propose a simple but efficient architecture for emotional speech synthesis called EMSpeech. Unlike those models which need additional reference audio as input, our model could predict emotion labels just from the input text and generate more expressive speech conditioned on the emotion embedding. In the experiment phase, we first validate the effectiveness of our dataset by an emotion classification task. Then we train our model on the proposed dataset and conduct a series of subjective evaluations. Finally, by showing a comparable performance in the emotional speech synthesis task, we successfully demonstrate the ability of the proposed model. | ['Zhou Zhao', 'Ming Lei', 'Rongjie Huang', 'Feiyang Chen', 'Jinglin Liu', 'Yi Ren', 'Chenye Cui'] | 2021-06-17 | null | null | null | null | ['emotional-speech-synthesis'] | ['speech'] | [ 1.85159333e-02 2.36492693e-01 2.78379738e-01 -6.38047099e-01
-8.77069533e-01 -2.09328607e-01 3.17014545e-01 -4.85893577e-01
-2.53313690e-01 7.88317382e-01 3.96042466e-01 -9.17728394e-02
5.10926008e-01 -3.10689539e-01 -4.96650159e-01 -6.80611312e-01
2.96382397e-01 6.17387183e-02 -2.66478807e-01 -2.23892897e-01
-4.60720003e-01 2.74324387e-01 -1.63719511e+00 2.94741571e-01
9.24362898e-01 1.49351740e+00 2.16730028e-01 8.02358925e-01
-7.94068724e-02 9.44214165e-01 -1.06309927e+00 -5.34614146e-01
-1.38670757e-01 -7.15714693e-01 -7.36847043e-01 7.60856718e-02
-2.08389059e-01 -2.52761513e-01 -3.55864167e-01 1.00306273e+00
1.07612360e+00 4.71590906e-01 5.38660705e-01 -1.41711664e+00
-7.33016431e-01 8.09832871e-01 5.32801449e-02 -3.14718813e-01
1.73475042e-01 2.58832122e-03 9.73579586e-01 -1.06452143e+00
5.34289062e-01 1.18754077e+00 3.35304797e-01 9.73170877e-01
-7.58973122e-01 -8.72729540e-01 6.35545477e-02 3.91808867e-01
-1.11564672e+00 -1.01290679e+00 1.23017764e+00 -3.06181684e-02
1.05929041e+00 3.61280739e-01 6.34065866e-01 1.74054170e+00
-3.53144407e-01 1.00699461e+00 1.15176773e+00 -3.59173506e-01
3.35088760e-01 4.29595947e-01 -2.93859363e-01 2.62537569e-01
-8.54660571e-01 1.70862347e-01 -4.36670780e-01 1.91970825e-01
-3.46150659e-02 -4.61763978e-01 -5.37246764e-01 3.34559292e-01
-1.02525997e+00 7.36614645e-01 1.81145504e-01 4.86694902e-01
-5.28003752e-01 3.12572196e-02 8.07233989e-01 4.07291502e-01
7.34069526e-01 2.55079120e-01 -4.06189710e-01 -6.27278805e-01
-1.01975322e+00 -1.93357959e-01 7.94510424e-01 8.38175535e-01
2.10983917e-01 8.47905815e-01 -1.95697948e-01 1.41248441e+00
5.41857108e-02 5.97776175e-01 7.16585755e-01 -8.12579870e-01
2.99365640e-01 2.97672432e-02 -1.26483694e-01 -8.07046533e-01
-1.77550361e-01 -4.24069166e-01 -1.17439818e+00 8.26318637e-02
-2.56176233e-01 -6.31195545e-01 -6.25068247e-01 1.85888243e+00
6.87704161e-02 1.36433437e-01 7.10899293e-01 8.83598387e-01
1.15302575e+00 1.38094628e+00 4.94341403e-02 -5.63875854e-01
1.11374760e+00 -1.37820458e+00 -1.48068941e+00 -9.15859081e-03
3.92659336e-01 -7.67170310e-01 1.32152951e+00 4.71549481e-01
-9.93361413e-01 -6.49846256e-01 -9.94066477e-01 -3.86434570e-02
-3.24005187e-01 6.16331697e-01 3.25880080e-01 6.39576793e-01
-1.11450505e+00 2.60908395e-01 -4.98967528e-01 -2.17386156e-01
1.46767125e-01 5.92384003e-02 -3.92282188e-01 5.54682910e-01
-1.64924991e+00 7.86181152e-01 2.75733531e-01 3.09480339e-01
-7.97399580e-01 -3.16173702e-01 -8.49791765e-01 3.28738630e-01
1.93269834e-01 -1.77161887e-01 1.50492978e+00 -1.42025936e+00
-2.32226324e+00 4.68079805e-01 -1.50280610e-01 -4.14431065e-01
2.84545332e-01 -7.36576840e-02 -1.05513513e+00 2.50939488e-01
-4.24583733e-01 8.90213311e-01 7.87856340e-01 -1.28361845e+00
-5.37394285e-01 1.26852423e-01 -3.05554479e-01 3.27145867e-02
-7.07846284e-01 3.17009121e-01 -2.01481193e-01 -9.63731766e-01
-4.54496533e-01 -9.58664656e-01 3.28448787e-02 -3.87970507e-01
-4.49686021e-01 -2.82726109e-01 9.25944865e-01 -7.93282866e-01
1.34135425e+00 -2.41347051e+00 1.66018277e-01 -1.63777247e-01
-2.16544047e-01 3.68538707e-01 -1.68003693e-01 3.83864611e-01
-3.80139470e-01 1.59878522e-01 -2.18361273e-01 -6.59838915e-01
4.49329376e-01 1.06566556e-01 -6.60551667e-01 -4.59825844e-02
4.18839514e-01 8.07076514e-01 -6.18876636e-01 -5.56962252e-01
2.14517266e-01 7.11186707e-01 -4.35390443e-01 6.72634840e-01
-1.25356257e-01 5.57372630e-01 -2.41348729e-01 5.16180634e-01
4.24670458e-01 1.48215562e-01 -1.59867346e-01 -2.16987625e-01
-1.19364798e-01 1.89786822e-01 -7.98436821e-01 1.52224517e+00
-8.16127658e-01 8.66963208e-01 2.09117278e-01 -1.00102198e+00
1.30404592e+00 1.00108004e+00 3.74255538e-01 -7.39357471e-01
4.38661784e-01 3.85371834e-01 -9.92541201e-03 -7.62063086e-01
6.18732929e-01 -5.42715371e-01 -2.66527772e-01 3.50050688e-01
2.79206127e-01 -5.03800869e-01 -1.85392007e-01 -2.47268647e-01
7.36919045e-01 -1.93232596e-01 7.37857148e-02 2.05558747e-01
7.57528245e-01 -4.72800672e-01 4.07839477e-01 6.67452067e-02
-3.34604859e-01 3.81642997e-01 3.85652184e-01 -4.77862060e-02
-9.36331332e-01 -6.70430422e-01 3.01879514e-02 1.00354326e+00
-3.55023414e-01 -3.48838955e-01 -9.82839823e-01 -6.13870263e-01
-7.49129057e-01 9.50173259e-01 -4.13106948e-01 -2.50019908e-01
-2.25666001e-01 -3.64936084e-01 9.44458425e-01 2.91447788e-01
6.20042384e-01 -1.58083975e+00 -1.74871966e-01 2.00643450e-01
-6.73964500e-01 -1.37555957e+00 -3.79901797e-01 2.46438250e-01
-2.52765059e-01 -2.39116460e-01 -9.89997029e-01 -1.12438166e+00
2.27827340e-01 -4.65435982e-01 8.55359375e-01 -2.95436054e-01
3.07657152e-01 5.66326417e-02 -7.06670821e-01 -5.09003341e-01
-9.33272421e-01 2.25030825e-01 5.75567856e-02 2.45801136e-01
-2.86124982e-02 -6.43153369e-01 -2.49157101e-01 1.35163009e-01
-8.81215096e-01 1.91628635e-01 5.04607856e-01 8.71638894e-01
4.29843873e-01 1.56797037e-01 1.13801992e+00 -3.18377435e-01
1.07028711e+00 -2.61185557e-01 -1.54187828e-01 1.68003872e-01
-8.42485502e-02 -1.70325264e-02 1.11265469e+00 -5.76207697e-01
-1.34139800e+00 2.17633415e-02 -9.91930008e-01 -6.02282703e-01
-2.29777992e-01 4.78106797e-01 -4.61140186e-01 3.55729669e-01
3.42011958e-01 2.23888695e-01 -1.98227733e-01 -2.73872823e-01
4.39034730e-01 1.53689802e+00 4.94730055e-01 -5.31826019e-01
3.40769440e-01 -5.96140213e-02 -4.66393322e-01 -1.05339539e+00
-6.43496096e-01 1.73049599e-01 -2.78537138e-03 -5.01199424e-01
9.48997915e-01 -9.63382542e-01 -8.15454602e-01 5.44226170e-01
-1.33335674e+00 -4.26073611e-01 -3.06262046e-01 6.12929344e-01
-7.45235503e-01 2.64789283e-01 -7.22972214e-01 -1.16286755e+00
-6.15700364e-01 -1.35452378e+00 1.06457722e+00 -1.11648597e-01
-2.57645875e-01 -6.35473490e-01 -1.90731173e-03 3.13051969e-01
5.80934942e-01 8.94111097e-02 7.61434495e-01 -7.71132946e-01
3.23139429e-02 -1.78294778e-02 1.52252629e-01 9.89089549e-01
8.50556493e-02 1.60493582e-01 -1.41710079e+00 1.34463459e-01
3.56784731e-01 -8.88823867e-01 4.81257021e-01 6.47701994e-02
1.43255806e+00 -3.39957327e-01 3.05063695e-01 5.86528420e-01
7.22682714e-01 6.80165231e-01 6.12950981e-01 -2.12629229e-01
3.14473897e-01 8.41641963e-01 6.51570559e-01 4.41877306e-01
2.32148737e-01 6.67475343e-01 1.73371643e-01 -3.05704683e-01
-7.40225092e-02 -2.10265100e-01 6.48971677e-01 1.82609677e+00
1.69967681e-01 -7.00571001e-01 -4.89999145e-01 3.57560635e-01
-1.60661077e+00 -1.01408136e+00 2.71642834e-01 1.63639140e+00
1.05541980e+00 -1.00735612e-01 -5.98420501e-02 3.99453968e-01
7.30905235e-01 3.70667666e-01 -4.93306875e-01 -8.74161124e-01
-3.05586219e-01 3.07060659e-01 -4.03924465e-01 4.35520440e-01
-9.06516075e-01 1.11671519e+00 6.07867432e+00 1.23251283e+00
-1.51690733e+00 4.92199399e-02 9.71993148e-01 -1.84094831e-01
-3.29567552e-01 -6.32417917e-01 -2.83242315e-01 5.95571339e-01
1.49244618e+00 -1.30020976e-01 6.53366148e-01 9.94154155e-01
4.61854815e-01 4.62680846e-01 -9.05197144e-01 1.27603555e+00
9.08244327e-02 -9.35906768e-01 -2.79208180e-02 -3.28349710e-01
6.40318215e-01 -2.80135036e-01 7.55625218e-02 8.37262988e-01
-1.27141118e-01 -1.09676957e+00 9.55637157e-01 2.66024351e-01
1.15591705e+00 -9.66666222e-01 7.60727644e-01 3.51687193e-01
-1.07480812e+00 9.46878642e-03 -7.27778077e-02 4.94216159e-02
3.95766288e-01 4.40796375e-01 -7.05894291e-01 4.62060839e-01
6.15542948e-01 4.52434450e-01 -3.85097810e-03 4.04747277e-01
-3.31896454e-01 8.31148863e-01 3.02581396e-03 -4.49276745e-01
2.62821287e-01 -9.88010317e-02 2.94770807e-01 1.37554252e+00
6.61155641e-01 1.49620056e-01 -1.63878858e-01 7.03638256e-01
-4.61600482e-01 5.19790232e-01 -4.01960969e-01 -7.03086376e-01
4.79981154e-01 1.32285035e+00 -2.62851298e-01 -5.19965827e-01
-3.15274268e-01 1.21997881e+00 2.01822549e-01 5.88614702e-01
-1.20212984e+00 -7.25143015e-01 6.49371803e-01 -6.71433389e-01
1.67975008e-01 7.44960904e-02 1.44678012e-01 -1.16898012e+00
8.29434842e-02 -1.32366800e+00 -2.74492919e-01 -1.26678443e+00
-1.17916811e+00 1.35243630e+00 -5.06383836e-01 -1.17806184e+00
-5.79567671e-01 -4.36981529e-01 -5.14300883e-01 8.81331027e-01
-1.39146626e+00 -9.07751560e-01 -2.02191070e-01 5.23498535e-01
8.87730718e-01 -4.34169084e-01 1.04876137e+00 5.02742529e-01
-6.16537988e-01 6.12535238e-01 -2.24667657e-02 1.50336042e-01
7.77267516e-01 -9.74079728e-01 2.22561792e-01 6.03730202e-01
2.07385849e-02 1.62644923e-01 7.50015259e-01 -1.95212573e-01
-9.65588868e-01 -1.03520739e+00 9.68221486e-01 2.07386658e-01
6.82335854e-01 -7.19351113e-01 -7.87326932e-01 4.40371424e-01
7.59297192e-01 -1.19228259e-01 7.92447388e-01 -2.83104628e-01
7.49733821e-02 -1.75836191e-01 -1.10484958e+00 7.88975418e-01
8.27438056e-01 -6.99010789e-01 -5.33439100e-01 -6.33904114e-02
1.30826020e+00 -2.12886661e-01 -9.61244822e-01 4.83585864e-01
4.60603893e-01 -8.75122964e-01 4.89478976e-01 -4.98705029e-01
6.32458031e-01 -9.03485715e-02 -4.51391518e-01 -1.81934643e+00
9.71461609e-02 -8.47309947e-01 1.11414954e-01 1.64743149e+00
5.84593654e-01 -4.21083748e-01 4.54522014e-01 2.52544224e-01
-5.69110096e-01 -9.64358926e-01 -9.25639033e-01 -6.25175774e-01
-6.17277436e-02 -7.79732645e-01 8.68269980e-01 9.69604671e-01
1.93362698e-01 5.98189116e-01 -8.04941654e-01 -6.38524219e-02
-1.40379727e-01 9.71141085e-02 6.85946763e-01 -7.26584435e-01
-3.03616077e-01 -4.61611450e-01 -5.26256636e-02 -1.03053987e+00
8.77786577e-01 -8.43863785e-01 4.37137127e-01 -1.33340538e+00
-2.54122376e-01 -8.43792632e-02 -1.77352369e-01 3.96975428e-01
-5.66260843e-03 -1.68138836e-02 1.88904151e-01 -4.00511354e-01
-4.05083418e-01 1.45914912e+00 1.24884117e+00 -1.06653333e-01
-5.59304208e-02 -1.76775813e-01 -4.84297544e-01 4.60600019e-01
8.39976072e-01 -2.86430091e-01 -5.53816736e-01 -5.68214133e-02
4.64344360e-02 4.44921345e-01 -1.58399791e-02 -9.53064919e-01
-4.75797951e-02 6.35314733e-02 1.17698044e-01 -4.81901795e-01
8.42755497e-01 -1.04353404e+00 1.88033655e-01 6.94035217e-02
-6.03365481e-01 -1.78956732e-01 2.37151459e-01 1.87078938e-01
-7.63363838e-01 -1.97504297e-01 7.71364808e-01 3.52288038e-01
-4.69950259e-01 2.70187765e-01 -5.50362289e-01 7.00041950e-02
7.99671352e-01 9.75091979e-02 -2.17608124e-01 -8.50262403e-01
-8.12037587e-01 5.44511825e-02 1.13927230e-01 6.99642301e-01
7.01028645e-01 -1.65788388e+00 -5.82557261e-01 2.42329553e-01
1.72523722e-01 -4.45175916e-01 5.52314341e-01 5.12401521e-01
-1.60571537e-03 2.76315063e-01 -6.53881431e-02 -2.55799115e-01
-1.25375986e+00 5.55101812e-01 4.37135667e-01 9.79411677e-02
-2.00951800e-01 5.74838817e-01 1.10776246e-01 -6.53641820e-01
4.24153060e-01 -4.45427269e-01 -1.50242954e-01 2.82998383e-02
3.44758034e-01 2.04024971e-01 3.86343449e-02 -8.32193613e-01
-1.74291074e-01 1.64620325e-01 4.77517396e-01 -5.69657028e-01
1.31971824e+00 -1.20179087e-01 3.78186293e-02 7.57997692e-01
1.39561796e+00 2.28029236e-01 -1.02872014e+00 1.36921003e-01
-3.32831621e-01 4.06835321e-03 1.27197415e-01 -9.19429123e-01
-1.28734243e+00 1.29373252e+00 4.58086997e-01 2.98425347e-01
1.45290589e+00 -1.45636454e-01 1.30520308e+00 5.15720487e-01
-5.97718954e-02 -1.47025120e+00 2.25653544e-01 6.11325979e-01
1.22338891e+00 -1.21361005e+00 -8.22698116e-01 -1.69424340e-01
-1.18567502e+00 9.80152607e-01 4.13596034e-01 3.75846326e-01
5.93893647e-01 4.52560633e-01 4.71906543e-01 2.48262525e-01
-1.06739831e+00 2.25406662e-02 6.78669140e-02 4.33980793e-01
6.16770208e-01 1.56665117e-01 -6.22626431e-02 1.13572574e+00
-7.55666137e-01 5.66791520e-02 3.20072889e-01 2.73122430e-01
-3.23330492e-01 -1.05754411e+00 -1.50054172e-01 3.70307937e-02
-6.23296976e-01 -1.15948990e-01 -5.45526206e-01 4.21331614e-01
-1.73984140e-01 1.39208639e+00 -5.57559468e-02 -6.62762403e-01
5.08114517e-01 5.46820164e-01 4.27846462e-02 -2.57906526e-01
-6.09919608e-01 2.67914504e-01 5.11005640e-01 -3.02939206e-01
-5.12258351e-01 -1.88464075e-01 -1.36653626e+00 -3.13675553e-02
-1.82839319e-01 4.73666221e-01 9.74107325e-01 7.64650166e-01
4.55910265e-01 8.36725533e-01 1.07745504e+00 -7.33296335e-01
-4.45584267e-01 -1.25500596e+00 -6.80048108e-01 4.40489799e-01
3.01585317e-01 -2.93397337e-01 -5.73207021e-01 1.36728525e-01] | [14.246064186096191, 6.158961772918701] |
645503b4-cf59-4964-9a8d-114fe41ce9c2 | model-based-convolutional-de-aliasing-network | 1908.02054 | null | https://arxiv.org/abs/1908.02054v1 | https://arxiv.org/pdf/1908.02054v1.pdf | Model-based Convolutional De-Aliasing Network Learning for Parallel MR Imaging | Parallel imaging has been an essential technique to accelerate MR imaging. Nevertheless, the acceleration rate is still limited due to the ill-condition and challenges associated with the undersampled reconstruction. In this paper, we propose a model-based convolutional de-aliasing network with adaptive parameter learning to achieve accurate reconstruction from multi-coil undersampled k-space data. Three main contributions have been made: a de-aliasing reconstruction model was proposed to accelerate parallel MR imaging with deep learning exploring both spatial redundancy and multi-coil correlations; a split Bregman iteration algorithm was developed to solve the model efficiently; and unlike most existing parallel imaging methods which rely on the accuracy of the estimated multi-coil sensitivity, the proposed method can perform parallel reconstruction from undersampled data without explicit sensitivity calculation. Evaluations were conducted on \emph{in vivo} brain dataset with a variety of undersampling patterns and different acceleration factors. Our results demonstrated that this method could achieve superior performance in both quantitative and qualitative analysis, compared to three state-of-the-art methods. | ['Shan-Shan Wang', 'Yanxia Chen', 'Taohui Xiao', 'Qiegen Liu', 'Cheng Li'] | 2019-08-06 | null | null | null | null | ['de-aliasing'] | ['computer-vision'] | [ 2.39829093e-01 -3.55160564e-01 6.72053695e-02 -4.73226100e-01
-6.92578435e-01 6.31913170e-02 1.82732970e-01 -6.65890649e-02
-7.44644105e-01 5.09102464e-01 2.29048625e-01 -2.39958555e-01
-5.19805133e-01 -2.26123691e-01 -5.12659788e-01 -7.95605302e-01
-6.15972221e-01 4.06115443e-01 2.32300863e-01 3.45837511e-02
1.44922048e-01 6.52772427e-01 -8.98961306e-01 3.23694050e-01
8.64801407e-01 9.07690704e-01 6.26760542e-01 3.86412442e-01
2.40871370e-01 1.06814432e+00 -1.21768966e-01 3.27707857e-01
1.35407940e-01 -3.59407008e-01 -1.04083633e+00 -1.65743485e-01
-7.42217824e-02 -6.80128515e-01 -4.65694159e-01 1.00328028e+00
1.12864947e+00 6.88047111e-02 3.28040332e-01 -5.22826254e-01
-7.03806356e-02 7.26790428e-01 -7.41542161e-01 9.33531582e-01
-2.00381026e-01 4.28149775e-02 1.88200414e-01 -9.91888225e-01
3.90890777e-01 5.71071744e-01 9.90988016e-01 2.12685049e-01
-1.17613983e+00 -8.87748778e-01 -2.65215188e-01 4.64000851e-01
-1.20184565e+00 -2.59629786e-01 8.63202751e-01 -4.45951939e-01
9.91841495e-01 9.31484774e-02 8.80558431e-01 6.26843512e-01
6.42187417e-01 5.55915833e-01 1.56747127e+00 -2.36972570e-01
1.08481392e-01 -5.65227747e-01 1.67618617e-01 5.35005867e-01
-1.99355595e-02 4.21816617e-01 -2.18879446e-01 -2.46749520e-01
9.59986985e-01 -1.07665300e-01 -5.42967618e-01 -4.20271158e-01
-1.54385400e+00 7.44964898e-01 6.06631994e-01 5.92980266e-01
-6.33531392e-01 6.38259724e-02 8.16284776e-01 -1.07922114e-01
4.34013247e-01 6.72260642e-01 -2.15273827e-01 1.03908017e-01
-1.35206234e+00 1.18939199e-01 2.08782256e-01 3.43850374e-01
1.37681216e-01 2.13488922e-01 -1.24643244e-01 8.02492976e-01
1.30736813e-01 4.03003931e-01 9.15898621e-01 -7.96061635e-01
1.32402271e-01 -7.72522166e-02 -1.14197180e-01 -9.51380551e-01
-1.27288437e+00 -8.96119952e-01 -1.35074699e+00 -6.70266151e-02
2.14033201e-01 -1.78451970e-01 -6.42286122e-01 1.42508554e+00
4.72239107e-01 3.45714808e-01 -3.10091794e-01 1.46456254e+00
7.61894405e-01 3.14771771e-01 7.18199760e-02 -5.11980236e-01
1.28550184e+00 -9.37738419e-01 -1.12102163e+00 1.02351069e-01
7.04937518e-01 -7.67352462e-01 7.62101412e-01 4.50545341e-01
-1.25944316e+00 -4.72101778e-01 -1.28817809e+00 3.91720414e-01
3.02685171e-01 1.41606331e-01 8.65741849e-01 5.55683255e-01
-9.74681199e-01 8.04662168e-01 -1.21083033e+00 4.64591861e-01
4.82941419e-01 4.85028684e-01 -8.95166546e-02 -2.45441973e-01
-1.38743401e+00 1.17549944e+00 4.38637137e-01 2.97894627e-01
-9.14090157e-01 -1.28781366e+00 -4.87349629e-01 -3.50614667e-01
5.32801449e-02 -6.01916969e-01 1.09732425e+00 -5.58545411e-01
-1.43803823e+00 5.16102493e-01 1.73774749e-01 -5.26990354e-01
6.40629470e-01 -7.37105310e-02 -5.09375751e-01 5.91949344e-01
9.48963091e-02 2.76750386e-01 6.51503205e-01 -7.05323398e-01
1.90570369e-01 -5.40943980e-01 -3.19561452e-01 3.12232643e-01
1.04873262e-01 2.67979234e-01 1.48709640e-01 -5.53027153e-01
6.06982529e-01 -8.82107615e-01 -5.83261847e-01 -1.81168735e-01
2.35116836e-02 4.99315262e-01 5.33558309e-01 -1.03428459e+00
1.06667852e+00 -1.83520353e+00 -5.97163998e-02 3.84262688e-02
6.64012194e-01 4.20445144e-01 3.73837389e-02 -7.98713714e-02
-6.62624180e-01 -4.73446935e-01 -3.40957254e-01 2.72360355e-01
-6.94067061e-01 -3.06723684e-01 4.69822474e-02 9.58647013e-01
-2.58729815e-01 7.45239675e-01 -1.20643032e+00 -3.84115398e-01
5.36259055e-01 7.87450314e-01 -4.49173898e-01 1.93118230e-01
5.61003327e-01 1.13346946e+00 -2.69983381e-01 6.69980943e-02
1.08361971e+00 -4.19286728e-01 3.69340271e-01 -8.31036866e-01
-2.93721974e-01 1.83409769e-02 -9.54488933e-01 1.95938885e+00
-6.07738733e-01 1.52365282e-01 2.73910850e-01 -1.27155745e+00
5.45808196e-01 5.48733056e-01 1.24404526e+00 -1.02253520e+00
2.13514388e-01 5.08932531e-01 3.91942054e-01 -7.44301796e-01
1.94108561e-01 -5.33955097e-01 4.75390285e-01 3.91527712e-01
-1.55959791e-02 -2.11917803e-01 -8.17279965e-02 2.49035079e-02
9.78705049e-01 -2.61116261e-03 9.52563584e-02 -8.64918530e-01
5.05208671e-01 -9.35571045e-02 4.16508019e-01 7.74375379e-01
-4.70355392e-01 5.74900031e-01 9.99309868e-03 -8.89867008e-01
-1.16784859e+00 -7.53679991e-01 -7.16894448e-01 3.89745712e-01
-2.55815927e-02 -7.48472735e-02 -6.95775986e-01 -2.16580659e-01
-6.82424963e-01 2.58257687e-01 -3.35669577e-01 -2.69227996e-02
-9.66156244e-01 -1.72201121e+00 4.59215760e-01 3.62832546e-01
7.56481528e-01 -5.76624632e-01 -1.03725433e+00 6.59659803e-01
-6.90115035e-01 -1.17633986e+00 -3.95645440e-01 3.41770679e-01
-1.12378490e+00 -1.06653249e+00 -1.00243592e+00 -4.22394812e-01
5.85170269e-01 3.89060050e-01 8.66929650e-01 -1.10505447e-01
-5.68389237e-01 -1.46388546e-01 -2.20640317e-01 -5.85528761e-02
-2.66312301e-01 -1.73317313e-01 8.57766718e-02 -2.97195762e-01
-1.16417352e-02 -6.92097008e-01 -1.14114964e+00 2.81948686e-01
-9.60870504e-01 5.64271092e-01 6.02727413e-01 1.23342335e+00
6.78194344e-01 6.19098134e-02 6.49182022e-01 -6.65363014e-01
4.87258494e-01 -4.60894138e-01 -5.21417499e-01 -3.78208086e-02
-6.18031323e-01 2.15187401e-01 6.63979948e-01 -4.84674454e-01
-8.24925840e-01 3.72585095e-02 -4.65560347e-01 -2.88637668e-01
1.60214096e-01 4.92356420e-01 4.30048615e-01 -5.50067008e-01
6.59663260e-01 5.03489971e-01 3.63624960e-01 -1.95110142e-01
1.83679964e-02 2.70260185e-01 3.89882535e-01 -4.59545434e-01
-1.63309216e-01 6.24880314e-01 3.89546424e-01 -7.87740707e-01
-6.56621218e-01 -4.05024558e-01 -7.73923337e-01 -6.03890061e-01
8.05466771e-01 -9.74373519e-01 -6.59403980e-01 7.93078065e-01
-1.03882122e+00 -2.83049196e-01 4.48479354e-02 1.14612341e+00
-5.81106901e-01 8.57827067e-01 -1.00048971e+00 -2.48104781e-01
-8.60764503e-01 -1.87393045e+00 4.95487869e-01 -2.43698224e-01
3.36435549e-02 -7.91901648e-01 4.58161123e-02 3.14944625e-01
1.05841053e+00 3.40481371e-01 9.16557193e-01 -2.26740628e-01
-1.15095571e-01 8.92532617e-02 -3.13833922e-01 1.21778980e-01
-3.88180725e-02 -9.29888010e-01 -6.56768203e-01 -5.71801245e-01
6.17770493e-01 -3.41370583e-01 3.91002953e-01 1.09138656e+00
1.66655433e+00 -2.06842851e-02 -9.20202509e-02 9.73018110e-01
1.41830397e+00 1.45034075e-01 5.86548984e-01 3.12340021e-01
4.94976193e-01 1.71519801e-01 2.65116781e-01 5.42669356e-01
1.95874780e-01 7.34940410e-01 3.15631270e-01 -3.11021179e-01
-4.43079084e-01 1.33277595e-01 -3.57757628e-01 1.30051935e+00
-1.12907864e-01 5.55485964e-01 -1.05672431e+00 3.25659156e-01
-1.52550447e+00 -6.68477654e-01 -5.50109267e-01 2.08028722e+00
8.09342623e-01 -3.01449209e-01 1.91011652e-02 1.19540378e-01
5.76783240e-01 1.08829446e-01 -6.66470885e-01 1.96629643e-01
1.44619988e-02 2.90248305e-01 6.92706108e-01 6.05969369e-01
-9.86025393e-01 4.00330454e-01 7.01352596e+00 8.15712631e-01
-1.51871061e+00 7.72208333e-01 7.98949778e-01 -1.61989823e-01
-4.63609546e-02 -3.56474936e-01 -3.70187581e-01 5.33425868e-01
1.02217638e+00 1.76703200e-01 4.82801855e-01 4.66459244e-01
5.06516337e-01 -1.67133540e-01 -5.10582924e-01 1.10328591e+00
-1.11367457e-01 -1.44886458e+00 -4.17931199e-01 -1.64846793e-01
6.07195020e-01 5.30450642e-01 -1.25122681e-01 -2.01893318e-02
-1.14529870e-01 -9.60486472e-01 4.26759720e-01 3.79343450e-01
9.63908672e-01 -7.09731758e-01 6.78655803e-01 3.89123678e-01
-9.35627282e-01 1.33782223e-01 -2.53270119e-01 -4.58803326e-02
3.78516197e-01 1.09826636e+00 -6.66153610e-01 7.25964427e-01
7.59369910e-01 3.94808531e-01 -1.84079304e-01 1.09095669e+00
6.39747381e-02 3.97030324e-01 -2.42975637e-01 3.09412003e-01
2.75926739e-01 -2.08065405e-01 3.06247115e-01 1.17222035e+00
1.96923599e-01 3.61056983e-01 8.07146952e-02 6.95720792e-01
3.87835145e-01 9.58923176e-02 -1.76783457e-01 5.71831524e-01
9.13950801e-02 1.38426220e+00 -7.54815638e-01 -3.51725161e-01
-3.95134270e-01 7.22771883e-01 2.84936190e-01 2.52321094e-01
-9.65169132e-01 -5.73524162e-02 -2.36404315e-01 4.93956059e-01
-7.01544434e-02 -3.98834258e-01 -2.74351150e-01 -1.23241389e+00
-1.75796807e-01 -8.42028677e-01 2.26030722e-01 -7.04683185e-01
-9.79717255e-01 9.38283443e-01 3.67332458e-01 -9.64042962e-01
-1.40881941e-01 -3.92342865e-01 -1.89193085e-01 9.56454515e-01
-1.67549157e+00 -8.68258357e-01 -2.77587503e-01 5.41475236e-01
1.69979826e-01 1.55885294e-02 7.59167910e-01 8.17148566e-01
-1.72452077e-01 2.00938836e-01 1.66641861e-01 -5.36167473e-02
3.93345267e-01 -6.92229033e-01 -8.68602097e-02 8.69996071e-01
-6.28513813e-01 5.10828078e-01 6.23458624e-01 -6.49773300e-01
-1.37454152e+00 -9.24165845e-01 3.82517785e-01 4.20442611e-01
6.97071910e-01 3.48087959e-02 -9.42593217e-01 3.44402701e-01
3.96657735e-02 6.66160047e-01 7.16621637e-01 -3.24629962e-01
1.70190826e-01 -1.13067821e-01 -1.26958680e+00 1.62510619e-01
7.14842200e-01 -1.37464628e-01 -2.74695516e-01 5.19683957e-01
4.30386603e-01 -9.06235218e-01 -1.32332742e+00 6.80058002e-01
5.40526330e-01 -9.24257040e-01 1.22435391e+00 -1.87186167e-01
4.67161685e-01 -2.25686625e-01 1.56667128e-01 -1.47063291e+00
-7.40588546e-01 -2.20893860e-01 -1.11329019e-01 2.33881071e-01
-9.15241987e-02 -5.65950751e-01 5.52387714e-01 3.85388643e-01
-5.29656827e-01 -1.00134635e+00 -1.35582995e+00 -5.50181448e-01
1.95577607e-01 -4.56972480e-01 5.41770995e-01 1.02369535e+00
6.50061071e-02 3.68010812e-02 -5.92155457e-01 2.15925336e-01
1.04359770e+00 -4.21399809e-02 -1.04255885e-01 -6.00519538e-01
-3.83213818e-01 -2.72428423e-01 -1.28702417e-01 -9.19932485e-01
2.96288254e-05 -1.09742832e+00 -1.51114896e-01 -1.08409154e+00
4.79431450e-01 -7.21563339e-01 -4.47711855e-01 2.68238336e-02
-1.72249123e-01 1.78119853e-01 -6.70587495e-02 3.83468777e-01
-3.20264429e-01 4.60854113e-01 1.74062490e+00 -1.04839483e-03
1.51738092e-01 -2.74392635e-01 -2.31189221e-01 5.16024768e-01
6.96433783e-01 -4.00825649e-01 -4.03242707e-01 -6.01053774e-01
2.82648373e-02 7.65864193e-01 4.08851594e-01 -1.37447476e+00
2.33270675e-01 2.45071307e-01 5.88940561e-01 -6.76472843e-01
7.20637143e-02 -7.30705082e-01 4.26697254e-01 9.21937585e-01
-4.12383527e-01 1.41531676e-01 3.33545178e-01 1.23251497e-03
-7.03899637e-02 -1.78147659e-01 1.36860740e+00 -4.37769502e-01
-3.85979712e-01 5.15395045e-01 -3.83499205e-01 -7.30499625e-02
5.83009303e-01 2.07482561e-01 1.75504938e-01 8.06403533e-02
-7.29842901e-01 -8.98569673e-02 -1.45997018e-01 -1.47844804e-02
8.69044900e-01 -1.37205517e+00 -7.57872403e-01 3.29267949e-01
-3.23770463e-01 -2.04888642e-01 1.20845866e+00 1.64298344e+00
-8.49653065e-01 4.92233664e-01 -6.02312088e-01 -9.22752261e-01
-6.92407310e-01 6.38530970e-01 9.27351594e-01 -8.45724463e-01
-9.93310034e-01 4.80037779e-01 2.03835536e-02 -6.40774369e-01
-1.59034997e-01 -2.80954868e-01 -1.22093834e-01 -4.56589878e-01
8.81355047e-01 3.57508987e-01 5.74481547e-01 -6.08550549e-01
-4.87865776e-01 5.10641456e-01 -2.24367961e-01 -6.09329492e-02
1.51454294e+00 -1.49545684e-01 -4.62778844e-02 1.27247766e-01
1.30758214e+00 -7.00732172e-01 -1.31439555e+00 -2.95546353e-01
-4.34728354e-01 -3.92675489e-01 8.44838262e-01 -8.81983817e-01
-1.31539083e+00 1.01049244e+00 1.25572228e+00 -4.74761814e-01
1.10741329e+00 -4.96360093e-01 9.78414178e-01 -1.62487626e-01
6.47777557e-01 -9.91561115e-01 4.64787614e-03 3.31574678e-01
9.53761995e-01 -1.23626077e+00 3.41977537e-01 -2.72571236e-01
-4.49910223e-01 1.15015841e+00 3.30776960e-01 -1.32524818e-01
7.81493425e-01 5.07130206e-01 -1.36507630e-01 -4.70488071e-01
-1.17938139e-01 5.19088030e-01 7.47684762e-02 5.07410526e-01
7.73628891e-01 5.73662221e-02 -6.89620614e-01 3.77970785e-01
5.21347374e-02 5.29899180e-01 4.79253590e-01 7.80243933e-01
-2.86025517e-02 -6.20556056e-01 -4.72864717e-01 5.80996871e-01
-8.07006419e-01 -2.24434048e-01 7.76329815e-01 5.53369641e-01
-7.42816627e-02 5.70802867e-01 -2.65702307e-01 -1.50951266e-01
9.05096903e-02 -3.56399179e-01 9.51991677e-01 -1.77301820e-02
-6.67353213e-01 2.91538894e-01 -1.99379176e-01 -6.91401601e-01
-6.38443708e-01 -5.48514009e-01 -1.45661402e+00 -1.09621368e-01
-2.74138182e-01 1.44883722e-01 9.36621726e-01 9.38159943e-01
3.91387582e-01 9.47549045e-01 6.43613935e-01 -1.07167220e+00
-6.95091724e-01 -1.08159637e+00 -7.19495296e-01 2.42084920e-01
2.60963649e-01 -6.68192625e-01 -1.65401846e-01 -5.00816464e-01] | [13.534424781799316, -2.3922982215881348] |
2a7a5331-292c-432b-983b-96ed1fafa048 | on-random-walk-based-graph-sampling | null | null | https://ieeexplore.ieee.org/document/7113345 | https://ronghuali.github.io/PaperFiles/On%20random%20walk%20based%20graph%20sampling.pdf | On Random Walk Based Graph Sampling | Random walk based graph sampling has been recognized as a fundamental technique to collect uniform node samples from a large graph. In this paper, we first present a comprehensive analysis of the drawbacks of three widely-used random walk based graph sampling algorithms, called re-weighted random walk (RW) algorithm, Metropolis-Hastings random walk (MH) algorithm and maximum-degree random walk (MD) algorithm. Then, to address the limitations of these algorithms, we propose two general random walk based algorithms, named rejection-controlled Metropolis-Hastings (RCMH) algorithm and generalized maximum-degree random walk (GMD) algorithm. We show that RCMH balances the tradeoff between the limitations of RW and MH, and GMD balances the tradeoff between the drawbacks of RW and MD. To further improve the performance of our algorithms, we integrate the so-called delayed acceptance technique and the non-backtracking random walk technique into RCMH and GMD respectively. We conduct extensive experiments over four real-world datasets, and the results demonstrate the effectiveness of the proposed algorithms. | ['Rong-Hua Li', 'Jeffrey Xu Yu', 'Tan Ji', 'Rui Mao', 'Lu Qin'] | 2020-05-13 | null | null | null | 2020-5 | ['graph-sampling'] | ['graphs'] | [ 1.04809918e-01 1.80853948e-01 -3.95936847e-01 -4.55189012e-02
-5.47599614e-01 -2.64953703e-01 5.53954840e-01 2.04947934e-01
-4.25770074e-01 1.06631291e+00 6.73222123e-03 -6.39223993e-01
-4.27175283e-01 -1.35527611e+00 -9.93516445e-02 -6.84769452e-01
-2.13915572e-01 6.03426814e-01 1.07282948e+00 7.55478293e-02
2.36313000e-01 3.56291473e-01 -8.79285693e-01 -6.18549049e-01
9.49526727e-01 1.82591766e-01 1.02617987e-01 7.65354276e-01
-2.79643565e-01 6.19646668e-01 -3.03829342e-01 -1.10752970e-01
1.18181884e-01 -7.83097148e-01 -7.24368095e-01 -7.60892555e-02
-5.60274422e-01 -3.40961426e-01 -3.23927909e-01 8.32459927e-01
5.64527929e-01 3.01836908e-01 5.59022069e-01 -1.43941963e+00
-2.17740804e-01 8.40277731e-01 -1.19544315e+00 3.19791064e-02
4.80674028e-01 -5.40351383e-02 1.00388479e+00 -5.79848409e-01
8.37409317e-01 1.13807666e+00 5.59030950e-01 3.19980323e-01
-8.28327715e-01 -5.40100038e-01 2.77336627e-01 9.71223339e-02
-1.65903759e+00 2.76430976e-02 7.63623357e-01 -7.10531650e-03
7.62217999e-01 4.57373470e-01 8.57102633e-01 3.93134832e-01
5.50283566e-02 9.01512861e-01 1.09370434e+00 -6.00808144e-01
6.08920395e-01 -2.99502313e-01 3.98046136e-01 8.21373522e-01
9.84545708e-01 2.24510720e-03 -8.55979323e-02 -8.60594273e-01
8.24994922e-01 1.35097086e-01 -2.85905153e-01 -6.60500884e-01
-8.86730969e-01 8.89866412e-01 2.83951312e-01 1.74514651e-02
-2.72313207e-01 2.48777524e-01 2.35289961e-01 1.38055727e-01
3.39701146e-01 -1.81569695e-01 1.87069438e-02 4.53787185e-02
-1.12079906e+00 4.51059580e-01 8.10585558e-01 1.12919271e+00
9.83806968e-01 -1.53578207e-01 -1.67809993e-01 5.93670607e-01
6.96741641e-01 4.29623395e-01 3.96071732e-01 -3.31050336e-01
4.24498469e-01 6.49011433e-01 2.75107473e-01 -1.07873404e+00
-2.18661964e-01 -1.56313017e-01 -1.03678727e+00 -3.79134864e-01
1.95071414e-01 -2.07592517e-01 -9.16865468e-01 1.54661846e+00
8.12076926e-01 2.41001502e-01 -4.15660888e-01 5.23149788e-01
5.94413459e-01 6.50390744e-01 -4.05140817e-02 -6.79214776e-01
6.56654894e-01 -9.23504412e-01 -4.93815660e-01 8.75275284e-02
5.99978387e-01 -4.55687344e-01 9.76126134e-01 -7.21687451e-03
-9.38389778e-01 -5.06261997e-02 -9.96809006e-01 6.39716566e-01
-5.97866587e-02 -5.11753678e-01 5.21076322e-01 9.04088199e-01
-1.00084245e+00 4.53382015e-01 -9.45085168e-01 -6.93117499e-01
1.46025792e-01 9.59254727e-02 6.99832812e-02 -2.74108142e-01
-8.89814675e-01 2.61299491e-01 6.30031228e-02 -1.47661462e-01
-8.28732967e-01 -1.02690831e-02 -4.87683266e-01 -1.05015710e-01
7.18418896e-01 -5.68934023e-01 8.26271892e-01 -1.62436739e-01
-1.33934593e+00 5.20382643e-01 -3.75575900e-01 -3.89257848e-01
5.52376330e-01 2.26659939e-01 -3.05998743e-01 2.67259926e-02
9.68699995e-03 -7.04233125e-02 3.20403337e-01 -1.12573624e+00
-3.76445949e-01 -3.02766800e-01 -3.31837595e-01 4.31149118e-02
-2.20427103e-02 -2.16179658e-02 -3.61100942e-01 -3.73110622e-01
4.00826156e-01 -9.65518117e-01 -7.66953170e-01 -3.50677937e-01
-8.85167718e-01 -4.51030463e-01 4.77357388e-01 -5.82542606e-02
1.78105295e+00 -1.64308226e+00 -8.84355605e-02 8.05924296e-01
4.49200720e-01 3.74195516e-01 -1.11956492e-01 1.14104378e+00
3.76946807e-01 3.08810622e-01 -3.22205573e-01 2.66305059e-01
-1.83227077e-01 2.25545660e-01 1.34793103e-01 5.38534343e-01
-4.15693641e-01 8.40481639e-01 -1.19781888e+00 -6.15580976e-01
1.03983618e-01 -1.94977950e-02 -3.86542112e-01 1.51996106e-01
-4.36598472e-02 -1.36965126e-01 -7.67190039e-01 4.27888781e-01
9.54230666e-01 -2.23269165e-01 6.87584400e-01 2.46674329e-01
7.77589306e-02 3.27756077e-01 -1.66093934e+00 7.40643740e-01
1.82952851e-01 -1.38257727e-01 -1.19541705e-01 -6.22745156e-01
9.54419672e-01 2.70100031e-02 2.04380333e-01 -2.05790162e-01
1.64178029e-01 2.40126505e-01 -1.88641667e-01 -1.72431394e-01
5.52975714e-01 -3.23403664e-02 2.25155279e-01 1.08626592e+00
-5.27279913e-01 1.03913574e-02 3.09336305e-01 6.76331699e-01
1.49554694e+00 -1.59647167e-01 8.75425816e-01 -4.73653167e-01
5.24410963e-01 -1.79971159e-01 5.83857775e-01 9.93458033e-01
-5.20649731e-01 4.80584830e-01 5.05336165e-01 -2.83181131e-01
-4.91338432e-01 -1.09253001e+00 6.66022539e-01 6.84166253e-01
4.32808667e-01 -6.15073800e-01 -5.20562232e-01 -8.16659451e-01
-1.29774556e-01 5.10657787e-01 -3.77453297e-01 -9.04081948e-03
-4.21900302e-01 -1.11400998e+00 2.00522885e-01 7.80520216e-02
6.71497107e-01 -8.65835607e-01 -3.87767673e-01 1.97020516e-01
-1.11637510e-01 -5.43828249e-01 -4.82283056e-01 -2.38480389e-01
-8.47074509e-01 -1.11632454e+00 -7.72105575e-01 -5.45197904e-01
8.58078301e-01 9.37249541e-01 9.35714245e-01 4.26974595e-01
-1.89380556e-01 9.39890370e-02 -7.32448161e-01 1.21755309e-01
-1.49041310e-01 6.32474303e-01 -1.65301204e-01 -2.55275778e-02
2.45109856e-01 -7.53260732e-01 -7.83885539e-01 4.92870986e-01
-7.90832639e-01 -3.62398922e-02 5.73905885e-01 4.47604001e-01
6.52231216e-01 9.43522826e-02 4.62799639e-01 -1.40439928e+00
1.05890512e+00 -6.36173427e-01 -5.06348848e-01 1.97423801e-01
-1.01126194e+00 8.24830979e-02 1.51035964e-01 -3.68931860e-01
-5.70216179e-01 -1.33121043e-01 -1.85849801e-01 2.87302524e-01
5.44002950e-01 6.19564474e-01 -1.73494264e-01 -2.14099124e-01
3.22150975e-01 2.24654108e-01 -1.41732097e-01 -1.70781240e-01
5.26482165e-01 6.76715434e-01 -2.11590871e-01 -3.04488152e-01
9.84011590e-01 3.64003688e-01 7.98005685e-02 -6.41545713e-01
-2.51147211e-01 -5.41793466e-01 -2.00203165e-01 -3.01337361e-01
3.62520993e-01 -3.28086406e-01 -4.87267762e-01 5.87205231e-01
-5.01185656e-01 -4.73252267e-01 -2.12059453e-01 3.27894390e-01
-1.62131160e-01 6.97960496e-01 -5.49587190e-01 -1.37697244e+00
-4.51696575e-01 -8.05770636e-01 4.06801909e-01 3.59499395e-01
-2.83063173e-01 -9.12010670e-01 3.84407014e-01 -3.20395052e-01
4.53892261e-01 6.13914192e-01 9.35989261e-01 -6.54851973e-01
-8.46099555e-01 -4.42845225e-01 -3.85358691e-01 -3.83079410e-01
2.16948569e-01 1.71998829e-01 -2.34554522e-02 -6.33955598e-01
-4.05161172e-01 -1.14847541e-01 6.38864756e-01 3.69487464e-01
4.54681963e-01 -1.47635952e-01 -7.51488388e-01 1.48457363e-01
1.80728018e+00 9.63354297e-03 8.53781521e-01 2.80370474e-01
5.87034881e-01 6.59665912e-02 6.88659310e-01 5.77551365e-01
6.82246029e-01 2.65327573e-01 3.26402009e-01 -8.03197175e-03
7.28573129e-02 -7.62069464e-01 1.15038939e-02 1.25069022e+00
2.10110079e-02 -6.67569935e-01 -7.52710581e-01 6.91009820e-01
-1.95034158e+00 -7.69220710e-01 -5.86418450e-01 2.39514613e+00
6.53155088e-01 3.06584507e-01 7.10385501e-01 4.38899040e-01
1.06607199e+00 3.94519925e-01 -3.91910613e-01 -3.07099164e-01
2.62991577e-01 2.02564299e-01 8.72359574e-01 6.96489871e-01
-3.70031506e-01 8.00256968e-01 6.23745537e+00 9.91073906e-01
-3.94732028e-01 3.80733609e-02 2.22144201e-01 2.93479234e-01
-6.58775866e-01 5.52036047e-01 -7.35662282e-01 3.39029312e-01
7.09666789e-01 -5.92020631e-01 5.49898982e-01 6.84380412e-01
1.91102043e-01 -4.69692677e-01 -4.85126942e-01 4.88613069e-01
-3.25425684e-01 -1.01790237e+00 3.45854312e-01 1.55989885e-01
7.07707226e-01 -5.90107217e-02 -5.63304961e-01 2.22924441e-01
1.16578269e+00 -5.48770785e-01 -6.09953189e-03 1.67994201e-01
5.10490060e-01 -8.25313330e-01 6.08970404e-01 5.24528563e-01
-1.58697128e+00 3.06017548e-01 -4.02729660e-01 -6.98521584e-02
4.05992568e-01 1.07823372e+00 -6.26781404e-01 9.30523932e-01
2.27471694e-01 1.43957511e-01 -3.42040777e-01 1.22319078e+00
-3.22283864e-01 7.83795357e-01 -5.13939977e-01 -6.88297987e-01
2.94802815e-01 -3.33355308e-01 5.90471268e-01 5.79865634e-01
8.54203701e-02 4.70962450e-02 2.87984192e-01 4.84547913e-01
1.04880720e-01 2.33863667e-01 -3.55586588e-01 8.63009766e-02
1.05938852e+00 1.12172592e+00 -1.13118970e+00 -2.53839195e-01
-1.15315251e-01 5.70789754e-01 4.02443379e-01 3.91405016e-01
-8.21989954e-01 -9.82792437e-01 -5.53279817e-02 7.40277708e-01
3.75061035e-01 -4.74393785e-01 2.11402565e-01 -8.20161164e-01
-1.93877265e-01 -7.14640141e-01 5.50350785e-01 -4.00539458e-01
-9.23945665e-01 4.32196259e-01 2.31574655e-01 -9.04438615e-01
-3.04267108e-02 2.47899592e-01 -1.00102556e+00 7.57481456e-01
-1.43132424e+00 -6.89888954e-01 -3.13425034e-01 5.63443005e-01
2.85380214e-01 2.23448217e-01 3.93209159e-01 1.03558317e-01
-4.44926322e-01 5.22430599e-01 4.73982394e-02 -2.92633045e-02
2.02891842e-01 -8.94091964e-01 6.78287089e-01 9.21413064e-01
-2.29394078e-01 8.27621460e-01 5.83882689e-01 -9.92053568e-01
-1.53232729e+00 -1.09726512e+00 9.42413449e-01 3.70396346e-01
3.92031372e-01 -2.61274040e-01 -6.56990886e-01 6.96589947e-01
-4.12749723e-02 -1.87082499e-01 5.02124786e-01 -5.47659770e-02
9.73822102e-02 -6.04034588e-02 -1.45014381e+00 8.75901937e-01
1.10392594e+00 -5.34430221e-02 -6.86521456e-02 3.76322627e-01
6.66887581e-01 3.18992846e-02 -4.44526672e-01 3.48776013e-01
4.24952537e-01 -1.11977398e+00 7.09309936e-01 -1.84293717e-01
-2.30361834e-01 -4.74030048e-01 6.31699711e-02 -1.04665911e+00
-5.32940447e-01 -1.14145827e+00 4.89964113e-02 1.42374432e+00
4.26592916e-01 -1.02638388e+00 1.01182890e+00 2.68798739e-01
7.71323323e-01 -1.09124303e+00 -7.96968818e-01 -7.16834068e-01
-1.52023539e-01 1.36541657e-03 9.60874021e-01 7.22500265e-01
-3.75959720e-03 4.40295130e-01 -5.28759539e-01 -8.52551982e-02
7.44192362e-01 9.09224153e-02 1.16298378e+00 -9.90106285e-01
-3.04746270e-01 1.89621784e-02 -1.62982434e-01 -1.21378815e+00
-5.10458112e-01 -7.09338188e-01 -9.61769223e-02 -1.95215571e+00
5.12826085e-01 -5.10491431e-01 -1.14578180e-01 -1.41900644e-01
-5.09273529e-01 -1.04164779e-01 -1.92149833e-01 3.67068201e-01
-8.88113916e-01 4.66037810e-01 1.11742604e+00 4.80023444e-01
-5.34546912e-01 3.25145394e-01 -4.82425839e-01 3.53075147e-01
8.00195873e-01 -7.02662587e-01 -8.00970376e-01 -8.14771354e-02
2.58635461e-01 2.51444697e-01 -2.64148176e-01 -6.34363115e-01
3.36244941e-01 -4.14136261e-01 -2.43981108e-01 -9.85013127e-01
-3.79879922e-01 -3.15464348e-01 3.62001687e-01 9.42736804e-01
-1.39480084e-01 1.93884611e-01 -5.65997064e-01 1.03201067e+00
2.24815890e-01 -1.78730756e-01 8.39985907e-01 -3.37800443e-01
-1.82640731e-01 5.15722573e-01 -6.62644744e-01 3.67512643e-01
1.20300925e+00 -4.19874400e-01 -2.09615752e-01 -6.76414728e-01
-3.75979751e-01 5.72603047e-01 4.63812143e-01 -1.05412915e-01
7.85439909e-01 -1.29007685e+00 -4.13037419e-01 1.41184241e-01
3.12934630e-02 -1.24266647e-01 -8.58305581e-03 9.62939024e-01
-7.79816508e-01 -1.14502767e-02 3.73269439e-01 -2.35317528e-01
-1.26122844e+00 4.91937935e-01 -1.22898281e-01 -7.46037245e-01
-4.37501192e-01 7.74573267e-01 -5.48827887e-01 -6.70625210e-01
8.21504220e-02 -1.97268292e-01 2.67447829e-01 -5.60067773e-01
2.24579155e-01 9.29554701e-01 -4.21275169e-01 -1.49714023e-01
-4.67015356e-01 4.21869189e-01 -1.68139219e-01 -2.75110304e-01
1.02338088e+00 -4.25943404e-01 -2.54319757e-01 3.63092929e-01
7.49156475e-01 2.73193508e-01 -4.79772359e-01 -3.33117872e-01
1.33181140e-01 -6.13942623e-01 -1.00974843e-01 -1.68278128e-01
-8.70371461e-01 2.14346066e-01 5.59536964e-02 7.82020867e-01
8.92246962e-01 -3.69055867e-01 1.19020772e+00 -1.47944512e-02
8.62405598e-01 -9.77829874e-01 -4.04098332e-01 -1.57950614e-02
4.70950156e-02 -6.74684882e-01 6.32254064e-01 -9.13461566e-01
-4.33184564e-01 5.95059574e-01 4.35414642e-01 -3.97150099e-01
9.08228874e-01 1.36013050e-02 -5.04677355e-01 -8.95065293e-02
-6.97967827e-01 -4.34990674e-01 -4.77502733e-01 6.02954090e-01
8.08126554e-02 2.33760461e-01 -8.81730855e-01 2.96568006e-01
9.13318768e-02 2.50216872e-01 6.72439039e-01 1.27925587e+00
-8.36096466e-01 -1.39166260e+00 1.45357415e-01 6.93046033e-01
-1.73178911e-02 4.23090495e-02 -6.68148458e-01 9.48151410e-01
-5.01111388e-01 9.63372588e-01 -4.50237781e-01 -4.96051282e-01
3.74153629e-02 -2.27468953e-01 3.53348762e-01 -4.39539820e-01
-1.78624824e-01 -8.02478492e-02 1.67159975e-01 -1.90380007e-01
-3.17061514e-01 -2.58040637e-01 -1.12339449e+00 -8.43284070e-01
-9.15303886e-01 5.62593639e-01 1.70894563e-01 6.59124434e-01
3.58104527e-01 1.24638885e-01 8.09865057e-01 -3.05872798e-01
-5.78623354e-01 -7.16855049e-01 -9.29381788e-01 -1.11409098e-01
-6.92562237e-02 -4.87292945e-01 -5.56754768e-01 -7.59792924e-01] | [7.0001726150512695, 5.282647132873535] |
a915e799-1a9e-403b-aa17-b7a7b956d7d2 | yolo-drone-airborne-real-time-detection-of | 2304.06925 | null | https://arxiv.org/abs/2304.06925v1 | https://arxiv.org/pdf/2304.06925v1.pdf | YOLO-Drone:Airborne real-time detection of dense small objects from high-altitude perspective | Unmanned Aerial Vehicles (UAVs), specifically drones equipped with remote sensing object detection technology, have rapidly gained a broad spectrum of applications and emerged as one of the primary research focuses in the field of computer vision. Although UAV remote sensing systems have the ability to detect various objects, small-scale objects can be challenging to detect reliably due to factors such as object size, image degradation, and real-time limitations. To tackle these issues, a real-time object detection algorithm (YOLO-Drone) is proposed and applied to two new UAV platforms as well as a specific light source (silicon-based golden LED). YOLO-Drone presents several novelties: 1) including a new backbone Darknet59; 2) a new complex feature aggregation module MSPP-FPN that incorporated one spatial pyramid pooling and three atrous spatial pyramid pooling modules; 3) and the use of Generalized Intersection over Union (GIoU) as the loss function. To evaluate performance, two benchmark datasets, UAVDT and VisDrone, along with one homemade dataset acquired at night under silicon-based golden LEDs, are utilized. The experimental results show that, in both UAVDT and VisDrone, the proposed YOLO-Drone outperforms state-of-the-art (SOTA) object detection methods by improving the mAP of 10.13% and 8.59%, respectively. With regards to UAVDT, the YOLO-Drone exhibits both high real-time inference speed of 53 FPS and a maximum mAP of 34.04%. Notably, YOLO-Drone achieves high performance under the silicon-based golden LEDs, with a mAP of up to 87.71%, surpassing the performance of YOLO series under ordinary light sources. To conclude, the proposed YOLO-Drone is a highly effective solution for object detection in UAV applications, particularly for night detection tasks where silicon-based golden light LED technology exhibits significant superiority. | ['Zhengnan Jiang', 'Hanzheng Hu', 'Feng Xiong', 'Jiahui Xiong', 'Li Zhu'] | 2023-04-14 | null | null | null | null | ['real-time-object-detection'] | ['computer-vision'] | [ 2.06406981e-01 -5.78235269e-01 3.49576265e-01 1.91492751e-01
-1.60848394e-01 -6.42575264e-01 4.11222458e-01 -3.16561937e-01
-6.71593130e-01 6.92206621e-01 -9.87071753e-01 3.82274836e-02
-7.10491464e-02 -9.45998490e-01 -6.10457361e-01 -8.95140409e-01
-1.14718117e-01 -1.98336318e-01 6.29519463e-01 -3.43696713e-01
-2.24675238e-01 5.66150784e-01 -2.03175187e+00 -3.71406227e-01
8.55519295e-01 1.51035106e+00 4.82306093e-01 6.03962779e-01
8.21955919e-01 1.44660354e-01 -6.81515992e-01 -7.81316608e-02
9.02824879e-01 1.69985667e-02 1.54136211e-01 -5.14023229e-02
8.51265848e-01 -6.98102951e-01 -2.02464804e-01 1.22137678e+00
4.92279172e-01 7.74891526e-02 3.56193691e-01 -1.37348115e+00
-2.12656736e-01 -1.07340135e-01 -9.04741585e-01 4.85547483e-01
-4.26629893e-02 7.29817927e-01 7.65087783e-01 -8.52034688e-01
1.98491156e-01 9.00409698e-01 6.23165905e-01 1.77389309e-01
-9.31242287e-01 -1.02319705e+00 -9.89541933e-02 -1.18871763e-01
-1.58807373e+00 -4.42950204e-02 2.41130203e-01 -5.18040121e-01
9.49954510e-01 7.44755864e-02 8.01399469e-01 5.53184330e-01
5.14157891e-01 2.46998787e-01 1.21540713e+00 -6.30073845e-02
1.37622818e-01 -3.73977199e-02 -1.99235171e-01 1.14506280e+00
9.41131294e-01 6.22253478e-01 -1.89779177e-01 9.63698551e-02
5.63389063e-01 1.84156567e-01 -5.03291130e-01 -3.08517087e-03
-1.28034937e+00 5.06333709e-01 9.09723401e-01 -1.29070118e-01
-3.59283000e-01 2.47562274e-01 1.95662469e-01 9.33530927e-03
2.48809829e-01 4.66082156e-01 -3.40547621e-01 4.88352388e-01
-8.67222548e-01 2.66977042e-01 3.29468906e-01 1.04761565e+00
8.55059922e-01 5.65571368e-01 -2.35645160e-01 2.87744284e-01
3.89234006e-01 1.26964676e+00 -4.37382143e-03 -3.66199017e-01
4.04965162e-01 7.34237850e-01 4.21508342e-01 -1.09773934e+00
-4.66147810e-01 -6.74042523e-01 -7.92399287e-01 7.31526375e-01
1.33198291e-01 -5.21302283e-01 -1.01076686e+00 1.33817506e+00
3.02099973e-01 8.46103430e-02 9.49426815e-02 1.21489632e+00
1.02816784e+00 8.04558754e-01 7.20523717e-03 -1.19936913e-01
1.64052355e+00 -5.69966793e-01 -3.35515440e-01 -4.66884494e-01
2.46244863e-01 -5.95175207e-01 6.57332361e-01 4.45081770e-01
-5.76104939e-01 -7.66945958e-01 -1.45667744e+00 2.76127040e-01
-4.79261190e-01 9.36544418e-01 5.65538824e-01 8.37820768e-01
-8.83783162e-01 -3.88409123e-02 -6.34823382e-01 -5.60179472e-01
5.52582443e-01 3.80527556e-01 -9.65107977e-02 -1.24687709e-01
-8.66226137e-01 6.07361615e-01 6.03570879e-01 1.94328159e-01
-1.38743579e+00 -7.73591697e-01 -6.38296425e-01 1.08751230e-01
7.17367887e-01 -6.76019013e-01 7.80504286e-01 -6.03550494e-01
-1.27348530e+00 7.67823160e-01 4.20678586e-01 -6.87288702e-01
2.88650066e-01 -3.36724758e-01 -4.55264181e-01 3.04260552e-01
2.95514073e-02 8.77673566e-01 1.26321518e+00 -9.47335005e-01
-1.24237502e+00 -5.29275060e-01 4.49365556e-01 1.46280810e-01
-2.62406677e-01 -2.26036999e-02 1.38906045e-02 -4.81921613e-01
-1.99700847e-01 -1.05117023e+00 1.21592529e-01 3.99242431e-01
-4.04215127e-01 -2.10881099e-01 1.26943159e+00 -1.98652089e-01
1.05622828e+00 -2.03802490e+00 -2.02208415e-01 -2.00748384e-01
1.72297046e-01 9.40021753e-01 -2.24806201e-02 2.31642108e-02
2.71234661e-01 -3.14617872e-01 -5.10993779e-01 1.24093354e-01
-4.23019707e-01 -2.19078600e-01 -3.24804574e-01 5.91600239e-01
3.94734323e-01 6.74003482e-01 -8.32801640e-01 -1.22475438e-01
4.96889532e-01 4.83366102e-01 -1.59579620e-01 1.16177537e-01
-2.99954265e-01 6.14074543e-02 -2.30857655e-01 1.19177914e+00
8.72668922e-01 1.93281636e-01 -3.02257657e-01 -4.38199699e-01
-8.07101905e-01 -4.49304998e-01 -9.16048110e-01 1.13168883e+00
-2.43883565e-01 7.53026843e-01 3.00969362e-01 -3.38892251e-01
1.28693163e+00 -7.61118382e-02 1.14249632e-01 -3.68701071e-01
5.71643174e-01 1.44707829e-01 -1.12002529e-01 -3.49694639e-01
6.20761871e-01 -4.37199995e-02 5.75115420e-02 -1.58886805e-01
4.09857789e-03 -5.02921700e-01 4.10197467e-01 -1.89062968e-01
9.09614980e-01 -7.80303627e-02 3.64911348e-01 -4.72544163e-01
5.79452872e-01 2.41442338e-01 5.61347961e-01 5.39784670e-01
-2.19958752e-01 2.22270831e-01 5.56535870e-02 -4.92986560e-01
-4.73831862e-01 -8.33813667e-01 -3.36964130e-01 6.27589107e-01
4.56307024e-01 -2.26055130e-01 -6.35661364e-01 -5.08897901e-01
1.52202874e-01 3.42303306e-01 -3.01019669e-01 -9.37356614e-03
-9.38918665e-02 -1.07762301e+00 7.25418925e-01 3.42295736e-01
1.36235356e+00 -6.66055381e-01 -1.65885210e+00 2.50252541e-02
2.38899872e-01 -1.43276238e+00 3.19095813e-02 9.53845493e-03
-6.46339357e-01 -1.32649279e+00 -5.89189529e-01 -4.02550817e-01
4.49997276e-01 1.04177284e+00 6.91757560e-01 -1.96924120e-01
-1.02693725e+00 5.07783711e-01 -2.47361973e-01 -8.54712486e-01
2.99470812e-01 -1.76612705e-01 3.07204813e-01 5.61779477e-02
3.47678334e-01 -8.19585025e-02 -8.39695036e-01 5.14771104e-01
-9.60528314e-01 -3.25522959e-01 8.45299542e-01 7.23957300e-01
5.24481177e-01 4.32485998e-01 1.76592946e-01 -2.80975938e-01
1.44492269e-01 -2.27076516e-01 -1.64889932e+00 8.36648345e-02
-5.70730090e-01 -7.14984953e-01 6.59993649e-01 -3.55891772e-02
-8.03837538e-01 3.49588573e-01 4.80138063e-01 -5.72340965e-01
-2.23604918e-01 8.42380896e-02 -1.18336335e-01 -5.87922454e-01
7.14012384e-01 7.02430680e-02 -1.79029584e-01 1.41443836e-03
-3.10517312e-03 6.11154735e-01 4.44537312e-01 -9.21832398e-02
1.17233729e+00 7.63343155e-01 3.61363411e-01 -1.29001689e+00
-5.77150941e-01 -3.99121344e-01 -2.92453021e-01 -4.00595129e-01
9.91913080e-01 -1.48888123e+00 -9.54899371e-01 8.54330897e-01
-1.01817465e+00 -6.00145273e-02 -9.59853157e-02 6.37355745e-01
-4.66958098e-02 1.33780137e-01 -1.32145241e-01 -9.84327316e-01
-6.10797584e-01 -1.16015172e+00 1.03167796e+00 6.71402812e-01
5.47808886e-01 -3.20692211e-01 -4.32202846e-01 8.19137096e-02
3.41171563e-01 5.26838839e-01 4.46564257e-01 1.28347710e-01
-1.10659909e+00 -3.75642240e-01 -8.42264533e-01 5.94537199e-01
4.01290506e-02 1.08381130e-01 -1.06736290e+00 -6.41245723e-01
-3.95956099e-01 -1.57304749e-01 1.00629091e+00 3.15607846e-01
5.67418575e-01 2.84832716e-01 -4.93175298e-01 8.19293141e-01
1.83099997e+00 4.93568450e-01 4.09122974e-01 1.91761270e-01
5.67508161e-01 1.73343122e-01 1.12008095e+00 7.16881216e-01
3.80932689e-02 7.38529086e-01 1.24324620e+00 -3.72775644e-02
-1.71188816e-01 1.77577347e-01 5.53544283e-01 -6.24297000e-02
-5.20079076e-01 -5.86025417e-01 -6.56835735e-01 3.12750965e-01
-1.38331795e+00 -8.04322720e-01 -3.29744697e-01 2.31752014e+00
-2.14910647e-03 -9.07633752e-02 6.76267743e-02 3.78574058e-02
7.01919019e-01 3.22535872e-01 -7.08619833e-01 3.05169165e-01
-2.71565914e-01 1.08770654e-01 9.61703122e-01 -1.88333496e-01
-1.58445430e+00 9.70915556e-01 4.80565214e+00 5.50245106e-01
-1.28756440e+00 -2.88086664e-02 2.72795055e-02 -5.45357587e-04
6.07751131e-01 -2.51554042e-01 -1.36899292e+00 4.52463359e-01
4.43052113e-01 9.72182602e-02 3.45343143e-01 9.13245499e-01
1.11999385e-01 -5.11717439e-01 -4.22890574e-01 1.09981644e+00
2.13523909e-01 -8.60278010e-01 3.51968035e-02 9.29455608e-02
6.69735253e-01 2.24919707e-01 2.72289217e-01 -5.07649854e-02
2.35392638e-02 -6.82098091e-01 6.13112628e-01 2.16752544e-01
1.29530954e+00 -7.58094013e-01 8.81338596e-01 9.79796797e-02
-1.61500967e+00 -4.08644766e-01 -8.54625165e-01 -1.42725244e-01
-1.40260994e-01 4.97549474e-01 -8.51011813e-01 6.65885031e-01
1.13091421e+00 8.81012797e-01 -5.54898620e-01 1.27228415e+00
-4.80454028e-01 4.70444798e-01 -5.19859552e-01 -1.50301591e-01
4.39590067e-01 -2.07765073e-01 1.04106247e+00 8.60967100e-01
7.60715008e-01 2.86849320e-01 3.48263025e-01 9.03187454e-01
-5.52297384e-02 -4.24944401e-01 -7.91864514e-01 5.29549411e-03
2.56375670e-01 1.72246444e+00 -7.36317635e-01 -2.81936023e-02
-4.10965532e-01 6.63643003e-01 -3.98843765e-01 3.20238650e-01
-1.11890960e+00 -8.85441899e-01 9.29626763e-01 1.83851257e-01
5.36436796e-01 -3.86775851e-01 2.65963972e-01 -8.64727199e-01
-1.43672913e-01 -4.60382491e-01 2.05577657e-01 -9.47791100e-01
-7.70996869e-01 9.45946217e-01 -9.38156918e-02 -1.67933619e+00
5.26505589e-01 -1.15749168e+00 -3.41568023e-01 5.58182120e-01
-1.97583830e+00 -1.32061720e+00 -1.13634431e+00 4.66955006e-01
5.53728044e-01 -4.75838840e-01 5.56763470e-01 2.31453925e-01
-9.65136945e-01 3.33001405e-01 -1.59288436e-01 1.47626009e-02
5.00601411e-01 -6.50645256e-01 2.97130402e-02 1.39073861e+00
-1.25977233e-01 2.21112117e-01 4.16988403e-01 -5.85921466e-01
-1.81294155e+00 -1.56455517e+00 -1.26548670e-02 -1.48037553e-01
4.23102200e-01 -1.21671990e-01 -3.78763199e-01 4.92959619e-01
3.12099084e-02 4.05226529e-01 1.24378018e-01 -8.51252377e-01
-9.23739076e-02 -6.98407829e-01 -1.34632385e+00 1.89681038e-01
9.60889399e-01 6.19961787e-03 -2.11027399e-01 3.00794572e-01
5.53495884e-01 -3.79951239e-01 -5.59152782e-01 7.50209212e-01
5.37481427e-01 -1.18538439e+00 9.26345766e-01 2.76296604e-02
-2.75559966e-02 -1.08865142e+00 -3.20083290e-01 -1.16158056e+00
-2.43180081e-01 -6.43569589e-01 2.92585850e-01 1.08269715e+00
-8.45116600e-02 -9.73724842e-01 3.90314728e-01 5.75219356e-02
-3.66590023e-01 -4.30930227e-01 -7.98668087e-01 -1.09322238e+00
-7.88306415e-01 -7.32762888e-02 5.73893070e-01 2.65844405e-01
-7.33145177e-01 1.29571855e-01 -2.83338875e-01 8.15687597e-01
8.72899473e-01 1.08320385e-01 8.76304150e-01 -1.64902270e+00
1.69256091e-01 -4.93994392e-02 -6.85972631e-01 -9.17709887e-01
-2.32153937e-01 -5.27940869e-01 3.11834037e-01 -1.24231195e+00
-2.64804959e-01 -3.43341500e-01 -1.00426137e-01 4.63002443e-01
-5.28472140e-02 8.03167820e-01 2.74395764e-01 2.78306641e-02
-4.77386862e-01 4.57692772e-01 9.81636941e-01 -2.01388881e-01
-1.86592713e-01 1.14523627e-01 -2.23412022e-01 6.92921937e-01
5.94523847e-01 -2.78395146e-01 -1.86587870e-01 -3.95992070e-01
2.48703256e-01 -3.07179242e-01 9.39618468e-01 -1.78792262e+00
3.19779187e-01 2.69784361e-01 4.84976232e-01 -6.33113921e-01
5.42427659e-01 -1.02630234e+00 -2.02107779e-03 9.43506360e-01
7.07098842e-01 -2.16502082e-02 5.99976480e-01 7.85286605e-01
-2.04488918e-01 8.33559632e-02 1.07628155e+00 -8.10779557e-02
-1.34207046e+00 4.41881806e-01 -4.92618471e-01 -1.52031407e-01
1.60779846e+00 -3.44600886e-01 -8.32587838e-01 2.97747433e-01
1.56438991e-01 1.00650951e-01 4.04129714e-01 2.55689710e-01
8.36856008e-01 -7.37037003e-01 -5.60884535e-01 5.46339750e-01
3.55983615e-01 1.20400876e-01 2.96587348e-01 9.38488305e-01
-7.21661150e-01 7.36763239e-01 -5.46662807e-01 -7.14044273e-01
-1.29765296e+00 2.94586569e-01 2.03574389e-01 3.60738635e-01
-3.98176432e-01 1.01110804e+00 2.61343509e-01 1.33111417e-01
-1.01359464e-01 -6.48188591e-01 -3.30609344e-02 1.88057736e-01
6.42372549e-01 8.52794111e-01 1.17519855e-01 -6.18192494e-01
-5.03827333e-01 9.72507179e-01 3.94479811e-01 6.74500406e-01
1.11515462e+00 -5.34407198e-02 -9.56597179e-02 -1.63074628e-01
2.98285842e-01 -3.09092849e-01 -1.52394009e+00 2.20201053e-02
-5.30619383e-01 -5.79662681e-01 5.25129199e-01 -7.96939552e-01
-1.10910690e+00 8.64104152e-01 1.02181470e+00 1.43257052e-01
1.39803541e+00 -3.90974671e-01 8.27423096e-01 5.11943758e-01
8.02582324e-01 -5.89105964e-01 -8.10984820e-02 4.59058821e-01
6.04441524e-01 -1.32692945e+00 3.78587395e-01 -5.45653403e-01
-3.36092949e-01 1.13224292e+00 8.85132909e-01 -2.36985475e-01
3.03715706e-01 2.11969599e-01 -1.12275466e-01 -2.81978548e-01
-3.12721252e-01 -8.24179292e-01 3.25993598e-01 6.26862764e-01
-2.36665040e-01 9.58233848e-02 1.42915875e-01 2.95943141e-01
8.71720612e-02 -2.50760950e-02 3.80064696e-01 7.49055326e-01
-7.98532069e-01 -2.24087089e-01 -4.96668816e-01 5.14931202e-01
-2.61354983e-01 -1.07123643e-01 -1.78367049e-01 1.05141449e+00
6.17747545e-01 1.06021321e+00 3.63135450e-02 -5.06313264e-01
5.16037643e-01 -4.55224782e-01 3.49178404e-01 -5.40827274e-01
-5.12188792e-01 -1.43342182e-01 -3.58881891e-01 -4.43367839e-01
-5.20528495e-01 -2.52497554e-01 -8.33195627e-01 5.22240177e-02
-7.39643455e-01 -3.37023914e-01 6.85973525e-01 5.64964175e-01
4.75452125e-01 4.77683038e-01 3.91068518e-01 -1.01045942e+00
-2.57232159e-01 -6.80475414e-01 -8.85737002e-01 -4.83886003e-01
3.25235397e-01 -1.07897961e+00 -4.70213711e-01 -1.73207149e-01] | [8.414885520935059, -0.9942310452461243] |
abf993c6-6c2b-4228-8365-593d5eac0ab5 | compartmental-and-cellular-automaton-seirs | 2112.02661 | null | https://arxiv.org/abs/2112.02661v1 | https://arxiv.org/pdf/2112.02661v1.pdf | Compartmental and cellular automaton $SEIRS$ epidemiology models for the COVID-19 pandemic with the effects of temporal immunity and vaccination | We consider the $SEIRS$ epidemiology model with such features of the COVID-19 outbreak as: abundance of unidentified infected individuals, limited time of immunity and a possibility of vaccination. Within a compartmental realization of this model, we found the disease-free and the endemic stationary states. They exist in their respective restricted regions of the le via linear stability analysis. The expression for the basic reproductive number is obtained as well. The positions and heights of a first peak for the fractions of infected individuals are obtained numerically and are fitted to simple algebraic forms, that depend on model rates. Computer simulations of a lattice-based realization for this model was performed by means of the cellular automaton algorithm. These allowed to study the effect of the quarantine measures explicitly, via changing the neighbourhood size. The attempt is made to match both quarantine and vaccination measures aimed on balanced solution for effective suppression of the pandemic. | ['Taras Patsahan', 'Jaroslav Ilnytskyi'] | 2021-12-05 | null | null | null | null | ['epidemiology'] | ['medical'] | [-7.12568983e-02 1.73411816e-01 7.68541172e-02 2.67394364e-01
2.56705582e-01 -3.68688017e-01 9.03714359e-01 4.44918603e-01
-6.93888426e-01 9.25269604e-01 -1.90260157e-01 -3.78038496e-01
-6.31981254e-01 -9.70561862e-01 -4.31452841e-01 -1.07705069e+00
-8.33451092e-01 8.47059488e-01 2.75937945e-01 -5.69110751e-01
-5.24630435e-02 6.17904067e-01 -1.51504230e+00 -4.89370555e-01
1.02758825e+00 4.18455452e-01 2.60017458e-02 8.90197754e-01
-1.17799612e-02 2.33109146e-01 -6.31470501e-01 1.20923221e-01
3.44253302e-01 -3.81071240e-01 -2.74621069e-01 1.22075088e-01
-7.07136750e-01 -1.19493259e-02 -2.76916474e-01 8.56759846e-01
2.00525418e-01 -5.24236746e-02 1.09622443e+00 -1.11855912e+00
-1.13122515e-01 2.04576924e-01 -5.69700420e-01 4.31226134e-01
2.69600779e-01 2.05813527e-01 1.64188504e-01 -1.98858663e-01
8.41097295e-01 1.04412830e+00 6.02320075e-01 3.86037201e-01
-1.45591092e+00 -2.16177180e-01 -5.08709610e-01 -2.71847457e-01
-1.75856662e+00 -2.08275378e-01 1.38186112e-01 -4.99567300e-01
1.01420414e+00 3.88805658e-01 9.90880013e-01 4.47111547e-01
7.41037011e-01 -2.80407220e-01 1.28866196e+00 -5.71506023e-01
4.43080723e-01 1.86123833e-01 1.02953948e-01 6.55417800e-01
5.88750303e-01 2.38635406e-01 3.45231712e-01 -6.90367460e-01
1.02839553e+00 1.58631355e-01 -1.78961083e-01 -2.93449998e-01
-8.98452997e-01 6.67718410e-01 1.90451235e-01 7.56342888e-01
-6.88712895e-01 -1.54920146e-01 6.24933094e-02 2.44868085e-01
5.38161457e-01 1.15589917e-01 -3.62861842e-01 3.28349948e-01
-5.03538430e-01 3.97061169e-01 8.34634066e-01 4.66447264e-01
6.00759566e-01 -6.26182929e-02 1.74487799e-01 1.85110793e-01
8.40166807e-02 1.17110193e+00 -4.69142720e-02 -5.97536027e-01
-7.76137412e-03 6.75103188e-01 5.56598663e-01 -9.24005330e-01
-6.91174448e-01 -3.83685172e-01 -1.22600913e+00 -1.37910441e-01
5.26979566e-01 -5.14419377e-01 -7.01874137e-01 1.83728123e+00
5.30117035e-01 3.51304352e-01 1.98710963e-01 2.34422833e-01
1.54422568e-02 9.61610794e-01 1.93506390e-01 -1.03133011e+00
1.31459916e+00 -3.73478211e-03 -8.17887604e-01 4.99294668e-01
5.55902839e-01 -4.23531920e-01 3.02325308e-01 -6.02733232e-02
-9.80284989e-01 -3.33006307e-02 -6.06563151e-01 1.09999406e+00
-5.32571197e-01 -3.04369479e-01 6.21395372e-02 6.46137118e-01
-1.34033120e+00 4.52828676e-01 -8.58640671e-01 -7.48929441e-01
-5.22277832e-01 5.55914104e-01 -1.00007534e-01 1.60968080e-01
-1.29461718e+00 8.45292449e-01 4.89272550e-02 3.37081164e-01
-5.30156612e-01 -1.81187928e-01 -4.22268569e-01 -8.17957819e-02
-8.56558979e-02 -8.46155226e-01 5.38117230e-01 -8.27759147e-01
-9.30312693e-01 8.35728645e-01 -2.09393948e-02 -4.53034073e-01
4.22332525e-01 5.68151593e-01 -5.25159121e-01 3.37912977e-01
-2.44902894e-02 2.39789829e-01 4.27963555e-01 -1.13529539e+00
-1.70115069e-01 -4.82044518e-01 -1.18563958e-01 -4.72944491e-02
2.36635491e-01 3.08837295e-01 3.18586469e-01 -1.47788286e-01
-1.30794927e-01 -8.37674856e-01 -8.35343182e-01 -7.05796897e-01
-8.16074237e-02 -1.67672262e-01 4.01374608e-01 -5.46906710e-01
1.14342213e+00 -1.97344041e+00 1.92837238e-01 6.46104932e-01
8.00478384e-02 2.93707639e-01 9.72243249e-02 9.73366797e-01
-1.06362044e-03 1.03924274e-01 -3.96945149e-01 3.58558625e-01
-2.39497319e-01 2.60526717e-01 7.19306841e-02 9.36622441e-01
2.48373851e-01 2.94168532e-01 -4.86684084e-01 -4.12361026e-01
-4.09968896e-03 5.96644521e-01 -6.86284781e-01 1.38532624e-01
-2.65225261e-01 4.63336378e-01 -5.74701965e-01 3.30849513e-02
7.88153052e-01 -9.53245256e-03 4.41949278e-01 5.58360517e-01
-6.00402176e-01 -2.87390530e-01 -1.15636146e+00 4.75269020e-01
-3.94554697e-02 -2.60909021e-01 5.06133139e-01 -1.07227886e+00
8.29421580e-01 6.48960233e-01 7.67326295e-01 -3.44691843e-01
6.34290040e-01 2.64685869e-01 2.70166129e-01 -3.90533119e-01
1.79735318e-01 -4.09982949e-01 -1.21746145e-01 5.11687279e-01
-2.18909055e-01 2.64629602e-01 3.02289099e-01 1.75056085e-01
8.68914664e-01 -6.26424551e-01 4.20454860e-01 -1.10476923e+00
6.78704202e-01 1.63953811e-01 1.10461749e-01 4.83639210e-01
-1.92112535e-01 -3.09437662e-01 6.41773522e-01 -1.05570331e-01
-1.18173766e+00 -1.01663613e+00 -4.25667435e-01 3.86848420e-01
1.60793617e-01 2.34861844e-04 -1.24217260e+00 2.57633597e-01
-1.95421502e-01 2.74916857e-01 -6.83404326e-01 -1.23192355e-01
-7.29559243e-01 -1.37266636e+00 3.60695750e-01 -4.54799294e-01
3.83358479e-01 -9.74518299e-01 -8.11749578e-01 2.70481348e-01
-5.29777147e-02 -5.81289828e-01 1.05204657e-01 1.42954528e-01
-9.43229795e-01 -1.19511640e+00 -9.64673638e-01 -7.39910185e-01
8.29955697e-01 -2.29142606e-01 6.11566663e-01 4.72514838e-01
-2.87283421e-01 6.36804521e-01 1.54593885e-01 -5.63161187e-02
-9.73886967e-01 -5.14924563e-02 6.35929108e-01 -1.51130378e-01
1.79033875e-01 -3.76765609e-01 -7.06087053e-01 4.26923156e-01
-9.92181480e-01 -6.60549700e-01 -1.72665752e-02 4.78896171e-01
3.46217364e-01 3.03583801e-01 7.80452549e-01 -1.22581221e-01
8.03644717e-01 -6.14213288e-01 -8.94790292e-01 1.51870489e-01
-3.51001084e-01 -5.87499216e-02 4.56550092e-01 -2.99350917e-01
-7.65672088e-01 -1.60453945e-01 -1.87535331e-01 2.70316035e-01
-2.11203113e-01 1.44564211e-01 1.54054046e-01 5.17107695e-02
4.59767908e-01 3.29426736e-01 3.72584850e-01 -3.28131139e-01
-1.07174879e-03 8.47567379e-01 9.16023031e-02 -4.19220299e-01
5.31861603e-01 4.60979164e-01 2.69013673e-01 -1.41744614e+00
1.52635425e-01 -4.35534030e-01 -3.72494221e-01 -3.44520986e-01
7.44372249e-01 -6.20105028e-01 -1.11607587e+00 7.82635808e-01
-1.04713297e+00 -3.15212429e-01 -3.69095862e-01 4.51153725e-01
-6.83605492e-01 2.50269204e-01 -7.42316425e-01 -1.39471471e+00
-1.07823044e-01 -8.14573288e-01 2.71401495e-01 1.16672076e-01
7.86003396e-02 -1.28346312e+00 6.73202574e-01 -3.19731593e-01
5.27660966e-01 3.70568782e-01 1.21405411e+00 -5.99587262e-01
-4.74606991e-01 -1.74656585e-01 8.38651732e-02 -2.69545227e-01
-6.85801506e-02 3.04269522e-01 -2.42976874e-01 -4.11714643e-01
2.79621065e-01 3.94074976e-01 5.25011778e-01 8.10489237e-01
-8.63600224e-02 -5.41743219e-01 -5.80945790e-01 4.63117659e-03
1.71706307e+00 5.20114005e-01 4.32669133e-01 -6.47157878e-02
-3.02811712e-01 8.58866215e-01 2.60914773e-01 5.25484443e-01
1.41822070e-01 5.62533796e-01 2.18837574e-01 5.29351123e-02
6.16484404e-01 2.03531817e-01 2.08982751e-01 9.73814905e-01
-4.77137595e-01 -3.29890996e-01 -1.18906355e+00 7.28036165e-01
-1.63249099e+00 -1.09295225e+00 -4.51853007e-01 2.45482826e+00
5.34783185e-01 6.13619797e-02 6.56586766e-01 -1.49165317e-02
9.91777003e-01 -1.09554574e-01 8.31975490e-02 -5.76080680e-01
-2.36640736e-01 1.49884015e-01 8.72319400e-01 1.03151393e+00
-6.70727432e-01 3.60964090e-01 7.84654617e+00 6.24332190e-01
-8.11628520e-01 1.63777754e-01 5.24978042e-01 3.15485418e-01
-3.08316886e-01 -2.00920328e-01 -7.65098512e-01 3.96249652e-01
1.41916680e+00 -2.04573959e-01 4.25910324e-01 9.18233860e-03
5.14062583e-01 -5.04776299e-01 5.87467663e-02 9.63085592e-02
-3.45385432e-01 -1.07649589e+00 -1.99339464e-01 4.53114867e-01
7.13920712e-01 -3.40067632e-02 -1.40918896e-01 -4.20741215e-02
1.99927650e-02 -5.70596814e-01 7.35643804e-02 6.95699513e-01
7.07203448e-01 -1.12904155e+00 6.79743290e-01 9.02194023e-01
-1.20333493e+00 -6.36863783e-02 -1.20245010e-01 -3.80526423e-01
3.35092247e-01 3.14619213e-01 -5.79494238e-01 2.07488999e-01
2.18059972e-01 -1.13584861e-01 -1.81238770e-01 9.24088955e-01
4.65904057e-01 3.32205087e-01 -8.20799947e-01 -3.78462821e-01
2.27850199e-01 -7.90717125e-01 6.38275087e-01 8.89012933e-01
3.77563268e-01 4.27291870e-01 -1.33501396e-01 6.28567278e-01
7.23366499e-01 1.93222880e-01 -7.83382416e-01 3.02110217e-03
7.99640194e-02 6.02101564e-01 -1.13463402e+00 -1.49125323e-01
3.48362297e-01 4.12053764e-01 -1.58847302e-01 4.34950113e-01
-6.54461324e-01 -4.09713298e-01 7.20871389e-01 7.92866170e-01
1.96421176e-01 -4.45019245e-01 1.92238316e-01 -9.69481409e-01
-5.17521739e-01 -3.96711171e-01 3.24610732e-02 -7.56550133e-02
-7.82607019e-01 7.03758180e-01 4.58083689e-01 -5.61026037e-01
-8.21422815e-01 -2.97903806e-01 -6.01302505e-01 8.88371348e-01
-7.88770258e-01 -4.98476654e-01 7.69832611e-01 8.44491780e-01
-3.38726699e-01 4.51880395e-02 8.68579805e-01 3.03346217e-01
-6.29783511e-01 -1.42559744e-02 7.17607856e-01 -3.65725249e-01
-2.22870097e-01 -6.49431825e-01 9.56565794e-03 7.19704390e-01
-6.81065679e-01 9.29904461e-01 1.36663294e+00 -1.06926823e+00
-1.06901789e+00 -6.40504479e-01 1.38429606e+00 -9.62866470e-03
8.69296312e-01 -5.53788841e-01 -7.01777220e-01 3.88111889e-01
5.43420494e-01 -5.02194107e-01 1.70442730e-01 -4.26296175e-01
4.13403243e-01 1.05109088e-01 -1.41689301e+00 7.03825235e-01
9.04679358e-01 -1.60203710e-01 -2.70878196e-01 5.83967209e-01
6.06268644e-01 3.41631234e-01 -7.13787079e-01 2.85563231e-01
1.08534321e-01 -9.32047963e-01 9.70100880e-01 -5.79868197e-01
-1.31154105e-01 -1.05921201e-01 1.64517328e-01 -1.01112568e+00
-8.87811556e-02 -8.27855349e-01 4.56387460e-01 8.54868233e-01
2.06099674e-01 -1.15107930e+00 1.82560816e-01 2.04035461e-01
6.66367590e-01 -7.30067849e-01 -1.41589689e+00 -8.02445531e-01
1.60047665e-01 4.60615814e-01 3.19204897e-01 5.71529210e-01
2.72259980e-01 1.05678663e-01 -3.59013200e-01 2.28396758e-01
7.34923959e-01 -3.06011051e-01 2.79263407e-01 -1.23563445e+00
-2.89121360e-01 -1.79958239e-01 -2.89714187e-01 -3.98819625e-01
-8.84286761e-02 -1.53878704e-01 -8.19084048e-02 -1.25014043e+00
-4.63106222e-02 -5.82849383e-01 -5.01252376e-02 -3.03083360e-01
3.25350493e-01 -1.73134487e-02 -1.34076148e-01 2.32889980e-01
-1.62581339e-01 1.62115350e-01 9.51594114e-01 5.15446424e-01
-4.40445274e-01 2.82712668e-01 -5.85481413e-02 6.73968732e-01
9.22272980e-01 -5.27205765e-01 -3.08309913e-01 4.34841245e-01
4.14421529e-01 7.89440036e-01 3.81800920e-01 -7.88998127e-01
-9.48767513e-02 -3.34040672e-01 -2.94936359e-01 -5.39667785e-01
2.48191416e-01 -9.04222906e-01 8.80025089e-01 1.27969241e+00
1.91887200e-01 4.76164281e-01 1.11815803e-01 8.02257121e-01
2.35089600e-01 -3.96087974e-01 9.13978279e-01 1.85515955e-01
2.83442169e-01 7.73378164e-02 -1.27911496e+00 -1.46520659e-01
1.33593595e+00 -5.04298918e-02 -1.54402345e-01 -3.13348651e-01
-8.54749978e-01 1.20688327e-01 4.91560310e-01 -3.79124671e-01
6.87771961e-02 -7.93648303e-01 -5.63337982e-01 3.82262528e-01
-4.07886118e-01 -6.61473870e-01 5.24951458e-01 1.06371117e+00
-8.67604434e-01 6.02071047e-01 -5.07840574e-01 -4.02422994e-01
-1.12079847e+00 1.00284410e+00 8.49716961e-01 -4.80552286e-01
-5.87472366e-03 -1.21685699e-01 -6.82682097e-02 -4.36750144e-01
-1.00943036e-01 -2.00291604e-01 -3.89461219e-01 -2.80225333e-02
4.65714246e-01 5.09386659e-01 -4.39497858e-01 -1.03485394e+00
-4.90153193e-01 5.95338106e-01 5.03008723e-01 -1.19715862e-01
1.18872166e+00 -3.85130703e-01 -5.56979835e-01 1.87885195e-01
6.86248004e-01 1.23849086e-01 -8.72658074e-01 2.66700555e-02
-1.29165292e-01 1.03394032e-01 -5.46432018e-01 -3.66041631e-01
-5.37166297e-01 5.35713792e-01 6.46421671e-01 9.87749457e-01
9.33531046e-01 9.61648896e-02 2.63879508e-01 6.07399903e-02
4.52050954e-01 -5.22937238e-01 -6.16694927e-01 4.98184025e-01
5.56434691e-01 -4.85042125e-01 -3.31861079e-01 -4.68687266e-01
2.67626084e-02 5.68756819e-01 6.33059293e-02 -5.94265342e-01
1.21584523e+00 4.94061440e-01 -1.66822284e-01 -2.19954982e-01
-7.61821866e-01 -2.78250396e-01 -2.88493395e-01 7.54248857e-01
8.32654685e-02 3.53253901e-01 -1.18525577e+00 2.69397080e-01
2.36858279e-01 -1.06628567e-01 6.16429090e-01 6.58085167e-01
-6.68636024e-01 -8.92316818e-01 -6.22211635e-01 2.12540388e-01
-4.06244069e-01 -2.57374137e-03 -2.44870856e-01 1.17208362e+00
2.43135020e-01 9.27912712e-01 4.83886719e-01 1.42491415e-01
1.48530737e-01 8.53305310e-02 3.60188454e-01 -1.05003729e-01
-8.10518682e-01 2.01596200e-01 -9.59607214e-02 1.01238228e-01
-5.40405393e-01 -6.60625756e-01 -1.20626414e+00 -7.72677064e-01
-3.05402189e-01 7.97694862e-01 4.74946499e-01 9.35289145e-01
2.19322398e-01 -7.90814124e-03 5.68502188e-01 -5.24643123e-01
-5.30887187e-01 -6.87912166e-01 -1.22750604e+00 -2.28685334e-01
3.32674950e-01 -3.89097631e-01 -6.30933821e-01 -3.92393351e-01] | [5.927765369415283, 4.3777947425842285] |
61e9b752-5d61-4ac2-8d1c-28cd9e8f3bf5 | multi-view-inference-for-relation-extraction | 2104.13579 | null | https://arxiv.org/abs/2104.13579v1 | https://arxiv.org/pdf/2104.13579v1.pdf | Multi-view Inference for Relation Extraction with Uncertain Knowledge | Knowledge graphs (KGs) are widely used to facilitate relation extraction (RE) tasks. While most previous RE methods focus on leveraging deterministic KGs, uncertain KGs, which assign a confidence score for each relation instance, can provide prior probability distributions of relational facts as valuable external knowledge for RE models. This paper proposes to exploit uncertain knowledge to improve relation extraction. Specifically, we introduce ProBase, an uncertain KG that indicates to what extent a target entity belongs to a concept, into our RE architecture. We then design a novel multi-view inference framework to systematically integrate local context and global knowledge across three views: mention-, entity- and concept-view. The experimental results show that our model achieves competitive performances on both sentence- and document-level relation extraction, which verifies the effectiveness of introducing uncertain knowledge and the multi-view inference framework that we design. | ['Shikun Zhang', 'Canming Huang', 'Wei Ye', 'Bo Li'] | 2021-04-28 | null | null | null | null | ['document-level-relation-extraction'] | ['natural-language-processing'] | [-4.02626783e-01 6.40946567e-01 -9.41425085e-01 -4.52348053e-01
-5.91283858e-01 -6.01908565e-01 6.83286905e-01 3.39576334e-01
2.46495735e-02 1.01366544e+00 3.86623889e-01 -3.02188903e-01
-3.41210604e-01 -1.33684635e+00 -8.08011591e-01 -1.15101635e-01
2.32097115e-02 5.92223048e-01 4.84726816e-01 -4.28833179e-02
-1.57260567e-01 2.81120211e-01 -1.11614060e+00 3.94540668e-01
8.53394866e-01 9.38664198e-01 -1.84558451e-01 1.72025412e-01
-1.52015820e-01 1.17554724e+00 -4.87700492e-01 -1.19230890e+00
-2.53658086e-01 1.79384440e-01 -1.02267540e+00 -4.00296748e-01
-1.24643706e-01 -2.10249379e-01 -3.98678154e-01 1.04449058e+00
1.40747875e-01 -1.62174180e-01 8.73136520e-01 -1.11058259e+00
-9.07085657e-01 1.50925469e+00 -6.76322639e-01 4.01871026e-01
3.82563949e-01 -5.42875290e-01 1.44764841e+00 -1.03034759e+00
9.05358553e-01 1.10044003e+00 5.45747578e-01 1.73216268e-01
-8.62131596e-01 -5.24096727e-01 6.58885539e-01 7.16750860e-01
-1.64413333e+00 -4.06070739e-01 5.33655465e-01 -1.87434964e-02
1.31832981e+00 2.63292462e-01 2.96015054e-01 9.61734116e-01
4.35335860e-02 9.02507961e-01 9.18381155e-01 -2.48825133e-01
1.83647871e-02 2.81104475e-01 1.69804856e-01 4.49288756e-01
7.85544097e-01 -2.61117369e-01 -7.57563412e-01 -5.21860039e-03
5.16940057e-01 -2.42983788e-01 -3.70851874e-01 -1.04859464e-01
-9.79422808e-01 3.42541635e-01 4.40459251e-01 1.92087442e-01
-3.73494864e-01 -4.81824279e-02 3.26223940e-01 -3.71680446e-02
5.62753379e-01 2.80233532e-01 -1.03136706e+00 1.45831630e-01
-5.07118225e-01 1.46616802e-01 1.05160975e+00 1.59873915e+00
5.44232249e-01 -4.76704240e-01 -4.08196718e-01 5.98118484e-01
4.89160895e-01 5.71916997e-01 -1.12123847e-01 -5.64505875e-01
8.83279920e-01 9.25434768e-01 -8.73303339e-02 -1.14676964e+00
-2.02019304e-01 -6.84344292e-01 -8.14375162e-01 -6.30630553e-01
-6.81980029e-02 -4.03234996e-02 -7.70057201e-01 1.63262403e+00
7.25642204e-01 3.82543474e-01 4.78850722e-01 5.04232585e-01
1.32160485e+00 3.36968124e-01 2.46270061e-01 -3.57097328e-01
1.61958337e+00 -8.29538822e-01 -1.06549144e+00 -1.32148117e-01
5.67662597e-01 -2.89440304e-01 4.54609036e-01 1.83702677e-01
-8.74051154e-01 -3.86383981e-02 -1.13971484e+00 -1.56105474e-01
-5.84865212e-01 1.69615559e-02 8.05729091e-01 4.44460958e-01
-3.70473295e-01 4.31004107e-01 -7.91983306e-01 1.72887042e-01
4.63182539e-01 7.22505972e-02 -5.65834761e-01 -2.45958582e-01
-1.75886655e+00 1.16874194e+00 8.69937062e-01 3.69400024e-01
-5.85166037e-01 -7.06070185e-01 -9.37112331e-01 4.13555384e-01
1.24672341e+00 -1.06322539e+00 8.58423054e-01 1.61293060e-01
-1.00645804e+00 5.31375706e-01 -5.02564490e-01 -4.00587171e-01
-5.35180457e-02 -2.74428368e-01 -8.96386027e-01 5.16352281e-02
1.25649810e-01 -8.50180015e-02 4.02099520e-01 -1.50257564e+00
-7.05540538e-01 -4.74542677e-01 4.39833850e-01 3.03257972e-01
-7.13156760e-02 4.00756933e-02 -8.81123066e-01 -5.65557003e-01
3.26922834e-01 -3.75598878e-01 1.68068688e-02 -9.17321980e-01
-1.18807280e+00 -5.07122338e-01 5.24894416e-01 -6.71283305e-01
1.93380582e+00 -1.46924067e+00 3.76730710e-01 5.83714485e-01
5.22036672e-01 1.35037586e-01 6.12828910e-01 5.19502640e-01
3.12754996e-02 3.87488753e-01 -4.27384712e-02 3.18279937e-02
1.12257689e-01 3.71432751e-01 -3.97581995e-01 -3.15795213e-01
2.66303331e-01 1.43511498e+00 -9.97871459e-01 -8.39334607e-01
-2.98637033e-01 1.52126417e-01 -1.46533072e-01 5.64321801e-02
-3.98315966e-01 -1.65942475e-01 -7.93398976e-01 9.15530145e-01
6.82816327e-01 -5.81287563e-01 7.33183682e-01 -6.37137830e-01
5.18336892e-01 5.74810863e-01 -1.19661331e+00 1.21957815e+00
-3.56507480e-01 -2.34671354e-01 -3.31246763e-01 -5.91200411e-01
7.97024071e-01 3.09343755e-01 -1.58222523e-02 -6.51365891e-02
-2.72301823e-01 6.55240146e-03 -2.76963711e-01 -4.95657623e-01
6.97132826e-01 -4.69526835e-02 -9.10629332e-02 2.10534126e-01
3.80885482e-01 1.29612833e-01 3.49423081e-01 8.93022895e-01
1.10862458e+00 3.51787865e-01 7.34080553e-01 -6.37315884e-02
5.93330979e-01 -2.65115261e-01 9.71765816e-01 4.09593284e-01
3.33802998e-01 6.96745515e-02 9.67127323e-01 6.30555451e-02
-5.05244792e-01 -1.17482638e+00 -1.10193163e-01 6.16214454e-01
2.86681652e-01 -1.03634143e+00 -2.99587429e-01 -1.49367726e+00
2.09135652e-01 8.60648513e-01 -6.00193501e-01 -1.46796852e-01
-2.96130031e-01 -7.67296493e-01 5.48000216e-01 9.43987668e-01
4.31090534e-01 -5.90986967e-01 3.53549272e-01 1.86846852e-01
-4.85567540e-01 -1.47586954e+00 -1.30683914e-01 3.12297326e-02
-7.00520098e-01 -1.31276822e+00 9.88303218e-03 -3.94341052e-01
4.03136104e-01 -1.44148380e-01 1.59290600e+00 -1.05812125e-01
3.82972032e-01 2.72508293e-01 -5.52577794e-01 -2.77482688e-01
-6.72865734e-02 3.55476737e-01 4.21404392e-02 -2.78223962e-01
8.10003817e-01 -8.89911413e-01 -4.52358752e-01 2.80445695e-01
-6.43712461e-01 4.79107611e-02 6.70229137e-01 5.22528708e-01
8.91411304e-01 3.38027745e-01 9.97100472e-01 -1.57242918e+00
4.96151119e-01 -8.49506676e-01 -1.73858598e-01 1.09871888e+00
-1.12600303e+00 1.81519896e-01 2.34690681e-01 1.42859295e-02
-1.43416762e+00 -4.73659664e-01 4.24739122e-02 -6.12134896e-02
4.04456183e-02 1.38427949e+00 -7.69096136e-01 5.96246004e-01
3.75789702e-01 -1.04373515e-01 -7.30108142e-01 -3.42003316e-01
8.51970673e-01 5.98711967e-01 6.43462360e-01 -1.08050752e+00
7.87984014e-01 3.67621362e-01 1.30855128e-01 -4.29795682e-02
-1.52382481e+00 -4.19348896e-01 -6.93629563e-01 4.99193668e-02
5.87270856e-01 -1.25379896e+00 -8.35855424e-01 2.09561884e-02
-1.03588879e+00 3.51214558e-01 -2.47033685e-01 3.92450392e-01
-1.00469142e-01 2.05189973e-01 -6.61887288e-01 -7.05322027e-01
-4.77118611e-01 -5.83498001e-01 9.71313477e-01 3.80958259e-01
2.00512093e-02 -1.08989346e+00 -1.61704734e-01 4.53570873e-01
1.24208014e-02 2.48395652e-01 1.00347984e+00 -8.19746852e-01
-9.11299527e-01 -1.33226603e-01 -5.74725449e-01 1.04337133e-01
2.94655383e-01 9.47366133e-02 -9.04221535e-01 2.79573977e-01
-4.00554329e-01 -3.12765628e-01 8.68199050e-01 -8.93757045e-02
1.00752354e+00 -3.88488233e-01 -9.58375216e-01 3.39867473e-01
1.34461689e+00 -1.24756597e-01 7.45715499e-01 1.65597171e-01
1.02225673e+00 5.75557709e-01 8.32830787e-01 3.26055110e-01
1.28916585e+00 5.23141444e-01 2.30711475e-01 3.07063550e-01
3.93680073e-02 -7.27795601e-01 1.04700141e-02 1.12463951e+00
-5.94102025e-01 -3.25096339e-01 -7.05559492e-01 5.99796593e-01
-2.05008030e+00 -1.13335383e+00 -1.37745515e-01 1.68462992e+00
1.39910352e+00 2.79810578e-01 -2.19346225e-01 -2.48235650e-02
6.16758943e-01 3.60871735e-03 -5.33746004e-01 8.38522241e-02
-3.68351340e-01 1.25578761e-01 3.68167460e-01 4.62713361e-01
-1.00219965e+00 9.86663222e-01 5.54758644e+00 1.17534637e+00
-3.80876899e-01 2.12308262e-02 4.05728012e-01 3.08447361e-01
-1.00719059e+00 2.36059830e-01 -1.40180624e+00 2.17087984e-01
7.96528697e-01 -6.51285946e-01 -4.11400124e-02 8.51933479e-01
-4.44320232e-01 -1.57331854e-01 -1.22465014e+00 5.15052736e-01
-1.44657984e-01 -1.56767249e+00 2.40061730e-01 3.37557904e-02
8.16119492e-01 -2.30166912e-01 -2.33917296e-01 5.85290849e-01
7.33094454e-01 -1.02338505e+00 3.77955288e-01 8.23622525e-01
7.74796486e-01 -9.56722558e-01 9.91729915e-01 3.10002446e-01
-1.63113320e+00 3.07630330e-01 -1.41458079e-01 4.86211419e-01
3.57298642e-01 1.05813670e+00 -8.85392189e-01 1.66641235e+00
6.08144820e-01 9.65442419e-01 -5.65475941e-01 3.89461398e-01
-9.36229348e-01 5.39156556e-01 -2.34914392e-01 1.86905041e-01
-2.89973885e-01 3.08837183e-02 2.78385192e-01 1.23567069e+00
3.41339670e-02 4.89806265e-01 -5.40036112e-02 9.45996940e-01
-3.51315796e-01 -8.46593231e-02 -5.64903080e-01 1.28996847e-02
9.67849672e-01 1.33720481e+00 -4.44401413e-01 -5.17875969e-01
-5.70358992e-01 4.43895251e-01 8.11156690e-01 4.49953347e-01
-6.80438340e-01 -5.34532249e-01 3.23660284e-01 -2.72656918e-01
6.87626719e-01 3.11568230e-02 -3.18062127e-01 -1.56938446e+00
3.66249114e-01 -4.94616032e-01 8.55917871e-01 -7.38404214e-01
-1.48298764e+00 5.36890984e-01 4.51202095e-01 -8.35709035e-01
-3.48291755e-01 -3.58060688e-01 -2.20898002e-01 8.87362242e-01
-1.77923965e+00 -1.56969154e+00 1.23762853e-01 4.48443860e-01
-1.59665838e-01 1.11433573e-03 8.50114405e-01 2.93076664e-01
-6.04840815e-01 7.37716734e-01 -3.93567801e-01 3.04797947e-01
5.12033105e-01 -1.48608613e+00 3.39155108e-01 8.93826902e-01
4.30741280e-01 1.23217082e+00 4.15932268e-01 -1.17402208e+00
-1.19571459e+00 -1.05076385e+00 1.23933434e+00 -9.16327834e-01
7.40714550e-01 -7.80342743e-02 -1.03205299e+00 1.23093903e+00
-4.52800281e-02 3.05324525e-01 8.89167666e-01 1.03854930e+00
-9.00557697e-01 -1.00079827e-01 -1.13857985e+00 3.75790149e-01
1.34136951e+00 -6.69517756e-01 -9.60898280e-01 5.95119968e-02
1.04847014e+00 -6.22770071e-01 -1.73705912e+00 1.06025708e+00
3.89101654e-01 -7.23277569e-01 1.03363836e+00 -9.22322631e-01
6.67391717e-01 -6.16483092e-01 -2.30001837e-01 -1.32123518e+00
-2.57743627e-01 -1.75378203e-01 -1.40736556e+00 1.81948888e+00
1.09336054e+00 -5.06716251e-01 6.72195673e-01 6.53436661e-01
2.75631323e-02 -1.20058274e+00 -7.22836196e-01 -7.57642925e-01
-3.28426510e-01 -4.83993351e-01 1.15203106e+00 1.07340145e+00
5.22265315e-01 7.45887995e-01 -1.24265522e-01 7.13247478e-01
4.52924937e-01 4.66138124e-01 4.50170726e-01 -1.12789917e+00
-4.10885066e-01 1.14265671e-02 -3.06326509e-01 -9.26517606e-01
2.19103441e-01 -1.06519842e+00 -3.97189856e-01 -1.87900567e+00
5.35544157e-01 -7.46870279e-01 -5.61200559e-01 4.81276304e-01
-7.18461037e-01 -2.94012487e-01 -2.56654322e-01 -1.90570694e-03
-9.40878034e-01 6.63941622e-01 1.20842254e+00 4.87428345e-02
2.02734426e-01 1.50911752e-02 -1.22519290e+00 6.72021031e-01
3.92522812e-01 -3.08008283e-01 -6.78778231e-01 -2.60809749e-01
9.88081396e-01 7.96490461e-02 6.28505573e-02 -2.98212707e-01
4.80055600e-01 -2.90939152e-01 2.27521107e-01 -8.50393772e-01
1.24074727e-01 -7.74896741e-01 1.68278098e-01 -3.61808658e-01
-8.99250284e-02 -1.74956143e-01 3.32375392e-02 1.12084818e+00
-4.75940347e-01 -9.63103678e-03 -6.27007112e-02 -3.73560786e-02
-6.86778367e-01 5.11779428e-01 4.26148802e-01 4.25855279e-01
8.93036306e-01 3.39082897e-01 -7.60555148e-01 -1.44166157e-01
-9.51195240e-01 5.69873571e-01 -2.34116931e-02 3.27097118e-01
7.13441730e-01 -1.49479127e+00 -6.59773707e-01 -3.13410699e-01
5.14775753e-01 4.19024795e-01 2.08036125e-01 6.31344318e-01
1.35907575e-01 4.12348896e-01 4.56318140e-01 -9.62212384e-02
-1.12878764e+00 7.73513258e-01 1.72185767e-02 -8.34446490e-01
-4.52155322e-01 1.12442589e+00 -5.74095808e-02 -3.96960855e-01
-4.98905554e-02 -2.61537164e-01 -6.79608762e-01 2.24005673e-02
4.83611673e-01 3.16972256e-01 1.88068673e-01 -3.98626000e-01
-5.78474224e-01 1.78262323e-01 -3.90852571e-01 -3.50788236e-02
1.20751548e+00 -3.41952294e-01 -5.11296630e-01 5.08670330e-01
4.38731045e-01 4.93907601e-01 -6.83835566e-01 -7.38090098e-01
5.08173466e-01 -3.17853093e-01 -1.50240520e-02 -1.23045349e+00
-9.81043339e-01 3.54034692e-01 -4.88230079e-01 1.41749114e-01
9.06950712e-01 5.13504744e-01 6.90794587e-01 4.52697456e-01
7.53482819e-01 -8.73613477e-01 -6.10067070e-01 4.52981561e-01
8.62973869e-01 -1.09144509e+00 5.04323900e-01 -1.22741532e+00
-9.20202971e-01 7.89341331e-01 8.69414389e-01 4.87214267e-01
8.03913951e-01 5.72311997e-01 -3.19448590e-01 -5.37165642e-01
-1.13016665e+00 -4.60205764e-01 6.16546392e-01 6.03503346e-01
3.66418809e-01 3.78617108e-01 -2.01164991e-01 1.19044125e+00
-5.29437140e-02 5.91249317e-02 2.51098007e-01 7.85657108e-01
-1.50253430e-01 -1.25305820e+00 3.13780695e-01 7.18852758e-01
-5.53652763e-01 -3.57264847e-01 -3.22919101e-01 5.74551105e-01
2.10160524e-01 8.63298714e-01 -5.74220121e-01 -6.86777413e-01
4.72431868e-01 1.53871387e-01 6.11467719e-01 -9.22661662e-01
-3.68648499e-01 -4.52490896e-01 8.40526104e-01 -3.14324647e-01
-5.85865676e-01 -4.92481649e-01 -1.34574974e+00 -3.88149858e-01
-8.69356692e-01 4.42850858e-01 2.37729192e-01 1.23063266e+00
5.96375883e-01 6.74539208e-01 3.21139097e-01 2.17514649e-01
-2.75590390e-01 -1.08951080e+00 -8.11985373e-01 -9.55272913e-02
-1.54348925e-01 -9.03906584e-01 7.89580343e-04 4.23069969e-02] | [9.168194770812988, 8.322436332702637] |
50a0beaa-6125-4654-a4da-5a120e76a741 | conceptfusion-open-set-multimodal-3d-mapping | 2302.07241 | null | https://arxiv.org/abs/2302.07241v2 | https://arxiv.org/pdf/2302.07241v2.pdf | ConceptFusion: Open-set Multimodal 3D Mapping | Building 3D maps of the environment is central to robot navigation, planning, and interaction with objects in a scene. Most existing approaches that integrate semantic concepts with 3D maps largely remain confined to the closed-set setting: they can only reason about a finite set of concepts, pre-defined at training time. Further, these maps can only be queried using class labels, or in recent work, using text prompts. We address both these issues with ConceptFusion, a scene representation that is (1) fundamentally open-set, enabling reasoning beyond a closed set of concepts and (ii) inherently multimodal, enabling a diverse range of possible queries to the 3D map, from language, to images, to audio, to 3D geometry, all working in concert. ConceptFusion leverages the open-set capabilities of today's foundation models pre-trained on internet-scale data to reason about concepts across modalities such as natural language, images, and audio. We demonstrate that pixel-aligned open-set features can be fused into 3D maps via traditional SLAM and multi-view fusion approaches. This enables effective zero-shot spatial reasoning, not needing any additional training or finetuning, and retains long-tailed concepts better than supervised approaches, outperforming them by more than 40% margin on 3D IoU. We extensively evaluate ConceptFusion on a number of real-world datasets, simulated home environments, a real-world tabletop manipulation task, and an autonomous driving platform. We showcase new avenues for blending foundation models with 3D open-set multimodal mapping. For more information, visit our project page https://concept-fusion.github.io or watch our 5-minute explainer video https://www.youtube.com/watch?v=rkXgws8fiDs | ['Antonio Torralba', 'Florian Shkurti', 'Liam Paull', 'Madhava Krishna', 'Celso Miguel de Melo', 'Joshua B. Tenenbaum', 'Ayush Tewari', 'Nikhil Keetha', 'Soroush Saryazdi', 'Ganesh Iyer', 'Shuang Li', 'Tao Chen', 'Mohd Omama', 'Qiao Gu', 'Alihusein Kuwajerwala', 'Krishna Murthy Jatavallabhula'] | 2023-02-14 | null | null | null | null | ['robot-navigation'] | ['robots'] | [-3.49521521e-03 2.17467979e-01 -1.70307949e-01 -5.27984262e-01
-1.04340672e+00 -8.84874821e-01 7.42633283e-01 1.78747565e-01
-2.30802149e-01 5.47880113e-01 3.07189971e-01 -4.86949861e-01
-1.48242086e-01 -7.92222977e-01 -8.76447082e-01 -2.15423316e-01
9.35342070e-03 7.45070338e-01 2.50604093e-01 -5.31581461e-01
1.69830665e-01 3.24382246e-01 -2.10806346e+00 9.03298780e-02
4.72058237e-01 9.99115944e-01 5.64818263e-01 6.74786925e-01
-2.77286708e-01 5.83180904e-01 -1.24368131e-01 1.23139851e-01
3.05640638e-01 1.81918949e-01 -8.24196398e-01 -4.53362241e-02
5.40679872e-01 -3.59947592e-01 -5.02190590e-01 8.68649423e-01
3.06328803e-01 4.32345301e-01 3.23845863e-01 -1.65338182e+00
-5.54471910e-01 8.02216753e-02 -8.45251083e-02 -1.47462441e-02
1.01874685e+00 1.56959221e-01 9.52606797e-01 -9.79580581e-01
7.36264646e-01 1.34027016e+00 5.58500886e-01 3.87242824e-01
-1.01926148e+00 -6.31751657e-01 2.95150638e-01 2.14745238e-01
-1.44531429e+00 -4.70592052e-01 4.47890460e-01 -4.57714438e-01
1.18714595e+00 2.18042850e-01 7.18759775e-01 1.02823579e+00
3.79472002e-02 6.48383915e-01 7.96194494e-01 -3.96666676e-03
3.20923597e-01 -1.37821212e-01 -4.36356291e-02 8.53135347e-01
-9.70701873e-02 2.05753818e-01 -1.09556425e+00 -1.12593286e-01
7.43894041e-01 2.27609992e-01 -3.60523432e-01 -7.69318342e-01
-1.60814273e+00 7.42609560e-01 6.22759759e-01 -8.72241631e-02
-1.08188890e-01 4.10564661e-01 -8.38163942e-02 2.34397292e-01
3.30331415e-01 4.70295906e-01 -6.38243794e-01 -3.66948009e-01
-2.89454103e-01 4.89452302e-01 7.63374865e-01 1.53698325e+00
1.29123557e+00 -5.65719128e-01 4.19233888e-01 3.97741705e-01
4.06370312e-01 9.97509778e-01 1.98220745e-01 -1.46719754e+00
5.79951227e-01 5.23835838e-01 1.77007481e-01 -6.69534683e-01
-5.80357969e-01 5.13016544e-02 -2.20696941e-01 3.25970411e-01
9.24821794e-02 -1.36711746e-01 -9.83665705e-01 1.68697786e+00
4.59124178e-01 1.59982845e-01 1.89163968e-01 1.00927746e+00
9.36214983e-01 5.00705719e-01 -2.00249404e-01 5.46004534e-01
1.32269108e+00 -7.95950055e-01 -3.62356126e-01 -4.19983804e-01
7.25847840e-01 -5.49260199e-01 1.11474776e+00 2.12599084e-01
-7.67511666e-01 -5.91473043e-01 -1.08566988e+00 -6.04995012e-01
-8.61531258e-01 -4.92893219e-01 8.39030087e-01 1.73876837e-01
-1.39556539e+00 2.18247503e-01 -9.02850449e-01 -8.88479352e-01
3.34608316e-01 3.06966543e-01 -8.99962544e-01 -4.73420322e-01
-1.00869739e+00 9.59874034e-01 4.23057258e-01 -3.99800211e-01
-8.80531967e-01 -6.45610034e-01 -1.30018318e+00 -2.12520331e-01
6.74989522e-01 -1.02990663e+00 1.39671099e+00 -1.70378357e-01
-1.18862224e+00 8.02033663e-01 -3.17505598e-01 -3.76526117e-01
2.15856507e-01 -4.85415041e-01 -2.55531251e-01 2.72623211e-01
5.66262186e-01 1.31746292e+00 3.91991049e-01 -1.50177836e+00
-8.01101208e-01 -5.38926780e-01 8.12971652e-01 7.74444818e-01
1.90976202e-01 -6.56338453e-01 -6.35134935e-01 -1.87476240e-02
7.51296163e-01 -1.04638541e+00 -1.21334046e-01 4.10473764e-01
-2.60745913e-01 -1.99274272e-02 1.05101097e+00 -1.63689852e-01
3.48556161e-01 -2.12187696e+00 1.68026462e-01 1.60290286e-01
1.95406020e-01 -5.15205503e-01 -4.10009502e-03 5.25532305e-01
2.03195572e-01 4.37128693e-02 -2.46609211e-01 -5.95591366e-01
3.18406254e-01 5.50577581e-01 -5.23430407e-01 2.43524358e-01
-8.90798867e-02 1.00161862e+00 -1.16581595e+00 -4.01897818e-01
7.56868958e-01 4.92954254e-01 -6.64360106e-01 -3.88823561e-02
-4.39090818e-01 5.35486877e-01 -3.97983193e-01 7.50925601e-01
4.72461045e-01 -2.44850621e-01 -1.42921880e-01 4.79633845e-02
-1.40953735e-01 2.72495717e-01 -1.09641325e+00 2.84222651e+00
-5.35994411e-01 6.44759297e-01 -1.82744469e-02 -6.28773868e-01
8.43456209e-01 1.44491926e-01 5.76420903e-01 -5.30282497e-01
-1.46597028e-01 1.80789858e-01 -6.78019464e-01 -5.18709123e-01
7.16822565e-01 1.73095837e-02 -5.32571316e-01 4.04221177e-01
3.63495767e-01 -9.44600046e-01 -1.93542585e-01 5.54293811e-01
1.27594900e+00 6.63862109e-01 1.77607253e-01 -5.39909676e-02
4.59495857e-02 2.79580504e-01 3.80667485e-02 8.81273270e-01
-1.26760587e-01 7.76649714e-01 8.19448754e-02 -2.78111249e-01
-7.25448787e-01 -1.32857311e+00 -1.28630430e-01 1.09332895e+00
8.38101387e-01 -5.80519259e-01 -2.82984138e-01 -4.60565597e-01
2.80687004e-01 7.93715000e-01 -5.38183868e-01 1.51499938e-02
-6.23953864e-02 -1.49198687e-02 3.98608893e-01 3.89974535e-01
5.49486220e-01 -5.93170166e-01 -1.01036978e+00 -1.11364745e-01
-5.52745938e-01 -1.34018469e+00 -9.78580713e-02 5.43633699e-01
-6.35596991e-01 -1.08506441e+00 -2.09965006e-01 -5.56107104e-01
4.66248691e-01 8.25502694e-01 1.06735480e+00 -1.41072333e-01
3.99586223e-02 1.10237682e+00 -5.22976279e-01 -5.25984228e-01
-7.18096420e-02 -5.86936660e-02 2.36956835e-01 -5.14102399e-01
4.50893879e-01 -6.66747153e-01 -3.25854778e-01 3.80441725e-01
-6.06260657e-01 2.84532517e-01 3.89477789e-01 4.09643948e-01
6.98005855e-01 -2.57317066e-01 2.23012045e-01 -3.52213264e-01
1.59569189e-01 -8.33961129e-01 -4.01785731e-01 -4.61933240e-02
-2.48792931e-01 8.27603638e-02 -1.43327966e-01 -1.58135071e-01
-8.56178820e-01 2.29252100e-01 -7.55018846e-04 -7.52647460e-01
-5.62565148e-01 5.81723690e-01 -1.66992307e-01 -1.03973411e-01
8.17558289e-01 5.27883060e-02 1.29886225e-01 -2.06122965e-01
8.59932542e-01 5.40024161e-01 8.67878497e-01 -6.30652726e-01
8.00653458e-01 9.42032635e-01 5.62501512e-02 -6.02145612e-01
-7.77385116e-01 -6.99754477e-01 -7.30685771e-01 -1.27228498e-01
9.60672915e-01 -1.31603110e+00 -7.71829903e-01 -4.54676803e-03
-1.08225989e+00 -5.14729440e-01 -3.32612485e-01 5.80384672e-01
-1.02285159e+00 1.07198164e-01 -6.00733049e-02 -5.02996027e-01
2.30383098e-01 -8.79963517e-01 1.52791345e+00 4.30235155e-02
-2.93756515e-01 -7.45856583e-01 -2.40022048e-01 4.71510410e-01
2.09525496e-01 2.40548283e-01 6.07344747e-01 -4.82950926e-01
-8.44007313e-01 -1.71858847e-01 -2.09988341e-01 -2.65493929e-01
1.40928134e-01 -5.36669016e-01 -1.09830451e+00 -3.40386573e-03
-1.13282345e-01 -5.59891462e-01 8.31661642e-01 6.87640300e-03
8.76326025e-01 2.84054309e-01 -5.73401451e-01 7.16131508e-01
1.25450456e+00 1.53018683e-01 2.17170894e-01 2.80578613e-01
6.36537611e-01 5.37317812e-01 8.72662604e-01 4.75003839e-01
1.10514855e+00 6.46548927e-01 9.21439707e-01 1.35218307e-01
-1.05771884e-01 -5.61637044e-01 7.78136551e-02 2.72503912e-01
1.25377432e-01 -2.84825355e-01 -1.14752734e+00 4.36207116e-01
-2.09978962e+00 -8.47378910e-01 2.09560633e-01 2.03410840e+00
3.71811211e-01 1.75817072e-01 -3.27844650e-01 -2.54262686e-01
3.23208511e-01 9.00692344e-02 -8.62048090e-01 1.03300072e-01
-2.61684150e-01 -9.86454263e-02 5.08852601e-01 8.36820960e-01
-1.14015937e+00 9.58013654e-01 5.54290628e+00 3.51126313e-01
-9.25960422e-01 3.39853913e-01 -2.40215566e-03 -5.39132357e-01
-7.67209291e-01 3.04980874e-01 -5.95900714e-01 1.24029638e-02
6.32488668e-01 1.19600119e-02 6.27470791e-01 8.11101139e-01
-1.33347496e-01 -5.39622307e-01 -1.26475704e+00 1.45247173e+00
2.80615449e-01 -1.48906577e+00 -2.58310288e-01 1.62737012e-01
4.52969968e-01 7.51242042e-01 -1.48436621e-01 4.61265385e-01
5.89427352e-01 -9.35665190e-01 1.20584226e+00 5.70314944e-01
8.97551119e-01 -3.74451786e-01 4.36645538e-01 5.51516056e-01
-1.28168762e+00 -1.78918898e-01 -6.91021830e-02 -2.79687345e-01
4.38251138e-01 9.02730301e-02 -7.33978093e-01 6.46668792e-01
9.39592183e-01 8.23356986e-01 -2.69875884e-01 7.24370122e-01
-2.09135532e-01 -3.12026113e-01 -8.38884771e-01 1.07258238e-01
2.12128341e-01 1.01837970e-01 6.72910511e-01 5.66169739e-01
6.96813166e-01 3.74986291e-01 2.25220859e-01 8.67273986e-01
1.33235499e-01 -5.00535071e-01 -1.07356083e+00 3.28783244e-01
8.28090012e-01 9.09117341e-01 -5.82442045e-01 -2.77782679e-01
-5.52884817e-01 9.66127574e-01 1.96801111e-01 5.28372705e-01
-8.03077161e-01 -3.44754547e-01 1.08963048e+00 6.73452690e-02
2.21187949e-01 -7.97112823e-01 -3.39284301e-01 -1.14216900e+00
7.44436635e-03 -3.26933950e-01 2.16488734e-01 -1.39094126e+00
-8.33161592e-01 4.10286218e-01 4.27954108e-01 -1.39691150e+00
-2.96557903e-01 -5.68215311e-01 -9.21041071e-02 6.88145995e-01
-1.54532671e+00 -1.27282071e+00 -7.30328798e-01 9.25751865e-01
5.21719813e-01 1.30392984e-01 1.05522335e+00 3.22894156e-02
3.65631908e-01 -2.05924764e-01 -2.98440725e-01 -3.60631436e-01
6.78463936e-01 -1.20365071e+00 6.63318872e-01 4.25162584e-01
3.85413319e-01 4.67815340e-01 6.25981092e-01 -5.66087782e-01
-1.65857792e+00 -7.64411628e-01 6.98824286e-01 -1.06656671e+00
4.65875477e-01 -7.36902833e-01 -5.03091335e-01 9.52722967e-01
-7.80597702e-02 7.65049458e-02 6.62070990e-01 3.35346103e-01
-5.26544333e-01 1.63753986e-01 -1.11943889e+00 5.04182994e-01
1.62479162e+00 -8.74870420e-01 -5.87907493e-01 4.31813627e-01
1.24598718e+00 -9.82333183e-01 -8.39042008e-01 4.09786105e-01
5.85206270e-01 -1.03287065e+00 1.06751525e+00 -1.99436441e-01
2.29450569e-01 -5.10437310e-01 -9.48248804e-01 -1.13918471e+00
-2.61840522e-02 -4.43868935e-01 -1.35759950e-01 5.98860741e-01
3.65808517e-01 -7.49272227e-01 7.46086240e-01 7.92437077e-01
-6.53971195e-01 -5.55601835e-01 -1.07825315e+00 -5.05471289e-01
-3.19749296e-01 -1.13035166e+00 9.40856457e-01 7.83963799e-01
2.47552365e-01 2.85519481e-01 3.64995375e-02 5.09674907e-01
4.76471752e-01 1.60077691e-01 1.08132827e+00 -1.17075646e+00
8.50545429e-03 -1.86661988e-01 -7.41611004e-01 -1.44891798e+00
1.57636195e-01 -8.18933785e-01 2.62472898e-01 -1.84860480e+00
-2.35626951e-01 -7.54564106e-01 -2.05800380e-03 9.23463404e-01
3.58056307e-01 1.72890484e-01 3.79387170e-01 3.55948031e-01
-8.18924725e-01 6.99191093e-01 1.02863431e+00 -1.43035799e-01
-1.16427541e-01 -3.18180293e-01 -7.49308586e-01 6.95829272e-01
7.29134381e-01 -1.52126685e-01 -8.77408147e-01 -9.22265530e-01
4.31125730e-01 1.89702764e-01 7.33955264e-01 -1.34529150e+00
5.03009677e-01 -2.14756191e-01 1.50049329e-01 -7.70829260e-01
1.11709440e+00 -8.95693839e-01 1.16935402e-01 -3.41407172e-02
-1.12596884e-01 -1.11352028e-02 4.27478313e-01 7.20466137e-01
-1.49099216e-01 2.20153645e-01 2.57596765e-02 -3.63278657e-01
-1.43722391e+00 4.06316161e-01 -1.22953407e-01 -1.82458252e-01
1.00685680e+00 -4.88971680e-01 -5.05961657e-01 -7.42920101e-01
-8.01625490e-01 6.40942633e-01 8.48057270e-01 8.82912934e-01
8.33523095e-01 -1.26770186e+00 -2.06869915e-01 3.46607089e-01
6.47111833e-01 6.52817726e-01 4.00448978e-01 6.07904255e-01
-3.17779660e-01 6.29958510e-01 -1.66621134e-01 -9.46286857e-01
-7.58688450e-01 4.23082530e-01 2.45848671e-01 6.14015520e-01
-6.82282031e-01 1.00500762e+00 3.94985795e-01 -1.14459896e+00
1.11690998e-01 -5.48410714e-01 1.76526085e-01 -2.71807313e-01
1.80410355e-01 -2.54039355e-02 -7.66469538e-02 -8.61356139e-01
-5.36452651e-01 7.31238782e-01 5.61063290e-01 -5.11605859e-01
1.03490055e+00 -6.73509479e-01 1.43431127e-01 7.61469901e-01
1.08810461e+00 -3.16475838e-01 -1.51250505e+00 -3.65827024e-01
-3.34278166e-01 -3.84933829e-01 -1.48074552e-02 -7.15627015e-01
-3.54961216e-01 7.70074487e-01 4.97274250e-01 -1.17213232e-02
7.90643871e-01 6.21343732e-01 5.69137990e-01 8.36991370e-01
1.14659047e+00 -7.98760414e-01 2.16948688e-01 7.33177185e-01
8.89892817e-01 -1.60771549e+00 -1.20499603e-01 -3.92363459e-01
-6.72529995e-01 9.74747539e-01 6.05242908e-01 1.95758134e-01
8.27239811e-01 1.41099289e-01 1.74308345e-01 -4.54048693e-01
-8.53380382e-01 -5.47883570e-01 4.42662761e-02 8.91241074e-01
-1.73582047e-01 1.69100255e-01 7.45246768e-01 2.30006024e-01
-5.62437177e-01 -8.35707411e-02 3.00354242e-01 1.22746682e+00
-6.38231039e-01 -7.05489993e-01 -2.62902230e-01 2.47444317e-01
3.82109344e-01 -3.13080885e-02 -7.51758516e-02 7.54727721e-01
4.25838411e-01 1.16262662e+00 3.82725924e-01 -6.03700519e-01
2.55632907e-01 3.02931666e-02 4.94981974e-01 -8.53903651e-01
2.44368613e-01 -3.15698296e-01 2.02062964e-01 -1.12044716e+00
-5.84503651e-01 -7.86383212e-01 -1.68273127e+00 -2.55307317e-01
-7.14975297e-02 -7.95385614e-02 1.02366567e+00 1.06849349e+00
7.31418610e-01 1.41003892e-01 2.38225404e-02 -1.32085478e+00
-3.73210162e-02 -7.08090246e-01 -3.42851639e-01 1.70394599e-01
6.07244730e-01 -1.05113304e+00 -3.00468922e-01 -1.21271119e-01] | [4.764964580535889, 0.315258264541626] |
ec259496-c797-48e4-b8a4-677ae49e92eb | wear-a-multimodal-dataset-for-wearable-and | 2304.05088 | null | https://arxiv.org/abs/2304.05088v2 | https://arxiv.org/pdf/2304.05088v2.pdf | WEAR: An Outdoor Sports Dataset for Wearable and Egocentric Activity Recognition | Though research has shown the complementarity of camera- and inertial-based data, datasets which offer both modalities remain scarce. In this paper, we introduce WEAR, an outdoor sports dataset for both vision- and inertial-based human activity recognition (HAR). The dataset comprises data from 18 participants performing a total of 18 different workout activities with untrimmed inertial (acceleration) and camera (egocentric video) data recorded at 10 different outside locations. Unlike previous egocentric datasets, WEAR provides a challenging prediction scenario marked by purposely introduced activity variations as well as an overall small information overlap across modalities. Provided benchmark results reveal that single-modality architectures each have different strengths and weaknesses in their prediction performance. Further, in light of the recent success of transformer-based temporal action localization models, we demonstrate their versatility by applying them in a plain fashion using vision, inertial and combined (vision + inertial) features as input. Results demonstrate both the applicability of vision-based transformers for inertial data and fusing both modalities by means of simple concatenation, with the combined approach (vision + inertial features) being able to produce the highest mean average precision and close-to-best F1-score. The dataset and code to reproduce experiments is publicly available via: https://mariusbock.github.io/wear/ | ['Hilde Kuehne', 'Kristof Van Laerhoven', 'Michael Moeller', 'Marius Bock'] | 2023-04-11 | null | null | null | null | ['egocentric-activity-recognition', 'action-localization', 'human-activity-recognition', 'wearable-activity-recognition', 'human-activity-recognition'] | ['computer-vision', 'computer-vision', 'computer-vision', 'time-series', 'time-series'] | [ 1.69974759e-01 -4.01590824e-01 -2.54963726e-01 -1.25633016e-01
-7.20241249e-01 -6.07687175e-01 1.02771723e+00 -1.14152946e-01
-4.61116612e-01 5.99507391e-01 7.35708654e-01 1.34304211e-01
-1.75776437e-01 -2.21073300e-01 -6.34089291e-01 -4.24136847e-01
-1.92417413e-01 6.82691485e-02 1.65412411e-01 -2.55569905e-01
1.54832363e-01 3.29960257e-01 -1.93711782e+00 2.38829479e-01
2.23600477e-01 1.02703094e+00 -5.42677417e-02 1.07175112e+00
6.25871122e-01 9.59993064e-01 -2.52536476e-01 -1.88824371e-01
4.59680825e-01 -3.26464027e-01 -5.65710664e-01 7.42918998e-02
7.39083052e-01 -3.27494353e-01 -7.60684609e-01 2.27460891e-01
8.22333574e-01 3.46615195e-01 1.95387051e-01 -1.31973398e+00
-4.00045365e-01 -1.35043323e-01 -2.92212248e-01 6.07692242e-01
1.32477343e+00 5.87497830e-01 7.71395445e-01 -7.79761791e-01
6.59050643e-01 7.11809397e-01 9.43364739e-01 2.51301974e-01
-1.21694148e+00 -7.32900500e-02 -2.47695073e-01 4.49101806e-01
-1.34571970e+00 -7.36927330e-01 6.13550007e-01 -5.30650795e-01
1.49060071e+00 3.37501764e-01 1.07345200e+00 1.69400704e+00
4.10550237e-01 8.83804977e-01 1.21816635e+00 -3.66992146e-01
9.86297876e-02 -1.84769675e-01 -5.10760322e-02 4.21171546e-01
-1.24724628e-02 3.21578085e-01 -1.24951601e+00 1.17397979e-01
6.05789959e-01 2.74780184e-01 -3.33968222e-01 -6.75744653e-01
-1.83583343e+00 1.68662071e-01 3.75987798e-01 2.37079591e-01
-4.94814575e-01 2.78084189e-01 4.88019675e-01 1.99358597e-01
2.92744368e-01 2.20665663e-01 -3.86054665e-01 -1.16731489e+00
-7.06446409e-01 3.51619869e-01 7.20748782e-01 7.78045475e-01
3.84208679e-01 -2.11294275e-02 -2.07952380e-01 5.40378392e-01
-1.16849495e-02 5.55171907e-01 6.55013144e-01 -9.03663397e-01
5.78132451e-01 4.66870934e-01 2.19675660e-01 -1.22365677e+00
-7.50621080e-01 -3.79821032e-01 -3.40299994e-01 1.49190677e-02
6.49492443e-01 -9.20870248e-03 -6.24108851e-01 1.58974314e+00
4.12891775e-01 6.37359142e-01 -2.14209318e-01 1.14501381e+00
4.90223616e-01 8.61731544e-02 8.32872093e-02 2.97020581e-02
1.40106153e+00 -8.73917043e-01 -4.91760582e-01 -4.67337847e-01
6.42382324e-01 -6.19427383e-01 1.12423980e+00 4.22444731e-01
-1.01618528e+00 -7.55164444e-01 -1.16741920e+00 -1.76069155e-01
-4.97267723e-01 1.55950189e-01 7.56077170e-01 7.89413571e-01
-9.99542236e-01 4.71421152e-01 -1.25656486e+00 -8.32804859e-01
4.63676080e-02 3.07204068e-01 -7.88786590e-01 -3.39020342e-02
-9.21706796e-01 1.13936329e+00 1.56996429e-01 1.83941394e-01
-7.86826432e-01 -6.22628987e-01 -9.99314010e-01 -4.13502276e-01
3.47323775e-01 -8.74511361e-01 1.17109120e+00 -6.14653766e-01
-1.46654332e+00 6.78114951e-01 -8.73413216e-03 -6.35671794e-01
8.71294677e-01 -8.20006669e-01 -5.21249592e-01 2.18680277e-01
-1.61998589e-02 4.05235291e-01 5.23370981e-01 -6.67965472e-01
-4.99652803e-01 -4.37407494e-01 1.90895498e-01 5.79018295e-01
-3.61689329e-01 -2.09683582e-01 -4.39265907e-01 -4.03093696e-01
3.06748282e-02 -1.12090337e+00 2.41658166e-01 -3.52839172e-01
-9.50238928e-02 1.36832148e-01 4.97170806e-01 -7.44475663e-01
1.20194876e+00 -2.00599742e+00 2.83605099e-01 -1.97678521e-01
1.37968749e-01 2.70709917e-02 1.94084987e-01 7.33715296e-01
-1.61788724e-02 -4.40256357e-01 1.18070297e-01 -6.02412760e-01
7.78145269e-02 3.16377133e-01 9.16143283e-02 9.12155926e-01
-8.16163048e-02 1.01409662e+00 -1.15038979e+00 -3.08435231e-01
9.12223935e-01 7.30009139e-01 -3.41086000e-01 -5.50038442e-02
4.09991205e-01 9.06477869e-01 -1.09210119e-01 9.01531100e-01
3.41391973e-02 5.43681607e-02 2.27455571e-02 -2.81484216e-01
-5.37700355e-02 2.18738228e-01 -1.33962739e+00 2.29743552e+00
-5.55534482e-01 7.50748813e-01 -3.22113603e-01 -5.99617660e-01
5.32711983e-01 3.47355634e-01 8.55463088e-01 -8.74134243e-01
2.09621683e-01 9.61841070e-05 -3.41330409e-01 -7.08194554e-01
8.42849374e-01 9.73546132e-02 -3.16849411e-01 2.65087754e-01
3.07733446e-01 2.45632008e-01 3.05060267e-01 1.04766920e-01
1.48451507e+00 7.21506655e-01 4.57351983e-01 1.41115651e-01
3.84546995e-01 6.08316099e-04 3.03761393e-01 8.67655456e-01
-6.34571731e-01 9.69494045e-01 7.43310479e-03 -3.77966404e-01
-8.65125537e-01 -1.27177656e+00 3.41544673e-02 1.08183861e+00
1.44796655e-01 -8.31955314e-01 -3.69852722e-01 -4.06889111e-01
5.37080318e-02 3.26697499e-01 -7.43736506e-01 -1.63533106e-01
-5.54800570e-01 -5.17604113e-01 7.32067525e-01 7.78025866e-01
4.92600054e-01 -7.83682048e-01 -1.02956009e+00 3.14991437e-02
-4.23404396e-01 -1.25407636e+00 -2.66664714e-01 -6.22174377e-03
-7.46290028e-01 -1.21350586e+00 -6.14585102e-01 -2.95298416e-02
7.20633492e-02 4.10894871e-01 1.06832659e+00 -2.30295330e-01
-3.19099933e-01 1.25967622e+00 -3.53555828e-01 -1.89277276e-01
2.31401712e-01 5.03808893e-02 5.73779166e-01 7.35445097e-02
4.68673825e-01 -6.78611636e-01 -8.49425852e-01 2.39225477e-01
-5.05689204e-01 -1.88678559e-02 3.98130834e-01 6.95082724e-01
3.89745802e-01 -6.68482542e-01 3.45332958e-02 -8.68267193e-02
4.92156535e-01 -6.71162724e-01 6.65252879e-02 -1.93395820e-02
-5.05562603e-01 -4.95903969e-01 1.30703136e-01 -3.29505205e-01
-8.91761541e-01 2.52273500e-01 1.25967592e-01 -5.02625227e-01
-4.79482234e-01 4.26766068e-01 1.53184637e-01 1.23444721e-02
9.92237270e-01 3.07266563e-01 8.65827501e-02 -5.17588317e-01
3.72435659e-01 4.59553182e-01 9.89009917e-01 -4.91188914e-01
4.08745766e-01 6.60005748e-01 4.15676720e-02 -7.71865308e-01
-5.33106565e-01 -7.47394919e-01 -9.43366528e-01 -7.01205969e-01
7.84030855e-01 -1.23406887e+00 -8.86487067e-01 6.90192878e-01
-5.27854860e-01 -1.89335331e-01 -4.77573901e-01 9.76878405e-01
-9.53462124e-01 4.99968350e-01 -5.04574597e-01 -8.49979281e-01
9.13794339e-02 -9.01911318e-01 1.37315774e+00 2.20428348e-01
-5.22035718e-01 -9.12704468e-01 4.20015305e-01 7.92905092e-01
3.26590747e-01 7.44610190e-01 -3.99609387e-01 -4.48552310e-01
-2.53213555e-01 -5.95831335e-01 2.13406727e-01 1.16125807e-01
1.00514434e-01 -4.15288717e-01 -1.11479926e+00 -3.67151678e-01
-1.48168728e-01 -3.51189256e-01 5.73858440e-01 2.73197740e-01
3.77362043e-01 1.70297906e-01 -1.43418401e-01 4.44731265e-01
1.21091568e+00 -3.12855780e-01 6.73373699e-01 7.47336984e-01
8.79967332e-01 2.14313343e-01 6.55383050e-01 6.47769868e-01
7.44292617e-01 1.12255204e+00 2.95535803e-01 2.05602214e-01
-2.49672279e-01 -2.77677268e-01 6.66574657e-01 5.63822746e-01
-5.95537186e-01 9.96662751e-02 -1.05640697e+00 6.01687551e-01
-2.02686524e+00 -1.31655729e+00 -2.41078451e-01 2.45041585e+00
3.87116849e-01 1.50123402e-01 5.34027994e-01 3.35595131e-01
1.95695549e-01 3.21346402e-01 -3.29324067e-01 -2.09464654e-01
-6.77837953e-02 -1.39140179e-02 6.35483384e-01 2.75377750e-01
-1.30687845e+00 3.66203278e-01 6.28162575e+00 3.30974191e-01
-8.98513019e-01 3.23353499e-01 -1.29330605e-02 -7.99947858e-01
3.28463584e-01 1.67692900e-02 -4.80329841e-01 5.81668377e-01
1.33950877e+00 5.59282191e-02 4.56266582e-01 7.50889719e-01
3.19330871e-01 -5.83882987e-01 -1.16986489e+00 1.18309474e+00
3.17660630e-01 -1.01522088e+00 -6.66859984e-01 2.13068023e-01
4.58970249e-01 5.16090035e-01 -1.66636765e-01 3.51209432e-01
-1.44343272e-01 -6.23715103e-01 1.05439126e+00 9.47326303e-01
4.94131058e-01 -3.63866568e-01 5.63371122e-01 2.61168242e-01
-1.35610425e+00 -3.23700905e-01 2.68651932e-01 -8.38095963e-01
3.55975837e-01 2.21155658e-01 -6.17850244e-01 8.85860145e-01
9.72800553e-01 1.23210371e+00 -8.80532384e-01 9.08137560e-01
-1.06950276e-01 3.92158806e-01 -5.26489615e-01 2.41171032e-01
2.98101325e-02 -1.10369205e-01 6.85592473e-01 1.25165260e+00
3.39844018e-01 -2.29643494e-01 2.65601546e-01 1.88866943e-01
4.92044121e-01 -2.17845231e-01 -1.01852071e+00 1.00795344e-01
2.42525011e-01 1.15700400e+00 -5.13658643e-01 -2.47420326e-01
-5.58639050e-01 1.01028848e+00 6.09591752e-02 1.50138751e-01
-1.01763272e+00 -2.84305250e-04 8.76540184e-01 5.93225993e-02
5.15811853e-02 -8.18472028e-01 -2.99936712e-01 -1.38230777e+00
3.67371619e-01 -6.76037490e-01 5.83382666e-01 -1.03524709e+00
-8.49137247e-01 1.35524839e-01 1.38824686e-01 -1.74560392e+00
-5.76468825e-01 -4.93260294e-01 -3.10252547e-01 6.42118514e-01
-1.05764437e+00 -1.36120605e+00 -6.99790359e-01 7.46866345e-01
4.70684171e-01 8.95216390e-02 6.87160313e-01 4.43566769e-01
-5.61619639e-01 3.20603430e-01 6.63791317e-03 -9.33841467e-02
7.68546879e-01 -1.31050491e+00 2.09762275e-01 1.00420439e+00
4.60876733e-01 7.60366559e-01 9.16512489e-01 -4.97785091e-01
-1.96706474e+00 -5.50884306e-01 8.36558998e-01 -1.38285863e+00
7.64727890e-01 -2.70495355e-01 -1.99081793e-01 9.25767124e-01
1.46920860e-01 1.34663969e-01 7.10072219e-01 1.84709042e-01
-1.23605840e-01 -4.18491587e-02 -8.43005598e-01 5.81163406e-01
1.43776715e+00 -7.97571599e-01 -7.17642725e-01 3.29155415e-01
2.85569131e-02 -8.63415956e-01 -1.24273682e+00 2.76677758e-01
1.15063310e+00 -1.43611050e+00 1.18561161e+00 -3.92596036e-01
2.68306106e-01 -4.37922984e-01 -4.26119953e-01 -1.04589474e+00
-2.47486144e-01 -4.02454197e-01 -5.61492562e-01 9.87212241e-01
-1.73597917e-01 -5.75512886e-01 6.14404678e-01 4.95354563e-01
-2.19326153e-01 -5.35814881e-01 -1.18745375e+00 -8.68893385e-01
-6.21476889e-01 -9.36544716e-01 2.58981735e-01 6.96975112e-01
4.72617149e-01 9.58505571e-02 -7.05126405e-01 -6.29137531e-02
2.81536698e-01 -2.35367492e-01 1.18732595e+00 -8.62333357e-01
-4.29882467e-01 -2.25397140e-01 -1.11718118e+00 -8.93190145e-01
-3.16960454e-01 -5.96545219e-01 -1.48361504e-01 -1.37933505e+00
-6.70460165e-02 1.07848480e-01 -4.85609949e-01 5.08832097e-01
1.00698583e-01 7.35144615e-01 4.06703711e-01 3.06741565e-01
-7.49947488e-01 3.34225029e-01 8.26632142e-01 9.83512402e-02
-1.68043032e-01 -7.41053224e-02 -4.08979952e-01 4.95813578e-01
5.63866615e-01 -1.37975868e-02 -5.07241964e-01 -3.93781722e-01
1.59596026e-01 3.02138436e-03 9.16794479e-01 -1.72528636e+00
2.25005031e-01 1.28461272e-01 6.23811841e-01 -2.70039052e-01
8.23432446e-01 -6.56789660e-01 4.52839196e-01 4.84320015e-01
-2.22108532e-02 3.11967015e-01 1.55437276e-01 8.53671312e-01
-1.37498066e-01 4.74438667e-01 7.70461634e-02 -1.19634286e-01
-1.19766116e+00 1.24228531e-02 -2.85059869e-01 -1.27110213e-01
1.09120631e+00 -8.75022709e-01 -3.67870688e-01 -4.10336822e-01
-9.80569184e-01 3.01531889e-03 6.39859676e-01 9.97699082e-01
2.84013897e-01 -1.36362779e+00 -3.84717286e-01 2.69116908e-01
4.78703499e-01 -6.79490149e-01 5.22807062e-01 1.70618665e+00
-2.75502712e-01 5.01712978e-01 -5.67898870e-01 -9.66848254e-01
-1.17902184e+00 2.98589975e-01 3.08133811e-01 -1.16616160e-01
-6.84091270e-01 4.90121126e-01 -4.61318582e-01 -3.92684937e-01
4.80134971e-03 -2.42127493e-01 -5.55985682e-02 1.47029951e-01
5.11597276e-01 7.92403519e-01 3.20797056e-01 -1.03535342e+00
-6.94629133e-01 4.46180314e-01 3.79505843e-01 -1.94721058e-01
1.14145374e+00 -4.68973607e-01 4.77961570e-01 7.71559834e-01
1.00739467e+00 -9.97951105e-02 -1.49515522e+00 -1.37054756e-01
-1.22020692e-01 -7.85852194e-01 -1.13248155e-02 -7.28614032e-01
-6.85974598e-01 6.77001715e-01 1.04129934e+00 1.83801085e-01
1.06976998e+00 -1.77300006e-01 5.36364496e-01 1.81682050e-01
6.16504431e-01 -1.16131663e+00 1.30471196e-02 4.12005454e-01
7.19222248e-01 -1.27291489e+00 3.64981472e-01 5.62606342e-02
-8.30408871e-01 8.26395869e-01 5.02739549e-01 -1.60366166e-02
4.52711225e-01 8.18982199e-02 -1.36263650e-02 -1.95295483e-01
-6.22393012e-01 -4.53888327e-01 2.58901179e-01 7.61176169e-01
4.57334369e-01 9.63210873e-03 -2.87559956e-01 3.39911461e-01
-2.06756368e-01 4.41836447e-01 4.07080352e-01 1.54736555e+00
9.08311754e-02 -7.54477918e-01 -3.96095991e-01 3.79386723e-01
-3.15679491e-01 4.28144395e-01 -1.78651169e-01 8.60654235e-01
1.17263623e-01 1.00070071e+00 -2.62076408e-02 -7.80283689e-01
6.33569479e-01 1.51136935e-01 6.57036483e-01 -2.24970832e-01
-8.06630492e-01 -3.05743575e-01 4.05479729e-01 -1.22556591e+00
-6.99540496e-01 -1.12931669e+00 -7.17351317e-01 -4.26643491e-01
7.02041462e-02 -4.19602245e-01 7.09572434e-01 1.07697642e+00
6.98847771e-01 4.71670568e-01 3.21653098e-01 -1.54573405e+00
-4.21738297e-01 -1.06373978e+00 -5.51109672e-01 6.37107015e-01
3.96995485e-01 -1.06080723e+00 -2.20167220e-01 2.99579650e-01] | [7.7265825271606445, 0.5440412759780884] |
6ac668c3-095e-47c0-81d4-a7bf4f740556 | occlusion-aware-video-object-inpainting | 2108.06765 | null | https://arxiv.org/abs/2108.06765v1 | https://arxiv.org/pdf/2108.06765v1.pdf | Occlusion-Aware Video Object Inpainting | Conventional video inpainting is neither object-oriented nor occlusion-aware, making it liable to obvious artifacts when large occluded object regions are inpainted. This paper presents occlusion-aware video object inpainting, which recovers both the complete shape and appearance for occluded objects in videos given their visible mask segmentation. To facilitate this new research, we construct the first large-scale video object inpainting benchmark YouTube-VOI to provide realistic occlusion scenarios with both occluded and visible object masks available. Our technical contribution VOIN jointly performs video object shape completion and occluded texture generation. In particular, the shape completion module models long-range object coherence while the flow completion module recovers accurate flow with sharp motion boundary, for propagating temporally-consistent texture to the same moving object across frames. For more realistic results, VOIN is optimized using both T-PatchGAN and a new spatio-temporal attention-based multi-class discriminator. Finally, we compare VOIN and strong baselines on YouTube-VOI. Experimental results clearly demonstrate the efficacy of our method including inpainting complex and dynamic objects. VOIN degrades gracefully with inaccurate input visible mask. | ['Chi-Keung Tang', 'Yu-Wing Tai', 'Lei Ke'] | 2021-08-15 | null | http://openaccess.thecvf.com//content/ICCV2021/html/Ke_Occlusion-Aware_Video_Object_Inpainting_ICCV_2021_paper.html | http://openaccess.thecvf.com//content/ICCV2021/papers/Ke_Occlusion-Aware_Video_Object_Inpainting_ICCV_2021_paper.pdf | iccv-2021-1 | ['texture-synthesis', 'video-inpainting'] | ['computer-vision', 'computer-vision'] | [ 9.88002867e-02 -1.49715438e-01 -3.49487782e-01 2.36221906e-02
-8.06236148e-01 -6.80714607e-01 2.65860468e-01 -6.65633619e-01
1.87525243e-01 9.06106532e-01 3.76706719e-01 2.26891354e-01
3.91758651e-01 -4.32793617e-01 -1.15328395e+00 -5.55688262e-01
6.13635108e-02 2.45320216e-01 4.85935241e-01 2.16613963e-01
-1.12071205e-02 6.04705870e-01 -1.43738163e+00 7.36168206e-01
8.17412496e-01 1.15128207e+00 3.02872866e-01 9.20385599e-01
-8.63507614e-02 1.14445233e+00 -5.62237322e-01 -1.23507276e-01
8.22439194e-01 -4.48635310e-01 -7.30174780e-01 6.30040407e-01
1.46272886e+00 -9.11047816e-01 -8.78839135e-01 7.11752415e-01
3.19612585e-02 2.49343410e-01 5.05205870e-01 -1.25518024e+00
-5.08038521e-01 1.23230211e-01 -9.01025891e-01 4.19281095e-01
3.83152544e-01 8.19710016e-01 8.26437593e-01 -1.12184393e+00
1.22228396e+00 1.50991607e+00 4.25860733e-01 5.27042270e-01
-1.39535975e+00 -5.83116114e-01 3.73015463e-01 1.21476606e-01
-1.14425933e+00 -6.77300692e-01 1.07524312e+00 -5.29086828e-01
6.52254820e-01 4.80787247e-01 7.82249510e-01 9.93263841e-01
3.68601054e-01 1.16481829e+00 7.89751291e-01 1.54434517e-02
-1.03037409e-01 -3.28745782e-01 -3.75103414e-01 5.46326339e-01
1.34986445e-01 3.79089236e-01 -6.48197949e-01 -1.52144711e-02
1.32546878e+00 1.69373423e-01 -7.50899673e-01 -2.88201958e-01
-1.18199861e+00 3.68684739e-01 1.37044623e-01 -4.89082299e-02
-2.91559279e-01 6.58485591e-01 2.12771773e-01 3.16406608e-01
8.08272481e-01 1.65457085e-01 -2.64929265e-01 3.21161747e-02
-1.36103010e+00 4.69153583e-01 4.24283415e-01 1.39256799e+00
7.49668360e-01 5.33364832e-01 -6.23559237e-01 6.11214280e-01
1.75972342e-01 5.00662327e-01 -5.48579767e-02 -1.66643524e+00
5.78883827e-01 5.54407984e-02 4.10873234e-01 -1.05839562e+00
2.31864065e-01 -1.94550633e-01 -6.49454534e-01 5.36576211e-01
5.23288071e-01 -1.64824240e-02 -9.71023619e-01 1.30912375e+00
6.54609442e-01 8.06791723e-01 -3.05157065e-01 1.33710933e+00
1.09106338e+00 9.21773374e-01 1.25398459e-02 -3.05372417e-01
1.13880086e+00 -1.60887277e+00 -1.06879175e+00 -3.23976398e-01
2.38745864e-02 -1.00254226e+00 9.07106042e-01 4.20198619e-01
-1.68175793e+00 -8.01108837e-01 -7.04401374e-01 -4.97622102e-01
5.92784107e-01 -2.23302558e-01 5.77778637e-01 2.77906358e-01
-8.63782465e-01 5.81301689e-01 -7.19216704e-01 1.96313158e-01
8.56129229e-01 -2.27871742e-02 -4.21899915e-01 -4.65750664e-01
-7.66625822e-01 2.24655405e-01 1.58031791e-01 8.17777887e-02
-1.48745573e+00 -1.24151826e+00 -9.23181713e-01 -4.76204343e-02
4.12023455e-01 -9.10533667e-01 8.19028616e-01 -1.37922204e+00
-1.30437624e+00 7.05604970e-01 -4.72964734e-01 -2.56798506e-01
9.94024754e-01 -3.25554729e-01 -6.04871847e-02 6.34708047e-01
1.91557974e-01 1.00647843e+00 1.44077408e+00 -1.55139828e+00
-6.01381063e-01 3.38947810e-02 -2.02337652e-01 2.02902079e-01
1.20623708e-01 -7.80313089e-02 -9.13440943e-01 -1.28875601e+00
-1.25349566e-01 -5.54163277e-01 1.25859991e-01 7.89714277e-01
-3.07633400e-01 1.12731867e-01 1.37803030e+00 -1.14812791e+00
1.10640872e+00 -2.16994834e+00 2.68841207e-01 -3.12792748e-01
5.06246388e-01 1.42084464e-01 -4.72161561e-01 1.79302844e-03
1.15400990e-02 -1.02267705e-01 -2.56601542e-01 -6.37568295e-01
-4.63040501e-01 5.03535926e-01 -6.16209805e-01 6.69904351e-01
4.25839573e-01 1.28101563e+00 -9.17835593e-01 -6.54377878e-01
5.28282523e-01 4.36680615e-01 -8.85385513e-01 4.86889482e-01
-5.22741854e-01 9.94246304e-01 -3.36998940e-01 1.13684702e+00
1.04793668e+00 -8.65961537e-02 -1.72897279e-01 -4.73758370e-01
2.86031254e-02 -2.05855742e-01 -1.14168572e+00 1.99283695e+00
-3.98375750e-01 1.02330709e+00 5.68589687e-01 -5.59249759e-01
4.75936234e-01 2.32520729e-01 8.51855278e-01 -5.38382530e-01
-2.93778349e-02 1.57526806e-01 -4.70032692e-01 -8.23263943e-01
5.08680463e-01 2.68623888e-01 5.49398065e-01 8.65427479e-02
6.20903820e-02 -2.51660973e-01 1.72372356e-01 2.63726741e-01
7.90422678e-01 6.39077425e-01 -3.45702022e-01 -2.30885386e-01
2.92573631e-01 -1.90471411e-01 8.01308274e-01 5.91348588e-01
-3.00316185e-01 1.46025574e+00 2.74330318e-01 -6.46481037e-01
-1.27935278e+00 -9.85955477e-01 -1.37966901e-01 7.54998803e-01
5.70139110e-01 -1.91975370e-01 -6.39825642e-01 -5.68158507e-01
1.33753344e-01 3.60808462e-01 -6.23286366e-01 2.68828511e-01
-1.02952647e+00 -1.40189350e-01 1.46938249e-01 4.42365766e-01
5.00991344e-01 -1.17693377e+00 -4.11987305e-01 4.30650979e-01
-8.14325631e-01 -1.36018562e+00 -1.32751417e+00 -4.91682053e-01
-9.94312644e-01 -1.13845432e+00 -1.03820109e+00 -8.48115563e-01
4.84265298e-01 6.66651726e-01 1.30378258e+00 3.20282638e-01
-5.52909374e-01 3.47585052e-01 -2.11659968e-01 2.30322123e-01
-4.46664393e-01 -4.56059247e-01 -4.73743737e-01 4.78017360e-01
-5.29072702e-01 -6.68212712e-01 -1.06177831e+00 4.27401721e-01
-1.29767847e+00 2.40489811e-01 1.58701390e-01 8.34821403e-01
6.73380315e-01 -3.75114471e-01 1.00374222e-01 -5.85299492e-01
-1.88581765e-01 -3.71903032e-01 -3.90102059e-01 2.45500296e-01
9.85245183e-02 -2.30630025e-01 4.04661328e-01 -7.27800965e-01
-1.20008409e+00 -3.43873873e-02 8.28292128e-03 -1.16830301e+00
-1.36946449e-02 -2.81216830e-01 -2.77393788e-01 -2.81675756e-01
3.72673661e-01 3.04834098e-01 -1.02751881e-01 -5.45238256e-01
3.98303807e-01 2.29227155e-01 6.58374250e-01 -6.83258176e-01
6.88457727e-01 1.01824415e+00 -1.98179916e-01 -7.90276170e-01
-7.41504252e-01 -4.94293988e-01 -3.52192193e-01 -4.18085545e-01
8.87572348e-01 -1.08079433e+00 -5.59570193e-01 4.52277064e-01
-1.47824299e+00 -8.31719875e-01 -7.10641861e-01 1.73478529e-01
-7.09705055e-01 7.60577798e-01 -9.22239423e-01 -3.40278596e-01
-3.93584728e-01 -1.30787098e+00 1.54777956e+00 -1.02719471e-01
-2.05894541e-02 -9.16920483e-01 -1.91104963e-01 5.10923088e-01
2.08219677e-01 4.44982022e-01 3.71310592e-01 3.94248277e-01
-1.34044409e+00 3.92616689e-01 -4.09086674e-01 2.85646379e-01
1.44377172e-01 1.51150405e-01 -9.94868279e-01 -4.46615487e-01
3.88192646e-02 -2.42119402e-01 1.13906825e+00 7.52336264e-01
1.16176879e+00 -4.80126709e-01 -2.06160605e-01 1.34576750e+00
1.38696468e+00 -7.36675709e-02 9.01537597e-01 1.53100937e-05
1.27891517e+00 6.39066279e-01 7.51114190e-01 3.19707096e-01
2.67429948e-02 9.61229801e-01 5.37134886e-01 -3.55307996e-01
-1.15123057e+00 -2.64146060e-01 3.53254288e-01 3.20777655e-01
-1.65801615e-01 -5.79367220e-01 -2.88944483e-01 6.93887770e-01
-1.76849580e+00 -1.07978570e+00 -3.38005304e-01 2.04519320e+00
1.00315428e+00 -3.71136755e-01 -2.35359743e-02 -2.95085579e-01
4.35923696e-01 3.97118986e-01 -5.16662538e-01 2.39341319e-01
-5.17924666e-01 1.38011158e-01 4.32821482e-01 1.06309831e+00
-1.05080187e+00 1.18630266e+00 5.66068888e+00 1.21874154e+00
-1.11273658e+00 5.71173429e-01 1.02414298e+00 -4.33404863e-01
-4.98986274e-01 5.12012541e-02 -5.68139076e-01 6.20273530e-01
1.35882959e-01 2.78794140e-01 3.84652644e-01 4.96023059e-01
5.73382556e-01 -2.60907829e-01 -8.91765714e-01 1.24043810e+00
1.36959344e-01 -1.81288540e+00 1.77044094e-01 -1.03998102e-01
1.35790682e+00 -1.79235145e-01 -2.52725296e-02 -2.77612627e-01
-1.76486701e-01 -8.74685824e-01 1.29832804e+00 5.71975946e-01
1.16364968e+00 -4.18052912e-01 5.17832600e-02 -3.00085336e-01
-1.32125223e+00 2.79776901e-02 -2.89926291e-01 2.41173044e-01
5.74793160e-01 4.53603119e-01 -1.23758808e-01 4.81034607e-01
7.68266380e-01 1.22845793e+00 -2.74126917e-01 1.22358644e+00
-1.81957662e-01 6.06931925e-01 -2.42659181e-01 8.58257294e-01
3.32619622e-02 -1.82703868e-01 9.78311598e-01 1.06579375e+00
1.03055209e-01 1.28857344e-01 4.57467794e-01 1.14725363e+00
2.88239103e-02 -1.04348511e-01 -4.16968495e-01 4.07213181e-01
1.45331919e-01 1.12993479e+00 -6.74523473e-01 -6.17041349e-01
-4.81734395e-01 1.41415262e+00 6.99283779e-02 9.17921722e-01
-8.83652031e-01 2.28068173e-01 8.90710294e-01 5.83731890e-01
5.47654033e-01 -2.32095376e-01 -2.84602612e-01 -1.57089019e+00
2.29468644e-01 -6.89837217e-01 2.45667920e-02 -8.63157928e-01
-1.16186750e+00 4.78696406e-01 -7.52813444e-02 -1.48995006e+00
-2.24345922e-02 -2.19886914e-01 -5.47003984e-01 6.84561968e-01
-1.66131401e+00 -1.35810804e+00 -5.99328995e-01 7.32038498e-01
1.22222233e+00 2.28063613e-01 7.24415341e-03 6.46547019e-01
-3.26788008e-01 4.75582302e-01 -3.41594988e-03 5.36699258e-02
9.23051178e-01 -6.69875145e-01 3.51191193e-01 1.06613111e+00
9.72740874e-02 1.35704324e-01 7.62531459e-01 -1.03126919e+00
-1.61486506e+00 -1.57886159e+00 3.21180522e-01 -5.13397753e-01
1.77124277e-01 -2.62425542e-01 -1.06758893e+00 7.82536268e-01
3.61615717e-01 7.58338809e-01 -2.37964481e-01 -9.55663681e-01
-1.68513432e-01 -1.10425875e-01 -1.05251384e+00 6.46503687e-01
1.22066295e+00 -9.07732025e-02 -1.21781556e-02 5.72104514e-01
8.63236845e-01 -6.86987579e-01 -7.77381003e-01 3.02562267e-01
5.38902879e-01 -1.09493148e+00 1.24859977e+00 -4.80612904e-01
6.71624303e-01 -5.89084625e-01 8.61200020e-02 -7.89040864e-01
-1.24900043e-01 -1.25446296e+00 -5.56811333e-01 1.18337727e+00
-2.24146783e-01 -1.75334483e-01 9.34760988e-01 5.84193349e-01
-4.00325954e-01 -6.06422484e-01 -8.29138160e-01 -6.66901886e-01
-1.53712630e-01 -5.34572661e-01 2.58010060e-01 9.43920612e-01
-7.26771593e-01 -3.21801692e-01 -9.76688027e-01 -1.19586587e-01
7.85829842e-01 3.21515054e-01 1.04223537e+00 -6.63409889e-01
-5.65017521e-01 -1.91139445e-01 -2.02010170e-01 -1.67528594e+00
1.19214922e-01 -5.64395905e-01 2.63350476e-02 -1.19711518e+00
1.23349652e-02 -4.05530185e-01 3.81598622e-01 2.47643545e-01
-4.32237506e-01 8.64374995e-01 3.49820971e-01 5.46578705e-01
-4.67363924e-01 7.17672288e-01 2.09236479e+00 -3.95597637e-01
-3.91236842e-01 -2.58266509e-01 -2.02227563e-01 5.56892514e-01
2.18472660e-01 -4.70162183e-01 -4.47542340e-01 -7.31544733e-01
-2.47129217e-01 4.64909881e-01 7.42664278e-01 -7.60717869e-01
-9.35913548e-02 -3.63149077e-01 5.69091618e-01 -4.37406778e-01
4.91177350e-01 -7.44751215e-01 5.45562744e-01 2.53345728e-01
-1.06101692e-01 -1.80488721e-01 2.64265567e-01 8.32696974e-01
-3.66357267e-01 1.38048762e-02 8.80955100e-01 -5.17876260e-02
-8.13553870e-01 9.73851979e-01 -1.18143573e-01 5.73023915e-01
1.07890940e+00 -5.31605005e-01 5.75533137e-03 -5.06787837e-01
-7.53165126e-01 9.83791128e-02 7.91305065e-01 6.21454298e-01
9.85682130e-01 -1.18800306e+00 -1.09081352e+00 4.82818991e-01
-1.58177659e-01 2.37583145e-01 7.28681386e-01 9.28716600e-01
-1.18864012e+00 2.17905529e-02 -2.25336194e-01 -8.14991355e-01
-1.16375923e+00 6.19833112e-01 3.35289240e-01 1.21556446e-01
-1.20797384e+00 9.39906538e-01 9.57082152e-01 3.00508320e-01
3.38541865e-01 -4.93382990e-01 3.33247870e-01 -3.59880090e-01
8.18645895e-01 2.63050258e-01 -2.68619359e-01 -6.88358068e-01
1.31433338e-01 7.46744692e-01 7.27055073e-02 -4.68043871e-02
9.58297729e-01 -4.56221163e-01 -5.92799038e-02 -1.47702023e-02
1.13663232e+00 2.60567218e-01 -2.31821775e+00 -1.29117623e-01
-7.35297322e-01 -1.23414028e+00 -7.52401054e-02 -2.09811583e-01
-1.65272272e+00 8.85099947e-01 2.92884409e-01 -3.11596811e-01
9.83071864e-01 -1.97279826e-01 1.27712023e+00 -4.81088877e-01
2.80569553e-01 -5.98608553e-01 3.13481688e-01 1.05549760e-01
1.19807327e+00 -1.12601030e+00 1.45546183e-01 -1.02324271e+00
-5.53754568e-01 1.12384236e+00 7.09098518e-01 -3.54082584e-01
3.83546233e-01 2.61451334e-01 2.59622708e-02 1.25609145e-01
-7.31142044e-01 1.88323647e-01 6.27486467e-01 6.47163868e-01
1.59924686e-01 -3.47416908e-01 -1.61615938e-01 -1.53571963e-01
4.71145362e-01 -1.54319331e-01 4.89034086e-01 5.36643386e-01
-1.32634103e-01 -8.33666265e-01 -5.77506125e-01 1.11262031e-01
-5.75447261e-01 -2.24540100e-01 1.22349903e-01 6.79418623e-01
1.11427851e-01 8.00922871e-01 1.85033306e-01 2.39985257e-01
-9.77638364e-02 -4.52033073e-01 7.25177348e-01 -3.77537727e-01
-5.01412094e-01 5.89580953e-01 -3.56357619e-02 -1.13484132e+00
-4.15929943e-01 -5.24304509e-01 -9.38101768e-01 -1.39652297e-01
-3.43007520e-02 -3.44590783e-01 6.47322461e-02 6.13271534e-01
5.39652288e-01 5.57627857e-01 4.68264520e-01 -1.50089777e+00
8.74856412e-02 -7.39624143e-01 -4.63728637e-01 8.33235264e-01
6.84752345e-01 -6.18830144e-01 -3.92831981e-01 6.02559328e-01] | [10.801901817321777, -1.3492519855499268] |
80b03680-7136-4dcf-a854-a002f565a28c | a-unified-technique-for-entropy-enhancement | null | null | https://www.sciencedirect.com/science/article/pii/S0010482522002165?dgcid=author | https://www.sciencedirect.com/science/article/pii/S0010482522002165?dgcid=author | A Unified Technique for Entropy Enhancement Based Diabetic Retinopathy Detection Using Hybrid Neural Network | In this paper, a unified technique for entropy enhancement-based diabetic retinopathy detection using a hybrid neural network is proposed for diagnosing diabetic retinopathy. Medical images play crucial roles in the diagnosis, but two images representing two different stages of a disease look alike. It, consequently, make the process of diagnosis extraneous and error-prone. Therefore, in this paper, a technique is proposed to address these issues. Firstly, a novel entropy enhancement technique is devised exploiting the discrete wavelet transforms to improve the visibility of the medical images by making the subtle features more prominent. Later, we designed a computationally efficient hybrid neural network that efficiently classifies diabetic retinopathy images. To examine the effectiveness of our technique, we have chosen three datasets: Ultra-Wide Filed (UWF) dataset, Asia Pacific Tele Ophthalmology Society (APTOS) dataset, and MESSIDOR-2 dataset. In the end, we performed extensive experiments to validate the performance of our technique. In addition, the comparison of the proposed scheme – in terms of accuracy, specificity, sensitivity, precision and recall curve, and area under the curve – with some of the best contemporary schemes shows the significant improvement of our techniques in terms of diabetic retinopathy classification. | ['R. Noor', 'M. Arif', 'A. Ullah', 'M. Imran', 'Fatima'] | 2020-07-01 | null | null | null | journal-2020-7 | ['diabetic-retinopathy-detection'] | ['medical'] | [ 3.41063082e-01 -2.87038743e-01 1.20532222e-01 -4.01490003e-01
-4.05531347e-01 7.36170486e-02 1.86175451e-01 1.51183203e-01
-5.70185840e-01 6.94863617e-01 3.47503662e-01 -1.98468715e-01
-6.07360959e-01 -7.86364675e-01 1.23069607e-01 -1.04861224e+00
-1.00957125e-03 -1.64252385e-01 6.85674399e-02 -2.20351201e-02
5.34104109e-01 6.30196393e-01 -1.80793190e+00 1.06047593e-01
1.30690539e+00 9.86323595e-01 1.19754806e-01 6.94904983e-01
1.40135556e-01 9.22752976e-01 -3.40834737e-01 -2.43100792e-01
3.41905951e-01 -4.16348815e-01 -4.58517432e-01 3.10245633e-01
4.69434291e-01 -5.17203212e-01 -3.15267205e-01 1.07059336e+00
9.73945916e-01 1.33669287e-01 7.36754239e-01 -4.46084291e-01
-5.95458508e-01 -1.40473887e-01 -9.00230348e-01 4.71500754e-01
5.48339188e-02 1.36714622e-01 3.16368699e-01 -3.26112032e-01
5.10315955e-01 8.62492800e-01 5.86916327e-01 2.29755029e-01
-6.49363518e-01 -4.46462065e-01 -4.73735869e-01 4.53401744e-01
-1.09363711e+00 -4.16132748e-01 5.54068685e-01 -3.29376101e-01
5.10685205e-01 3.92097384e-01 6.68881655e-01 3.05979490e-01
3.64402384e-01 4.67728853e-01 1.57181978e+00 -6.12222910e-01
-6.58852234e-02 2.19325274e-01 2.79352635e-01 8.24215710e-01
4.81368512e-01 4.05229688e-01 1.04912683e-01 -1.04328498e-01
6.53468966e-01 9.73608047e-02 -5.95201969e-01 1.34620622e-01
-7.22930193e-01 7.41572082e-01 2.49076009e-01 2.38789901e-01
-4.93072569e-01 -5.02790689e-01 5.57140827e-01 1.15169555e-01
3.78790468e-01 1.18373379e-01 -1.89191207e-01 7.27928989e-03
-6.80484712e-01 -1.32324934e-01 4.84973133e-01 3.11908424e-01
1.33544326e-01 -2.36716598e-01 -2.88015664e-01 9.63633478e-01
3.81376058e-01 5.68018854e-02 6.22573912e-01 -8.99621129e-01
7.14313686e-02 6.69232011e-01 -4.52031121e-02 -8.58694851e-01
-4.63288516e-01 -5.70393622e-01 -1.23834646e+00 5.92780054e-01
1.88405737e-01 -2.73936510e-01 -1.23319650e+00 9.11929548e-01
2.66334891e-01 6.11460842e-02 6.04209602e-01 9.11651254e-01
1.05115378e+00 5.72957039e-01 -2.42227856e-02 -5.47813773e-01
1.44271207e+00 -6.80528104e-01 -7.68669426e-01 3.79937559e-01
3.12976658e-01 -8.97749424e-01 4.71673012e-01 4.89837646e-01
-1.03652930e+00 -4.50739413e-01 -7.73043454e-01 -1.60723682e-02
-9.56721976e-03 4.75979537e-01 5.41016996e-01 5.39271474e-01
-8.04402292e-01 4.14788216e-01 -5.17840028e-01 -6.00127697e-01
5.27360976e-01 3.58933687e-01 -2.82330483e-01 -1.97572529e-01
-8.23233485e-01 9.31923568e-01 4.86026406e-01 1.12776108e-01
1.53631628e-01 -2.54381478e-01 -5.34887969e-01 -1.11317836e-01
-2.09362581e-01 -6.38302267e-01 8.76652718e-01 -8.17817867e-01
-1.29589117e+00 7.60844231e-01 -1.24095716e-01 -5.14972389e-01
4.45592344e-01 1.59129664e-01 -6.26286864e-01 4.99892861e-01
-4.37808067e-01 3.09727818e-01 6.41822875e-01 -9.09281492e-01
-1.14643764e+00 -4.55152929e-01 -7.11750761e-02 1.71536878e-01
-4.94247079e-02 3.96111965e-01 -1.63884789e-01 -5.23960173e-01
1.91528141e-01 -5.72798371e-01 -2.11364031e-01 1.43796250e-01
-8.85117948e-02 -1.67150840e-01 7.58239508e-01 -1.02778232e+00
1.09260440e+00 -2.35163951e+00 -2.79272050e-01 2.65832156e-01
3.78541619e-01 8.59847009e-01 8.82340446e-02 -1.10960856e-01
-3.83063182e-02 -2.35925838e-02 -2.11340472e-01 1.29507571e-01
-5.69717050e-01 1.75288320e-01 2.66119301e-01 5.03998935e-01
2.38822252e-01 2.97469586e-01 -3.38498831e-01 -7.51634359e-01
3.57329100e-01 8.68136823e-01 -3.29179168e-01 -1.32293152e-02
5.64145565e-01 3.09881181e-01 -3.67990911e-01 8.91094148e-01
9.72434103e-01 -2.90105622e-02 -9.14459080e-02 -4.94933963e-01
-3.95257473e-01 -3.06415111e-01 -9.85255957e-01 7.31218815e-01
-1.22035541e-01 7.69122601e-01 -3.70150566e-01 -9.38959181e-01
9.19630885e-01 5.97630799e-01 4.18155193e-01 -8.23928654e-01
3.72331560e-01 1.89570144e-01 8.54652449e-02 -1.18739188e+00
2.38220185e-01 -4.10704575e-02 9.05708611e-01 3.08802333e-02
-1.28715560e-01 5.49916089e-01 3.54540914e-01 -3.54773045e-01
7.62943089e-01 -1.49205267e-01 7.10252702e-01 3.43074292e-01
7.25960314e-01 -1.03127256e-01 6.22635305e-01 3.96547228e-01
-6.65271699e-01 6.18406296e-01 3.23164910e-01 -4.93759692e-01
-8.95069957e-01 -6.25288963e-01 -6.08915746e-01 9.73330364e-02
1.50669232e-01 9.72783044e-02 -5.03836572e-01 -5.38819432e-01
-2.18206584e-01 3.13494772e-01 -4.05378550e-01 -5.05006462e-02
-3.56109053e-01 -1.27284718e+00 3.76685113e-01 3.30855280e-01
1.13473582e+00 -8.06805670e-01 -9.35757458e-01 -9.41805169e-02
-6.02400526e-02 -9.39779699e-01 -1.47147655e-01 -2.11163104e-01
-1.26388597e+00 -1.25107014e+00 -1.04664814e+00 -1.05113101e+00
6.36206567e-01 2.84122318e-01 6.30932033e-01 5.36854230e-02
-5.30987084e-01 -1.69258088e-01 -3.56369793e-01 -3.87655467e-01
-4.04148787e-01 -7.28138387e-01 -2.27497175e-01 1.94713354e-01
6.56822324e-01 -4.88430291e-01 -9.03003454e-01 -1.59002393e-02
-7.57685542e-01 -1.51025712e-01 1.16970861e+00 6.49050832e-01
6.08512461e-01 6.17672682e-01 3.25814903e-01 -6.46749556e-01
6.87730849e-01 -1.48413077e-01 -6.55169249e-01 2.37278670e-01
-8.89766991e-01 -2.12179512e-01 3.25537145e-01 -2.99724489e-01
-1.25520658e+00 -2.61818748e-02 -2.53348406e-02 -1.53722018e-01
-4.50692743e-01 5.83357215e-01 2.76720077e-01 -5.03449798e-01
6.44669831e-01 3.41821879e-01 3.05210829e-01 -7.42462814e-01
-1.81906104e-01 1.28860366e+00 6.55298173e-01 2.30087087e-01
3.57588768e-01 4.46640760e-01 1.57840267e-01 -7.92032480e-01
-5.13033926e-01 -6.81572080e-01 -2.92391062e-01 -1.47073627e-01
1.01249909e+00 -8.34177792e-01 -7.53187656e-01 7.98589766e-01
-9.23847854e-01 5.08530140e-01 -4.49895971e-02 1.06059957e+00
-3.37701142e-01 6.47329509e-01 -4.94412273e-01 -9.51177478e-01
-5.64395785e-01 -1.05684829e+00 3.74575645e-01 7.39751816e-01
2.02134117e-01 -8.82498682e-01 1.65550523e-02 5.49572587e-01
4.82157767e-01 5.17050922e-01 9.80730832e-01 -2.93579966e-01
-3.15962583e-01 -3.30660462e-01 -7.58852065e-01 8.23895931e-01
3.82404983e-01 3.50846708e-01 -8.89623344e-01 -1.09253183e-01
2.72002220e-01 9.85736474e-02 1.07359183e+00 7.73047924e-01
1.15238142e+00 -1.66042998e-01 -8.71268436e-02 8.29099894e-01
1.80288458e+00 7.72965431e-01 1.09323514e+00 4.25612837e-01
7.87884444e-02 4.65160221e-01 5.39663255e-01 4.26296324e-01
1.53368443e-01 2.38525793e-01 2.69549400e-01 -4.97075766e-01
-3.88054103e-01 4.68212187e-01 7.12508518e-06 6.78631842e-01
-9.10651445e-01 -2.86794543e-01 -6.66913033e-01 4.80744153e-01
-1.46984994e+00 -1.02285028e+00 -3.34610462e-01 2.21630359e+00
7.70600736e-01 -1.52923077e-01 -2.68201940e-02 4.20824766e-01
8.29885840e-01 -2.13775158e-01 -2.60580420e-01 -5.60725391e-01
-1.50006473e-01 4.27334726e-01 5.98827243e-01 1.06339283e-01
-1.21893919e+00 2.28321413e-03 6.16478109e+00 5.77520072e-01
-1.33801568e+00 -6.55279830e-02 4.53835785e-01 4.16930392e-02
4.10201579e-01 -3.47360939e-01 -4.28155750e-01 6.61837161e-01
7.48487771e-01 -4.89958674e-02 1.26787558e-01 3.20580006e-01
4.72051799e-01 -5.14799416e-01 -4.07976568e-01 1.08444321e+00
1.27525866e-01 -1.15551412e+00 -3.39374132e-02 1.07257746e-01
4.92835343e-01 -1.34603932e-01 2.48077750e-01 -2.65807658e-01
-2.70779014e-01 -8.95692527e-01 -1.76482648e-01 9.34281230e-01
8.25400770e-01 -8.94004047e-01 1.30013931e+00 2.24602725e-02
-6.39189184e-01 -2.11382955e-01 -2.49818966e-01 4.08044569e-02
-5.02906216e-04 6.12156630e-01 -5.50117731e-01 6.12615466e-01
6.00791574e-01 7.12562978e-01 -5.28427064e-01 1.80375361e+00
2.22427964e-01 5.62233567e-01 -1.12354808e-01 2.39791647e-01
1.25930995e-01 -3.30147356e-01 5.99740565e-01 1.00544691e+00
5.58932722e-01 5.02381623e-01 -3.43564361e-01 4.31740940e-01
1.94311947e-01 3.50118697e-01 -4.15407658e-01 -2.79316325e-02
-1.17251426e-02 8.98142993e-01 -3.73735040e-01 -2.03819573e-01
-6.70093358e-01 6.31022394e-01 -3.29351395e-01 4.00659442e-01
-5.57603300e-01 -8.74157786e-01 4.20325577e-01 -1.51676521e-01
1.22038126e-01 1.74145043e-01 -3.02300990e-01 -8.79126787e-01
-6.70218915e-02 -6.87941074e-01 4.35890794e-01 -8.03153694e-01
-9.74082410e-01 4.99682695e-01 -3.17554116e-01 -1.55108893e+00
2.43591130e-01 -3.81662220e-01 -5.76484442e-01 1.06806898e+00
-1.91239297e+00 -7.37475514e-01 -5.26389062e-01 4.62503433e-01
3.09287816e-01 -5.03979445e-01 7.15398788e-01 5.90586185e-01
-7.05231488e-01 4.06251013e-01 3.66957933e-01 8.14969763e-02
7.51901746e-01 -9.87169743e-01 -2.87627101e-01 9.21537578e-01
-5.75980961e-01 3.28607351e-01 5.53853869e-01 -5.00098109e-01
-7.18020320e-01 -1.11896586e+00 9.23994064e-01 4.19350594e-01
5.75876608e-03 8.22700977e-01 -8.49192679e-01 2.51026779e-01
2.33210072e-01 -6.18555769e-02 7.79011965e-01 -2.95870423e-01
1.39002070e-01 -1.51833639e-01 -1.65020621e+00 3.88126016e-01
4.80256528e-01 -2.49425873e-01 -6.85291529e-01 3.24194819e-01
2.50032812e-01 -3.86049837e-01 -1.11929142e+00 6.93749726e-01
7.30193198e-01 -1.30285847e+00 7.65324235e-01 -4.16275769e-01
6.01584971e-01 -4.25972313e-01 -9.25829932e-02 -9.44754958e-01
-3.27342451e-01 -3.59665036e-01 -1.46502182e-01 9.38648820e-01
2.23779716e-02 -9.52033520e-01 3.05159718e-01 1.20780706e-01
1.45185873e-01 -9.46363151e-01 -5.55724740e-01 -4.37764287e-01
-4.47426885e-01 2.40361124e-01 2.38942623e-01 7.31996775e-01
-4.18644845e-01 -1.69437394e-01 -2.17762589e-01 3.72891366e-01
6.79062426e-01 -2.59873457e-02 1.97218657e-01 -1.38255811e+00
4.41338606e-02 -2.29137748e-01 -9.65769470e-01 -3.58976960e-01
-5.89401066e-01 -4.27759498e-01 -3.10059905e-01 -1.79273808e+00
3.63050938e-01 -2.87725061e-01 -5.02088606e-01 2.67318934e-01
-1.83690175e-01 4.35202569e-01 -5.93802296e-02 2.87992775e-01
6.38294443e-02 1.47920743e-01 1.44942021e+00 -1.22250561e-02
-4.16142166e-01 2.82998502e-01 -8.39986682e-01 7.35542715e-01
9.77993429e-01 -1.84138581e-01 -3.03359479e-01 -2.56225497e-01
-2.78221697e-01 7.88386464e-02 4.26314920e-01 -1.20756221e+00
1.35354429e-01 -8.40415433e-02 3.57406557e-01 -5.51541090e-01
1.35992005e-01 -7.61990726e-01 -3.51322778e-02 5.25100470e-01
-1.34378970e-01 -3.79592925e-01 6.95080450e-03 4.14126813e-01
-4.99874800e-01 -2.45709643e-01 1.20134854e+00 2.18768809e-02
-7.62560964e-01 1.38624936e-01 -2.97811568e-01 -3.08794379e-01
1.09143984e+00 -6.70382380e-01 -5.45746505e-01 3.13736983e-02
-6.71721399e-01 -1.00442454e-01 3.26790571e-01 -7.99066275e-02
8.41417491e-01 -8.52324486e-01 -9.11866248e-01 2.93314606e-01
1.40563786e-01 -1.36662364e-01 4.16900069e-01 1.34630239e+00
-8.80562842e-01 3.18427116e-01 -4.76837963e-01 -3.92627805e-01
-1.92015469e+00 2.45077729e-01 5.50180316e-01 1.21305317e-01
-8.57612789e-01 6.04030073e-01 -1.74944326e-01 1.79348037e-01
3.11066121e-01 -4.57668751e-01 -8.49169552e-01 4.85878326e-02
7.86242187e-01 6.69547975e-01 -2.25476082e-03 -5.12626231e-01
7.71021321e-02 9.52093482e-01 -2.32854113e-01 4.03687537e-01
1.28884780e+00 -3.08377504e-01 -3.82244349e-01 -1.17908977e-01
1.00147974e+00 -1.84887908e-02 -8.37875128e-01 -2.05641657e-01
-3.83026898e-01 -6.67653024e-01 6.58281505e-01 -1.15901172e+00
-1.18832624e+00 7.04794765e-01 1.55442047e+00 5.97128049e-02
1.90493608e+00 -5.26902795e-01 9.56723332e-01 2.15031847e-01
-2.23135278e-01 -8.97948861e-01 -5.54453790e-01 9.53785628e-02
4.88149643e-01 -1.35017741e+00 1.50424335e-02 -4.60497141e-01
-6.24025166e-01 1.41073954e+00 2.47462004e-01 1.20731346e-01
5.69651961e-01 -5.45347780e-02 4.24535841e-01 -1.08332701e-01
-4.61044967e-01 -4.36859727e-01 3.31370682e-01 7.26894975e-01
3.31712931e-01 -1.31568119e-01 -7.90319264e-01 6.82090148e-02
2.68508732e-01 8.16264093e-01 7.79435754e-01 1.00368547e+00
-7.81377614e-01 -7.19414473e-01 -3.79918367e-01 7.55812407e-01
-7.17087746e-01 1.11273136e-02 -4.65837568e-02 6.71462357e-01
4.33368653e-01 9.17672455e-01 -6.15406819e-02 -2.68005371e-01
1.58137232e-01 -8.49074274e-02 4.30604219e-01 -3.61868292e-02
-2.37845272e-01 3.38666365e-02 1.39058635e-01 -2.02414379e-01
-1.04371333e+00 -3.81539613e-01 -1.07010806e+00 -2.44837701e-01
-3.73699099e-01 -9.35768883e-04 7.11804926e-01 7.91262567e-01
2.87500441e-01 6.00644827e-01 7.40109324e-01 -2.01691210e-01
-4.46103483e-01 -9.64869857e-01 -7.98236370e-01 3.16113941e-02
7.70959616e-01 -4.18374717e-01 -3.46135736e-01 1.87663421e-01] | [15.817502975463867, -3.946977138519287] |
46a68eb8-5e02-4c5d-864a-74b9bea77ac7 | end-to-end-multi-modal-multi-task-vehicle | 1801.06734 | null | http://arxiv.org/abs/1801.06734v2 | http://arxiv.org/pdf/1801.06734v2.pdf | End-to-end Multi-Modal Multi-Task Vehicle Control for Self-Driving Cars with Visual Perception | Convolutional Neural Networks (CNN) have been successfully applied to
autonomous driving tasks, many in an end-to-end manner. Previous end-to-end
steering control methods take an image or an image sequence as the input and
directly predict the steering angle with CNN. Although single task learning on
steering angles has reported good performances, the steering angle alone is not
sufficient for vehicle control. In this work, we propose a multi-task learning
framework to predict the steering angle and speed control simultaneously in an
end-to-end manner. Since it is nontrivial to predict accurate speed values with
only visual inputs, we first propose a network to predict discrete speed
commands and steering angles with image sequences. Moreover, we propose a
multi-modal multi-task network to predict speed values and steering angles by
taking previous feedback speeds and visual recordings as inputs. Experiments
are conducted on the public Udacity dataset and a newly collected SAIC dataset.
Results show that the proposed model predicts steering angles and speed values
accurately. Furthermore, we improve the failure data synthesis methods to solve
the problem of error accumulation in real road tests. | ['Jiebo Luo', 'Jerry Yu', 'Yixuan Zhang', 'Zhengyuan Yang', 'Junjie Cai'] | 2018-01-20 | null | null | null | null | ['steering-control'] | ['computer-vision'] | [ 1.02230839e-01 -2.42425531e-01 -3.06652337e-01 -1.02178335e+00
-5.23017287e-01 -3.22287261e-01 5.40733039e-01 -6.18747771e-01
-5.14591038e-01 4.58313942e-01 -2.94454664e-01 -5.37392437e-01
1.13719650e-01 -7.92180181e-01 -9.70608234e-01 -5.78669965e-01
3.08115989e-01 2.07963154e-01 3.40184957e-01 -6.75609171e-01
3.82277757e-01 5.18302441e-01 -1.87726736e+00 1.76337540e-01
8.30220520e-01 1.19251919e+00 3.36665511e-01 1.08924139e+00
2.20540509e-01 6.56240821e-01 -1.57456589e-03 2.41782852e-02
4.98326927e-01 3.08011752e-02 -4.85304475e-01 -6.29762709e-02
7.68591046e-01 -6.78191841e-01 -5.80582976e-01 5.53634524e-01
5.92463732e-01 -2.85299812e-02 5.56665421e-01 -1.73938847e+00
-2.20542535e-01 -1.20658763e-02 -1.64360151e-01 -1.32127345e-01
-1.90480992e-01 7.01463163e-01 6.71747744e-01 -7.73778021e-01
4.76475924e-01 9.58173633e-01 6.35321736e-01 5.27180552e-01
-9.48298275e-01 -7.15236485e-01 1.24519281e-01 7.50802994e-01
-1.12819850e+00 -6.59949958e-01 7.58914351e-01 -4.99329478e-01
8.24063241e-01 -1.09382197e-01 4.16130990e-01 1.13367569e+00
3.76088619e-01 8.87431920e-01 8.05015862e-01 8.41197744e-02
-2.70428751e-02 6.97314814e-02 -3.13121200e-01 6.71737194e-01
-2.70649999e-01 4.15502459e-01 -4.15715963e-01 8.03350508e-01
4.88431752e-01 -1.49196208e-01 -7.60647878e-02 -6.06637120e-01
-1.46101654e+00 7.81364620e-01 6.70233309e-01 -4.31447983e-01
-2.98844516e-01 5.97054183e-01 6.33687735e-01 5.03766716e-01
1.39543176e-01 1.22338742e-01 -7.62006640e-01 -4.55674261e-01
-6.98105335e-01 5.80336630e-01 4.07688648e-01 1.17692077e+00
9.07241106e-01 3.64101142e-01 -8.66757035e-02 7.87472546e-01
2.48955190e-01 1.12007785e+00 2.01892808e-01 -1.08994305e+00
9.10134435e-01 1.04582943e-01 4.61977422e-01 -1.26327109e+00
-8.72277558e-01 5.63545804e-03 -7.51576006e-01 5.05383372e-01
5.82148492e-01 -3.08083534e-01 -9.22915518e-01 1.68918741e+00
6.48573115e-02 -6.09369911e-02 7.53139406e-02 1.29478467e+00
5.10425270e-01 7.35636830e-01 -1.85659945e-01 1.29872531e-01
9.53676999e-01 -1.24918473e+00 -6.80560350e-01 -6.58040226e-01
7.55243480e-01 -5.01507699e-01 1.09505999e+00 6.54240191e-01
-7.43274927e-01 -1.01054871e+00 -1.18107557e+00 -7.10986927e-02
-6.08888805e-01 8.23290646e-01 3.45195740e-01 5.08360326e-01
-9.64497030e-01 5.23765683e-01 -8.12993526e-01 -1.45031111e-02
2.90018499e-01 3.63208532e-01 -2.03075692e-01 8.49210192e-03
-1.34294140e+00 1.16579962e+00 -1.88814811e-02 4.02282506e-01
-1.03275084e+00 -7.87579596e-01 -9.17848170e-01 -3.14137310e-01
5.64977109e-01 -4.17941540e-01 1.39318490e+00 -8.40049207e-01
-1.91958439e+00 1.36957020e-01 -2.61484832e-01 -4.44536090e-01
8.32581699e-01 -4.40993875e-01 -4.58302051e-01 -2.26399079e-01
3.19669098e-02 1.06991589e+00 9.05772328e-01 -1.28318238e+00
-1.00367904e+00 -1.31841779e-01 -8.41650963e-02 4.34785746e-02
-1.85914844e-01 -7.69662857e-01 -4.91448998e-01 -3.39192338e-02
-1.23167790e-01 -1.22826433e+00 -3.57427895e-01 2.86127716e-01
-3.98729146e-01 -1.58542812e-01 1.21474588e+00 -4.03014749e-01
8.67303073e-01 -1.95366490e+00 3.55153275e-03 -1.48940757e-02
1.17922463e-01 2.49596745e-01 -4.48670477e-01 3.77090834e-02
1.81299731e-01 -2.80522287e-01 1.81829736e-01 -3.03448260e-01
-4.40247841e-02 3.06356639e-01 -3.41423422e-01 4.68782842e-01
4.82196629e-01 9.54858303e-01 -7.82418370e-01 -1.66805938e-01
6.45788252e-01 3.95043105e-01 -5.22926152e-01 2.39256114e-01
-1.74010053e-01 3.47234190e-01 -3.97842795e-01 2.95377016e-01
6.88432097e-01 7.57198706e-02 -1.95121616e-01 -4.72881794e-01
-5.06556809e-01 9.50918123e-02 -1.03537917e+00 1.49509478e+00
-1.03598022e+00 1.23164773e+00 -1.38102517e-01 -1.11089349e+00
1.25972176e+00 6.41748682e-02 4.47027415e-01 -1.25006771e+00
1.77801579e-01 2.66911566e-01 2.50755288e-02 -7.14498520e-01
8.05065453e-01 2.21510738e-01 -3.95765215e-01 -1.34023622e-01
-2.08177045e-01 -4.93653238e-01 1.81528285e-01 -2.29411662e-01
7.24288881e-01 4.27078247e-01 -2.12115243e-01 1.52222335e-01
6.11458063e-01 2.62568295e-01 5.43370247e-01 4.53618258e-01
-3.34397882e-01 7.11186826e-01 4.59013969e-01 -8.22927058e-01
-1.66012669e+00 -9.11877990e-01 7.13364109e-02 1.05042434e+00
2.87320942e-01 -2.46518385e-02 -3.39736104e-01 -4.54726279e-01
1.48463979e-01 7.48140931e-01 -6.48206055e-01 -2.15334505e-01
-7.26985276e-01 -2.28304535e-01 4.20364738e-01 8.87571216e-01
5.81589878e-01 -7.16078699e-01 -9.18295324e-01 3.59927833e-01
-2.58753657e-01 -1.44845581e+00 -6.02364421e-01 1.32995680e-01
-2.98271984e-01 -9.07944083e-01 -4.73035038e-01 -5.21829844e-01
2.41024658e-01 2.73601115e-01 5.83734155e-01 -3.69656116e-01
-1.01639785e-01 -1.33749679e-01 -1.18035786e-01 -5.56566417e-01
-2.73090482e-01 2.21247330e-01 2.69409537e-01 1.12750515e-01
8.22511315e-02 -2.63936013e-01 -7.98182845e-01 8.29616487e-01
-2.05925450e-01 6.13179743e-01 6.86937928e-01 7.42296755e-01
3.47944230e-01 -6.54167950e-01 1.04243457e+00 -6.25101507e-01
2.32945502e-01 -3.34304780e-01 -1.00354075e+00 -1.55099347e-01
-9.97125745e-01 1.64527044e-01 1.21296024e+00 -2.26711497e-01
-8.68152678e-01 5.33483922e-01 -2.08252594e-01 -6.38660550e-01
-2.27271304e-01 2.25195065e-01 -1.35029927e-02 5.55416942e-02
6.04741991e-01 2.95293331e-01 4.32437152e-01 1.51936278e-01
5.73980987e-01 7.78178453e-01 8.04870069e-01 2.01203581e-02
6.49241149e-01 3.88733625e-01 2.74603039e-01 -7.28619754e-01
-6.01643026e-01 -4.53227609e-01 -8.29225063e-01 -7.38433242e-01
9.11711514e-01 -1.13742352e+00 -1.12088561e+00 5.80252230e-01
-1.13820720e+00 -6.83910549e-01 3.23821455e-01 3.82831931e-01
-1.13471043e+00 1.16602937e-02 3.69942822e-02 -5.51444292e-01
-2.02853203e-01 -1.46821749e+00 1.04969549e+00 2.03906015e-01
1.24615662e-01 -7.82992423e-01 -1.09713905e-01 4.04815108e-01
8.14456999e-01 2.46435776e-01 3.59065622e-01 -6.84197918e-02
-5.74903667e-01 -3.13871026e-01 -4.70545769e-01 2.69535720e-01
-3.08779199e-02 1.09328374e-01 -8.22749376e-01 -1.43305212e-01
-8.14001858e-01 -6.61006391e-01 8.10516059e-01 5.48695087e-01
1.38296330e+00 -6.91598281e-02 -3.18405777e-01 7.52821088e-01
1.50484288e+00 3.02439600e-01 6.29982948e-01 5.73280513e-01
9.74225879e-01 4.75708485e-01 9.29255426e-01 4.67217505e-01
8.59275103e-01 9.69459593e-01 6.36936843e-01 -1.46896839e-02
-1.45091683e-01 -7.41070360e-02 4.87064183e-01 6.34422064e-01
5.66024054e-03 -3.60018224e-01 -1.04906869e+00 7.39122510e-01
-1.94154513e+00 -8.73931289e-01 -6.29549325e-01 1.91223514e+00
5.02110004e-01 3.68339628e-01 5.95527589e-02 3.39900516e-02
2.99537152e-01 1.73585579e-01 -8.89830112e-01 -7.78643012e-01
1.43546209e-01 -2.92896539e-01 1.05391204e+00 4.62059051e-01
-1.08923852e+00 8.64656448e-01 5.86832428e+00 6.14927530e-01
-1.70250940e+00 -3.32258403e-01 5.35079122e-01 -2.93531120e-01
-1.70835778e-01 -3.53795111e-01 -1.01075971e+00 3.67590070e-01
1.31243563e+00 7.52857924e-02 4.34305429e-01 1.01011539e+00
9.36025918e-01 -3.53644826e-02 -9.82455373e-01 9.08500969e-01
-1.16222985e-01 -1.33018887e+00 -2.60687143e-01 -2.28325799e-01
6.87180161e-01 3.30988646e-01 1.40197635e-01 4.58288193e-01
8.92699435e-02 -1.08264577e+00 9.12488997e-01 7.58707225e-01
1.22577405e+00 -9.15599465e-01 7.59782612e-01 4.62336600e-01
-1.32755852e+00 -2.51691937e-01 -1.43818319e-01 -1.21272177e-01
4.53632772e-01 2.16183886e-01 -1.16908669e+00 4.95607048e-01
4.87430274e-01 1.04397082e+00 -5.51863194e-01 8.86501193e-01
-1.12516582e-01 2.74077892e-01 -2.41255373e-01 -5.40006042e-01
6.09932065e-01 -2.08605044e-02 1.39323860e-01 1.14312506e+00
4.18440163e-01 -1.88520014e-01 8.84665772e-02 6.98977530e-01
2.81320691e-01 -2.26702586e-01 -7.28950441e-01 3.05759549e-01
2.69489974e-01 1.32530189e+00 -1.75056115e-01 -1.50830850e-01
-6.30725384e-01 6.68862224e-01 6.32834211e-02 3.64783317e-01
-1.12034857e+00 -8.10368717e-01 9.71716404e-01 8.48096982e-02
5.27271271e-01 -5.08551598e-01 -6.00387037e-01 -5.48758268e-01
-3.39247659e-02 -3.32443863e-01 -4.67281073e-01 -9.78961110e-01
-5.96028030e-01 4.04895484e-01 -1.52757376e-01 -1.57617712e+00
-5.28502464e-01 -9.14489448e-01 -5.12615979e-01 8.38802814e-01
-1.89048135e+00 -9.69966471e-01 -7.45624065e-01 4.46184874e-01
9.98943329e-01 -3.78956556e-01 2.62441993e-01 4.50577587e-01
-4.65712398e-01 5.21821260e-01 1.29001856e-01 -5.13749458e-02
9.49037075e-01 -1.14130330e+00 5.12636960e-01 5.74404776e-01
-6.35670066e-01 -2.71912962e-01 9.90025878e-01 -1.56521812e-01
-1.75940228e+00 -1.45032966e+00 6.40208662e-01 -3.73610437e-01
3.93545717e-01 -2.77571499e-01 -5.47698319e-01 3.63619477e-01
1.68895740e-02 3.59592527e-01 -1.06199190e-01 -2.25695491e-01
6.40437305e-02 -5.01547337e-01 -6.07589006e-01 5.79790115e-01
8.85246992e-01 -1.71874896e-01 1.67340606e-01 7.05481321e-02
4.21884507e-01 -6.59180939e-01 -5.40592790e-01 5.38595557e-01
8.28284919e-01 -7.03992009e-01 9.56554413e-01 -4.02863920e-01
9.54235792e-01 -4.41739947e-01 -5.70429228e-02 -1.66464913e+00
-1.30254835e-01 -2.64294267e-01 1.50213704e-01 5.97320557e-01
7.73428261e-01 -2.71922797e-01 9.17127073e-01 3.26753974e-01
-5.46281397e-01 -8.77733350e-01 -9.41684365e-01 -6.17154002e-01
4.89109010e-02 -1.04310572e+00 3.63346428e-01 3.91319364e-01
-2.70136982e-01 4.26789373e-01 -9.15770531e-01 2.60748327e-01
3.63102704e-01 7.40439370e-02 1.36766613e+00 -8.53603959e-01
3.29311788e-01 -4.33844805e-01 -3.55251849e-01 -1.42700887e+00
1.15282282e-01 -5.57272732e-01 7.69652426e-01 -1.42069423e+00
-4.50800508e-01 -4.63933319e-01 2.70638078e-01 4.44555521e-01
8.39140117e-02 2.09078081e-02 8.00809935e-02 -2.64150947e-01
-5.59345901e-01 7.36035645e-01 1.54432118e+00 -3.04751188e-01
-2.55208492e-01 2.25309759e-01 -1.44407645e-01 2.79559374e-01
9.69058394e-01 5.95038831e-02 -6.15840971e-01 -6.29769444e-01
3.08814079e-01 4.43602145e-01 4.53086495e-01 -1.08009100e+00
4.60085630e-01 -4.40917730e-01 3.70076448e-01 -1.06814325e+00
3.90182942e-01 -7.02434123e-01 -3.28334540e-01 4.64975387e-01
-2.67670900e-01 1.18803203e-01 3.19688708e-01 5.55918217e-01
-7.54542276e-02 3.80931884e-01 8.24938655e-01 5.41898608e-01
-1.36785853e+00 5.82486808e-01 -5.44519663e-01 -3.39407593e-01
1.07736850e+00 -2.41722763e-01 -2.22662121e-01 -5.29361129e-01
-5.44943750e-01 8.47298741e-01 -3.54698524e-02 1.03173053e+00
8.92500699e-01 -1.55407727e+00 -8.19694161e-01 3.72789949e-01
5.02493322e-01 -1.85147133e-02 4.36819732e-01 7.80570328e-01
-5.90614438e-01 8.11140180e-01 -5.30003667e-01 -9.94111836e-01
-8.72624397e-01 4.17399406e-01 4.32660967e-01 3.13232809e-01
-4.19476509e-01 4.39575672e-01 -1.50603071e-01 -7.01670110e-01
-7.34775960e-02 -5.42293787e-01 -2.99018860e-01 -3.83326598e-02
3.16975832e-01 2.90977329e-01 1.58370987e-01 -5.41878879e-01
-3.19549918e-01 7.42898643e-01 9.97264236e-02 2.57250071e-02
1.30384779e+00 -5.29752553e-01 9.02471364e-01 5.49469948e-01
1.50655830e+00 -6.19898915e-01 -1.98880684e+00 5.96403005e-03
-3.79282683e-01 -4.29086387e-01 3.83524597e-01 -6.82123780e-01
-1.24837470e+00 1.01265883e+00 7.59199619e-01 -1.30379051e-01
8.04427266e-01 -5.32755136e-01 9.47344601e-01 6.75748408e-01
1.80424169e-01 -1.58946347e+00 1.74248397e-01 7.83292472e-01
9.31572199e-01 -1.87068641e+00 -3.51030380e-01 -1.12967528e-01
-9.58410740e-01 1.46728742e+00 1.10987473e+00 2.55198032e-02
5.50050080e-01 2.93965191e-01 2.59196520e-01 1.62775069e-01
-1.12090826e+00 -9.54229236e-02 3.12396646e-01 6.41930878e-01
1.55750081e-01 1.27713338e-01 3.60290706e-02 2.59029061e-01
-3.27849090e-01 3.11486423e-01 7.38025248e-01 5.04064977e-01
-5.42590499e-01 -5.66588938e-01 -4.12819386e-02 5.13506353e-01
3.52797985e-01 3.76441479e-01 2.87322760e-01 7.89304733e-01
6.12622462e-02 1.00659847e+00 3.03753406e-01 -9.07525361e-01
8.59209061e-01 -1.08427040e-01 5.24995439e-02 -1.36051290e-02
7.48668760e-02 -5.63548028e-01 2.97903001e-01 -8.99265468e-01
-7.38562942e-02 -7.06995606e-01 -1.26201725e+00 -4.99870270e-01
-3.50347571e-02 -5.96084177e-01 1.01982582e+00 9.79818106e-01
2.92218029e-01 7.85367727e-01 1.22868240e+00 -1.21597350e+00
-5.99565864e-01 -8.44851613e-01 -1.03300020e-01 2.30975166e-01
8.53982866e-01 -7.59602964e-01 -1.44266292e-01 1.59116432e-01] | [8.04005241394043, -1.39430832862854] |
a7eb7330-55ef-47d0-bcc9-fa2d6e3b4d9e | m-2-3dlanenet-multi-modal-3d-lane-detection | 2209.05996 | null | https://arxiv.org/abs/2209.05996v2 | https://arxiv.org/pdf/2209.05996v2.pdf | M^2-3DLaneNet: Multi-Modal 3D Lane Detection | Estimating accurate lane lines in 3D space remains challenging due to their sparse and slim nature. In this work, we propose the M^2-3DLaneNet, a Multi-Modal framework for effective 3D lane detection. Aiming at integrating complementary information from multi-sensors, M^2-3DLaneNet first extracts multi-modal features with modal-specific backbones, then fuses them in a unified Bird's-Eye View (BEV) space. Specifically, our method consists of two core components. 1) To achieve accurate 2D-3D mapping, we propose the top-down BEV generation. Within it, a Line-Restricted Deform-Attention (LRDA) module is utilized to effectively enhance image features in a top-down manner, fully capturing the slenderness features of lanes. After that, it casts the 2D pyramidal features into 3D space using depth-aware lifting and generates BEV features through pillarization. 2) We further propose the bottom-up BEV fusion, which aggregates multi-modal features through multi-scale cascaded attention, integrating complementary information from camera and LiDAR sensors. Sufficient experiments demonstrate the effectiveness of M^2-3DLaneNet, which outperforms previous state-of-the-art methods by a large margin, i.e., 12.1% F1-score improvement on OpenLane dataset. | ['Zhen Li', 'Shuguang Cui', 'Tang Kun', 'Shuqi Mei', 'Chao Zheng', 'Chaoda Zheng', 'Xu Yan', 'Yueru Luo'] | 2022-09-13 | null | null | null | null | ['3d-lane-detection', 'lane-detection'] | ['computer-vision', 'computer-vision'] | [-3.58285271e-02 -3.33598495e-01 -1.12365082e-01 -4.65718538e-01
-1.04282117e+00 -5.00871778e-01 4.29784626e-01 -2.75559962e-01
-2.31284276e-01 4.34498042e-01 3.33367229e-01 -1.34013623e-01
-1.07294299e-01 -9.32020366e-01 -9.12159145e-01 -4.61725652e-01
2.34603509e-01 1.48426205e-01 5.18111706e-01 -5.88683248e-01
2.38849282e-01 9.68582094e-01 -1.85672522e+00 2.69611347e-02
1.14231670e+00 1.22664416e+00 2.01553896e-01 4.52336341e-01
-1.63398713e-01 4.21440005e-01 1.10940039e-01 -3.25922549e-01
2.63967961e-01 2.80312121e-01 -1.88433662e-01 4.32429284e-01
1.04840195e+00 -6.95287228e-01 -5.58926225e-01 8.69231999e-01
3.59560072e-01 1.18524998e-01 5.77849925e-01 -1.24939728e+00
-5.37503541e-01 -1.49979174e-01 -1.07158351e+00 -4.73421961e-02
2.19547555e-01 3.53299916e-01 9.25276399e-01 -1.33541322e+00
4.35542822e-01 1.49414277e+00 7.01418877e-01 5.14517613e-02
-1.19193029e+00 -7.12702334e-01 3.87227029e-01 1.24623425e-01
-1.53502655e+00 -3.29963535e-01 1.09405029e+00 -4.71660346e-01
6.88012362e-01 6.73475266e-02 4.53335106e-01 6.37765169e-01
2.30331719e-01 8.73224258e-01 1.04531276e+00 -1.78459045e-02
-4.13015068e-01 -8.68367404e-02 1.87476024e-01 8.94168019e-01
1.49211228e-01 1.57632992e-01 -5.45984805e-01 3.02895695e-01
7.87756920e-01 2.08141014e-01 -1.61028773e-01 -5.62648356e-01
-1.07496619e+00 8.54104340e-01 8.44014227e-01 -2.44807065e-01
-3.24507028e-01 -9.89219993e-02 1.45430401e-01 -2.48783469e-01
4.68690127e-01 -8.65048096e-02 -2.69191504e-01 2.81244963e-01
-6.71626985e-01 4.46119457e-01 -1.02673173e-01 1.10362542e+00
1.29303932e+00 -1.39898404e-01 -1.85883567e-01 8.06756556e-01
5.24708271e-01 1.00975323e+00 -7.33001456e-02 -1.05496144e+00
6.88445628e-01 7.76482940e-01 1.29248664e-01 -1.05056369e+00
-5.90246141e-01 -3.65386844e-01 -6.36631906e-01 3.47272635e-01
3.20193991e-02 1.68386716e-02 -8.67980719e-01 1.60104084e+00
4.79378879e-01 1.72974721e-01 -6.43531606e-02 1.11279738e+00
7.73937047e-01 6.90258503e-01 -1.67374134e-01 1.91933021e-01
1.29260683e+00 -1.11945796e+00 -3.20644259e-01 -4.87650275e-01
3.58237505e-01 -7.35978663e-01 1.01694441e+00 7.06574395e-02
-8.11517775e-01 -8.45540643e-01 -1.01141274e+00 -5.09482861e-01
-4.55275267e-01 4.31996524e-01 4.98527706e-01 2.13021711e-01
-7.63644874e-01 -1.83172673e-01 -3.45544368e-01 -1.24717288e-01
5.75415254e-01 8.37722123e-02 -6.21968865e-01 -5.03792167e-01
-1.00979173e+00 6.59731269e-01 2.31111273e-01 3.99263054e-01
-7.22677588e-01 -8.01238656e-01 -1.41838396e+00 -1.31323501e-01
6.84477150e-01 -7.27152944e-01 6.14663661e-01 4.55554388e-02
-1.12624586e+00 8.67518008e-01 -5.89921594e-01 -2.90700227e-01
3.76886606e-01 -5.26162386e-01 -3.85016948e-01 4.65405686e-03
3.05658728e-01 8.42003822e-01 7.83182442e-01 -1.56550598e+00
-1.04814947e+00 -6.31060958e-01 6.60583600e-02 3.82026494e-01
4.91454378e-02 -5.27231276e-01 -7.39557445e-01 -4.05838907e-01
3.99204791e-01 -8.10050905e-01 -5.50724715e-02 1.05244689e-01
-5.75307369e-01 -1.83823064e-01 1.01721323e+00 -5.98306298e-01
1.10393393e+00 -2.21944213e+00 1.02100137e-03 1.53947413e-01
4.99907911e-01 2.09032550e-01 -1.86688885e-01 2.03474581e-01
3.47590625e-01 -1.71425551e-01 -2.97631562e-01 -4.94257152e-01
8.60825479e-02 -9.20310710e-03 -4.57300156e-01 4.25058544e-01
5.12102902e-01 1.10478866e+00 -6.60441995e-01 -4.45439398e-01
8.10271084e-01 5.44585526e-01 -5.27277231e-01 5.36781214e-02
8.64813775e-02 2.97952712e-01 -5.07879078e-01 8.59239101e-01
1.20650864e+00 1.63766012e-01 -5.68675995e-01 -6.97408795e-01
-6.53803587e-01 -1.79900616e-01 -1.01533258e+00 1.73489702e+00
-7.99476445e-01 6.11082494e-01 7.53118992e-02 -5.95047355e-01
1.24262416e+00 -6.35746837e-01 3.32470447e-01 -9.67684746e-01
-1.16164260e-01 2.91658103e-01 -5.56521118e-01 -2.88911223e-01
6.16338968e-01 1.97253868e-01 -5.29135823e-01 -1.70430675e-01
-1.36016577e-01 -3.03377330e-01 -2.67086104e-02 1.44732669e-01
6.95045948e-01 2.46197850e-01 -4.21327204e-02 6.28811494e-02
1.04888356e+00 -1.02454960e-01 6.04758978e-01 4.93927419e-01
-2.58726060e-01 5.14847159e-01 2.16558903e-01 -3.27055335e-01
-9.86262500e-01 -1.28413200e+00 -1.57939926e-01 7.03484237e-01
6.59041643e-01 -1.31662562e-01 -3.93638164e-01 -7.29241967e-01
5.88995218e-01 6.23985350e-01 -5.60860693e-01 -1.62027076e-01
-5.30440569e-01 -2.38637671e-01 2.86650479e-01 6.25157416e-01
8.55342984e-01 -3.91509235e-01 -6.44815922e-01 1.43541023e-01
-3.02252527e-02 -1.51444197e+00 -6.98603153e-01 -1.56018391e-01
-6.65713191e-01 -9.10832763e-01 -5.03762841e-01 -4.24642235e-01
4.20920461e-01 8.73551667e-01 8.01995397e-01 -2.60588735e-01
-2.66024292e-01 2.39133298e-01 -1.52686805e-01 -3.27193856e-01
3.91231775e-01 -6.78008348e-02 -1.88633669e-02 3.95336092e-01
4.01857257e-01 -3.99102241e-01 -6.43210292e-01 3.76218736e-01
-3.90543729e-01 2.55431443e-01 1.09823990e+00 8.59955847e-01
9.30047631e-01 -1.03941746e-01 1.79570034e-01 -5.59882820e-01
1.11830279e-01 -2.65706986e-01 -9.51290727e-01 -6.72672018e-02
-2.85870433e-01 -5.06111793e-02 3.74617636e-01 1.37999505e-01
-1.07942760e+00 2.45866731e-01 -3.17480773e-01 -8.51475120e-01
-3.76752734e-01 2.07043633e-01 -5.66867828e-01 -2.20301792e-01
1.96685612e-01 4.27509755e-01 -5.91785945e-02 -4.69018072e-01
5.99337041e-01 7.97946393e-01 7.75752544e-01 -2.67200053e-01
1.19592667e+00 8.32119107e-01 1.95920616e-01 -9.55701888e-01
-1.13046873e+00 -4.69985634e-01 -8.24100375e-01 -3.93981457e-01
9.89950061e-01 -1.19179773e+00 -7.37562597e-01 6.33281410e-01
-9.57614720e-01 -1.33113489e-01 1.39358069e-03 4.55632925e-01
-5.02929449e-01 2.83434331e-01 -2.36963153e-01 -8.14084768e-01
-1.98951647e-01 -1.28636682e+00 1.72640252e+00 2.94991732e-01
5.07154822e-01 -5.38839698e-01 -1.24693997e-01 8.52304399e-01
1.62826672e-01 3.62595469e-01 6.94040954e-01 -1.23894721e-01
-9.77457404e-01 -2.31170282e-01 -8.42554510e-01 3.63291830e-01
-1.98672980e-01 -2.30823264e-01 -9.89121377e-01 2.81023793e-02
-7.14630127e-01 -2.17143953e-01 1.16135442e+00 3.96202207e-01
1.07756591e+00 2.05637455e-01 -3.76537830e-01 8.26261699e-01
1.46378839e+00 -8.82143900e-02 4.15436447e-01 1.99665412e-01
1.08730960e+00 5.29291093e-01 1.06955707e+00 3.04571390e-01
9.99095976e-01 7.73988307e-01 7.97859669e-01 -2.39976332e-01
-3.28091085e-01 -6.00624442e-01 2.80354291e-01 4.75921839e-01
2.26119831e-01 -8.74104723e-03 -7.36078739e-01 5.92213273e-01
-1.78949523e+00 -7.47702420e-01 -3.72113258e-01 2.04336691e+00
9.74897593e-02 4.62474287e-01 1.76084891e-01 -1.30877972e-01
5.43122292e-01 4.82792974e-01 -7.28737891e-01 -7.77140334e-02
-2.75848269e-01 -4.31899756e-01 8.06380868e-01 6.37692750e-01
-1.29752254e+00 1.08866405e+00 4.35067034e+00 9.69483376e-01
-1.10308516e+00 -5.12218624e-02 3.07341576e-01 1.49352968e-01
-5.88413417e-01 -4.80820611e-02 -1.29960978e+00 5.57115912e-01
4.20300782e-01 9.24999490e-02 2.09773958e-01 6.53801978e-01
2.68224955e-01 -1.94175139e-01 -6.03349924e-01 1.02631128e+00
2.55584508e-01 -1.37668777e+00 -2.02655364e-02 3.14055115e-01
7.33541846e-01 2.83419073e-01 4.55590636e-02 4.50182170e-01
-2.06578374e-02 -6.63640559e-01 9.48755801e-01 8.23155463e-01
9.33315158e-01 -1.06671095e+00 7.10129976e-01 5.55279315e-01
-1.67667007e+00 -3.75554115e-01 -2.91345209e-01 3.00892115e-01
4.62922782e-01 9.36822474e-01 -3.70729178e-01 9.02994394e-01
6.87152445e-01 9.57183719e-01 -5.57340086e-01 9.56224620e-01
-1.01382494e-01 1.34003595e-01 -5.07672369e-01 3.24181825e-01
5.37683308e-01 -3.06632847e-01 6.34308755e-01 1.05340230e+00
4.90335077e-01 -2.02321168e-03 3.79078925e-01 9.20699596e-01
-4.88616042e-02 -1.04973003e-01 -8.07673454e-01 5.28016806e-01
7.06032038e-01 1.37586367e+00 -1.84480935e-01 -8.43205675e-02
-7.12025046e-01 8.05454135e-01 2.76910007e-01 2.26638079e-01
-1.02004421e+00 -5.57301462e-01 8.08931828e-01 3.09896529e-01
6.56898975e-01 -4.35291559e-01 -5.43398112e-02 -1.12507427e+00
1.34610385e-01 -2.33317837e-01 5.46223409e-02 -1.02874005e+00
-1.20748460e+00 4.77110744e-01 -1.03539657e-02 -1.38209558e+00
2.55067557e-01 -6.84061587e-01 -3.59055847e-01 9.75838602e-01
-2.15175009e+00 -1.87959194e+00 -8.43456924e-01 6.04614198e-01
7.04414964e-01 -3.35006863e-02 1.99493378e-01 4.09838259e-01
-6.15511537e-01 3.50789398e-01 -1.75763980e-01 8.10911134e-02
6.55964077e-01 -9.51320529e-01 4.31262463e-01 9.09806013e-01
-5.94159774e-02 2.08662838e-01 1.62821069e-01 -5.45731008e-01
-1.53938317e+00 -1.45937693e+00 7.21670091e-01 -3.72857273e-01
5.25579691e-01 -4.53231871e-01 -7.09742963e-01 5.50436080e-01
-3.73200089e-01 3.37123603e-01 1.66474447e-01 -3.22932273e-01
-4.47728336e-01 -5.86652994e-01 -9.52294350e-01 4.85331357e-01
1.43187749e+00 -6.62622631e-01 -4.70901966e-01 -5.86132035e-02
7.51460671e-01 -5.72661817e-01 -8.73548865e-01 7.20065653e-01
5.40030956e-01 -1.19122672e+00 1.31438351e+00 -3.58806513e-02
3.45556587e-01 -7.28334427e-01 -5.75734735e-01 -1.05542588e+00
-3.24655950e-01 -2.62531996e-01 -1.76723033e-01 1.20741427e+00
1.22772932e-01 -6.12177968e-01 5.92533350e-01 -3.67130060e-03
-7.42173135e-01 -8.94985497e-01 -9.25805271e-01 -4.74415332e-01
-8.22775811e-02 -7.36405730e-01 6.91347480e-01 3.73701841e-01
-7.59418368e-01 5.10270238e-01 -4.40858632e-01 5.44499815e-01
1.01557434e+00 5.39852560e-01 1.12898242e+00 -1.32199109e+00
2.74452031e-01 -3.49904150e-01 -5.59121192e-01 -1.59072042e+00
4.09421176e-01 -7.23125935e-01 1.33132944e-02 -1.49912906e+00
-1.11321732e-01 -3.30005556e-01 -5.35726100e-02 2.71331042e-01
-1.87505871e-01 4.39461321e-01 3.24263573e-01 -2.83734892e-02
-5.97087383e-01 8.46379638e-01 1.30729473e+00 -1.47156447e-01
-2.28867866e-02 -2.42579237e-01 -6.02817535e-01 7.68044412e-01
3.83159578e-01 1.79918110e-01 -2.98290372e-01 -5.24222136e-01
-7.37302676e-02 -9.64438170e-03 7.23269463e-01 -1.23505557e+00
1.82148203e-01 -1.20186940e-01 5.17566860e-01 -1.45567679e+00
6.45722389e-01 -7.95636833e-01 -3.65978420e-01 8.78601819e-02
1.17745988e-01 -1.30283207e-01 3.19188625e-01 7.66776919e-01
-3.14466357e-01 2.46245280e-01 7.23324656e-01 1.54308364e-01
-1.56480110e+00 6.10230029e-01 -2.22237259e-02 -4.89875413e-02
1.25356662e+00 -3.17734897e-01 -4.69675958e-01 -1.59808800e-01
-3.20048958e-01 7.07961261e-01 5.71786702e-01 4.31991726e-01
9.45853114e-01 -1.57761860e+00 -7.30881870e-01 7.26171434e-01
5.52402377e-01 3.49562824e-01 9.46975231e-01 1.03998804e+00
-4.29665238e-01 5.79480946e-01 -2.56513953e-01 -9.20394003e-01
-1.04357886e+00 4.78381693e-01 2.40848288e-01 3.80886421e-02
-7.32239723e-01 8.20250213e-01 3.16255510e-01 -6.03327394e-01
-2.01680928e-01 -1.48763984e-01 -2.23987430e-01 1.94355831e-01
3.81826818e-01 4.05869067e-01 2.15270044e-03 -1.27681673e+00
-5.32155514e-01 1.43227232e+00 1.05880359e-02 1.40910402e-01
1.28141582e+00 -6.13268375e-01 2.24560037e-01 2.84642488e-01
1.29326940e+00 4.08843875e-01 -1.73707902e+00 -2.12916225e-01
-4.60716128e-01 -7.92208016e-01 4.19644356e-01 -3.98540884e-01
-1.03002036e+00 1.19372821e+00 6.03424728e-01 -2.16111377e-01
1.20244443e+00 1.14120655e-02 1.12600660e+00 1.19153693e-01
4.92261678e-01 -7.99556196e-01 -7.23298416e-02 5.45220017e-01
7.44702041e-01 -1.49035299e+00 -6.84308857e-02 -6.46797538e-01
-7.46649206e-01 1.06144321e+00 7.77266145e-01 -1.85827002e-01
5.47283053e-01 5.69375604e-02 -1.06468633e-01 -2.26007551e-01
-3.93805593e-01 -6.05534136e-01 6.29373789e-01 5.90068638e-01
-1.60696805e-01 -4.45355736e-02 1.62140280e-01 5.38076043e-01
1.50663108e-01 -3.00770998e-02 2.47489855e-01 6.40940487e-01
-6.09135985e-01 -6.05135262e-01 -3.76058489e-01 4.22827959e-01
4.71634924e-01 1.65880114e-01 -1.82818770e-02 9.17607963e-01
4.95985299e-01 7.13345289e-01 2.91932046e-01 -7.37403870e-01
7.75712132e-01 -2.44175285e-01 1.99674100e-01 -2.52062112e-01
2.27354631e-01 2.09690109e-01 -7.48289656e-03 -9.49982762e-01
-1.42312616e-01 -7.72442639e-01 -1.16641057e+00 -3.20291400e-01
-1.64872095e-01 -3.40878665e-01 5.07502019e-01 8.15537035e-01
8.05770338e-01 4.52415019e-01 9.32145596e-01 -1.30205762e+00
-2.67793447e-01 -4.95182544e-01 -4.79795039e-01 2.92098880e-01
6.52698874e-01 -1.19530773e+00 -2.56342113e-01 -3.33386004e-01] | [8.03225326538086, -1.8401490449905396] |
a466c063-01c0-414c-8192-797858cd5e4d | baller2vec-a-look-ahead-multi-entity | 2104.11980 | null | https://arxiv.org/abs/2104.11980v2 | https://arxiv.org/pdf/2104.11980v2.pdf | baller2vec++: A Look-Ahead Multi-Entity Transformer For Modeling Coordinated Agents | In many multi-agent spatiotemporal systems, agents operate under the influence of shared, unobserved variables (e.g., the play a team is executing in a game of basketball). As a result, the trajectories of the agents are often statistically dependent at any given time step; however, almost universally, multi-agent models implicitly assume the agents' trajectories are statistically independent at each time step. In this paper, we introduce baller2vec++, a multi-entity Transformer that can effectively model coordinated agents. Specifically, baller2vec++ applies a specially designed self-attention mask to a mixture of location and "look-ahead" trajectory sequences to learn the distributions of statistically dependent agent trajectories. We show that, unlike baller2vec (baller2vec++'s predecessor), baller2vec++ can learn to emulate the behavior of perfectly coordinated agents in a simulated toy dataset. Additionally, when modeling the trajectories of professional basketball players, baller2vec++ outperforms baller2vec by a wide margin. | ['Anh Nguyen', 'Michael A. Alcorn'] | 2021-04-24 | null | https://openreview.net/forum?id=p2XgjS3Qp4X | https://openreview.net/pdf?id=p2XgjS3Qp4X | neurips-2021-12 | ['trajectory-modeling'] | ['time-series'] | [-8.50169063e-01 -2.61967033e-01 1.19240163e-02 8.67502764e-02
-5.51346004e-01 -7.26104021e-01 9.03592825e-01 2.37654403e-01
-7.58305073e-01 6.50578439e-01 5.38946748e-01 3.06058601e-02
-5.35492674e-02 -7.58520544e-01 -9.47596133e-01 -6.07549250e-01
-3.95416617e-01 1.12002409e+00 1.60189852e-01 -4.38246489e-01
-2.64920235e-01 1.96418136e-01 -1.01478815e+00 -6.38124496e-02
2.59895205e-01 2.78422415e-01 2.90199935e-01 1.20665467e+00
2.30247065e-01 1.43000555e+00 -6.58774674e-01 -5.30785024e-01
2.80518979e-01 -2.95104086e-01 -3.89042675e-01 5.41353645e-03
-1.79634213e-01 -4.70777899e-01 -1.28996718e+00 7.28709757e-01
1.48475513e-01 5.35851955e-01 7.36866832e-01 -1.69930720e+00
-9.14883256e-01 7.37597883e-01 -3.38710964e-01 4.58703995e-01
1.49224341e-01 6.86876297e-01 1.24731719e+00 -4.50348139e-01
6.07617438e-01 9.88656878e-01 5.92196524e-01 2.79301554e-01
-1.12553990e+00 -3.73431355e-01 6.06949627e-01 3.24007928e-01
-1.30393207e+00 -3.53139341e-02 5.60295880e-01 -7.16446817e-01
1.08979130e+00 1.88045576e-02 6.51964605e-01 1.40729535e+00
3.35681379e-01 1.13032639e+00 3.05971593e-01 2.36251250e-01
3.14409286e-01 1.24898925e-02 3.36811781e-01 4.39382583e-01
4.59723622e-02 3.08671653e-01 -3.39352250e-01 -3.89898539e-01
7.57846832e-01 4.88148183e-01 6.55027702e-02 -2.79726803e-01
-1.42150605e+00 8.39732289e-01 3.40669215e-01 2.52267450e-01
-5.81114948e-01 7.64355123e-01 3.41304272e-01 2.46318519e-01
3.27696770e-01 3.28881592e-01 -3.67196769e-01 -3.15364331e-01
-3.82845521e-01 9.21716511e-01 7.70687938e-01 1.12409842e+00
6.64916575e-01 2.11184040e-01 -3.71025622e-01 3.29896837e-01
2.75342315e-01 5.81913650e-01 6.07534826e-01 -8.02634060e-01
5.29649019e-01 3.31828833e-01 6.96431398e-01 -8.58837903e-01
-5.79833746e-01 -2.95102745e-01 -6.31019831e-01 1.47912398e-01
5.85402966e-01 -7.83914924e-01 -6.16888583e-01 2.00661254e+00
8.93506333e-02 7.04502761e-01 2.16480136e-01 9.60373759e-01
5.17693222e-01 8.35149884e-01 6.11687116e-02 1.48316294e-01
9.82875049e-01 -1.23827612e+00 -5.49816906e-01 -2.13750765e-01
4.60821360e-01 -1.96470805e-02 5.03038824e-01 -3.79798561e-02
-1.16738641e+00 -5.34194112e-01 -5.74907660e-01 2.39329860e-01
-3.99612069e-01 -2.10833579e-01 7.57569134e-01 9.43366438e-02
-7.92103946e-01 2.89420635e-01 -1.44448864e+00 -1.31794810e-01
2.34841689e-01 2.73662925e-01 -5.13862848e-01 8.90242234e-02
-9.09791529e-01 8.48095953e-01 -2.95226239e-02 -2.26186216e-01
-1.57628226e+00 -7.07242012e-01 -9.68972683e-01 1.23275250e-01
4.22141463e-01 -6.91363454e-01 1.45264709e+00 -5.09907246e-01
-1.23950613e+00 1.78709120e-01 -1.53865308e-01 -8.15882266e-01
6.08520210e-01 6.55412525e-02 -3.85110438e-01 -3.68527234e-01
3.24497670e-01 6.97727352e-02 2.76061028e-01 -1.21099901e+00
-7.87863970e-01 -3.74014586e-01 2.06208006e-01 4.93859559e-01
2.07209080e-01 -4.19102199e-02 -6.39481783e-01 -5.08140922e-01
-6.09505177e-01 -1.18257284e+00 -7.34289348e-01 -4.33058411e-01
-4.02136743e-01 -5.12229204e-01 3.31277877e-01 -3.26173037e-01
7.86769569e-01 -2.11483264e+00 6.25954390e-01 -8.76893550e-02
5.70295036e-01 -5.77203755e-04 -4.20219034e-01 9.05216873e-01
4.33608532e-01 -7.78372958e-02 5.22833169e-02 -8.78500462e-01
3.62289041e-01 3.34326893e-01 -4.89257909e-02 9.91690576e-01
-2.58469850e-01 1.17343903e+00 -1.04333055e+00 3.55123505e-02
3.20368558e-01 3.37397426e-01 -5.84213018e-01 1.59345865e-01
-5.33760071e-01 4.10165876e-01 -5.86837411e-01 -7.01684505e-03
1.97231025e-01 -4.29444164e-01 2.09690318e-01 4.78963286e-01
-8.19596276e-02 -1.60979420e-01 -1.16108608e+00 1.42230916e+00
-3.84112775e-01 8.56586337e-01 2.37815976e-02 -6.54666543e-01
2.98992723e-01 5.26051462e-01 9.26916659e-01 -5.07106125e-01
3.19341242e-01 -3.83667529e-01 2.19206050e-01 -4.53323156e-01
7.03374326e-01 4.75791022e-02 -4.08976167e-01 7.83717036e-01
7.82987624e-02 2.61019915e-01 2.82935739e-01 4.50795710e-01
1.22421849e+00 -1.84543774e-01 1.67523086e-01 3.71761024e-01
-1.51940919e-02 3.18535179e-01 8.26408088e-01 1.11092556e+00
-4.51167375e-01 4.97547537e-01 3.76664221e-01 -6.92027271e-01
-1.34462821e+00 -1.05749810e+00 5.92552304e-01 1.13107467e+00
3.91288161e-01 -3.03922355e-01 -2.68080950e-01 -4.21330661e-01
4.75298971e-01 6.16361737e-01 -1.03491139e+00 5.41724898e-02
-6.83115065e-01 -9.55352128e-01 4.10954088e-01 6.70244336e-01
-8.46621171e-02 -1.08596313e+00 -5.80380678e-01 5.83217442e-01
5.40095456e-02 -1.05952930e+00 -7.43735790e-01 -1.68177709e-01
1.37920693e-01 -1.19469404e+00 -7.42121100e-01 -4.84417707e-01
3.55883867e-01 3.03339839e-01 1.14393902e+00 -3.27495076e-02
-1.95332423e-01 6.60348356e-01 -3.92792523e-01 -4.92457151e-01
-2.16978341e-01 -2.25096121e-02 3.40393335e-01 2.19963774e-01
4.72316146e-01 -3.69613588e-01 -5.04806399e-01 8.15041550e-03
-5.79531252e-01 -1.53512269e-01 1.53957367e-01 6.34344101e-01
2.84094989e-01 1.92074999e-01 3.50936830e-01 -5.11295497e-01
6.78168595e-01 -1.08398008e+00 -5.00381052e-01 1.03362590e-01
1.00557357e-01 -3.01436722e-01 9.60016549e-01 -7.88097382e-01
-6.07478023e-01 -3.01046550e-01 3.42105441e-02 -8.11319411e-01
-3.51143897e-01 5.19325793e-01 1.66273177e-01 5.13085783e-01
3.47661436e-01 4.68159705e-01 9.05607492e-02 -2.24029094e-01
5.05935550e-01 2.32955903e-01 5.29609442e-01 -3.10725600e-01
8.00426602e-01 6.60028160e-01 -4.02250320e-01 -7.84111679e-01
-2.45764464e-01 -4.69726294e-01 -2.71066189e-01 -1.77461892e-01
1.02258062e+00 -1.24152780e+00 -1.39169824e+00 7.71219492e-01
-1.09169936e+00 -1.03263617e+00 -3.99389803e-01 9.87847328e-01
-6.85103595e-01 -1.18635729e-01 -7.34619677e-01 -8.63276184e-01
4.28220302e-01 -1.18834138e+00 6.29598618e-01 2.39168316e-01
-1.55336812e-01 -1.42279732e+00 6.91630065e-01 2.30774313e-01
1.14067197e-01 8.94053876e-02 4.73826915e-01 -1.13730586e+00
-7.54640818e-01 -2.69161731e-01 3.19027752e-01 -2.38378450e-01
9.46153551e-02 6.74248848e-04 -3.31581444e-01 -4.21463698e-01
-4.77769554e-01 2.61311419e-02 6.07968271e-01 7.20226645e-01
6.15079463e-01 -3.50788683e-01 -6.00200593e-01 4.90654588e-01
1.28952765e+00 4.50269163e-01 8.61895978e-02 1.98450997e-01
8.09703946e-01 2.93408394e-01 1.60228893e-01 7.82999277e-01
1.17193794e+00 7.50409126e-01 6.25790000e-01 9.09492001e-02
2.95138717e-01 -4.29476172e-01 3.35986912e-01 4.08534855e-01
-6.32716529e-03 -8.69344056e-01 -8.73561084e-01 1.03131413e+00
-2.21904421e+00 -1.55791378e+00 -9.56355110e-02 1.88752353e+00
1.05914436e-01 -3.27384531e-01 6.97920442e-01 -7.33082235e-01
5.09374499e-01 4.52031136e-01 -1.03442371e+00 2.06650749e-01
-3.49757850e-01 -3.49102676e-01 9.23318028e-01 8.17441285e-01
-1.13862967e+00 1.06652474e+00 6.10302401e+00 2.87587017e-01
-6.20370686e-01 3.93954754e-01 1.67284146e-01 -4.89675730e-01
-2.85186768e-01 -3.07488710e-01 -6.92398012e-01 7.53669620e-01
8.15592051e-01 -6.50298297e-01 5.70216238e-01 5.47324419e-01
3.35696429e-01 2.76273280e-01 -1.14854336e+00 7.95190334e-01
5.12602553e-02 -1.59999537e+00 -3.52528512e-01 3.17117631e-01
8.34809601e-01 6.68766677e-01 2.93952286e-01 6.12980068e-01
1.56165290e+00 -1.27473259e+00 8.48312914e-01 4.87313718e-01
5.26410006e-02 -8.62565577e-01 6.35365665e-01 6.55244291e-01
-1.10117173e+00 -3.26323479e-01 -2.30548531e-01 -2.76604414e-01
7.57621586e-01 -1.57454699e-01 -6.22444332e-01 5.84804595e-01
5.29541075e-01 9.35963690e-01 -1.26462564e-01 9.03701663e-01
5.01020551e-02 4.06474710e-01 -2.43897542e-01 -1.26345620e-01
7.60532796e-01 -3.94660503e-01 8.18794131e-01 8.75998139e-01
2.18204901e-01 3.40181112e-01 6.98477209e-01 7.41135955e-01
-4.51197438e-02 -2.39282534e-01 -7.48949707e-01 -2.29828194e-01
5.95673144e-01 9.24145520e-01 -2.00656369e-01 -4.46614414e-01
-8.42818439e-01 9.24183309e-01 6.79004014e-01 6.53697193e-01
-1.12978196e+00 9.27837938e-02 1.49987853e+00 -2.21175224e-01
5.95976055e-01 -5.78487754e-01 2.83881932e-01 -1.26766920e+00
-3.14982027e-01 -7.09387779e-01 2.26524830e-01 -6.01646125e-01
-1.55999494e+00 7.70586610e-01 -3.86821806e-01 -1.03449655e+00
-4.62009877e-01 -2.86505073e-01 -9.55258608e-01 7.71537662e-01
-9.67964888e-01 -1.19855237e+00 4.66072969e-02 8.30470383e-01
6.44546986e-01 -6.70430064e-01 4.67970610e-01 3.37219574e-02
-9.28928852e-01 2.74685651e-01 5.28112233e-01 6.87119722e-01
1.82949394e-01 -1.41719699e+00 9.72029328e-01 6.74327135e-01
6.50807083e-01 4.47819054e-01 9.88047183e-01 -6.63460732e-01
-1.45043862e+00 -1.35824347e+00 5.20346582e-01 -7.59341776e-01
9.40951884e-01 -3.57075840e-01 -7.45075166e-01 1.51056910e+00
4.74560380e-01 7.05628097e-02 8.31286371e-01 7.52426982e-02
-1.16560347e-01 4.06386703e-01 -8.23288202e-01 1.03238952e+00
7.55497038e-01 -4.72079337e-01 -5.20517588e-01 6.32319689e-01
7.94989586e-01 -3.43808562e-01 -8.15855980e-01 -2.79208779e-01
4.00206208e-01 -9.78626013e-01 1.00649428e+00 -1.34767580e+00
1.16276436e-01 -3.33503067e-01 -2.88714349e-01 -1.86920071e+00
-6.26460791e-01 -6.07799768e-01 -2.19141543e-01 6.73162162e-01
4.86259788e-01 -5.85261822e-01 8.08649063e-01 7.94116139e-01
1.58411488e-02 -3.76061648e-01 -8.45083177e-01 -8.20042491e-01
3.46217662e-01 -5.18467784e-01 1.02932358e+00 7.53079414e-01
3.57358973e-03 -4.41826694e-02 -8.64956796e-01 7.05588281e-01
6.40123069e-01 -3.53373773e-03 1.16382527e+00 -8.38094771e-01
-6.41160965e-01 -5.11702657e-01 -5.10674298e-01 -1.32725263e+00
6.47274196e-01 -7.07978904e-01 5.17895212e-03 -1.54453695e+00
2.60683477e-01 -5.13330042e-01 -3.29775870e-01 -6.97545186e-02
-2.32705489e-01 -4.58290726e-02 4.23488051e-01 1.21482521e-01
-7.01496542e-01 8.20215046e-01 1.01352203e+00 -3.27929765e-01
-3.89131367e-01 1.59797490e-01 -4.38596368e-01 5.82877934e-01
5.61818182e-01 -3.53062868e-01 -4.80775326e-01 -6.99137926e-01
8.80754367e-02 5.16954839e-01 5.80655098e-01 -7.56273925e-01
7.72467792e-01 -6.21272802e-01 -1.89616796e-04 -6.05352819e-01
7.35655069e-01 -6.98136568e-01 4.46013659e-01 1.68373942e-01
-4.38867986e-01 5.32531023e-01 -1.58913389e-01 1.06821167e+00
-2.07262710e-01 1.44634947e-01 1.87746957e-01 -3.76284868e-01
-6.48604155e-01 1.15377772e+00 -9.02777076e-01 2.49813974e-01
1.43053126e+00 1.62365332e-01 -3.65478724e-01 -8.04090381e-01
-8.88635635e-01 8.83076787e-01 5.46232283e-01 4.08535749e-01
4.39541012e-01 -1.14967394e+00 -8.85073304e-01 4.66503687e-02
1.26690669e-02 -1.05220368e-02 7.70553589e-01 5.85460901e-01
-1.38830200e-01 2.64267385e-01 -2.04870459e-02 -2.85934150e-01
-9.55673218e-01 9.79924440e-01 5.87696671e-01 -2.62517661e-01
-7.62030363e-01 8.03593576e-01 3.90707314e-01 -5.58389962e-01
3.73415090e-02 6.12628832e-03 -1.90336704e-01 -6.57735392e-02
6.49941504e-01 4.70607311e-01 -5.86170137e-01 -9.89901900e-01
-2.34353542e-01 8.67576897e-02 -2.36928448e-01 -6.08089149e-01
1.55701888e+00 -6.00607805e-02 2.21981108e-01 7.53604472e-01
1.03323829e+00 -9.08067971e-02 -1.69787335e+00 -3.62832993e-01
-3.85121942e-01 -3.48788261e-01 -3.06921035e-01 -2.84408242e-01
-1.09523416e+00 3.47316235e-01 -2.36104369e-01 3.26628238e-01
1.04247570e-01 2.49464720e-01 7.82535553e-01 9.85640958e-02
6.64139092e-01 -7.73004174e-01 -8.77501890e-02 7.05536187e-01
4.75676209e-01 -1.27603018e+00 -4.72442091e-01 3.59815478e-01
-1.20483792e+00 3.67365271e-01 6.08549297e-01 -5.82630754e-01
7.84651935e-01 3.67101520e-01 1.31080657e-01 -2.94400156e-01
-1.35841238e+00 -4.34464127e-01 -1.44389540e-01 7.61909962e-01
-1.05838530e-01 4.47211772e-01 4.85634595e-01 7.37808228e-01
-1.82620913e-01 -2.94251442e-01 1.00687730e+00 4.76489544e-01
-4.38752212e-02 -6.92282200e-01 -7.83285871e-02 1.90330535e-01
-2.34789386e-01 2.82169849e-01 1.04620688e-01 9.08956528e-01
1.26861349e-01 1.06813681e+00 6.87442243e-01 -3.20865929e-01
4.03500080e-01 -3.78125697e-01 1.97333589e-01 -6.18916631e-01
-8.51558506e-01 -3.01488549e-01 -2.12174267e-01 -4.72848147e-01
-1.69289574e-01 -1.04950583e+00 -1.05990624e+00 -7.09318399e-01
2.55533457e-01 3.07796538e-01 2.38213524e-01 9.67132032e-01
3.12563658e-01 6.62770033e-01 6.48248255e-01 -8.65013659e-01
-4.13455993e-01 -8.69706094e-01 -8.30384672e-01 6.50935471e-01
8.38585138e-01 -6.28060281e-01 -2.41481155e-01 -2.12547421e-01] | [5.853166103363037, 0.7291544675827026] |
3760f5a4-c513-454f-8f0e-aa8d2b83ae31 | lanesnns-spiking-neural-networks-for-lane | 2208.02253 | null | https://arxiv.org/abs/2208.02253v1 | https://arxiv.org/pdf/2208.02253v1.pdf | LaneSNNs: Spiking Neural Networks for Lane Detection on the Loihi Neuromorphic Processor | Autonomous Driving (AD) related features represent important elements for the next generation of mobile robots and autonomous vehicles focused on increasingly intelligent, autonomous, and interconnected systems. The applications involving the use of these features must provide, by definition, real-time decisions, and this property is key to avoid catastrophic accidents. Moreover, all the decision processes must require low power consumption, to increase the lifetime and autonomy of battery-driven systems. These challenges can be addressed through efficient implementations of Spiking Neural Networks (SNNs) on Neuromorphic Chips and the use of event-based cameras instead of traditional frame-based cameras. In this paper, we present a new SNN-based approach, called LaneSNN, for detecting the lanes marked on the streets using the event-based camera input. We develop four novel SNN models characterized by low complexity and fast response, and train them using an offline supervised learning rule. Afterward, we implement and map the learned SNNs models onto the Intel Loihi Neuromorphic Research Chip. For the loss function, we develop a novel method based on the linear composition of Weighted binary Cross Entropy (WCE) and Mean Squared Error (MSE) measures. Our experimental results show a maximum Intersection over Union (IoU) measure of about 0.62 and very low power consumption of about 1 W. The best IoU is achieved with an SNN implementation that occupies only 36 neurocores on the Loihi processor while providing a low latency of less than 8 ms to recognize an image, thereby enabling real-time performance. The IoU measures provided by our networks are comparable with the state-of-the-art, but at a much low power consumption of 1 W. | ['Muhammad Shafique', 'Guido Masera', 'Maurizio Martina', 'Alberto Marchisio', 'Alberto Viale'] | 2022-08-03 | null | null | null | null | ['lane-detection'] | ['computer-vision'] | [ 1.77815378e-01 -2.00316206e-01 1.52376369e-01 -4.68424857e-01
-1.34197965e-01 -2.48392463e-01 4.77180302e-01 -4.19188216e-02
-1.12823808e+00 6.78489864e-01 -6.53736353e-01 -1.78440303e-01
-1.29632965e-01 -9.34175372e-01 -9.65613425e-01 -7.26179302e-01
-1.12241365e-01 7.29981586e-02 8.42401743e-01 2.42561270e-02
1.95423439e-01 5.79306483e-01 -2.02679300e+00 -8.83778855e-02
7.00728297e-01 1.64180982e+00 5.54662943e-01 4.27719563e-01
1.50291622e-01 5.82540333e-01 -4.32294011e-01 -1.74503118e-01
9.71545726e-02 -8.76323357e-02 6.51447102e-03 -6.07307136e-01
8.70360583e-02 -3.24737817e-01 -6.38101220e-01 1.19546664e+00
4.17531669e-01 -1.36860237e-01 3.68492872e-01 -1.39847791e+00
2.93906056e-03 4.00774986e-01 -9.86349061e-02 3.05609792e-01
-1.86213434e-01 3.60721797e-01 4.17440891e-01 -7.23239899e-01
3.94357026e-01 9.82570708e-01 5.82262158e-01 5.79633832e-01
-1.06877100e+00 -1.04864752e+00 -2.00889811e-01 5.62329710e-01
-1.24676621e+00 -6.92828000e-01 4.62071747e-01 -1.13081865e-01
1.19926882e+00 -2.90742993e-01 8.40411246e-01 1.00567234e+00
8.58881772e-01 3.71344358e-01 1.11128855e+00 -1.19738765e-01
7.51468122e-01 -1.32201344e-01 3.27597886e-01 8.11340511e-01
6.95571661e-01 1.90849917e-03 -6.92599297e-01 1.90154314e-01
5.03806889e-01 1.91445395e-01 5.00209071e-02 -1.55459866e-01
-1.10873055e+00 4.80111450e-01 5.05845129e-01 1.83162987e-01
-1.64288789e-01 7.93379068e-01 4.29011405e-01 -3.51001620e-02
-4.52722639e-01 2.94360388e-02 6.07914990e-03 -3.26482028e-01
-6.73718393e-01 -1.30549058e-01 7.98642337e-01 8.96188140e-01
8.02984297e-01 2.01262936e-01 2.24130005e-01 4.71955895e-01
3.86631638e-01 1.06924927e+00 6.30224228e-01 -1.01399636e+00
1.06391110e-01 6.43587530e-01 -2.42711976e-01 -7.54639566e-01
-6.97128057e-01 -1.76542997e-01 -8.93352747e-01 4.44756359e-01
2.14125693e-01 1.60673186e-02 -8.85414302e-01 1.68319523e+00
-2.07271889e-01 -3.93067971e-02 3.38041335e-02 6.76891923e-01
4.08249885e-01 7.89678097e-01 -3.57383713e-02 1.19891251e-02
1.36020470e+00 -5.48024714e-01 -6.37510657e-01 -5.32735705e-01
2.23343521e-01 -8.75900686e-02 7.40486562e-01 3.92827958e-01
-8.79261076e-01 -5.33556461e-01 -1.75722980e+00 1.50241936e-02
-4.71785605e-01 1.38171345e-01 4.16520774e-01 6.85821176e-01
-1.12037909e+00 6.61716640e-01 -1.22738945e+00 -4.71643478e-01
5.13340533e-01 8.43731463e-01 -2.23384082e-01 1.71620727e-01
-1.02537930e+00 9.08958614e-01 2.08408669e-01 9.99566242e-02
-1.08454061e+00 -3.15051705e-01 -6.91317260e-01 2.87165135e-01
-8.38627517e-02 -2.63280272e-01 1.12263298e+00 -7.52202034e-01
-1.62461925e+00 6.10767186e-01 -5.10672480e-02 -9.74610806e-01
-6.29356317e-03 9.01451558e-02 -4.80777293e-01 2.20708713e-01
-1.04821227e-01 8.84180844e-01 6.06033087e-01 -7.28622377e-01
-8.90907109e-01 -4.85875070e-01 -1.50208935e-01 -3.80325973e-01
-7.28298545e-01 -1.57481641e-01 -2.85424620e-01 9.76315215e-02
3.94084724e-03 -1.00395465e+00 -1.75346896e-01 4.65475559e-01
1.54974699e-01 -1.02575332e-01 1.08319521e+00 -1.71683058e-02
9.08052146e-01 -2.31796336e+00 -5.01702130e-01 2.28327900e-01
2.31576916e-02 3.45650882e-01 -6.23937100e-02 -1.94058314e-01
5.00699937e-01 -3.88713479e-01 -3.51601034e-01 8.30644369e-03
-1.18529253e-01 3.02740932e-01 -2.60873795e-01 5.67812264e-01
2.98731834e-01 6.94123089e-01 -6.90942705e-01 -1.62838444e-01
2.20516160e-01 3.80833119e-01 -3.02330077e-01 -9.08323303e-02
6.09170049e-02 -1.03476301e-01 -1.97097868e-01 5.19605875e-01
4.10460085e-01 4.27201837e-02 2.67271340e-01 -1.84253320e-01
-4.81720358e-01 3.68682861e-01 -9.75837171e-01 1.62786651e+00
-4.82212156e-01 9.60356951e-01 7.52559453e-02 -8.35009217e-01
1.31745231e+00 -1.53643325e-01 2.67340183e-01 -1.31671512e+00
5.24493396e-01 7.08131731e-01 1.51885077e-01 -1.65344924e-01
1.07350014e-01 2.58507937e-01 -2.69207567e-01 2.06681192e-01
3.92448992e-01 3.67383510e-01 1.61258757e-01 -1.42046690e-01
1.53816307e+00 -1.06797338e-01 3.40513736e-02 -5.94177186e-01
3.28127921e-01 -1.84140235e-01 7.87642002e-01 6.50996447e-01
-4.31817114e-01 -4.78201406e-03 4.34111208e-01 -4.14777011e-01
-9.64820325e-01 -1.15214372e+00 -2.84219295e-01 5.18635094e-01
5.40633738e-01 -8.08776617e-02 -9.24041808e-01 -1.20500386e-01
-7.32804090e-02 5.37067354e-01 -1.63302794e-01 -4.18455780e-01
-6.28104985e-01 -7.56222367e-01 8.28126073e-01 6.93128467e-01
8.99838626e-01 -9.62583482e-01 -1.68035960e+00 4.84512240e-01
3.68438751e-01 -1.35972238e+00 1.23284094e-01 9.33030427e-01
-8.54765236e-01 -6.22085750e-01 -2.30167300e-01 -9.13450897e-01
5.32440245e-01 1.10068351e-01 5.57928145e-01 -4.65362847e-01
-4.36864018e-01 9.65027511e-03 7.90851191e-02 -5.15891969e-01
7.53841102e-02 3.94387543e-02 4.71806824e-01 -3.53951938e-02
6.50320590e-01 -8.30876887e-01 -8.25523615e-01 3.85302246e-01
-7.79694557e-01 -7.81837478e-02 7.81574309e-01 8.06592345e-01
8.03353846e-01 -1.06327057e-01 7.12492108e-01 -2.92073727e-01
1.14042193e-01 -2.47480065e-01 -1.14453483e+00 -4.16780785e-02
-8.58711481e-01 3.18144828e-01 9.94792938e-01 -4.16718721e-01
-7.57379413e-01 4.71600026e-01 -3.54259387e-02 -2.91985571e-01
4.59664501e-02 -3.58395092e-02 -2.47175038e-01 -4.25287962e-01
5.73711574e-01 2.19028935e-01 1.56374663e-01 5.46614639e-02
-1.33135468e-01 8.92578065e-01 7.25625932e-01 -1.42036587e-01
2.94607371e-01 7.41829693e-01 3.60996783e-01 -7.45660663e-01
-1.41654164e-01 -1.51367575e-01 -1.05913751e-01 -5.15733421e-01
6.38008535e-01 -1.01388955e+00 -1.13210678e+00 7.29109824e-01
-1.16947281e+00 -3.66492599e-01 -2.61983350e-02 4.94857758e-01
-6.41928136e-01 -1.28355309e-01 -4.94403839e-01 -8.26326251e-01
-6.15180790e-01 -1.17397058e+00 6.31715357e-01 7.05621779e-01
-6.57899631e-03 -2.87228614e-01 -1.42710969e-01 -1.96726605e-01
6.49051130e-01 -3.08523979e-02 8.48315239e-01 -3.78245324e-01
-7.03839421e-01 -2.92756200e-01 -4.66978639e-01 2.27809757e-01
-2.35896841e-01 -2.21437901e-01 -1.32618666e+00 -2.07496420e-01
1.63434982e-01 -2.50502080e-01 9.88081992e-01 2.51398742e-01
1.28418136e+00 1.28810465e-01 -5.47752976e-01 6.76688254e-01
1.77870107e+00 6.37220502e-01 8.50780070e-01 3.86192322e-01
2.16569275e-01 3.24247360e-01 3.92140418e-01 4.60278898e-01
3.01522315e-01 5.26959240e-01 8.11457992e-01 3.35001141e-01
1.55250687e-04 -5.81140770e-03 7.45886147e-01 8.82843435e-01
3.24015021e-01 -1.90415442e-01 -9.20098662e-01 6.11051857e-01
-1.98329997e+00 -8.22481215e-01 -6.41212910e-02 2.41850853e+00
3.98589104e-01 5.83807945e-01 -1.91830307e-01 1.24443509e-01
7.52188206e-01 -1.83282480e-01 -1.00881243e+00 -5.58455408e-01
-1.49400055e-01 4.16097641e-01 1.01499593e+00 -1.59789309e-01
-9.70919311e-01 5.39554417e-01 5.52256680e+00 6.92293108e-01
-1.43197966e+00 1.68675229e-01 3.70577246e-01 -4.78229344e-01
1.30026415e-01 -1.48797676e-01 -1.15136766e+00 9.04659331e-01
1.74270487e+00 -1.56837944e-02 5.09200811e-01 1.02805734e+00
1.13837101e-01 -3.79754484e-01 -9.82698381e-01 1.34414446e+00
-2.46600248e-02 -1.26206338e+00 -2.90094733e-01 1.93170719e-02
3.60204250e-01 3.60533506e-01 -7.18274415e-02 1.47993505e-01
-1.25647038e-01 -7.13157415e-01 9.70528364e-01 5.82864761e-01
9.22772706e-01 -1.01936078e+00 7.64846742e-01 2.08665088e-01
-1.05379510e+00 -5.70465505e-01 -6.65212035e-01 -1.18852593e-01
1.53557223e-03 8.42567265e-01 -1.71160594e-01 -1.74586147e-01
1.12907946e+00 5.65881968e-01 -2.49897137e-01 1.14763761e+00
7.46144503e-02 4.54278946e-01 -6.31357610e-01 -7.68242776e-01
1.71034768e-01 -2.09272206e-02 2.55985856e-01 1.13821995e+00
6.26429141e-01 -7.04551786e-02 -3.95863682e-01 8.32740068e-01
-3.54393035e-01 -4.37469304e-01 -7.24946558e-01 2.18081012e-01
6.58883631e-01 1.51884425e+00 -8.92338157e-01 -2.03255881e-02
-4.14577574e-01 8.96328747e-01 2.12566748e-01 -2.15229258e-01
-9.59450126e-01 -1.03650546e+00 6.58161879e-01 1.78180993e-01
3.98472130e-01 -3.28941256e-01 -5.31052589e-01 -6.95427179e-01
3.30910474e-01 -3.66844118e-01 -3.06208014e-01 -5.32028556e-01
-8.05385172e-01 6.72598362e-01 -4.87669468e-01 -1.26978528e+00
-7.77332485e-02 -1.13275719e+00 -5.36534071e-01 2.96806991e-01
-1.55392063e+00 -4.22368705e-01 -4.96568054e-01 3.54857653e-01
2.34308332e-01 -1.44334763e-01 7.34058142e-01 5.24609566e-01
-7.28159130e-01 5.20019114e-01 3.38102996e-01 -6.90609887e-02
5.95890760e-01 -7.20031917e-01 3.88608962e-01 8.92547786e-01
-2.63140738e-01 3.83356661e-01 3.34056199e-01 -2.62052149e-01
-1.91189110e+00 -1.25016189e+00 6.93622947e-01 1.22294836e-01
5.86394310e-01 -6.55838847e-01 -6.11468852e-01 1.75715610e-01
1.16517901e-01 1.47257820e-01 3.27861905e-01 -5.25316417e-01
-1.78641453e-01 -9.00874853e-01 -1.20046639e+00 5.53474426e-01
1.19530487e+00 -3.78022224e-01 -1.41554013e-01 -1.46894425e-01
3.24343711e-01 1.12706050e-01 -4.32129830e-01 4.53085154e-01
8.84980857e-01 -9.32032824e-01 4.06425297e-01 1.76523849e-01
1.54299781e-01 -6.05223656e-01 -2.96310157e-01 -8.36489379e-01
-2.06931621e-01 -4.16548669e-01 -1.32491112e-01 1.04926944e+00
4.48451251e-01 -8.04225028e-01 7.09724963e-01 4.50861365e-01
-2.27171212e-01 -8.48388910e-01 -1.41430449e+00 -1.13910472e+00
-4.94996578e-01 -4.35897559e-01 2.60326028e-01 1.94088984e-02
1.70446768e-01 1.58168778e-01 1.38502717e-01 1.62763186e-02
7.36867309e-01 -1.90618962e-01 2.02943072e-01 -1.25277317e+00
1.29451603e-01 -3.47986341e-01 -1.20716834e+00 -8.02562475e-01
1.15169667e-01 -7.04802513e-01 5.79353392e-01 -1.26546907e+00
2.66670972e-01 -3.19216311e-01 -4.93017197e-01 6.07920229e-01
5.02528071e-01 2.15912238e-01 4.41421429e-03 7.57904444e-03
-8.46667647e-01 5.73969007e-01 2.69789487e-01 -2.43177146e-01
1.88067168e-01 -4.14255232e-01 -1.44358486e-01 8.73449445e-01
8.38113844e-01 -6.53133094e-01 -2.13513434e-01 -5.83430111e-01
1.56531706e-01 -3.36519629e-01 5.07991195e-01 -2.07024932e+00
9.00839269e-01 3.46589833e-01 5.15191615e-01 -3.69238883e-01
4.30260628e-01 -1.00895560e+00 1.48169398e-01 1.02679074e+00
8.22139978e-02 4.96385545e-02 3.37184995e-01 5.97262621e-01
-3.46271284e-02 -3.28539997e-01 1.03388786e+00 2.52471298e-01
-1.00693977e+00 6.13712817e-02 -1.00299907e+00 -3.74230325e-01
1.26923800e+00 -3.74610096e-01 -6.43121541e-01 -2.88483594e-02
-8.18778761e-03 1.72452077e-01 4.29678053e-01 1.78480268e-01
7.89295137e-01 -1.16243935e+00 2.10466441e-02 4.01043773e-01
1.54351607e-01 -3.29324335e-01 5.31243496e-02 5.99346578e-01
-6.21219575e-01 6.86135232e-01 -8.42667341e-01 -6.86519861e-01
-7.89699018e-01 1.81043714e-01 4.02252585e-01 1.13417290e-01
-3.20404857e-01 4.80266005e-01 -1.26253262e-01 -5.92009537e-03
2.83493429e-01 -2.74853677e-01 9.09514800e-02 -1.66607425e-01
5.91874301e-01 5.79708099e-01 3.89955550e-01 -1.58479333e-01
-6.38766706e-01 4.77076322e-01 1.31912455e-01 -2.18581885e-01
1.29780960e+00 1.25562117e-01 -9.41092446e-02 4.88232464e-01
1.00163019e+00 -6.83246553e-01 -1.44087374e+00 3.47590558e-02
1.13255814e-01 1.49351433e-01 2.51678348e-01 -5.33378363e-01
-1.06074810e+00 9.78492022e-01 1.30066526e+00 -1.67705476e-01
1.31110823e+00 -2.89407611e-01 1.08727336e+00 8.15293968e-01
9.90756750e-01 -1.36186755e+00 -1.55196354e-01 6.88469112e-01
3.27284247e-01 -8.65998209e-01 -3.68121058e-01 2.30555590e-02
-1.77376211e-01 1.24579394e+00 9.00912106e-01 -3.22457284e-01
6.93667591e-01 7.86264360e-01 -2.96497941e-01 -7.27992447e-04
-9.66209590e-01 -1.60520494e-01 -3.11886787e-01 5.00947475e-01
-1.72647938e-01 5.80702648e-02 -3.62869620e-01 8.34095716e-01
1.19858474e-01 2.27958590e-01 4.19044137e-01 9.27010536e-01
-9.02323544e-01 -7.44534492e-01 -3.73291038e-02 7.76887596e-01
-2.68368214e-01 1.40503526e-01 -3.53534110e-02 2.46787608e-01
2.18229339e-01 8.58779132e-01 5.65974236e-01 -7.69791007e-01
4.36740160e-01 -4.37560380e-02 3.48708272e-01 -1.18781514e-01
-3.70729655e-01 -4.49488699e-01 -1.57950252e-01 -8.41863215e-01
-9.92968380e-02 -4.52954262e-01 -1.62962699e+00 -2.88884103e-01
-1.81596905e-01 -4.49845344e-01 1.33743441e+00 7.93331146e-01
7.72533178e-01 4.07978505e-01 7.22698987e-01 -8.14967155e-01
-3.24698687e-01 -4.98266399e-01 -4.87951577e-01 -2.34439671e-01
-4.21755202e-02 -7.30307937e-01 -2.18024448e-01 -1.59211263e-01] | [8.19087028503418, 2.4010822772979736] |
87f7eea6-e073-4832-9a7f-7e198e4626cc | a-model-based-active-testing-approach-to | 1401.3850 | null | http://arxiv.org/abs/1401.3850v1 | http://arxiv.org/pdf/1401.3850v1.pdf | A Model-Based Active Testing Approach to Sequential Diagnosis | Model-based diagnostic reasoning often leads to a large number of diagnostic
hypotheses. The set of diagnoses can be reduced by taking into account extra
observations (passive monitoring), measuring additional variables (probing) or
executing additional tests (sequential diagnosis/test sequencing). In this
paper we combine the above approaches with techniques from Automated Test
Pattern Generation (ATPG) and Model-Based Diagnosis (MBD) into a framework
called FRACTAL (FRamework for ACtive Testing ALgorithms). Apart from the inputs
and outputs that connect a system to its environment, in active testing we
consider additional input variables to which a sequence of test vectors can be
supplied. We address the computationally hard problem of computing optimal
control assignments (as defined in FRACTAL) in terms of a greedy approximation
algorithm called FRACTAL-G. We compare the decrease in the number of remaining
minimal cardinality diagnoses of FRACTAL-G to that of two more FRACTAL
algorithms: FRACTAL-ATPG and FRACTAL-P. FRACTAL-ATPG is based on ATPG and
sequential diagnosis while FRACTAL-P is based on probing and, although not an
active testing algorithm, provides a baseline for comparing the lower bound on
the number of reachable diagnoses for the FRACTAL algorithms. We empirically
evaluate the trade-offs of the three FRACTAL algorithms by performing extensive
experimentation on the ISCAS85/74XXX benchmark of combinational circuits. | ['Alexander Feldman', 'Arjan van Gemund', 'Gregory Provan'] | 2014-01-16 | null | null | null | null | ['sequential-diagnosis'] | ['medical'] | [ 4.74919140e-01 7.61829078e-01 -1.89998552e-01 -1.67792261e-01
-6.67305946e-01 -7.14322507e-01 4.17135835e-01 2.69893050e-01
2.75438368e-01 8.40165436e-01 -6.11348808e-01 -8.43723357e-01
-6.30312443e-01 -1.22223163e+00 -4.85525668e-01 -5.03825366e-01
-3.18481296e-01 1.12806201e+00 8.43978524e-01 9.40387473e-02
1.95982546e-01 5.43422878e-01 -1.67193055e+00 3.54231030e-01
1.01065958e+00 7.30097175e-01 -1.77942276e-01 7.54291236e-01
6.49288818e-02 7.02102780e-01 -6.56009138e-01 -1.89439237e-01
1.65301174e-01 -1.07830882e+00 -9.46141481e-01 4.70919639e-01
-4.20105398e-01 -1.49786860e-01 2.01215103e-01 1.01568210e+00
2.83313096e-01 -3.57666820e-01 3.67529660e-01 -1.78033102e+00
1.90973043e-01 9.91610646e-01 -1.48421764e-01 6.68338686e-02
6.01043046e-01 4.06943142e-01 7.30509400e-01 -2.44583189e-01
6.26188397e-01 1.22134566e+00 3.68196756e-01 5.66108763e-01
-1.51388526e+00 -2.48749003e-01 -4.52187173e-02 2.92809367e-01
-1.25250089e+00 -2.45885029e-01 3.52537423e-01 -5.36653757e-01
1.26445794e+00 7.29491651e-01 8.13675821e-01 8.34514916e-01
2.70347595e-01 4.90395516e-01 1.17175126e+00 -7.40391731e-01
1.06169355e+00 5.41076101e-02 2.09057853e-01 8.69729102e-01
7.62264609e-01 4.74256665e-01 8.45440105e-02 -7.29703903e-01
5.04953504e-01 -1.43671051e-01 -1.15619600e-01 -3.41780931e-01
-9.59314942e-01 6.69893682e-01 -3.58513564e-01 2.60457367e-01
-1.76354900e-01 1.46758720e-01 3.52277189e-01 8.32808673e-01
-1.99714974e-01 7.92976558e-01 -6.69185579e-01 -1.03402354e-01
-8.43436003e-01 2.87648946e-01 1.27885938e+00 7.75193274e-01
6.95967019e-01 -1.29388208e-02 -3.39698106e-01 4.52216789e-02
2.47387394e-01 3.25070113e-01 4.40771878e-01 -6.07860446e-01
2.64905989e-01 1.26872039e+00 7.99269453e-02 -4.40925032e-01
-3.21897775e-01 -4.33470756e-01 -2.73388416e-01 2.98679531e-01
4.51610684e-02 -2.64718354e-01 -9.41674411e-01 1.44440019e+00
2.45992258e-01 2.74930894e-01 -8.49166662e-02 2.77688771e-01
-1.55086266e-02 3.44631523e-01 -6.27070963e-01 -6.71238899e-01
7.10053802e-01 -8.04918528e-01 -1.97618335e-01 -4.95185442e-02
1.09783459e+00 -1.02074653e-01 6.11638546e-01 8.43088150e-01
-1.12744915e+00 -1.45353660e-01 -1.46010554e+00 1.02757168e+00
-1.57231748e-01 -2.23532259e-01 5.40036738e-01 9.15151417e-01
-1.45372117e+00 7.09804058e-01 -1.02811563e+00 -2.97952473e-01
1.12976441e-02 9.62108791e-01 -1.24844924e-01 -3.94670755e-01
-7.44760096e-01 5.69508851e-01 4.35716599e-01 -1.66306823e-01
-1.20294154e+00 -2.53867358e-01 -7.81778932e-01 1.45027384e-01
7.35469759e-01 -5.36210358e-01 1.10999656e+00 -8.03916872e-01
-1.40654171e+00 4.69657123e-01 1.41253680e-01 -6.09821022e-01
6.39066637e-01 6.78280473e-01 -6.56971097e-01 5.41546121e-02
-9.91445407e-02 3.85233104e-01 4.52214330e-01 -8.99446130e-01
-6.72395706e-01 -1.72275692e-01 4.10388261e-01 -6.21140778e-01
2.73412764e-01 -3.22857559e-01 -6.42492548e-02 1.59792900e-01
2.84917355e-01 -1.04337049e+00 -5.11505008e-01 -4.57346678e-01
-8.91672194e-01 7.24364221e-02 6.33173347e-01 -7.77446851e-03
1.38897932e+00 -1.76899147e+00 3.34815681e-01 8.65045547e-01
9.76937860e-02 -1.27850138e-02 -2.36150593e-01 7.47751713e-01
-2.14636236e-01 1.52203083e-01 -4.18776274e-01 2.70523399e-01
1.83348164e-01 5.88129520e-01 -7.90657252e-02 3.42917383e-01
3.65741640e-01 8.94879103e-01 -8.49126697e-01 -3.42708468e-01
8.31531510e-02 -4.11479205e-01 -8.51362526e-01 -4.22822535e-02
-7.01220334e-01 1.70629188e-01 -5.67969501e-01 9.55250740e-01
4.50778186e-01 -2.66983449e-01 5.64669013e-01 5.66910923e-01
1.54047802e-01 2.45578989e-01 -1.15595043e+00 8.70444298e-01
-2.44762480e-01 2.12799713e-01 -4.61132228e-01 -1.09038162e+00
9.76457417e-01 7.16000319e-01 2.24321678e-01 -7.13804901e-01
1.82354748e-01 5.44866562e-01 5.07542670e-01 -5.72563529e-01
-2.29799926e-01 7.30841188e-03 -2.00099722e-01 7.75165617e-01
-5.38458712e-02 -1.53627964e-02 6.70785964e-01 -1.66781411e-01
2.19061089e+00 -3.73705387e-01 4.39468533e-01 -4.27081227e-01
6.97457790e-01 3.24391574e-01 6.44173980e-01 8.48584890e-01
2.39796251e-01 4.98093963e-02 1.28111744e+00 -5.52578717e-02
-7.40635574e-01 -8.99937570e-01 7.00057223e-02 -8.03861171e-02
-1.77566126e-01 -5.07041335e-01 -7.51206100e-01 -1.11047316e+00
-4.97611519e-03 9.86176431e-01 -8.08850169e-01 -5.52024901e-01
-2.62469620e-01 -5.87555230e-01 4.79410499e-01 3.34124804e-01
1.50281027e-01 -9.50088680e-01 -1.00235963e+00 4.05329704e-01
4.21550304e-01 -5.21483004e-01 3.02722991e-01 7.51360834e-01
-1.07974482e+00 -1.45921803e+00 1.47030875e-01 -3.22753906e-01
1.02915812e+00 -3.35122019e-01 1.18885875e+00 5.03162205e-01
-3.13506782e-01 3.20611537e-01 -6.62393272e-01 -2.60885239e-01
-9.20249999e-01 -2.10017890e-01 -3.42150062e-01 -2.50233114e-01
9.92658511e-02 -6.53347850e-01 -3.59144132e-03 5.78726470e-01
-1.06886816e+00 -1.23610109e-01 8.43372643e-01 8.77758384e-01
6.98713064e-01 4.73240852e-01 2.98506528e-01 -1.27431548e+00
3.99537057e-01 -5.16116202e-01 -9.98619318e-01 3.59914690e-01
-8.96668017e-01 3.78396720e-01 5.80478728e-01 -6.65096104e-01
-2.78671086e-01 4.53300700e-02 -1.40757859e-01 -3.52224290e-01
1.65408675e-03 6.36892974e-01 -5.25409043e-01 -1.85918167e-01
6.07809246e-01 7.92938098e-02 -2.95393139e-01 -9.46609974e-02
-3.53319049e-01 2.02939451e-01 1.19178779e-01 -4.79233027e-01
8.00896287e-01 6.03968352e-02 4.69971180e-01 -1.12340175e-01
-2.09747642e-01 1.26365632e-01 -2.66497731e-01 -4.65652719e-02
9.82355177e-02 -4.96872775e-02 -6.68988705e-01 -8.02442208e-02
-5.75476766e-01 -5.67162156e-01 -7.76606083e-01 1.19121678e-01
-6.12290680e-01 -4.66672182e-02 -3.97384614e-01 -9.14032400e-01
5.55033721e-02 -1.60617292e+00 8.57118726e-01 -3.06343347e-01
-5.53665400e-01 -8.20748866e-01 2.86721289e-01 -3.18697691e-01
8.83139744e-02 8.24253380e-01 1.54462349e+00 -1.09175026e+00
-4.93016154e-01 -7.42590010e-01 5.89864433e-01 3.54462415e-02
-5.83269335e-02 9.81998593e-02 -6.72581613e-01 -3.34491611e-01
4.25754845e-01 1.00314237e-01 2.45391533e-01 2.11895108e-01
8.54619205e-01 -5.10730445e-01 -6.87005520e-01 7.21642673e-02
1.70142281e+00 8.06227982e-01 9.25414026e-01 1.67107552e-01
5.24675436e-02 3.74309868e-01 7.81379700e-01 4.44827914e-01
-2.68031895e-01 7.26119757e-01 5.73017895e-01 1.73009694e-01
3.28554511e-01 1.13809660e-01 3.66757840e-01 3.86086136e-01
2.61662602e-01 -4.75778162e-01 -1.17014444e+00 4.56311584e-01
-1.66542101e+00 -7.83299744e-01 -2.97363937e-01 2.47262430e+00
6.58976197e-01 7.73033977e-01 2.50917792e-01 1.18641663e+00
6.03924811e-01 -8.27663898e-01 -4.92400199e-01 -8.28409910e-01
2.87121296e-01 4.16473508e-01 2.41611287e-01 5.60215712e-01
-3.63169879e-01 3.25597450e-02 6.27307272e+00 6.09319568e-01
-1.06206858e+00 9.70412120e-02 5.17297804e-01 -1.89077053e-02
-5.23217022e-01 4.01517421e-01 -5.37313998e-01 3.94173414e-01
1.52252614e+00 -2.60727942e-01 2.30768979e-01 9.07428086e-01
-9.22521949e-02 -2.89781898e-01 -1.86323941e+00 3.88059169e-01
-3.77895176e-01 -1.05741227e+00 -1.95545256e-01 5.40359199e-01
8.30762863e-01 -3.08151990e-01 -2.40768239e-01 3.07568163e-01
3.68744284e-01 -7.76549399e-01 9.53070462e-01 1.97069511e-01
7.72167385e-01 -8.34396839e-01 9.72919524e-01 3.33897084e-01
-7.09019423e-01 -4.76471305e-01 9.55903903e-02 -1.75145850e-01
-2.32648835e-01 9.75657463e-01 -1.11129320e+00 4.94010538e-01
7.10766464e-02 8.55770111e-02 -6.48263156e-01 1.19999468e+00
-3.18075955e-01 8.07894289e-01 -2.14583442e-01 -1.18600979e-01
1.14896730e-01 1.41465053e-01 6.00069165e-01 6.46279752e-01
4.75378364e-01 -2.24656120e-01 1.31725296e-01 1.02042270e+00
5.29425800e-01 -5.86443901e-01 -4.07523781e-01 -9.63978916e-02
6.04725957e-01 8.77048910e-01 -1.19532692e+00 -4.91363853e-01
-1.35155976e-01 6.94460034e-01 -2.33736560e-01 -1.46614220e-02
-8.76585126e-01 -1.06858052e-01 5.63331470e-02 4.14643288e-01
3.21596920e-01 2.25024879e-01 -1.57320604e-01 -5.72973788e-01
4.42656316e-02 -1.06658053e+00 3.56622815e-01 -4.30955291e-01
-6.62827671e-01 6.29156411e-01 2.83455521e-01 -1.30166769e+00
-8.36393237e-01 -3.64183724e-01 -7.21704662e-01 5.05101204e-01
-7.14346826e-01 -3.42797756e-01 3.64597254e-02 4.91623849e-01
3.22042108e-01 2.45802552e-01 1.01556754e+00 -8.58839601e-03
-6.61804795e-01 6.00144565e-01 -2.17566282e-01 -5.87576330e-01
-6.41313419e-02 -1.36075449e+00 3.68291706e-01 1.08397293e+00
-2.60172993e-01 1.28620207e-01 9.01653945e-01 -7.43887484e-01
-1.78904843e+00 -1.14098597e+00 8.14302683e-01 -2.86524761e-02
5.10958791e-01 -3.40100169e-01 -4.89834338e-01 7.26007283e-01
-2.94030100e-01 -9.99565870e-02 4.24376637e-01 -1.58339575e-01
4.19533066e-02 -1.30835891e-01 -1.70093119e+00 5.68593323e-01
9.33444023e-01 -1.36911169e-01 -1.06032483e-01 4.22066063e-01
5.95472634e-01 -2.08905235e-01 -6.94378972e-01 4.37268049e-01
-1.53794913e-02 -1.41093552e+00 2.58356124e-01 -7.74046034e-02
1.88596845e-01 -4.50757235e-01 -1.04523590e-03 -1.04490316e+00
-2.51863152e-01 -8.13338697e-01 -3.25856328e-01 9.79996622e-01
7.21890330e-01 -1.06730402e+00 6.01998568e-01 2.88622826e-01
-3.19990754e-01 -1.31075799e+00 -9.48600054e-01 -1.02909601e+00
-4.91020769e-01 -3.43886733e-01 9.02127385e-01 6.92322731e-01
4.05427456e-01 4.04892527e-02 4.34020907e-01 2.84676161e-02
2.05510214e-01 1.80151641e-01 3.02590281e-01 -1.40162766e+00
-1.29278910e+00 -3.69198501e-01 -9.93195236e-01 -1.98458061e-02
-2.44125530e-01 -7.20095634e-01 -5.28212115e-02 -1.23339605e+00
8.35258663e-02 -3.74525338e-01 -1.06014565e-01 8.83324325e-01
3.53392690e-01 -5.28484136e-02 -4.87522632e-01 -2.11224109e-01
-3.96928787e-01 -2.79358357e-01 8.72013986e-01 -1.94838420e-01
-2.95615196e-01 1.80581793e-01 -3.53224456e-01 2.93490559e-01
5.66684008e-01 -6.67554021e-01 -7.02387094e-01 2.33512148e-01
7.16774285e-01 8.42237592e-01 2.13597476e-01 -1.23985541e+00
1.91152722e-01 -1.02923408e-01 -5.71362004e-02 -4.17797714e-01
-2.02603161e-01 -8.45126450e-01 8.58930588e-01 1.36477315e+00
-1.96975663e-01 3.73894811e-01 1.96840510e-01 3.21564347e-01
-1.59094885e-01 -6.27044022e-01 4.23362106e-01 2.99737658e-02
-4.47723567e-01 -6.10807128e-02 -7.36169040e-01 -3.01926851e-01
1.37618971e+00 -5.75010300e-01 -3.42741370e-01 1.59725487e-01
-9.38127577e-01 9.74863768e-02 6.51527584e-01 -2.41181552e-01
5.96115828e-01 -8.97677064e-01 -3.11793476e-01 3.85983855e-01
1.83560804e-01 -1.78717628e-01 -5.57694845e-02 1.26096237e+00
-5.27042329e-01 5.36373317e-01 -8.87275562e-02 -5.52972913e-01
-1.12844276e+00 8.94141972e-01 4.42397743e-01 -6.45025551e-01
-4.44002926e-01 5.73265016e-01 -2.36156523e-01 -1.49986967e-01
-1.43024951e-01 -5.89877784e-01 2.07088128e-01 -4.01851147e-01
3.62600058e-01 5.23425400e-01 6.23543441e-01 4.77647185e-01
-4.83463466e-01 -7.81185701e-02 1.03377454e-01 -3.80672187e-01
1.37112808e+00 3.27823937e-01 -3.41282576e-01 4.53137696e-01
6.47594452e-01 3.15460153e-02 -7.29796708e-01 3.81831676e-01
4.01561797e-01 -6.69952929e-02 -1.17270328e-01 -1.10920465e+00
-9.87034559e-01 4.50196654e-01 5.28672993e-01 7.65836596e-01
1.58856606e+00 -1.66274831e-01 3.61547545e-02 4.67511833e-01
1.00626409e+00 -5.03568828e-01 -2.55456150e-01 1.19216070e-01
5.51078916e-01 -2.94082165e-01 -1.66685104e-01 -5.60679436e-01
-4.61202860e-02 1.03534794e+00 4.49562788e-01 -2.61248499e-01
2.32149750e-01 1.09832954e+00 -5.16790390e-01 -2.81359076e-01
-1.36531663e+00 -6.23097196e-02 -1.50387242e-01 5.81544161e-01
-1.87257707e-01 6.45678490e-02 -3.95115286e-01 6.48213923e-01
2.34953705e-02 1.51783466e-01 6.85546219e-01 1.33798456e+00
-4.00225163e-01 -1.53621018e+00 -5.07232308e-01 7.88738728e-01
7.62709826e-02 1.76702306e-01 -8.07060122e-01 1.15504217e+00
4.03549582e-01 1.12578547e+00 2.94688940e-02 -7.02746689e-01
2.81897366e-01 3.55504304e-02 9.45065439e-01 -8.86905015e-01
-5.60418427e-01 3.57206725e-02 3.95954132e-01 -7.33762801e-01
3.11186276e-02 -7.21625865e-01 -1.23228323e+00 -1.20075479e-01
-6.46065474e-01 3.99762064e-01 2.64604449e-01 1.24960685e+00
1.97467148e-01 8.86771500e-01 6.62186265e-01 -3.30132216e-01
-5.74391484e-01 -7.31684804e-01 -7.50008285e-01 -4.31217641e-01
-1.77856922e-01 -6.91101670e-01 -4.04987961e-01 -5.14010727e-01] | [5.389865875244141, 2.7074971199035645] |
d00cd722-c8c7-4d4f-bf31-a200d7c22c3c | real-time-covid-19-diagnosis-from-x-ray | 2106.01435 | null | https://arxiv.org/abs/2106.01435v1 | https://arxiv.org/pdf/2106.01435v1.pdf | Real-Time COVID-19 Diagnosis from X-Ray Images Using Deep CNN and Extreme Learning Machines Stabilized by Chimp Optimization Algorithm | Real-time detection of COVID-19 using radiological images has gained priority due to the increasing demand for fast diagnosis of COVID-19 cases. This paper introduces a novel two-phase approach for classifying chest X-ray images. Deep Learning (DL) methods fail to cover these aspects since training and fine-tuning the model's parameters consume much time. In this approach, the first phase comes to train a deep CNN working as a feature extractor, and the second phase comes to use Extreme Learning Machines (ELMs) for real-time detection. The main drawback of ELMs is to meet the need of a large number of hidden-layer nodes to gain a reliable and accurate detector in applying image processing since the detective performance remarkably depends on the setting of initial weights and biases. Therefore, this paper uses Chimp Optimization Algorithm (ChOA) to improve results and increase the reliability of the network while maintaining real-time capability. The designed detector is to be benchmarked on the COVID-Xray-5k and COVIDetectioNet datasets, and the results are verified by comparing it with the classic DCNN, Genetic Algorithm optimized ELM (GA-ELM), Cuckoo Search optimized ELM (CS-ELM), and Whale Optimization Algorithm optimized ELM (WOA-ELM). The proposed approach outperforms other comparative benchmarks with 98.25% and 99.11% as ultimate accuracy on the COVID-Xray-5k and COVIDetectioNet datasets, respectively, and it led relative error to reduce as the amount of 1.75% and 1.01% as compared to a convolutional CNN. More importantly, the time needed for training deep ChOA-ELM is only 0.9474 milliseconds, and the overall testing time for 3100 images is 2.937 seconds. | ['Tarik A. Rashid', 'Sarkhel H. Taher Karim', 'Gholam-Reza Parvizi', 'Mokhtar Mohammadi', 'Mohammad Khishe', 'Hu Tianqing'] | 2021-05-14 | null | null | null | null | ['covid-19-detection'] | ['medical'] | [-9.17725265e-02 -8.78148153e-02 1.94476992e-01 -2.04216525e-01
-5.15042722e-01 7.85099044e-02 1.12477586e-01 3.72475684e-01
-1.07786763e+00 6.06027007e-01 -5.69728494e-01 -4.12200093e-01
-5.83417356e-01 -6.72838986e-01 -3.50510240e-01 -1.08025587e+00
-3.54228020e-02 7.08591163e-01 9.93234441e-02 1.02620348e-02
1.24226153e-01 6.88367844e-01 -1.24554718e+00 1.16732813e-01
8.41777563e-01 1.19362247e+00 4.54755872e-01 9.02953923e-01
8.21629837e-02 7.76992679e-01 -6.29265547e-01 -1.92607448e-01
-2.71312390e-02 -2.41085589e-01 -4.80934411e-01 -4.71106768e-01
-1.54271960e-01 -1.17278792e-01 -5.96681312e-02 8.86695385e-01
1.20869887e+00 -8.53979066e-02 8.96150708e-01 -9.75849569e-01
-8.24460164e-02 4.45595324e-01 -8.22020411e-01 3.52549046e-01
-2.28994161e-01 1.48418799e-01 4.88216102e-01 -8.11720371e-01
3.52913231e-01 7.59880126e-01 9.55203593e-01 5.46184599e-01
-5.01115143e-01 -6.06350839e-01 -5.51526785e-01 4.58787411e-01
-1.53271389e+00 2.25335076e-01 4.48616892e-01 -2.98008591e-01
1.24917269e+00 3.96904379e-01 6.66431487e-01 7.32249379e-01
8.88191462e-01 6.26374006e-01 8.13783348e-01 -7.08566606e-01
4.74063814e-01 2.24373341e-01 1.71483994e-01 9.78914618e-01
4.83981967e-01 1.10347539e-01 -1.01732448e-01 -8.25969651e-02
5.11556029e-01 1.93554580e-01 -1.31764650e-01 1.16245061e-01
-7.03857899e-01 9.77999330e-01 6.58401787e-01 6.14048600e-01
-5.88244617e-01 -8.74847639e-03 6.33629978e-01 -1.67642850e-02
6.45067245e-02 6.56486392e-01 -2.69679725e-01 4.45349403e-02
-8.68458211e-01 -6.24357350e-02 6.13975585e-01 4.05535966e-01
-4.60002683e-02 1.26767248e-01 -1.10149920e-01 8.69329154e-01
2.07246691e-01 3.49210709e-01 1.00148451e+00 -2.42804125e-01
1.46622837e-01 7.15627968e-01 -3.32002312e-01 -9.07015026e-01
-1.04786026e+00 -8.71087611e-01 -1.21169841e+00 3.16673934e-01
1.40506387e-01 -2.98729897e-01 -1.15993428e+00 1.28236973e+00
2.14431152e-01 -1.67914599e-01 7.46400058e-02 9.78181005e-01
9.03973639e-01 6.64616346e-01 6.68381229e-02 -2.03959972e-01
1.62233770e+00 -8.28539431e-01 -6.01426840e-01 2.96772178e-02
8.19718897e-01 -6.60610080e-01 7.24850118e-01 6.04470909e-01
-9.22812760e-01 -5.86619198e-01 -1.43851495e+00 3.83323640e-01
-3.35496932e-01 4.09173518e-01 4.14409697e-01 7.16701031e-01
-7.98659027e-01 3.75401348e-01 -8.03358912e-01 -1.90049469e-01
4.17755038e-01 7.80490518e-01 3.54611408e-03 7.90909454e-02
-1.12584519e+00 1.02610040e+00 8.23431432e-01 3.19486439e-01
-8.59043479e-01 -5.82604229e-01 -3.99291843e-01 1.88066795e-01
1.95499182e-01 -4.46295470e-01 1.04245639e+00 -7.93951929e-01
-1.32434714e+00 7.77482092e-01 4.76898998e-01 -6.20833397e-01
5.50998032e-01 1.19859315e-01 -4.90980893e-01 2.53275037e-01
-3.17013144e-01 5.46336412e-01 6.67049766e-01 -8.00574541e-01
-7.23560333e-01 -3.12274814e-01 -2.89085239e-01 6.80780858e-02
-4.29961920e-01 1.18554130e-01 -3.49909961e-01 -4.40017819e-01
7.26006925e-02 -7.89395750e-01 -2.72145092e-01 -2.65600532e-01
-3.67252767e-01 -2.64727920e-01 9.35474992e-01 -6.10531509e-01
1.24888074e+00 -2.01288915e+00 -2.13046715e-01 3.28910977e-01
3.94817024e-01 7.02403069e-01 3.36863279e-01 -3.23345289e-02
-4.00216311e-01 -2.58486450e-01 -1.98921502e-01 -2.51047742e-02
-3.01992118e-01 2.05371499e-01 5.65427423e-01 6.18151724e-01
1.27251923e-01 9.28373575e-01 -4.22683865e-01 -9.74551260e-01
2.35650599e-01 5.36238194e-01 -4.51931477e-01 2.37653792e-01
-2.50693504e-02 1.61977615e-02 -3.56035054e-01 7.34157622e-01
6.10484481e-01 -3.12676251e-01 6.48776144e-02 -1.99644372e-01
-2.50993446e-02 -5.35307527e-01 -1.33887970e+00 1.15423262e+00
-5.33381760e-01 5.08058548e-01 -2.47427389e-01 -1.19452310e+00
9.44200277e-01 4.09812659e-01 6.74521148e-01 -9.02712643e-01
8.35181594e-01 4.53938186e-01 3.38404834e-01 -9.63751554e-01
1.21417388e-01 -1.14606149e-01 -5.95656037e-02 4.03183937e-01
1.27622321e-01 3.46505523e-01 -2.62984242e-02 -1.16325155e-01
8.58734429e-01 -3.91795367e-01 4.70208406e-01 -2.50659496e-01
7.68260300e-01 2.07045317e-01 5.22267759e-01 8.71632934e-01
1.89673789e-02 3.88131738e-01 3.23535651e-01 -7.90721655e-01
-9.83639359e-01 -6.94607437e-01 -5.08178949e-01 5.95904946e-01
-8.22307393e-02 4.60070297e-02 -8.63741457e-01 -4.84707147e-01
-3.21910292e-01 7.49449015e-01 -5.37678123e-01 -4.17177796e-01
-8.26813161e-01 -1.44261169e+00 6.38089299e-01 5.96302629e-01
5.90880394e-01 -1.20814323e+00 -1.24623501e+00 3.23394358e-01
2.43603587e-01 -7.53694654e-01 1.92130357e-01 5.35988092e-01
-7.12450325e-01 -1.01138878e+00 -7.52373040e-01 -8.33087027e-01
6.83143139e-01 -5.00054061e-01 8.64704072e-01 4.19577569e-01
-1.06058991e+00 -1.10803440e-01 -2.35013708e-01 -7.65425801e-01
-4.49545056e-01 2.17645735e-01 -3.40552926e-02 -1.03018448e-01
4.13243771e-01 -5.47022969e-02 -5.78846812e-01 7.90031105e-02
-8.25994551e-01 -1.54630402e-02 8.59903157e-01 1.22377062e+00
5.78142583e-01 2.64129698e-01 4.28609490e-01 -6.22517467e-01
6.27191365e-01 -3.43168557e-01 -7.11534798e-01 3.14208955e-01
-1.05743182e+00 -8.07517841e-02 8.58619094e-01 -4.69224602e-01
-8.52217138e-01 1.16534472e-01 -3.39855254e-01 -4.60775077e-01
8.43651146e-02 4.45601255e-01 1.85102433e-01 -3.11708838e-01
7.32928336e-01 4.52999212e-02 3.72623801e-02 -2.26782247e-01
-3.13406080e-01 1.00706041e+00 5.24498403e-01 -1.05637968e-01
2.63383359e-01 1.71161041e-01 1.84501365e-01 -6.50491476e-01
-4.96743649e-01 -3.18538100e-01 -3.53116810e-01 -4.69334275e-01
1.18238926e+00 -5.06693661e-01 -1.03086805e+00 6.89072251e-01
-8.42418194e-01 7.25332350e-02 -1.42595276e-01 8.84068310e-01
-3.87111664e-01 7.21816719e-02 -6.81269169e-01 -7.99440980e-01
-9.96688783e-01 -1.54373085e+00 5.98472714e-01 4.02303636e-01
5.08251749e-02 -7.95617104e-01 -2.43783370e-01 7.25970343e-02
3.96903276e-01 4.94205534e-01 1.14148438e+00 -8.75277340e-01
-1.99770287e-01 -6.31272554e-01 -1.08041808e-01 2.59081393e-01
-3.87010485e-01 2.68916935e-02 -7.90687025e-01 -3.92649144e-01
4.51167434e-01 -3.31745565e-01 6.10453308e-01 6.57978833e-01
1.24309349e+00 8.12257454e-02 -4.65967894e-01 9.15050805e-01
1.90435016e+00 7.63080359e-01 5.77081263e-01 6.78645849e-01
4.40693796e-01 6.51251376e-02 6.08416021e-01 3.79419297e-01
-1.45248950e-01 4.34047282e-01 6.90500557e-01 -4.73623961e-01
6.48178086e-02 3.37236196e-01 -3.58070672e-01 8.72144461e-01
-6.06788844e-02 -4.91733849e-01 -1.31395280e+00 3.33787382e-01
-1.55494487e+00 -6.10297143e-01 -3.48113596e-01 1.78875744e+00
5.05710721e-01 3.72263640e-01 -3.22091669e-01 5.25777340e-01
6.77533388e-01 -2.73051232e-01 -6.32047534e-01 -7.16001809e-01
4.78660129e-02 4.63330269e-01 5.47250152e-01 1.79343104e-01
-9.59178746e-01 2.57356673e-01 4.97865629e+00 9.96847689e-01
-1.46433628e+00 2.76198745e-01 6.53163612e-01 -3.31383437e-01
5.85354507e-01 -5.63857973e-01 -8.11538815e-01 5.42121589e-01
9.99929607e-01 2.72061020e-01 -3.72833177e-03 9.04763997e-01
-3.92247662e-02 -2.79719383e-01 -7.86837876e-01 1.34521544e+00
2.09989280e-01 -1.15655649e+00 -4.39565748e-01 -8.75895992e-02
5.06124377e-01 9.99801755e-02 1.74705870e-02 3.27585071e-01
-2.33199999e-01 -1.13841271e+00 4.12485093e-01 4.99265254e-01
7.45524943e-01 -1.08127773e+00 1.42570722e+00 5.02314091e-01
-7.86169291e-01 -4.42386329e-01 -5.70502758e-01 3.77817363e-01
-3.25404690e-03 5.77223182e-01 -1.32861495e+00 5.75019717e-01
9.60270286e-01 -2.20446140e-01 -4.42821413e-01 1.14215076e+00
4.60229889e-02 5.31696856e-01 -5.84536791e-01 -6.27239645e-01
5.45315087e-01 2.52821296e-01 3.18663985e-01 1.20053434e+00
2.76960075e-01 5.58441691e-02 -1.53468847e-01 4.86148447e-01
1.90634429e-01 3.73717606e-01 4.98895235e-02 2.19641596e-01
1.68864295e-01 1.37657309e+00 -9.70870197e-01 -3.97818297e-01
-2.46240087e-02 6.52031720e-01 8.29363167e-02 -1.79705739e-01
-1.28011096e+00 -6.60473466e-01 -4.48326729e-02 -8.00055787e-02
2.57018358e-01 2.27734864e-01 -3.77285510e-01 -6.13760412e-01
-1.39219165e-01 -8.31755519e-01 7.77779520e-01 -7.15100765e-01
-7.66876698e-01 9.84441757e-01 -1.60441458e-01 -1.17600644e+00
-3.66135299e-01 -8.47626150e-01 -5.13734341e-01 8.74282658e-01
-1.34155667e+00 -7.40153670e-01 -5.50549328e-01 4.48465407e-01
4.31843340e-01 -3.88376594e-01 7.79005170e-01 3.31147999e-01
-8.93959165e-01 8.43209386e-01 2.69187033e-01 1.67112529e-01
2.90625632e-01 -1.04155302e+00 -1.94551095e-01 5.89443147e-01
-2.50665694e-01 8.36058259e-02 4.70719784e-01 -2.99897879e-01
-1.23640692e+00 -9.32670712e-01 7.24787474e-01 9.45274830e-02
2.14582458e-02 -2.50974387e-01 -6.88730955e-01 1.15962356e-01
1.28373340e-01 2.82482207e-02 4.53442007e-01 -4.94457632e-01
3.32331032e-01 -1.63057044e-01 -1.36463106e+00 2.30035782e-01
4.92821842e-01 2.26969328e-02 -4.87526089e-01 4.93505388e-01
3.98009896e-01 -6.26158714e-01 -9.58791196e-01 7.36217558e-01
4.77643818e-01 -8.66551876e-01 8.70505154e-01 -4.22263831e-01
2.46482745e-01 -9.80071202e-02 1.57485858e-01 -9.10226047e-01
-3.34913313e-01 3.40793747e-03 9.09535959e-02 5.66179574e-01
4.53816384e-01 -7.82133937e-01 6.92785084e-01 1.48448735e-01
-1.35961771e-01 -1.54365158e+00 -8.68663430e-01 -5.70704162e-01
-2.66261965e-01 -3.52579176e-01 4.69687641e-01 9.49403286e-01
-4.38107222e-01 6.58229366e-02 -6.45751879e-02 1.91390485e-01
4.70748931e-01 4.82866392e-02 2.35348627e-01 -1.07032263e+00
-5.83736598e-01 -3.68435532e-01 -6.69242144e-01 -7.98754394e-02
-4.97008741e-01 -8.96452785e-01 -2.17462659e-01 -1.36560762e+00
1.88177794e-01 -5.06334364e-01 -5.87028742e-01 3.96859109e-01
1.80726647e-02 1.81631327e-01 3.11027616e-02 1.73738018e-01
-3.01735282e-01 1.81477487e-01 9.74588215e-01 2.86820848e-02
-1.08234212e-01 -2.04463745e-03 -1.81284502e-01 8.06845367e-01
9.68815684e-01 -9.03420806e-01 -2.73233086e-01 -3.45195025e-01
1.98456734e-01 2.59630233e-01 2.17307627e-01 -1.42810154e+00
4.92232502e-01 1.98773026e-01 9.14647579e-01 -9.59929347e-01
2.54452676e-01 -8.06641757e-01 2.44794592e-01 1.18517339e+00
3.38759907e-02 2.86066025e-01 8.98858905e-02 9.63064209e-02
-1.80606142e-01 -8.46614480e-01 1.17991734e+00 -2.22438231e-01
-7.54801273e-01 7.26435035e-02 -3.27110529e-01 -1.76252261e-01
1.57684183e+00 -4.93909478e-01 -1.23975024e-01 1.55207559e-01
-6.89941406e-01 1.08295120e-01 -8.32568035e-02 9.70911086e-02
7.71521032e-01 -9.22186553e-01 -6.57425284e-01 4.11150098e-01
-1.05514608e-01 1.05767183e-01 4.75591809e-01 1.14282882e+00
-1.13097262e+00 4.18440849e-01 -2.71732420e-01 -9.37082171e-01
-1.26638019e+00 6.01898849e-01 8.86691272e-01 -4.81184632e-01
-7.30032325e-01 1.11082494e+00 -2.35643491e-01 -1.71178013e-01
4.22786653e-01 -3.32228430e-02 -3.62075925e-01 -3.20582315e-02
3.38697493e-01 4.36491966e-01 5.53736627e-01 -2.62557089e-01
-5.28094828e-01 5.59283257e-01 -3.15214545e-01 3.14831644e-01
1.56877661e+00 5.12398601e-01 -1.21149652e-01 -1.46474525e-01
1.47380447e+00 -4.61171687e-01 -6.38802290e-01 1.21080302e-01
-1.25679579e-02 1.06117487e-01 4.01600778e-01 -1.02329791e+00
-1.14957333e+00 8.38438809e-01 1.28288817e+00 -7.40026832e-02
1.56083310e+00 -6.52926713e-02 7.61321783e-01 4.32738453e-01
9.92728546e-02 -1.17714286e+00 6.78813532e-02 2.10645244e-01
6.72498703e-01 -8.88289392e-01 1.84946790e-01 1.70988783e-01
-5.09774983e-01 1.35440350e+00 7.84739792e-01 -1.59914017e-01
7.08328605e-01 6.55922592e-01 5.78183867e-03 -6.29083872e-01
-6.47103906e-01 2.63684332e-01 -2.97552533e-02 1.49669066e-01
1.48257449e-01 7.77880251e-02 -5.79872727e-01 5.64423978e-01
-1.50573075e-01 4.46665622e-02 2.22697034e-01 1.06187904e+00
-5.06697536e-01 -5.45414746e-01 -5.70943356e-01 6.68799818e-01
-8.27828288e-01 1.36075750e-01 1.02903612e-01 1.07329214e+00
4.10111308e-01 4.67450947e-01 1.07567526e-01 -3.48058283e-01
3.05296451e-01 -8.29171315e-02 2.67732531e-01 -7.27418214e-02
-7.98051000e-01 1.06569707e-01 -2.14603633e-01 -3.06662083e-01
-1.41145423e-01 -2.77062923e-01 -1.53387272e+00 -4.25374471e-02
-7.38083959e-01 2.24657446e-01 1.10045981e+00 7.56207228e-01
-5.22620557e-03 9.40208316e-01 6.04466915e-01 -2.57329226e-01
-7.84778297e-01 -8.74727011e-01 -4.61308032e-01 -4.33469824e-02
4.92789922e-03 -6.00333154e-01 -1.27171785e-01 -4.33595121e-01] | [14.951973915100098, -2.518519639968872] |
02cc0d32-49a6-4b98-8e47-1971e0f35a12 | a-fast-and-accurate-physics-informed-neural | 2009.11990 | null | https://arxiv.org/abs/2009.11990v2 | https://arxiv.org/pdf/2009.11990v2.pdf | A fast and accurate physics-informed neural network reduced order model with shallow masked autoencoder | Traditional linear subspace reduced order models (LS-ROMs) are able to accelerate physical simulations, in which the intrinsic solution space falls into a subspace with a small dimension, i.e., the solution space has a small Kolmogorov n-width. However, for physical phenomena not of this type, e.g., any advection-dominated flow phenomena, such as in traffic flow, atmospheric flows, and air flow over vehicles, a low-dimensional linear subspace poorly approximates the solution. To address cases such as these, we have developed a fast and accurate physics-informed neural network ROM, namely nonlinear manifold ROM (NM-ROM), which can better approximate high-fidelity model solutions with a smaller latent space dimension than the LS-ROMs. Our method takes advantage of the existing numerical methods that are used to solve the corresponding full order models. The efficiency is achieved by developing a hyper-reduction technique in the context of the NM-ROM. Numerical results show that neural networks can learn a more efficient latent space representation on advection-dominated data from 1D and 2D Burgers' equations. A speedup of up to 2.6 for 1D Burgers' and a speedup of 11.7 for 2D Burgers' equations are achieved with an appropriate treatment of the nonlinear terms through a hyper-reduction technique. Finally, a posteriori error bounds for the NM-ROMs are derived that take account of the hyper-reduced operators. | ['Youngsoo Choi', 'David Widemann', 'Youngkyu Kim', 'Tarek Zohdi'] | 2020-09-25 | null | null | null | null | ['physical-simulations'] | ['miscellaneous'] | [-1.27803847e-01 -2.01860338e-01 1.58468008e-01 1.75217673e-01
-3.43782604e-01 3.38068642e-02 5.31518102e-01 -3.04281980e-01
-2.80036211e-01 8.83424401e-01 -7.93271586e-02 -4.50952321e-01
-4.27415401e-01 -9.21362638e-01 -7.55826354e-01 -1.04594183e+00
-2.01833695e-01 8.33904445e-01 -1.00731671e-01 -4.22429651e-01
1.50049385e-02 8.68554235e-01 -1.36507499e+00 -3.78130138e-01
1.12069738e+00 6.47855043e-01 -1.75016209e-01 7.61568844e-01
-1.49734914e-01 4.26280111e-01 5.57065122e-02 4.09875184e-01
5.15586913e-01 -5.04769564e-01 -5.48126578e-01 -3.03119004e-01
4.63070303e-01 -4.36105847e-01 -9.20704544e-01 1.11896896e+00
2.91015416e-01 7.91881442e-01 8.52950692e-01 -1.19254386e+00
-3.38205665e-01 2.14051113e-01 -3.06702554e-01 3.93446870e-02
-3.98736387e-01 2.84089327e-01 4.72328126e-01 -1.00994265e+00
6.69977069e-01 1.47262132e+00 8.38011980e-01 7.34007061e-01
-1.43110454e+00 -5.13287008e-01 -4.39146072e-01 -5.68594784e-02
-1.54042876e+00 -2.93286443e-01 9.53914106e-01 -5.96696615e-01
9.23144519e-01 4.45155799e-01 5.48236370e-01 6.12891912e-01
5.09051144e-01 3.90384704e-01 1.10927451e+00 -3.85815382e-01
4.03606892e-01 9.77523476e-02 6.05279624e-01 6.31731451e-01
5.74958146e-01 4.80346620e-01 -3.11949193e-01 -5.58995903e-01
8.79646063e-01 -1.56236947e-01 -3.83564532e-01 -6.03664279e-01
-1.05462551e+00 1.04470170e+00 3.00220132e-01 1.98523685e-01
-4.79617983e-01 5.58412969e-02 2.29272440e-01 -1.58038035e-01
6.98518634e-01 4.53907758e-01 -1.56564802e-01 -8.06023106e-02
-1.14198589e+00 6.00239635e-01 1.06480730e+00 7.11197436e-01
9.26183045e-01 4.22398627e-01 2.38223299e-01 4.27994192e-01
3.56973201e-01 9.20747280e-01 2.18784720e-01 -1.51014102e+00
2.04462901e-01 2.04916552e-01 4.13870782e-01 -8.47519040e-01
-5.33017337e-01 -4.48916018e-01 -1.64868605e+00 4.72104549e-01
3.78558815e-01 -2.91247796e-02 -6.08668268e-01 1.76826906e+00
6.09324574e-01 2.94691443e-01 5.22073030e-01 1.08056426e+00
4.04681951e-01 1.23703945e+00 -4.04764205e-01 -6.20133042e-01
1.00783801e+00 -8.60941350e-01 -8.16044629e-01 6.13812171e-02
8.92173529e-01 -3.81518126e-01 8.75612438e-01 2.24578097e-01
-1.20574868e+00 -4.61634427e-01 -9.81333017e-01 -6.36001676e-02
-1.99356303e-01 -2.89241135e-01 5.20079136e-01 3.54620755e-01
-1.05077970e+00 1.32213521e+00 -1.14399838e+00 -6.60376623e-02
8.07903055e-03 1.40331015e-01 -2.74503708e-01 2.38120303e-01
-1.37558830e+00 7.12376177e-01 1.62811339e-01 4.37111467e-01
-6.30387068e-01 -1.01407540e+00 -7.32774258e-01 -5.01624644e-02
-8.22017565e-02 -9.71051872e-01 7.90913999e-01 -2.18929142e-01
-1.53304529e+00 2.00847805e-01 -4.18981791e-01 -4.64618593e-01
5.54871678e-01 -2.76482582e-01 -2.65721798e-01 3.17352325e-01
6.70787022e-02 2.29317442e-01 7.29001105e-01 -1.10759628e+00
9.89927277e-02 -1.13543153e-01 -2.46180534e-01 7.65522271e-02
-4.46598083e-01 -4.44949210e-01 1.28256157e-01 -4.22786206e-01
2.96810269e-01 -1.51542032e+00 -6.82226419e-01 1.54991476e-02
-2.08840251e-01 -8.15877095e-02 1.00494444e+00 -5.96741498e-01
1.03024030e+00 -2.06417298e+00 5.34156859e-01 1.41962558e-01
4.80426192e-01 4.28132445e-01 1.98676616e-01 4.74567324e-01
-1.28793523e-01 8.71118531e-02 -6.72486246e-01 -5.79747438e-01
1.02829657e-01 3.54510993e-01 -6.91538155e-01 7.25112319e-01
-1.03275456e-01 5.47227740e-01 -6.42029166e-01 -3.16761941e-01
2.32775524e-01 7.78453171e-01 -7.09515810e-01 2.67254598e-02
7.77904466e-02 6.39948130e-01 -4.26947564e-01 -2.90987641e-01
1.05514944e+00 -4.21916068e-01 -3.88867497e-01 -1.07614689e-01
-4.98622030e-01 9.89486650e-02 -1.28531098e+00 1.38884437e+00
-6.05166912e-01 7.45827019e-01 5.18809974e-01 -9.32753384e-01
5.99425316e-01 1.73777878e-01 6.11077964e-01 -2.45178297e-01
-1.64547428e-01 4.58007455e-01 -6.58390895e-02 -3.27984184e-01
6.23569787e-01 -4.22977537e-01 3.20854247e-01 4.35489893e-01
-4.72997129e-01 -1.60713777e-01 2.04567671e-01 5.33959031e-01
9.47800398e-01 -6.37850314e-02 -2.02829018e-02 -7.15070844e-01
6.71996474e-01 3.18079799e-01 5.35823047e-01 7.55500972e-01
-4.66003232e-02 2.54322767e-01 9.74148735e-02 -5.71374893e-01
-1.30962896e+00 -8.27198565e-01 -6.22226536e-01 1.93998918e-01
5.53408787e-02 -3.62811744e-01 -7.65892625e-01 1.38230726e-01
1.18648216e-01 8.00147951e-01 -2.21065030e-01 -4.47385460e-01
-1.05973101e+00 -8.22982609e-01 4.35914218e-01 2.53429949e-01
7.05581546e-01 -7.14288235e-01 -3.62763077e-01 9.34230760e-02
-1.25334924e-02 -1.09980738e+00 -2.08476171e-01 -1.47877932e-01
-1.37169230e+00 -7.05434740e-01 -6.05939984e-01 -2.69344926e-01
5.70766389e-01 1.32023588e-01 6.83871806e-01 -1.44456819e-01
-2.00915843e-01 2.17511028e-01 3.48946124e-01 2.13553444e-01
-8.30184340e-01 -6.46710917e-02 8.32691312e-01 -2.53411025e-01
9.11888480e-02 -7.89549828e-01 -4.01190788e-01 9.54605415e-02
-8.77330363e-01 3.44795644e-01 4.10820782e-01 9.11394835e-01
4.39955354e-01 4.39809680e-01 4.77746278e-02 -5.49376011e-01
5.27783811e-01 -4.80909169e-01 -9.28556919e-01 -3.85154665e-01
-7.91547716e-01 4.94728923e-01 9.77615952e-01 -6.30050659e-01
-1.01157260e+00 -1.19555429e-01 -8.51903781e-02 -9.62119699e-01
-6.34797439e-02 4.22979802e-01 2.00461030e-01 -2.85097450e-01
7.00813115e-01 3.63648713e-01 2.26196364e-01 -6.65547609e-01
1.47378668e-01 5.66430926e-01 3.69894117e-01 -6.78769290e-01
1.24012160e+00 6.81957722e-01 8.87857437e-01 -1.36970234e+00
-4.49115783e-01 -4.07127231e-01 -4.38189566e-01 -1.11857332e-01
6.58296883e-01 -6.19243264e-01 -8.89756620e-01 5.57833254e-01
-1.21440279e+00 -4.28801149e-01 -6.86872840e-01 1.11484873e+00
-6.23455763e-01 5.91463685e-01 -1.22754908e+00 -1.20314741e+00
-2.94523239e-01 -1.20497978e+00 6.66366875e-01 -8.15411471e-03
2.44487021e-02 -1.18290353e+00 3.96889299e-01 1.55783772e-01
7.27888882e-01 2.25755185e-01 1.07803273e+00 -1.57853648e-01
-7.34871805e-01 1.01224594e-01 -3.65767591e-02 3.09525311e-01
-2.44540364e-01 -1.26557812e-01 -6.89112663e-01 -4.56667840e-01
6.43298388e-01 2.19831824e-01 9.30061281e-01 6.30694568e-01
6.96003497e-01 -5.93276501e-01 -4.70587969e-01 8.37861538e-01
1.18619370e+00 -1.71489820e-01 2.09676266e-01 -8.06518123e-02
9.58775818e-01 7.26023853e-01 2.08590895e-01 1.71951801e-01
-4.85060364e-02 6.03111744e-01 1.04364745e-01 -5.00479862e-02
1.28073143e-02 4.60538529e-02 4.13253874e-01 1.46645057e+00
-1.83030277e-01 2.01319888e-01 -1.00909054e+00 2.25361362e-01
-1.80940259e+00 -6.72201335e-01 -8.58389735e-01 2.26211715e+00
7.05339909e-01 5.79980165e-02 -9.47304145e-02 1.11848705e-01
6.02800310e-01 3.67188156e-02 -8.63086820e-01 -3.30340236e-01
-7.36633092e-02 -3.27391066e-02 7.30699599e-01 1.01028955e+00
-8.28605652e-01 5.85314691e-01 6.31808233e+00 8.96427393e-01
-1.00250280e+00 3.15823555e-01 1.22738548e-01 5.34960534e-03
-2.62670666e-01 2.63907671e-01 -8.46421301e-01 3.06178182e-01
1.36942923e+00 -2.31176153e-01 6.29615426e-01 8.13696921e-01
5.91772556e-01 2.19773248e-01 -6.73346221e-01 1.06274641e+00
-3.10015261e-01 -1.54726231e+00 1.25563428e-01 6.32298231e-01
8.54932547e-01 8.69622007e-02 -1.33628234e-01 2.14299798e-01
1.98128950e-02 -7.58101225e-01 3.26882213e-01 7.13140666e-01
8.20984483e-01 -6.56111062e-01 4.37685817e-01 8.40593338e-01
-1.11446977e+00 2.92084038e-01 -5.57129860e-01 -2.69055843e-01
5.60358465e-01 9.85709488e-01 -5.59786800e-03 3.84214610e-01
4.05535102e-01 7.38420963e-01 8.95056203e-02 6.52973950e-01
3.86349648e-01 7.04471588e-01 -7.58050919e-01 2.01075003e-01
2.36183271e-01 -9.38512266e-01 1.29000628e+00 8.46184969e-01
4.82378602e-01 4.78684098e-01 -5.35152070e-02 1.33597255e+00
2.40580350e-01 -6.49565309e-02 -6.96476042e-01 -6.89063147e-02
-1.14504257e-02 1.18593574e+00 -5.26392639e-01 -4.94527966e-01
1.06793933e-01 5.21887004e-01 -9.34964567e-02 7.83349454e-01
-7.39910245e-01 -3.21756154e-01 1.01816809e+00 2.89225101e-01
-5.94251715e-02 -8.46934497e-01 -1.13951713e-02 -1.36727214e+00
-8.61799940e-02 -4.42510992e-01 -8.86697024e-02 -5.90043247e-01
-1.28157759e+00 5.60174763e-01 4.15328026e-01 -1.21022260e+00
-4.04136568e-01 -9.92783427e-01 -4.70529616e-01 1.10466230e+00
-1.27499330e+00 -6.63809836e-01 -1.73579276e-01 4.29723352e-01
1.14268437e-01 -4.50734468e-03 7.46192694e-01 3.50098133e-01
-7.40168273e-01 -2.00796183e-02 7.28646159e-01 -1.75731763e-01
1.26465976e-01 -1.18998849e+00 4.33782578e-01 9.03679252e-01
-3.48878682e-01 9.13102984e-01 1.10018194e+00 -8.02802682e-01
-1.73123705e+00 -1.02016282e+00 7.50277877e-01 -5.43779954e-02
7.37680733e-01 -2.13175014e-01 -1.37104452e+00 3.99387240e-01
-1.65267006e-01 1.94359228e-01 3.36044997e-01 -2.05398187e-01
2.56209467e-02 5.89946192e-03 -9.17413712e-01 5.19550920e-01
1.14398897e+00 -6.59577429e-01 -3.55107069e-01 7.16002166e-01
7.30303407e-01 -5.61424911e-01 -8.41129422e-01 4.63400275e-01
2.86002487e-01 -6.74153030e-01 1.22660995e+00 -8.01102042e-01
3.03884894e-01 -3.36341709e-01 7.20709711e-02 -1.13973582e+00
-3.38467717e-01 -1.04034615e+00 -6.97463810e-01 7.26578355e-01
-1.23161122e-01 -9.75747883e-01 6.31344080e-01 7.20041871e-01
-1.70655727e-01 -6.43193901e-01 -1.16712034e+00 -1.11280787e+00
6.67965353e-01 -5.64633131e-01 3.60101879e-01 9.70522285e-01
-1.96628109e-01 1.14953682e-01 -4.85855937e-01 4.50106770e-01
1.07756007e+00 2.52245277e-01 5.99622309e-01 -1.44955707e+00
-2.27408633e-01 -3.94936711e-01 -6.04237989e-02 -1.31816149e+00
6.90218389e-01 -9.58374619e-01 3.54378559e-02 -1.26301229e+00
-3.83729823e-02 -6.63682282e-01 -1.02748731e-02 -2.22726166e-01
1.81082755e-01 1.63457006e-01 8.60024244e-02 7.20928192e-01
-2.77405493e-02 8.51283729e-01 1.15227842e+00 1.17801979e-01
-3.95398051e-01 -1.28443003e-01 1.68127492e-01 8.67109001e-01
7.65994608e-01 -6.37782633e-01 -1.73433810e-01 -2.64748894e-02
2.22254276e-01 3.61624956e-01 5.62393367e-01 -1.24239576e+00
4.80683416e-01 -2.09047437e-01 5.77265210e-02 -6.67263746e-01
6.79893911e-01 -5.80331564e-01 3.71473670e-01 6.87517941e-01
-3.32110852e-01 -1.60366729e-01 4.24369127e-01 5.66367090e-01
3.81819718e-02 -1.45758778e-01 8.62194359e-01 -9.40757617e-03
-6.14342541e-02 2.63516128e-01 -8.71739209e-01 4.30482626e-02
5.66772580e-01 6.97039440e-02 -2.12921590e-01 -3.83029491e-01
-6.20852232e-01 -8.90238956e-03 4.51989204e-01 -1.45726681e-01
5.23900151e-01 -1.64438927e+00 -7.04232514e-01 3.83763462e-01
-4.51398849e-01 2.32121274e-01 5.63972533e-01 1.04129899e+00
-8.19017708e-01 7.14125335e-01 -4.87504117e-02 -8.04193377e-01
-7.15727925e-01 4.60482121e-01 5.60322046e-01 -4.03754413e-01
-8.31762910e-01 3.81390661e-01 4.40379351e-01 -6.10024393e-01
-3.80819112e-01 -3.38341445e-01 1.18540630e-01 -3.89386117e-01
4.10594583e-01 1.05953610e+00 -1.78415418e-01 -1.18738306e+00
-5.22248931e-02 7.14728713e-01 3.74160409e-01 -2.37224296e-01
1.21032929e+00 -1.04975514e-02 -4.69699055e-01 6.47225678e-01
1.50902784e+00 3.07371560e-02 -1.25611532e+00 -2.50719279e-01
-5.49145401e-01 -1.73350781e-01 5.28030217e-01 2.03422293e-01
-8.72918785e-01 1.05124903e+00 4.69022363e-01 2.43242919e-01
7.06166923e-01 -3.33790958e-01 1.16758740e+00 6.46318376e-01
2.29109466e-01 -8.35718513e-01 -4.87322390e-01 7.01787233e-01
7.67708004e-01 -8.91775608e-01 -2.73850039e-02 -3.72558028e-01
1.45504564e-01 1.27903306e+00 3.43224198e-01 -2.57861078e-01
8.57119262e-01 1.79978058e-01 -3.03053200e-01 -1.27399769e-02
-5.39169192e-01 2.66227692e-01 3.85030180e-01 -8.07561651e-02
-1.52161540e-02 1.25891874e-02 -2.49249384e-01 1.36537263e-02
-2.20620573e-01 -9.84865203e-02 5.57917058e-01 7.06664741e-01
-1.68179885e-01 -8.27025056e-01 -5.45020998e-01 3.25577885e-01
1.36020169e-01 -2.41187476e-02 1.33887693e-01 8.06287169e-01
-2.68812060e-01 6.14864528e-01 8.27616379e-02 -6.22246005e-02
4.45491821e-02 2.07140610e-01 6.09834455e-02 -1.91331521e-01
2.81028878e-02 -3.15598994e-02 -2.95144945e-01 -7.24666476e-01
-3.46625596e-01 -6.76636398e-01 -1.48372209e+00 -8.11582804e-01
-3.02592725e-01 6.40657723e-01 5.38460016e-01 9.74368989e-01
3.77021581e-01 3.75545502e-01 3.22508454e-01 -1.37212992e+00
-8.75631511e-01 -8.42234433e-01 -8.04211795e-01 2.48820558e-01
7.45945930e-01 -7.21651673e-01 -1.03159618e+00 -3.86012435e-01] | [6.489818572998047, 3.496568202972412] |
a49675f8-4a45-4656-8510-5f4b1905979b | tensor-networks-for-multi-modal-non-euclidean | 2103.14998 | null | https://arxiv.org/abs/2103.14998v1 | https://arxiv.org/pdf/2103.14998v1.pdf | Tensor Networks for Multi-Modal Non-Euclidean Data | Modern data sources are typically of large scale and multi-modal natures, and acquired on irregular domains, which poses serious challenges to traditional deep learning models. These issues are partially mitigated by either extending existing deep learning algorithms to irregular domains through graphs, or by employing tensor methods to alleviate the computational bottlenecks imposed by the Curse of Dimensionality. To simultaneously resolve both these issues, we introduce a novel Multi-Graph Tensor Network (MGTN) framework, which leverages on the desirable properties of graphs, tensors and neural networks in a physically meaningful and compact manner. This equips MGTNs with the ability to exploit local information in irregular data sources at a drastically reduced parameter complexity, and over a range of learning paradigms such as regression, classification and reinforcement learning. The benefits of the MGTN framework, especially its ability to avoid overfitting through the inherent low-rank regularization properties of tensor networks, are demonstrated through its superior performance against competing models in the individual tensor, graph, and neural network domains. | ['Danilo P. Mandic', 'Kriton Konstantinidis', 'Yao Lei Xu'] | 2021-03-27 | null | null | null | null | ['tensor-networks'] | ['methodology'] | [-8.33436996e-02 -1.15254313e-01 -1.31303236e-01 1.01940721e-01
-5.51096678e-01 -3.70672077e-01 6.40559852e-01 1.09702855e-01
8.70666206e-02 5.89521050e-01 5.14489770e-01 -1.99527174e-01
-7.26592541e-01 -7.95963109e-01 -5.04470050e-01 -6.38651252e-01
-4.83408213e-01 2.20390484e-01 -4.17089574e-02 -1.84598744e-01
1.10518336e-01 7.80444503e-01 -1.19139028e+00 -7.40436465e-02
7.89581120e-01 1.21197724e+00 -3.28712016e-01 2.77316600e-01
-7.16624260e-02 9.26420391e-01 -2.98352800e-02 -4.52773899e-01
3.50768507e-01 2.89439380e-01 -6.12960994e-01 4.04003084e-01
7.21131802e-01 -4.24483538e-01 -8.10213566e-01 8.72695565e-01
1.14264086e-01 1.63192406e-01 5.82671046e-01 -1.08964372e+00
-9.62423682e-01 1.59039840e-01 -8.03451002e-01 1.94350690e-01
-5.81124499e-02 8.32320377e-02 1.41135848e+00 -9.83727217e-01
3.59042436e-01 1.33744192e+00 1.03153229e+00 8.62603784e-02
-1.65156484e+00 -2.36146256e-01 2.38712624e-01 -1.69190973e-01
-1.20231318e+00 -3.16616267e-01 1.19837940e+00 -7.90173292e-01
6.63276851e-01 -4.38366868e-02 5.08940876e-01 1.12082374e+00
1.76163852e-01 7.17516899e-01 9.32580411e-01 -1.63995642e-02
1.28591061e-01 -6.35037720e-01 4.22165655e-02 9.19684947e-01
4.58184153e-01 1.12997271e-01 -7.26821423e-01 -3.44108701e-01
1.03674603e+00 1.31942630e-01 -4.52889428e-02 -7.56708443e-01
-1.36371028e+00 1.00394714e+00 6.62146986e-01 2.22255692e-01
-4.46186692e-01 4.83313352e-01 4.46401328e-01 1.34214282e-01
7.55238175e-01 4.65039253e-01 -4.14526641e-01 8.79901201e-02
-8.19560707e-01 2.80711144e-01 4.61463541e-01 6.38280392e-01
7.92088449e-01 5.30790031e-01 2.33261481e-01 8.59621823e-01
5.33811092e-01 4.27363575e-01 8.03343281e-02 -9.34585333e-01
5.29450834e-01 9.07257378e-01 -2.61273116e-01 -1.63463426e+00
-6.91843390e-01 -8.10025871e-01 -1.01770675e+00 4.82092276e-02
4.69007760e-01 1.52949631e-01 -7.33062446e-01 1.71376836e+00
5.00670135e-01 -9.01308358e-02 -2.20519736e-01 8.83796751e-01
6.61322832e-01 2.16089800e-01 -1.05840623e-01 2.97388643e-01
1.12859929e+00 -7.14435160e-01 -3.67034286e-01 -2.70562172e-01
7.70802319e-01 -6.01652145e-01 1.11562228e+00 3.61244768e-01
-7.87816226e-01 -2.69920796e-01 -1.04899514e+00 -4.07114595e-01
-3.03515702e-01 -6.09586649e-02 1.20662487e+00 3.18890125e-01
-9.73640621e-01 6.16571009e-01 -1.00709546e+00 -1.30905792e-01
7.79303551e-01 2.53682375e-01 -5.01710176e-01 -2.93010026e-01
-9.43476319e-01 4.95787889e-01 4.06798273e-02 3.57932627e-01
-5.21620929e-01 -9.35608447e-01 -9.96545374e-01 -1.82014704e-02
4.02003020e-01 -6.77445054e-01 6.58810377e-01 -4.89495873e-01
-1.06516373e+00 3.62280041e-01 3.85726303e-01 -2.80016184e-01
3.81485552e-01 -1.81208536e-01 -3.19794089e-01 2.39308819e-01
1.68154463e-01 2.12174103e-01 1.01571405e+00 -9.79461730e-01
7.03144073e-02 -6.31182909e-01 1.02432765e-01 1.09496317e-03
-5.71369588e-01 -4.02485222e-01 -1.34405404e-01 -8.76875162e-01
3.42889994e-01 -9.37652409e-01 -4.58710104e-01 1.71382129e-01
-4.94012415e-01 -1.23952493e-01 1.09914911e+00 -5.03377974e-01
7.10176349e-01 -2.22016859e+00 4.20422941e-01 2.41765603e-01
9.18916166e-01 1.21864356e-01 -3.03950131e-01 4.46253955e-01
7.83639476e-02 1.45661272e-03 -1.32261157e-01 -3.95019978e-01
-8.98324326e-03 3.74086291e-01 -1.21245645e-01 6.01651549e-01
5.57436407e-01 1.05841279e+00 -8.54805231e-01 -3.34794432e-01
1.18817054e-02 6.18312359e-01 -5.61538696e-01 1.37930699e-02
-2.88142353e-01 4.96164322e-01 -7.78225541e-01 8.06065440e-01
5.38613260e-01 -7.18416393e-01 2.23987371e-01 -4.88321513e-01
1.37315840e-01 2.26784259e-01 -1.15511036e+00 1.83363712e+00
-3.05680633e-01 3.58593881e-01 5.23430765e-01 -1.27158284e+00
7.96432018e-01 1.11648969e-01 8.87681723e-01 -8.04044962e-01
-8.37927982e-02 3.67665470e-01 -8.74159038e-02 -5.15339494e-01
5.30243039e-01 -1.53905720e-01 5.24650998e-02 4.99940306e-01
1.58344179e-01 6.06991611e-02 2.09826484e-01 3.91892612e-01
1.27388966e+00 2.68781424e-01 -2.39030585e-01 -5.33612430e-01
7.93094561e-02 -1.44867867e-01 4.53613132e-01 3.89041394e-01
8.71737581e-03 3.85486692e-01 6.98088169e-01 -6.70745254e-01
-9.78396416e-01 -1.11389697e+00 -1.77649394e-01 1.02794373e+00
-3.41473818e-01 -3.67074013e-01 -1.70297176e-01 -6.04125679e-01
5.31132996e-01 -1.39507607e-01 -7.03653276e-01 -2.40369543e-01
-6.24368489e-01 -9.04567182e-01 4.78262603e-01 5.95043004e-01
3.82246196e-01 -4.63850886e-01 -1.02355227e-01 2.45579898e-01
-3.15518714e-02 -1.48748326e+00 -2.97452331e-01 1.57226756e-01
-1.25220060e+00 -1.06679773e+00 -3.77132028e-01 -2.93076366e-01
4.53560352e-01 4.88541812e-01 1.18636000e+00 1.70798764e-01
-2.45155483e-01 4.89002019e-01 -1.16281748e-01 1.67957067e-01
-5.02169207e-02 1.36585355e-01 2.28371695e-01 3.47453594e-01
5.22641949e-02 -1.09801781e+00 -4.93185490e-01 -2.14609392e-02
-1.19105327e+00 -1.00251034e-01 7.23527551e-01 1.01873791e+00
3.57833385e-01 -4.87020612e-02 7.10117221e-01 -7.41336763e-01
7.47574687e-01 -6.76601470e-01 -7.07430303e-01 -3.88688706e-02
-5.19559860e-01 4.08724695e-01 7.10335493e-01 -3.53120953e-01
-5.81377923e-01 -2.30267674e-01 2.58497536e-01 -6.14997387e-01
2.99208581e-01 1.02277887e+00 1.21042371e-01 -7.44468570e-01
6.16112649e-01 -1.91736668e-01 3.24186951e-01 -7.32252479e-01
5.07214606e-01 -7.84164574e-03 1.99828237e-01 -1.08438456e+00
1.05533314e+00 4.88024622e-01 6.54871345e-01 -9.61558223e-01
-1.08572865e+00 -2.16412887e-01 -8.51263881e-01 -1.19906306e-01
5.77577174e-01 -9.01506245e-01 -5.27467370e-01 4.59179044e-01
-7.97297955e-01 -1.46774277e-01 -1.88385278e-01 4.57465738e-01
-1.46939814e-01 7.24525869e-01 -7.99150467e-01 -4.80707109e-01
-1.30754530e-01 -1.03737855e+00 1.06907475e+00 -1.58688024e-01
1.42990887e-01 -1.45115387e+00 -2.19700336e-01 6.82524443e-01
7.04736173e-01 7.61463940e-01 1.26064610e+00 -3.42317730e-01
-9.39722836e-01 -2.50383317e-01 -7.70911157e-01 3.89370948e-01
1.82222143e-01 -1.38908535e-01 -7.43591964e-01 -3.67189109e-01
-1.66548073e-01 -7.10714579e-01 7.88317621e-01 2.07203984e-01
9.80909884e-01 -1.57688662e-01 1.12784326e-01 7.81624019e-01
1.36078393e+00 -5.42665243e-01 6.32136092e-02 2.28578985e-01
1.35353982e+00 5.69394767e-01 -1.23580597e-01 3.29480112e-01
5.97663701e-01 6.50758326e-01 7.83491373e-01 -2.70541251e-01
-2.89348662e-01 -1.50106385e-01 -6.57943189e-02 1.10841548e+00
-1.09564655e-01 2.13767648e-01 -1.06638396e+00 5.62941015e-01
-1.91087890e+00 -5.30745864e-01 -2.82149076e-01 1.91452551e+00
5.49690247e-01 1.34170741e-01 1.73918068e-01 1.84665367e-01
1.61478013e-01 7.03050971e-01 -6.22909665e-01 -1.72495648e-01
-3.51911873e-01 3.25481296e-02 4.46308672e-01 1.86003074e-01
-1.19608426e+00 5.55848718e-01 6.38013840e+00 5.11891246e-01
-1.23571241e+00 -5.56886718e-02 4.12562937e-01 -3.71182114e-02
-4.43730086e-01 -1.33769587e-01 -2.83169031e-01 1.49569616e-01
8.53926480e-01 2.18940109e-01 8.58149111e-01 5.54564238e-01
9.22683403e-02 2.89410502e-01 -1.15647388e+00 8.68983388e-01
-1.15107119e-01 -1.63953173e+00 1.34079054e-01 6.06899023e-01
6.56150162e-01 5.04886866e-01 3.60118479e-01 1.80675149e-01
4.08895254e-01 -1.13868630e+00 4.48894590e-01 3.68372053e-01
6.03706896e-01 -4.42618161e-01 3.12812090e-01 1.71575975e-02
-1.21804941e+00 -5.65639846e-02 -3.13609272e-01 -1.62734538e-01
-1.85930774e-01 1.05348265e+00 -4.56682712e-01 9.13580537e-01
4.96905893e-01 1.10490024e+00 -7.36419439e-01 7.90926099e-01
2.63679743e-01 4.80574071e-01 -4.72405046e-01 3.98451269e-01
6.43391132e-01 -5.03764272e-01 5.92340112e-01 9.26333547e-01
8.38256702e-02 -3.11859429e-01 5.10022163e-01 1.00079560e+00
-2.99260736e-01 -8.25198516e-02 -9.24016058e-01 -4.40588474e-01
6.39188886e-02 1.44653857e+00 -6.10358596e-01 2.75459588e-01
-7.15999782e-01 4.07427132e-01 8.06771696e-01 6.57294214e-01
-4.37570363e-01 -1.62177920e-01 6.83670580e-01 2.05438152e-01
3.29159677e-01 -9.70274925e-01 -4.51682359e-01 -1.37485719e+00
4.60579455e-01 -1.01428056e+00 2.92135447e-01 -3.88456821e-01
-1.65379655e+00 4.75919634e-01 -2.31793821e-01 -1.10776389e+00
7.19405487e-02 -9.94205296e-01 -3.15426201e-01 6.06618941e-01
-1.67377615e+00 -1.57136905e+00 -1.41716346e-01 8.73995006e-01
-1.31994307e-01 -1.21745855e-01 5.66131592e-01 5.69553077e-01
-5.71417212e-01 3.26214880e-01 2.78915137e-01 3.13779980e-01
1.75057471e-01 -1.30184352e+00 3.34123671e-01 8.26691031e-01
2.39337355e-01 7.09849536e-01 2.74152607e-01 -3.78529549e-01
-2.03514719e+00 -1.24962449e+00 3.27116579e-01 -4.08197373e-01
1.37918210e+00 -5.33102930e-01 -1.08124352e+00 5.75393021e-01
-4.10098910e-01 7.38954723e-01 6.12743497e-01 6.52938604e-01
-9.82268393e-01 -1.81208462e-01 -8.37070584e-01 5.60354233e-01
1.05544746e+00 -9.10052538e-01 -2.11202651e-01 7.40645111e-01
7.10986197e-01 -2.35163867e-01 -1.41684639e+00 4.44925964e-01
3.05405855e-01 -8.51803184e-01 1.22849154e+00 -8.72547746e-01
5.20262599e-01 -7.39513487e-02 -5.11124432e-01 -1.23253000e+00
-4.57344681e-01 -7.86148608e-01 -4.83377159e-01 1.06953394e+00
3.55953187e-01 -6.27542078e-01 7.47622907e-01 5.42510927e-01
-2.29323700e-01 -9.54755902e-01 -1.19948220e+00 -8.58118236e-01
2.08730102e-01 -5.21911263e-01 5.63211501e-01 1.08835232e+00
-2.63341933e-01 5.28610468e-01 -5.72550595e-01 1.50570899e-01
1.05522490e+00 1.02372229e-01 6.18462503e-01 -1.55367792e+00
-2.97494590e-01 -3.79002512e-01 -5.95443606e-01 -8.76257598e-01
5.00115275e-05 -9.97435629e-01 -4.02803391e-01 -1.39964890e+00
-1.82565704e-01 -7.35842526e-01 -3.76299173e-01 3.65398645e-01
1.55605331e-01 4.00240004e-01 5.88315167e-02 4.99511808e-01
-6.38966382e-01 8.10307741e-01 1.51431286e+00 -8.86548236e-02
1.36554599e-01 -2.78698027e-01 -7.57816434e-01 7.45329440e-01
3.89328569e-01 -2.12919936e-01 -3.62209678e-01 -9.85234439e-01
4.32427406e-01 9.64083746e-02 6.38180912e-01 -9.47090030e-01
-3.75406141e-03 -1.34477228e-01 3.39557439e-01 -1.14396468e-01
4.77830797e-01 -7.74640679e-01 -8.79260451e-02 3.81788909e-02
-2.03051969e-01 2.35712245e-01 2.45015338e-01 1.02914476e+00
-1.21338680e-01 4.12991256e-01 7.29937911e-01 6.58608899e-02
-3.13899249e-01 7.22243965e-01 1.41190141e-02 2.34504923e-01
6.67570233e-01 -1.15853928e-01 -5.50608039e-01 -1.60396278e-01
-6.00549400e-01 1.27910465e-01 2.84907460e-01 5.10128617e-01
4.98225242e-01 -1.62666941e+00 -4.93605316e-01 3.84309828e-01
1.43290743e-01 1.52118012e-01 1.27485856e-01 1.20467854e+00
-4.25747275e-01 3.06871653e-01 -5.82372770e-02 -6.39214277e-01
-5.85020483e-01 4.77250248e-01 3.49766254e-01 -5.50135314e-01
-1.00178552e+00 6.43975198e-01 1.26067638e-01 -6.13099992e-01
5.58019280e-02 -3.52072805e-01 2.32521482e-02 -7.04436973e-02
4.94028479e-02 3.48634541e-01 3.81280601e-01 -7.32531905e-01
-2.50136584e-01 5.87244213e-01 -1.87054589e-01 3.83258909e-01
1.64308763e+00 3.24356891e-02 -3.03679794e-01 3.03334028e-01
1.21630132e+00 1.27918154e-01 -1.52849555e+00 -7.04989731e-01
3.85824561e-01 -4.30584460e-01 4.72119123e-01 -4.66063678e-01
-1.39777362e+00 8.23659301e-01 9.07051936e-02 7.10902929e-01
7.92235255e-01 -7.64661580e-02 1.02227592e+00 4.41810787e-01
2.99705386e-01 -9.06508148e-01 4.86061573e-01 6.29186392e-01
8.73747051e-01 -1.23027515e+00 2.40857691e-01 -2.66687691e-01
-3.61713439e-01 1.27285302e+00 2.99986482e-01 -4.99287456e-01
8.76975477e-01 7.35344067e-02 -1.31526321e-01 -7.41288662e-01
-6.50539935e-01 -5.15926443e-02 5.04067361e-01 5.61424077e-01
4.60998148e-01 1.20903607e-02 1.66470990e-01 1.00193344e-01
6.01747073e-02 -1.71765715e-01 4.03557241e-01 8.76787782e-01
-1.09652869e-01 -9.28810775e-01 -2.10234016e-01 7.20844805e-01
-5.87876022e-01 -7.66575038e-02 -2.05407828e-01 8.00396025e-01
-2.68454522e-01 8.71011376e-01 -2.34834909e-01 -4.19062257e-01
1.45252660e-01 -2.73752123e-01 2.66115487e-01 -3.85614812e-01
-2.39065200e-01 1.35835916e-01 2.26727009e-01 -8.10426891e-01
-4.76328254e-01 -5.41399300e-01 -7.78270721e-01 -5.23525059e-01
-1.09766655e-01 -1.07839657e-02 5.88232219e-01 1.20476210e+00
7.39892542e-01 5.58512509e-01 5.58171213e-01 -1.05544221e+00
-8.39968979e-01 -7.46248960e-01 -7.90678501e-01 5.14127433e-01
7.20664561e-01 -1.08485389e+00 -4.93263572e-01 -3.75030816e-01] | [6.882254600524902, 5.846897602081299] |
5acfd885-902f-45ea-b47d-d213bbd217ca | improved-cross-lingual-transfer-learning-for | 2306.00789 | null | https://arxiv.org/abs/2306.00789v1 | https://arxiv.org/pdf/2306.00789v1.pdf | Improved Cross-Lingual Transfer Learning For Automatic Speech Translation | Research in multilingual speech-to-text translation is topical. Having a single model that supports multiple translation tasks is desirable. The goal of this work it to improve cross-lingual transfer learning in multilingual speech-to-text translation via semantic knowledge distillation. We show that by initializing the encoder of the encoder-decoder sequence-to-sequence translation model with SAMU-XLS-R, a multilingual speech transformer encoder trained using multi-modal (speech-text) semantic knowledge distillation, we achieve significantly better cross-lingual task knowledge transfer than the baseline XLS-R, a multilingual speech transformer encoder trained via self-supervised learning. We demonstrate the effectiveness of our approach on two popular datasets, namely, CoVoST-2 and Europarl. On the 21 translation tasks of the CoVoST-2 benchmark, we achieve an average improvement of 12.8 BLEU points over the baselines. In the zero-shot translation scenario, we achieve an average gain of 18.8 and 11.9 average BLEU points on unseen medium and low-resource languages. We make similar observations on Europarl speech translation benchmark. | ['James Glass', 'Victoria Mingote', 'Pablo Gimeno', 'Luis Vicente', 'Antoine Laurent', 'Nauman Dawalatabad', 'Sameer Khurana'] | 2023-06-01 | null | null | null | null | ['speech-to-text-translation', 'cross-lingual-transfer'] | ['natural-language-processing', 'natural-language-processing'] | [ 1.36530548e-01 1.20489292e-01 -4.22420621e-01 -3.15599024e-01
-2.03953099e+00 -8.01020503e-01 7.58542240e-01 -3.66954237e-01
-3.37922722e-01 1.05943871e+00 3.72658163e-01 -8.85704339e-01
5.80192685e-01 -2.35500753e-01 -1.28461945e+00 -4.06536371e-01
5.84691823e-01 1.01997161e+00 -1.63112640e-01 -4.46460217e-01
-4.99981523e-01 -3.29135776e-01 -5.79522431e-01 6.78070188e-01
1.11513925e+00 5.90962231e-01 3.58956605e-01 5.02685249e-01
-3.49328667e-02 5.37732661e-01 -3.20538670e-01 -8.58861923e-01
1.32568508e-01 -7.48641610e-01 -1.11840594e+00 -1.16663091e-01
4.30095375e-01 5.69897443e-02 -1.19702443e-01 9.25568879e-01
7.25500464e-01 -3.25548738e-01 6.50159299e-01 -9.14539993e-01
-1.05164051e+00 1.03899455e+00 -2.70077705e-01 2.95913927e-02
2.60354489e-01 -1.40894160e-01 1.16296017e+00 -1.30289102e+00
7.71521986e-01 1.42400968e+00 4.55366969e-01 5.33114314e-01
-1.26682746e+00 -6.88586056e-01 -3.35200220e-01 7.66265243e-02
-1.37935126e+00 -1.07152021e+00 8.11148137e-02 -1.92608967e-01
1.54679763e+00 -4.59242016e-02 -2.98656560e-02 1.52025390e+00
2.47118309e-01 1.04544091e+00 1.25747025e+00 -8.01163554e-01
-1.02248602e-01 5.05706489e-01 -5.13988018e-01 5.79919517e-01
-1.73596770e-01 1.13650732e-01 -8.74147594e-01 8.97938311e-02
2.54518896e-01 -6.04091465e-01 -1.87882423e-01 1.71362720e-02
-1.54256880e+00 7.63730228e-01 1.71474777e-02 4.12873149e-01
-5.89017160e-02 5.39835021e-02 7.61213601e-01 9.87904131e-01
9.97167468e-01 4.26187009e-01 -9.08780932e-01 -3.12583476e-01
-8.08427572e-01 -2.26462603e-01 9.15460348e-01 1.36935842e+00
7.14020729e-01 1.77855060e-01 -2.27204546e-01 1.17228484e+00
7.35291047e-03 1.41108012e+00 7.37896621e-01 -6.02765381e-01
9.87679243e-01 -1.20183751e-01 -2.19456013e-02 6.26011640e-02
2.92630613e-01 -5.22799730e-01 -3.97738755e-01 -5.27473092e-01
3.17314342e-02 -3.57366771e-01 -8.38245749e-01 1.69851935e+00
-1.30973130e-01 4.01452631e-02 9.00343657e-01 5.37729144e-01
4.94855285e-01 9.76032376e-01 6.67846180e-04 -3.31121534e-01
1.34668279e+00 -1.48028648e+00 -9.28576052e-01 -4.52425987e-01
1.06869960e+00 -1.37079573e+00 1.32710373e+00 -2.21667066e-01
-1.25506747e+00 -4.18018669e-01 -8.24247599e-01 -1.53865978e-01
-4.17505741e-01 6.06563330e-01 1.36115238e-01 3.12707245e-01
-1.32989752e+00 2.02772960e-01 -8.64057779e-01 -5.98032296e-01
-7.91130215e-02 8.99453163e-02 -3.77728879e-01 -1.53723046e-01
-1.51703835e+00 1.41086888e+00 3.22539955e-01 -4.34757680e-01
-1.18011022e+00 -7.36182153e-01 -8.55828464e-01 -7.29168952e-02
1.34036243e-01 -8.29793930e-01 1.72486365e+00 -1.17291915e+00
-1.79649949e+00 1.09859800e+00 -4.76270407e-01 -6.22699082e-01
4.08793181e-01 -2.23993510e-01 -5.54408491e-01 -1.26321211e-01
5.37547529e-01 5.70428073e-01 5.65507054e-01 -9.57126260e-01
-6.25382543e-01 -1.69564471e-01 -4.40078825e-01 6.17065728e-01
-2.05696300e-01 5.42669535e-01 -3.92763704e-01 -7.08575845e-01
-3.96673203e-01 -1.25924122e+00 2.77060837e-01 -6.65681422e-01
-1.69981658e-01 -2.30009809e-01 7.80634522e-01 -1.14128447e+00
8.46097052e-01 -1.86175811e+00 3.88053775e-01 -4.20812994e-01
-6.43149972e-01 5.50128400e-01 -4.11182970e-01 6.33583724e-01
-5.25445677e-02 -1.22304842e-01 -2.62942910e-01 -7.62857378e-01
5.69057241e-02 3.22363883e-01 -6.68953955e-01 1.25565439e-01
2.07159430e-01 1.51791024e+00 -9.55065131e-01 -3.76037478e-01
-1.80999525e-02 5.67950785e-01 -1.04678951e-01 2.77533233e-01
-2.78720349e-01 5.55993795e-01 -1.67052582e-01 6.20325506e-01
4.59248424e-02 -1.80567130e-01 1.48737118e-01 1.00577340e-01
1.57115117e-01 9.68112290e-01 1.80523507e-02 2.39315772e+00
-1.11777091e+00 7.47531831e-01 -2.65249342e-01 -7.97549307e-01
8.51932168e-01 1.05165923e+00 1.45997778e-01 -9.45982933e-01
-1.55869961e-01 1.01955485e+00 -2.94598311e-01 -1.96481660e-01
3.00619632e-01 -5.36090016e-01 -2.77332693e-01 5.17677665e-01
6.54750049e-01 -2.85984844e-01 -9.81900543e-02 1.63286075e-01
9.07366574e-01 4.32329237e-01 1.45823970e-01 -2.95990109e-01
3.00987512e-01 1.23491392e-01 2.27676585e-01 3.18144292e-01
1.15433738e-01 2.64204621e-01 -3.31369527e-02 1.13759637e-02
-1.27763343e+00 -1.39996111e+00 6.01950996e-02 1.31272376e+00
-3.95488262e-01 -4.24154013e-01 -9.79416072e-01 -8.05947602e-01
-2.55721420e-01 1.12485683e+00 -1.11337081e-01 -1.70117751e-01
-7.00309455e-01 -6.30873442e-01 9.55196619e-01 3.15426379e-01
3.67762893e-01 -7.64916241e-01 4.98599648e-01 3.85311574e-01
-9.53004062e-01 -1.74169540e+00 -9.58779633e-01 8.51250365e-02
-7.36428440e-01 -3.22496980e-01 -9.89110053e-01 -1.15990424e+00
2.98202634e-01 1.97166547e-01 1.32363439e+00 -7.13924646e-01
3.48702878e-01 1.77625865e-01 -4.42010462e-01 -3.39732111e-01
-1.05594409e+00 5.63967526e-01 2.63656586e-01 -1.28117099e-01
5.33139408e-01 -4.01276469e-01 -5.35403900e-02 4.35831755e-01
-2.66000003e-01 3.05270016e-01 7.45654821e-01 9.32662666e-01
5.88550746e-01 -7.49643207e-01 8.75188768e-01 -5.54969251e-01
5.63801348e-01 -3.41675431e-01 -4.30217236e-01 8.34101737e-01
-5.95480084e-01 3.22965324e-01 7.19657123e-01 -1.48998350e-01
-1.08854890e+00 -1.18597887e-01 -6.83448985e-02 -4.49208379e-01
2.33113974e-01 4.78872418e-01 -1.30260780e-01 1.49790794e-01
6.89532340e-01 7.62945414e-01 -1.70419261e-01 -5.15598416e-01
6.48096979e-01 1.28046381e+00 5.66868544e-01 -7.74178207e-01
5.89733899e-01 -3.11157219e-02 -6.67902708e-01 -5.01901507e-01
-9.85423028e-01 -4.23956275e-01 -5.48042595e-01 1.80595160e-01
1.05183029e+00 -1.56511736e+00 -7.99398422e-02 2.93117046e-01
-1.54301310e+00 -6.09568179e-01 -6.89087585e-02 6.86635375e-01
-1.04397428e+00 -9.61841047e-02 -8.59982669e-01 -2.19720453e-01
-7.67855406e-01 -1.42595589e+00 1.68749869e+00 -6.22011006e-01
-9.84183773e-02 -1.26971006e+00 3.14772993e-01 7.91924179e-01
4.81549859e-01 -4.42109287e-01 7.51287401e-01 -7.92887270e-01
-4.01132971e-01 2.40283698e-01 -9.93397534e-02 6.58690989e-01
3.71101916e-01 -6.18695676e-01 -8.46264422e-01 -6.63292468e-01
-2.25596771e-01 -8.07970107e-01 7.38844097e-01 -1.75003540e-02
2.49536216e-01 -5.06443799e-01 -2.62037843e-01 4.56694543e-01
1.27754402e+00 1.58169463e-01 4.36676800e-01 7.72676542e-02
6.48042023e-01 2.93663591e-01 5.85170746e-01 -3.67697507e-01
7.37848699e-01 1.13238597e+00 -3.15190613e-01 -1.54035404e-01
-6.63069546e-01 -5.80437362e-01 1.17411160e+00 1.82052302e+00
1.91142902e-01 -2.34142393e-01 -1.04505038e+00 7.92273402e-01
-1.85933578e+00 -5.85806966e-01 2.56078303e-01 2.16080165e+00
1.47760832e+00 -2.93722004e-01 -1.38468996e-01 -6.73657358e-01
6.49662673e-01 -2.17033386e-01 -2.17456281e-01 -6.06247723e-01
-2.87035316e-01 4.89134908e-01 6.81484222e-01 1.07830822e+00
-8.15457582e-01 1.85681498e+00 5.79003191e+00 1.17854273e+00
-1.23221290e+00 9.34246421e-01 3.77509892e-01 2.16327664e-02
-3.39878142e-01 1.28007486e-01 -1.03527105e+00 4.02532697e-01
1.65042698e+00 -6.05127215e-01 7.19079673e-01 5.54884493e-01
2.48700172e-01 5.08382320e-01 -1.17326868e+00 8.42639148e-01
3.10327798e-01 -1.19267023e+00 1.82803825e-01 -1.84479773e-01
1.19465899e+00 7.60047913e-01 9.20514911e-02 5.74076056e-01
7.47186542e-01 -9.58770931e-01 6.14572883e-01 -1.26329720e-01
1.46934080e+00 -6.35434449e-01 6.05441272e-01 2.12663785e-01
-1.00763321e+00 6.05534375e-01 -2.75812536e-01 4.14617091e-01
5.10519743e-01 2.85720259e-01 -1.38785064e+00 9.45645511e-01
3.58829170e-01 9.21756148e-01 -1.63079128e-02 1.64070040e-01
-5.47493696e-01 9.10370648e-01 -5.17219938e-02 1.93038717e-01
5.70248425e-01 -1.20360829e-01 4.36134696e-01 1.47254491e+00
6.98860824e-01 -5.43063998e-01 1.73865452e-01 4.46036607e-01
-4.97138292e-01 5.00014365e-01 -8.11699212e-01 -3.20233464e-01
4.53134596e-01 6.82299793e-01 2.57105008e-02 -7.33170092e-01
-5.91432214e-01 1.61518168e+00 4.60410893e-01 5.44041574e-01
-7.80950010e-01 -1.87832698e-01 7.11033762e-01 -1.82185084e-01
2.78227329e-01 -3.56689900e-01 4.61086892e-02 -1.48581827e+00
1.51411235e-01 -1.35163546e+00 -1.71131827e-02 -9.39343333e-01
-1.15557659e+00 8.42312634e-01 -3.14980835e-01 -1.07200491e+00
-6.96998894e-01 -4.67981160e-01 -1.05907664e-01 1.27066875e+00
-1.72956097e+00 -1.81852484e+00 5.41580617e-01 6.71517193e-01
1.05027461e+00 -7.13526905e-01 1.24327290e+00 5.49242675e-01
-8.33828002e-02 8.74261200e-01 4.91047531e-01 2.76044220e-01
1.32491338e+00 -1.10443282e+00 9.69442070e-01 7.19443202e-01
5.01769364e-01 3.86377811e-01 6.15319312e-01 -6.39193237e-01
-1.41469955e+00 -1.50006807e+00 1.74819160e+00 -7.93247461e-01
1.09035122e+00 -6.04948103e-01 -7.32188046e-01 1.30033457e+00
8.73136282e-01 -2.91530967e-01 6.22604966e-01 6.87288493e-02
-5.47793686e-01 -1.26530482e-02 -6.45663977e-01 6.96503580e-01
9.49785769e-01 -1.14479756e+00 -7.28805542e-01 8.89402330e-01
1.31796396e+00 -4.59082603e-01 -1.02855229e+00 2.72799492e-01
3.25003117e-01 -9.05401707e-02 8.07216108e-01 -8.14215958e-01
4.32240427e-01 -7.73907751e-02 -5.19647539e-01 -2.05371881e+00
1.92131922e-01 -8.86896253e-01 2.99744219e-01 9.26028073e-01
1.05762124e+00 -7.26231933e-01 1.10275224e-01 -3.83370429e-01
-5.17038345e-01 -3.92492771e-01 -1.28167915e+00 -1.35561419e+00
6.56861603e-01 -1.76203862e-01 3.34491402e-01 1.12490499e+00
2.62773633e-01 1.09978127e+00 -6.12596571e-01 -4.04459089e-02
5.27589262e-01 -1.28376946e-01 7.14406252e-01 -4.17540550e-01
-6.14579916e-01 -1.12731822e-01 2.01137420e-02 -1.21845102e+00
7.46995151e-01 -1.62906420e+00 1.31856948e-01 -1.47812152e+00
3.26256573e-01 -1.57653332e-01 -8.73543099e-02 8.26329052e-01
-8.12344402e-02 1.79245859e-01 -1.44881522e-02 2.38664120e-01
-5.31105757e-01 8.37540567e-01 1.24554300e+00 -2.24175066e-01
1.35618553e-01 -2.25446612e-01 -4.18142408e-01 9.22524258e-02
7.10779011e-01 -5.84354162e-01 -4.62290108e-01 -1.05358934e+00
-2.29868349e-02 3.49324346e-01 -1.14812329e-01 -5.76016307e-01
6.14578277e-02 1.20956309e-01 -4.35730577e-01 -2.01215521e-01
4.12052453e-01 -5.01363873e-01 4.14310284e-02 2.93369323e-01
-4.22879636e-01 2.08878577e-01 3.51328164e-01 9.57128927e-02
-5.68540812e-01 1.80037409e-01 6.89908087e-01 -1.90867081e-01
-3.20449531e-01 1.46944001e-01 -1.77781552e-01 3.64648819e-01
7.95131326e-01 4.24766213e-01 -6.10794842e-01 -4.62433726e-01
-7.06456125e-01 7.16328695e-02 3.66513401e-01 9.45008039e-01
2.12157726e-01 -1.66766465e+00 -1.37362969e+00 9.54353362e-02
5.05843759e-01 -6.72986269e-01 -3.47199321e-01 1.05967271e+00
-4.78588790e-02 1.05090833e+00 7.63634667e-02 -7.17972755e-01
-1.21229482e+00 2.18363628e-01 4.34487134e-01 -3.76994461e-01
-1.88549832e-01 6.94669366e-01 1.03209451e-01 -9.77553546e-01
-2.05343813e-01 -5.86795583e-02 6.96149290e-01 -3.57497603e-01
2.96026587e-01 1.01357251e-01 4.93050992e-01 -8.56319368e-01
-3.61754775e-01 5.70801437e-01 -1.92613602e-01 -7.40791321e-01
9.81294096e-01 -3.43044013e-01 -1.13584980e-01 5.99369884e-01
1.63690233e+00 1.25242714e-02 -7.22140253e-01 -7.75361121e-01
5.42347431e-02 -5.80731109e-02 -1.23927936e-01 -1.18887162e+00
-6.39304280e-01 1.09492517e+00 2.48763531e-01 -5.27364314e-01
7.62351155e-01 3.20751250e-01 1.16828036e+00 6.13819420e-01
7.06746042e-01 -1.11292338e+00 -5.79954386e-02 9.93415058e-01
9.47252274e-01 -1.53672278e+00 -7.16256917e-01 -2.21935973e-01
-9.24421310e-01 7.16655374e-01 1.00127853e-01 4.48425800e-01
2.50835866e-01 4.00496215e-01 5.25584877e-01 1.94594800e-01
-1.07145679e+00 -1.59807339e-01 3.32360864e-01 3.39209467e-01
7.65910983e-01 5.12863576e-01 -1.90234184e-01 1.27280340e-01
-4.52301860e-01 5.72005995e-02 -3.87525564e-04 5.76698661e-01
-3.88109326e-01 -1.47974586e+00 -1.19971922e-02 -1.69345915e-01
-6.87297165e-01 -7.64872134e-01 -4.27977860e-01 3.24026495e-01
-4.85933870e-01 1.10317385e+00 -1.86546892e-01 -2.73687899e-01
2.24220455e-01 5.33751488e-01 5.93403220e-01 -7.90368319e-01
-4.63113308e-01 1.59170568e-01 5.44660211e-01 -4.09275115e-01
-2.46705383e-01 -5.93060553e-01 -1.05380929e+00 -2.65558600e-01
-1.60456628e-01 4.71092492e-01 1.05846620e+00 1.06821597e+00
5.27964592e-01 3.77230287e-01 6.13810658e-01 -1.89765498e-01
-7.79850960e-01 -1.15791309e+00 -5.62806837e-02 1.46128088e-01
2.12969080e-01 -2.07595661e-01 -1.83935001e-01 3.98832500e-01] | [14.438764572143555, 7.291724681854248] |
7a5dcde4-62a5-45b6-ae52-ca4a7aca34d5 | slide-single-image-3d-photography-with-soft | 2109.01068 | null | https://arxiv.org/abs/2109.01068v1 | https://arxiv.org/pdf/2109.01068v1.pdf | SLIDE: Single Image 3D Photography with Soft Layering and Depth-aware Inpainting | Single image 3D photography enables viewers to view a still image from novel viewpoints. Recent approaches combine monocular depth networks with inpainting networks to achieve compelling results. A drawback of these techniques is the use of hard depth layering, making them unable to model intricate appearance details such as thin hair-like structures. We present SLIDE, a modular and unified system for single image 3D photography that uses a simple yet effective soft layering strategy to better preserve appearance details in novel views. In addition, we propose a novel depth-aware training strategy for our inpainting module, better suited for the 3D photography task. The resulting SLIDE approach is modular, enabling the use of other components such as segmentation and matting for improved layering. At the same time, SLIDE uses an efficient layered depth formulation that only requires a single forward pass through the component networks to produce high quality 3D photos. Extensive experimental analysis on three view-synthesis datasets, in combination with user studies on in-the-wild image collections, demonstrate superior performance of our technique in comparison to existing strong baselines while being conceptually much simpler. Project page: https://varunjampani.github.io/slide | ['Ce Liu', 'Brian Curless', 'David Salesin', 'William T. Freeman', 'Dominik Kaeser', 'Michael Krainin', 'Richard Tucker', 'Abhishek Kar', 'Kyle Sargent', 'Huiwen Chang', 'Varun Jampani'] | 2021-09-02 | null | http://openaccess.thecvf.com//content/ICCV2021/html/Jampani_SLIDE_Single_Image_3D_Photography_With_Soft_Layering_and_Depth-Aware_ICCV_2021_paper.html | http://openaccess.thecvf.com//content/ICCV2021/papers/Jampani_SLIDE_Single_Image_3D_Photography_With_Soft_Layering_and_Depth-Aware_ICCV_2021_paper.pdf | iccv-2021-1 | ['image-matting'] | ['computer-vision'] | [ 2.68548399e-01 2.94118375e-01 -6.47676829e-03 -5.46519339e-01
-5.54998875e-01 -5.81282020e-01 5.48722148e-01 -5.42602301e-01
-1.14177644e-01 3.61268580e-01 1.99766174e-01 -2.42307216e-01
3.94646615e-01 -4.65866268e-01 -9.35696542e-01 -4.88489389e-01
3.84719133e-01 1.68124691e-01 3.48853886e-01 -5.09070493e-02
1.97937246e-02 7.34447241e-01 -1.57601535e+00 4.77660596e-01
6.54298484e-01 8.80605459e-01 3.72442752e-01 9.04167235e-01
-6.48668036e-02 6.88048899e-01 -5.01402378e-01 -5.84012806e-01
7.51340330e-01 -2.81287611e-01 -4.88705993e-01 6.28661752e-01
1.29599357e+00 -1.08105767e+00 -2.32095063e-01 7.97364235e-01
4.99468923e-01 -6.02516867e-02 2.75682509e-01 -9.96874392e-01
-5.33986986e-01 9.06376690e-02 -8.35759282e-01 -2.17205077e-01
5.36580622e-01 3.59031200e-01 8.72827590e-01 -8.59370649e-01
9.67443049e-01 1.37415957e+00 6.54942751e-01 8.41936707e-01
-1.38154054e+00 -5.43094754e-01 4.86771107e-01 -2.31250282e-03
-1.01492858e+00 -5.74347377e-01 9.88685787e-01 -1.29527554e-01
1.07706892e+00 1.85838699e-01 7.42825925e-01 1.15176570e+00
1.33497044e-01 9.27060544e-01 1.17339253e+00 -3.46874267e-01
1.42116368e-01 2.26434171e-01 -4.18893129e-01 7.33455181e-01
-2.78519131e-02 1.00828089e-01 -5.02714276e-01 2.35787421e-01
1.05124927e+00 3.87770236e-01 -5.13095379e-01 -7.17303753e-01
-8.12584817e-01 3.47817272e-01 5.09786844e-01 -2.62186816e-03
-1.80601135e-01 2.13860631e-01 -2.24629715e-02 2.50473827e-01
8.84603620e-01 1.71355799e-01 -3.59721899e-01 4.80549969e-02
-1.17484570e+00 3.96771610e-01 5.12469351e-01 1.00299942e+00
8.23374927e-01 9.24698189e-02 3.48375261e-01 7.33062863e-01
1.55317172e-01 3.51424485e-01 -1.80461109e-02 -1.56877911e+00
4.85382766e-01 6.49831831e-01 1.34326965e-02 -7.16802418e-01
-2.05144987e-01 -1.86471894e-01 -7.96503723e-01 1.13213623e+00
2.70106882e-01 1.29006922e-01 -1.15949214e+00 1.51316905e+00
5.87143540e-01 3.42351906e-02 -2.23252743e-01 8.64354372e-01
8.43066573e-01 6.67962313e-01 -4.86865550e-01 1.92016229e-01
1.06024921e+00 -1.18885267e+00 -5.13769507e-01 -2.85174191e-01
3.31602484e-01 -8.83353770e-01 1.39776194e+00 6.87899530e-01
-1.70529342e+00 -4.01231319e-01 -1.14261484e+00 -7.47599483e-01
-6.80073798e-02 -1.14224225e-01 5.88071942e-01 5.78674078e-01
-1.48774242e+00 6.75095618e-01 -7.17299581e-01 -2.29360506e-01
5.60036421e-01 2.14019626e-01 -6.16768420e-01 -3.77561837e-01
-4.66846317e-01 7.32832015e-01 3.40063088e-02 -5.81316687e-02
-7.64462829e-01 -1.04264462e+00 -1.00683725e+00 -5.81824891e-02
4.15808648e-01 -1.11578536e+00 1.41944110e+00 -1.04713821e+00
-1.85562766e+00 1.16724193e+00 -1.96389318e-01 -2.99137563e-01
8.60440671e-01 -4.49894100e-01 1.47223547e-01 6.71498299e-01
-2.87764072e-01 9.41151500e-01 1.18069828e+00 -1.51980686e+00
-4.46282893e-01 -3.42498362e-01 5.71363628e-01 5.12084126e-01
-8.53852406e-02 -2.83628792e-01 -7.10990489e-01 -5.36278963e-01
1.73322279e-02 -6.70388699e-01 -1.42516419e-01 7.69391835e-01
-2.24077672e-01 2.36192808e-01 1.03354418e+00 -7.48355448e-01
7.02960372e-01 -2.18145013e+00 3.03968936e-01 -2.07772031e-01
6.31713390e-01 3.17369968e-01 -1.06764235e-01 5.32348871e-01
-1.27364853e-02 -6.57952130e-02 -2.92101920e-01 -1.31864798e+00
-1.59269139e-01 3.12378198e-01 -2.95993686e-01 3.85286242e-01
2.09571928e-01 9.09451187e-01 -5.43971062e-01 -3.34391505e-01
7.19616413e-01 7.17764735e-01 -7.55191267e-01 3.61768544e-01
-4.01408941e-01 2.84463286e-01 2.15769827e-01 7.49587417e-01
1.02895069e+00 -3.42184931e-01 -1.35918036e-01 -2.85848022e-01
-1.89420342e-01 3.21850955e-01 -1.03517258e+00 2.23374557e+00
-7.16740549e-01 6.11893833e-01 4.59565610e-01 -3.96704406e-01
5.88943660e-01 7.23236352e-02 8.23949724e-02 -7.36633539e-01
3.89332324e-02 6.50366023e-02 -5.34893155e-01 -3.70340437e-01
4.79826242e-01 -1.69608697e-01 3.47226620e-01 5.20186305e-01
4.98950593e-02 -6.33782864e-01 -9.41114314e-03 4.46363956e-01
7.72942305e-01 6.47443116e-01 1.23546412e-02 2.66216218e-01
1.78136215e-01 -4.33369011e-01 3.10552716e-01 3.92001361e-01
7.06148073e-02 1.34340310e+00 3.04040760e-01 -5.09366870e-01
-1.33963513e+00 -1.17568135e+00 1.88021392e-01 5.27597904e-01
2.70793021e-01 -4.17870313e-01 -7.06133306e-01 -6.11297011e-01
-1.75828040e-01 5.49105108e-01 -7.06514716e-01 2.40010411e-01
-5.50078034e-01 -4.93274592e-02 1.41921237e-01 3.16332668e-01
6.44761860e-01 -7.68201590e-01 -7.79960930e-01 -1.68570489e-01
3.98897752e-02 -1.28172469e+00 -5.54793298e-01 -1.33746490e-02
-1.09057891e+00 -8.85591865e-01 -1.05965805e+00 -5.99237025e-01
6.20570540e-01 8.33154619e-01 1.15672588e+00 5.48892058e-02
-3.58302981e-01 4.37152267e-01 -2.02909634e-01 -1.12764329e-01
-3.62125486e-01 -3.16522978e-02 -3.55621248e-01 -5.69798648e-02
-2.77817249e-01 -1.16117311e+00 -1.03541744e+00 -1.21398838e-02
-1.24726200e+00 6.68310940e-01 7.55433202e-01 6.55265808e-01
3.35665733e-01 -3.86718929e-01 -1.44401059e-01 -9.39763069e-01
4.93570752e-02 3.95947807e-02 -6.34276986e-01 5.86590134e-02
-5.83505511e-01 -2.33861819e-01 5.73894441e-01 -4.31583256e-01
-1.40384066e+00 -6.07905351e-02 -3.29838187e-01 -7.87134767e-01
-2.91183591e-01 -1.68024629e-01 -3.00182492e-01 -2.39522859e-01
5.11808157e-01 1.69947177e-01 2.34201059e-01 -7.25698411e-01
6.20992482e-01 3.02471936e-01 4.93464410e-01 -1.93914428e-01
1.07915747e+00 9.52525437e-01 -1.22438639e-01 -8.82396758e-01
-8.41188788e-01 -1.80614069e-01 -7.14022100e-01 -1.03971973e-01
7.61771023e-01 -1.12788045e+00 -6.68065667e-01 6.22369468e-01
-1.07119596e+00 -7.15692818e-01 -3.54395866e-01 7.58821964e-02
-5.88799715e-01 6.26848161e-01 -7.92889059e-01 -4.41291213e-01
-3.41471255e-01 -9.92037714e-01 1.54058230e+00 2.78860182e-01
-1.07317328e-01 -9.97304201e-01 -5.59660979e-02 7.01740086e-01
2.92637646e-01 4.61178333e-01 7.66163886e-01 3.26079220e-01
-1.04312789e+00 4.83908271e-03 -3.35516781e-01 8.36506605e-01
-4.27671187e-02 6.59809262e-02 -1.31310117e+00 -2.77232558e-01
4.40566577e-02 -3.26538682e-01 7.88910627e-01 3.14062744e-01
1.04349828e+00 -3.47374141e-01 6.19828030e-02 1.30755627e+00
1.50095356e+00 -5.03377169e-02 7.92515635e-01 2.55000055e-01
9.93809342e-01 4.20035273e-01 1.82437390e-01 2.08717600e-01
4.97511566e-01 7.66198397e-01 6.01123393e-01 -7.39495456e-01
-6.72579527e-01 -3.72087449e-01 4.50073481e-01 4.96342152e-01
-2.36218069e-02 -2.08145767e-01 -3.83488268e-01 3.46125126e-01
-1.46027195e+00 -9.11930621e-01 7.81666040e-02 1.98670149e+00
8.33172083e-01 1.11789599e-01 2.97148749e-02 4.35606763e-02
6.77360073e-02 4.17774171e-01 -6.12802148e-01 -4.97121006e-01
-2.64004827e-01 3.19385052e-01 4.36568677e-01 8.82999659e-01
-6.93101168e-01 8.51946294e-01 5.74435568e+00 5.06627977e-01
-1.19155049e+00 3.40988152e-02 6.41457498e-01 -7.75548279e-01
-5.60699821e-01 3.16106230e-02 -6.50727034e-01 2.36062393e-01
2.95524836e-01 1.45916373e-01 3.38071197e-01 6.83847427e-01
3.26688439e-01 -4.26504523e-01 -1.13711154e+00 1.23885155e+00
3.20203811e-01 -1.54182935e+00 2.58238941e-01 8.72508511e-02
9.02713001e-01 9.63159353e-02 2.17830628e-01 -3.71462643e-01
7.77152330e-02 -8.18278790e-01 8.32605004e-01 2.96108633e-01
8.68400097e-01 -6.60943270e-01 3.00347120e-01 2.99163610e-01
-6.52861714e-01 1.10972315e-01 -2.10667446e-01 -3.19242656e-01
5.13043821e-01 7.13889778e-01 -7.39457548e-01 5.05798042e-01
9.93180931e-01 9.61842537e-01 -6.07283831e-01 9.16137338e-01
-4.25136954e-01 7.65834823e-02 -3.93619150e-01 5.26518583e-01
2.75396705e-01 -2.09570408e-01 5.51044047e-01 9.60761189e-01
1.61507651e-01 -1.98747478e-02 -1.83266133e-01 9.39835727e-01
-2.12927893e-01 -2.23524660e-01 -9.09267962e-01 4.99858707e-01
9.94700938e-02 1.23293197e+00 -6.01270735e-01 -2.88530082e-01
-4.98085767e-01 1.53582191e+00 3.55407834e-01 4.46072012e-01
-5.13345122e-01 -2.20367134e-01 7.22557902e-01 4.92325991e-01
5.02897859e-01 -4.72380638e-01 -2.87780046e-01 -1.44497025e+00
4.04861450e-01 -9.22217727e-01 -4.15317491e-02 -1.14808524e+00
-1.05862129e+00 5.56831896e-01 1.32262215e-01 -1.28542221e+00
-1.73223630e-01 -5.40933967e-01 -5.61120987e-01 8.08918953e-01
-1.76510429e+00 -1.39520943e+00 -4.80086416e-01 6.99881732e-01
8.22397351e-01 2.81789660e-01 6.23568475e-01 3.20634961e-01
-2.68584162e-01 4.96490479e-01 -1.23271514e-02 -4.11973715e-01
9.33740377e-01 -1.31066895e+00 7.37750947e-01 1.06512463e+00
6.02140417e-03 5.06667435e-01 5.34135997e-01 -3.51863295e-01
-1.27127242e+00 -7.92901874e-01 7.16402173e-01 -5.96588314e-01
-1.05768489e-02 -8.47632229e-01 -7.87383199e-01 6.84302628e-01
5.91975331e-01 -3.04950386e-01 4.31451946e-01 -2.03861147e-01
-6.19072855e-01 -1.81380704e-01 -1.12988865e+00 9.66154277e-01
1.19285858e+00 -4.77803707e-01 -4.15680677e-01 1.61821380e-01
8.98485482e-01 -6.13536775e-01 -5.66926122e-01 1.10315450e-01
7.56662726e-01 -1.69312823e+00 1.24448919e+00 7.75346830e-02
8.83840322e-01 -5.05462289e-01 5.17395139e-02 -1.09102452e+00
-5.46417758e-02 -1.07996333e+00 -1.85680807e-01 1.03136241e+00
1.31214574e-01 -5.59938967e-01 8.67779016e-01 6.91542983e-01
-2.14621142e-01 -7.82724679e-01 -6.51858747e-01 -5.57310283e-01
-1.92728773e-01 -3.30110788e-01 2.91728407e-01 6.68703496e-01
-5.18147230e-01 3.53991449e-01 -7.34569013e-01 -8.39411560e-03
9.80861485e-01 1.54337287e-01 1.28062332e+00 -8.83587182e-01
-6.18465304e-01 -3.60619366e-01 -2.62874633e-01 -1.47077811e+00
-1.78954229e-01 -6.49437189e-01 -2.34259993e-01 -1.58896470e+00
-6.23745807e-02 -1.01329125e-01 5.36800742e-01 3.16525221e-01
2.32513696e-02 6.41791582e-01 5.30508339e-01 1.78388074e-01
-3.12457085e-01 6.01154864e-01 1.56956339e+00 1.56766936e-01
-3.65373522e-01 7.87569955e-03 -7.07714140e-01 9.30352271e-01
6.03535831e-01 -2.76594520e-01 -7.56374240e-01 -9.36386824e-01
2.99460620e-01 -8.50156322e-02 6.52296484e-01 -9.06251490e-01
3.64861377e-02 6.93619102e-02 7.28598773e-01 -7.92752385e-01
7.94793129e-01 -8.93833518e-01 3.14210951e-01 2.39631325e-01
-1.80602968e-01 1.43539175e-01 3.34988773e-01 5.04633605e-01
-5.46192862e-02 8.69668573e-02 8.93783271e-01 -3.98935944e-01
-5.92850864e-01 4.08792675e-01 4.21176441e-02 -1.62748352e-01
8.55894804e-01 -6.33355081e-01 -1.85327694e-01 -5.92401743e-01
-5.91485560e-01 4.36237343e-02 1.21596909e+00 5.01135647e-01
8.76376688e-01 -9.82586384e-01 -5.35511136e-01 4.46160704e-01
-5.83475716e-02 4.63361204e-01 7.05122948e-01 5.68047583e-01
-1.01296937e+00 -4.58957665e-02 -2.89825499e-01 -7.72348881e-01
-1.50897300e+00 3.07036370e-01 4.00721192e-01 -1.56334355e-01
-1.28655231e+00 8.96675527e-01 7.59901464e-01 -2.56389529e-01
4.27646250e-01 -5.69657028e-01 4.64679748e-01 -3.06575775e-01
7.65542746e-01 4.41921838e-02 -4.71157059e-02 -1.10596754e-01
-6.30743578e-02 7.62369275e-01 -3.68701667e-01 -2.74719387e-01
1.68760109e+00 -6.39087737e-01 2.05117539e-02 4.44567800e-01
1.35934103e+00 1.08722366e-01 -1.89059865e+00 -2.52736956e-01
-7.30066895e-01 -9.05252934e-01 -1.90864652e-02 -9.17890728e-01
-1.03421879e+00 1.09213138e+00 4.34113830e-01 -2.18718797e-01
1.33569479e+00 -5.69464229e-02 9.93792593e-01 9.84126925e-02
1.73831239e-01 -7.40164399e-01 1.36613190e-01 1.51603431e-01
1.04552245e+00 -1.21336210e+00 2.04221264e-01 -5.34040511e-01
-3.78330708e-01 1.12802780e+00 6.09160900e-01 -3.11168104e-01
4.33465570e-01 3.89656246e-01 3.65737230e-01 -1.10930353e-01
-8.90818119e-01 2.74072937e-03 2.67855406e-01 7.17221558e-01
1.15270190e-01 -3.48216891e-01 1.60116211e-01 -6.83918893e-02
-1.36195987e-01 -9.79138836e-02 7.88840830e-01 7.81135499e-01
-7.73083195e-02 -1.28511846e+00 -1.17919967e-01 -5.96009158e-02
-3.48583579e-01 -1.58930019e-01 -5.53089976e-01 9.36232030e-01
1.96056929e-03 5.96135139e-01 -1.36933446e-01 -1.08955823e-01
3.98473531e-01 -1.59766302e-01 8.64300847e-01 -7.15210378e-01
-5.90969741e-01 2.28311494e-01 -7.08310232e-02 -1.08140922e+00
-5.56124866e-01 -4.48484927e-01 -6.53707385e-01 -5.39220214e-01
-7.12488825e-03 -6.29632175e-01 6.73296273e-01 6.60561383e-01
6.78816259e-01 2.58882195e-01 5.93683600e-01 -1.52117860e+00
-1.72870979e-01 -6.15054965e-01 -5.24719596e-01 5.23938119e-01
7.07602322e-01 -3.92718822e-01 -4.39648002e-01 2.87767202e-01] | [9.260276794433594, -3.0565261840820312] |
10cd6d92-3b06-4c9d-b07d-aeebb42c63f8 | almost-no-label-no-cry | null | null | http://papers.nips.cc/paper/5453-almost-no-label-no-cry | http://papers.nips.cc/paper/5453-almost-no-label-no-cry.pdf | (Almost) No Label No Cry | In Learning with Label Proportions (LLP), the objective is to learn a supervised classifier when, instead of labels, only label proportions for bags of observations are known. This setting has broad practical relevance, in particular for privacy preserving data processing. We first show that the mean operator, a statistic which aggregates all labels, is minimally sufficient for the minimization of many proper scoring losses with linear (or kernelized) classifiers without using labels. We provide a fast learning algorithm that estimates the mean operator via a manifold regularizer with guaranteed approximation bounds. Then, we present an iterative learning algorithm that uses this as initialization. We ground this algorithm in Rademacher-style generalization bounds that fit the LLP setting, introducing a generalization of Rademacher complexity and a Label Proportion Complexity measure. This latter algorithm optimizes tractable bounds for the corresponding bag-empirical risk. Experiments are provided on fourteen domains, whose size ranges up to 300K observations. They display that our algorithms are scalable and tend to consistently outperform the state of the art in LLP. Moreover, in many cases, our algorithms compete with or are just percents of AUC away from the Oracle that learns knowing all labels. On the largest domains, half a dozen proportions can suffice, i.e. roughly 40K times less than the total number of labels. | ['Giorgio Patrini', 'Tiberio Caetano', 'Paul Rivera', 'Richard Nock'] | 2014-12-01 | null | null | null | neurips-2014-12 | ['style-generalization'] | ['computer-vision'] | [ 2.57488072e-01 4.38916266e-01 -3.17472875e-01 -6.32619619e-01
-1.30572259e+00 -1.12193036e+00 2.82452703e-01 6.08807147e-01
-6.75060928e-01 7.59014666e-01 -1.84575886e-01 -2.72693813e-01
-2.95225829e-01 -5.17428219e-01 -9.83189523e-01 -1.13452816e+00
-1.19362846e-01 5.84411502e-01 -2.28536859e-01 4.63268697e-01
3.52468155e-02 4.70013201e-01 -1.39095747e+00 5.77080697e-02
5.91760397e-01 1.20160878e+00 -5.75587094e-01 5.50766468e-01
2.75623053e-01 5.06588042e-01 -2.14503914e-01 -7.49370813e-01
8.05977345e-01 -2.17845023e-01 -8.78399491e-01 2.00747833e-01
9.53673065e-01 -3.00980657e-01 -1.68808326e-01 1.28500605e+00
3.97987157e-01 1.28191665e-01 1.07520962e+00 -1.37540734e+00
-4.09248620e-01 5.98732352e-01 -5.11834562e-01 -2.67357886e-01
9.61541105e-03 -1.32676899e-01 1.47286987e+00 -7.28336751e-01
3.02596986e-01 1.03678572e+00 8.48457992e-01 5.14485300e-01
-1.61985111e+00 -5.53019643e-01 3.14211659e-02 -2.85815775e-01
-1.49761021e+00 -4.31810617e-01 1.20731600e-01 -6.77170932e-01
4.13582593e-01 5.46892881e-01 2.33138409e-02 5.57996213e-01
-1.08591080e-01 6.46814406e-01 1.23058212e+00 -5.57169497e-01
4.49233770e-01 8.07286799e-01 7.31525660e-01 8.54652286e-01
4.60000485e-01 9.37337950e-02 -5.29171705e-01 -7.25654900e-01
2.47474819e-01 1.14870079e-01 -2.91653067e-01 -1.00243354e+00
-1.02087271e+00 1.12315857e+00 5.50390705e-02 -2.68696994e-01
2.04186022e-01 1.14807941e-01 3.95864367e-01 4.95101869e-01
6.41040266e-01 4.30462360e-01 -6.16963029e-01 3.98862720e-01
-8.06937814e-01 2.02126339e-01 1.20388377e+00 1.15275979e+00
8.10702622e-01 -7.37417042e-01 -2.11497933e-01 6.84939265e-01
4.50626276e-02 7.03121543e-01 -3.06284800e-02 -1.13111079e+00
5.66309988e-01 2.59580404e-01 4.48284209e-01 -7.12244153e-01
-3.77569258e-01 -3.05408478e-01 -7.71653056e-01 1.09813519e-01
1.00552011e+00 -1.04015619e-01 -3.85459006e-01 1.99622309e+00
5.52609742e-01 -5.90566508e-02 -3.76415998e-02 6.40583873e-01
-8.26548412e-02 4.45914328e-01 -2.03726888e-02 -5.05644977e-01
1.43263566e+00 -6.34772778e-01 -4.97227818e-01 8.78230035e-02
1.12020397e+00 -2.50000656e-01 1.13783181e+00 5.93733907e-01
-7.58849263e-01 -6.19311258e-02 -9.34486210e-01 -1.24060199e-01
-4.48810428e-01 1.34180367e-01 7.57977366e-01 1.21544838e+00
-8.98609579e-01 7.95548916e-01 -6.29346490e-01 -2.92832345e-01
7.05144346e-01 4.58878845e-01 -5.92183411e-01 -1.07852727e-01
-8.86593580e-01 5.85609078e-01 3.76197428e-01 -3.26486140e-01
-7.91018248e-01 -9.49161589e-01 -7.85651445e-01 2.59507187e-02
3.02064061e-01 -4.03529316e-01 1.19023418e+00 -6.53442264e-01
-1.08337426e+00 1.23502707e+00 -1.68712568e-02 -7.30597556e-01
8.33051026e-01 -2.18753338e-01 8.80488008e-02 6.76493421e-02
-3.48123275e-02 4.02130514e-01 8.37595105e-01 -1.08286631e+00
-8.94767284e-01 -5.56108177e-01 1.28538243e-03 1.40370518e-01
-7.54594564e-01 -2.32095703e-01 -2.89347172e-02 -2.64830679e-01
1.07452758e-02 -9.45725262e-01 -1.62934005e-01 4.71925259e-01
-4.40979362e-01 -2.91568816e-01 4.53976929e-01 -5.02289832e-01
9.95561898e-01 -2.49638534e+00 -2.07178205e-01 4.48953003e-01
3.60330403e-01 -9.49435607e-02 1.65435925e-01 3.08260351e-01
-9.27983969e-02 6.39297962e-02 -5.69862604e-01 -6.34095848e-01
5.68102062e-01 1.89840980e-02 -6.67193711e-01 1.25764215e+00
-9.73571688e-02 6.36874914e-01 -7.16462851e-01 -3.04043949e-01
-7.81133175e-02 -4.66297045e-02 -6.54274821e-01 5.56852221e-02
-9.25140604e-02 2.26496696e-01 -9.35691670e-02 4.44923818e-01
9.62303579e-01 -4.46172595e-01 1.03545226e-01 1.06861539e-01
4.04999644e-01 -1.07853137e-01 -1.32796252e+00 1.35591733e+00
-2.01682299e-01 3.06108534e-01 2.48909250e-01 -1.11608613e+00
5.48421204e-01 1.35620713e-01 5.24922132e-01 2.61558145e-01
1.49136081e-01 3.68741453e-01 -6.67942345e-01 -1.60280406e-01
1.63930766e-02 -3.93362671e-01 -2.93989718e-01 6.07960761e-01
1.05334833e-01 1.90031156e-01 -9.92684960e-02 9.36399624e-02
1.24054050e+00 -5.57804942e-01 5.72573125e-01 -4.40631419e-01
4.52993333e-01 -3.26201707e-01 5.14097035e-01 1.00172400e+00
-2.26572707e-01 5.54046273e-01 8.36840689e-01 -3.03491145e-01
-9.00138378e-01 -1.12026966e+00 -6.61911547e-01 1.23649871e+00
-3.86127084e-02 -1.63279280e-01 -9.19499934e-01 -1.16217482e+00
4.85820085e-01 7.09061861e-01 -7.64222324e-01 -1.77139506e-01
5.94528839e-02 -9.98795271e-01 5.50483644e-01 1.71243131e-01
2.24931180e-01 -4.70365554e-01 -1.63119227e-01 -1.60246044e-01
1.28963202e-01 -8.64187777e-01 -7.74564803e-01 5.21546125e-01
-8.38185370e-01 -1.24513113e+00 -5.72640657e-01 -5.98797977e-01
9.18469965e-01 -1.29774651e-02 8.46311867e-01 -3.75612915e-01
-1.28338531e-01 5.79997659e-01 -8.79721344e-02 -3.50438535e-01
-2.62680441e-01 1.82180479e-01 2.71526426e-01 2.92114079e-01
6.55757725e-01 -5.11299849e-01 -2.61540711e-01 2.15065747e-01
-9.27373409e-01 -4.88912642e-01 4.58076775e-01 9.30776656e-01
8.72248471e-01 1.98760584e-01 3.59899521e-01 -1.62014413e+00
2.67093271e-01 -6.19457364e-01 -1.07519007e+00 4.50089872e-01
-8.70807111e-01 1.77430764e-01 7.74741769e-01 -4.08873379e-01
-5.44154048e-01 5.29379487e-01 4.43278491e-01 -3.63415420e-01
-1.81351587e-01 -1.99021101e-01 -4.93368328e-01 -2.58420408e-01
7.64463127e-01 3.42650376e-02 8.97690207e-02 -5.23018837e-01
6.21333003e-01 8.36476862e-01 4.28515017e-01 -6.95722640e-01
8.49525809e-01 7.03088760e-01 2.93717384e-01 -4.30741221e-01
-1.27146614e+00 -6.84195220e-01 -7.28226364e-01 2.90116072e-01
5.78186095e-01 -8.02109599e-01 -1.03062415e+00 4.44196761e-01
-7.32446194e-01 -2.77973264e-01 -8.05982232e-01 3.75399947e-01
-7.80559063e-01 4.84307289e-01 -5.75472891e-01 -1.12587094e+00
-1.13461345e-01 -6.36949718e-01 1.03254247e+00 -1.45023018e-01
1.81066617e-01 -1.07854283e+00 4.23196405e-02 3.54158163e-01
-8.79072100e-02 2.52924025e-01 8.83210659e-01 -1.29279733e+00
-3.33123326e-01 -5.04878163e-01 -2.61213303e-01 7.50855207e-01
3.78064322e-03 -7.41207182e-01 -1.32316136e+00 -7.19424665e-01
2.63840109e-01 -5.06279409e-01 9.67003286e-01 2.22161546e-01
1.53767276e+00 -7.75417268e-01 -2.89023697e-01 7.67698944e-01
1.52169931e+00 -3.94492745e-01 1.04877859e-01 -9.77711529e-02
7.20636606e-01 8.57662380e-01 7.08008230e-01 7.19760716e-01
1.06085859e-01 4.93435055e-01 1.49655476e-01 2.12124035e-01
3.46581101e-01 -2.88011581e-01 4.23940718e-01 3.83655459e-01
6.52361453e-01 -6.28262535e-02 -6.10779166e-01 4.66463417e-01
-1.75070977e+00 -5.92321873e-01 -7.09669665e-02 2.91075730e+00
1.11146545e+00 -3.47382203e-02 2.83457786e-01 1.14915788e-01
7.36753285e-01 -1.18049152e-01 -9.00443137e-01 -2.44458660e-01
-1.51542976e-01 1.24513820e-01 1.39175558e+00 7.46894658e-01
-1.56903374e+00 5.74914455e-01 6.18462706e+00 1.01171398e+00
-3.46455902e-01 3.80022019e-01 9.10585582e-01 -1.42043397e-01
-1.20340124e-01 -6.73435675e-03 -1.17300403e+00 4.97637928e-01
1.08377564e+00 -1.24735914e-01 6.92988455e-01 1.19093323e+00
-3.26919973e-01 -4.25327048e-02 -1.76600790e+00 1.11653626e+00
3.89335603e-02 -7.70692050e-01 -3.59989226e-01 6.08549476e-01
8.42957020e-01 -2.75339186e-01 3.45130384e-01 1.46354496e-01
6.67906284e-01 -1.07113266e+00 5.61278880e-01 3.65248650e-01
1.13261425e+00 -8.49004865e-01 7.60841846e-01 5.29414296e-01
-7.48584032e-01 -3.09309423e-01 -7.51550138e-01 1.45344719e-01
-2.29077980e-01 1.04391634e+00 -7.05080986e-01 3.13238084e-01
4.17168587e-01 4.81386006e-01 -4.04995203e-01 9.93740916e-01
7.40108714e-02 7.03369141e-01 -7.69064248e-01 1.07989147e-01
-9.64562073e-02 -2.10669816e-01 3.37615013e-01 1.34260535e+00
1.33474901e-01 2.01696768e-01 1.08385891e-01 4.22087759e-01
-5.30854762e-01 4.63890493e-01 -7.56495655e-01 1.79926038e-01
7.20280707e-01 1.25137794e+00 -5.43191314e-01 -1.91834584e-01
-3.10999662e-01 1.02007353e+00 5.83653569e-01 1.94375813e-01
-6.07570350e-01 -3.36664110e-01 7.48823404e-01 9.34471935e-03
3.68666261e-01 2.43376657e-01 -3.65425050e-01 -9.92218614e-01
1.76516071e-01 -7.69468069e-01 8.40134978e-01 4.05149944e-02
-2.03470588e+00 -5.76760955e-02 3.89460027e-02 -1.24020052e+00
8.01185593e-02 -9.41908121e-01 6.64586574e-02 7.33496726e-01
-1.24389088e+00 -7.41658211e-01 1.03289761e-01 5.32535076e-01
-1.76255852e-01 -1.22556739e-01 1.05745423e+00 2.34094441e-01
-5.01240134e-01 1.19563663e+00 7.64533401e-01 4.41834331e-02
8.98609281e-01 -1.76808882e+00 -1.66474164e-01 6.58724606e-01
3.92208725e-01 3.32837045e-01 6.16272748e-01 -1.16093583e-01
-1.16435623e+00 -1.31978214e+00 7.53993034e-01 -8.34456563e-01
5.72676778e-01 -7.03583598e-01 -7.97413707e-01 9.11918283e-01
-4.39970255e-01 4.75913554e-01 1.06113982e+00 1.22206941e-01
-7.94336438e-01 -5.48099518e-01 -1.63277316e+00 1.06673412e-01
9.74543512e-01 -5.72827518e-01 -1.20335728e-01 9.80460048e-01
6.71936810e-01 -4.02498811e-01 -9.20941651e-01 -1.89446528e-02
3.89968067e-01 -6.82005227e-01 7.20197141e-01 -7.68820643e-01
-1.42049670e-01 -1.83843717e-01 -5.00187933e-01 -9.68241632e-01
-1.35801658e-01 -8.46947372e-01 -2.80358464e-01 1.38941348e+00
5.04168570e-01 -1.13912153e+00 9.56327975e-01 1.06493700e+00
4.93498951e-01 -7.38616824e-01 -1.14489102e+00 -1.03856003e+00
3.31823975e-01 -2.97295898e-01 4.81645614e-01 1.03884327e+00
-7.91825633e-03 7.66719058e-02 -4.25106734e-01 4.63180751e-01
1.13336551e+00 -1.00333340e-01 6.90327823e-01 -1.38595247e+00
-5.50067842e-01 -1.80508450e-01 -5.38480759e-01 -1.04769754e+00
3.84225577e-01 -1.31126714e+00 -2.76331473e-02 -7.58137763e-01
4.64542121e-01 -7.22931206e-01 -4.59544122e-01 6.27097487e-01
-7.28162304e-02 -3.32696550e-02 1.68264620e-02 6.81627318e-02
-7.43399501e-01 2.36336961e-01 6.53588533e-01 -6.97726309e-02
2.69284043e-02 3.66411954e-01 -9.29847062e-01 6.25774503e-01
7.64957130e-01 -8.11799288e-01 -4.44298267e-01 2.37574577e-02
1.79702386e-01 -3.58355463e-01 3.97061795e-01 -8.43276918e-01
2.22086161e-01 1.22300684e-02 1.25739723e-01 -2.64558822e-01
1.48977041e-01 -1.12894392e+00 5.09037171e-03 2.84512192e-01
-8.86801660e-01 -6.15239978e-01 -2.61792272e-01 1.04950678e+00
1.97961003e-01 -5.14148355e-01 1.00605822e+00 2.38999203e-01
4.48549539e-02 5.33674538e-01 9.96481627e-02 4.27685559e-01
1.25827408e+00 1.50236115e-01 -1.37411147e-01 -3.13950270e-01
-6.41182303e-01 3.29122275e-01 4.85895276e-01 -3.22126240e-01
7.01723248e-02 -1.33626783e+00 -6.68092072e-01 3.78862023e-02
2.86227614e-01 1.54034391e-01 -3.25177647e-02 8.21543515e-01
-2.72217870e-01 4.88730907e-01 3.99807274e-01 -3.99639964e-01
-1.22793460e+00 9.07804787e-01 3.54248673e-01 -3.35316360e-01
-5.45126557e-01 1.23582089e+00 5.05219102e-01 -6.72356904e-01
7.49399304e-01 -2.40761116e-01 3.05100471e-01 1.60961464e-01
7.13149488e-01 5.33248007e-01 1.39247969e-01 -2.76989728e-01
-2.14660972e-01 3.43741745e-01 -2.16477901e-01 -2.01955125e-01
1.15977311e+00 -1.74387604e-01 -1.89285100e-01 6.32886350e-01
1.64279056e+00 1.04782008e-01 -1.53911471e+00 -5.29892981e-01
1.87160075e-01 -6.39961541e-01 -3.27415109e-01 -6.30207658e-01
-1.04040873e+00 8.85381103e-01 7.73504555e-01 3.72164279e-01
1.06918752e+00 2.55285114e-01 6.28624558e-01 7.07664549e-01
6.01346314e-01 -9.65021610e-01 -4.40292746e-01 6.39217570e-02
3.98079813e-01 -1.23574793e+00 1.30752161e-01 -5.22751033e-01
-5.36525667e-01 7.58567512e-01 2.09896073e-01 -9.72501189e-02
9.07700837e-01 7.83730075e-02 -1.30260304e-01 5.46688326e-02
-6.38735056e-01 1.85903057e-01 1.63579792e-01 5.94400465e-01
-1.10211261e-01 3.56050938e-01 -2.40563720e-01 8.21274102e-01
-2.94335186e-01 -4.10857588e-01 2.27682680e-01 4.65185970e-01
-4.73687649e-01 -1.01962268e+00 -3.80689859e-01 9.06382382e-01
-6.54892623e-01 9.94804315e-03 -2.57678807e-01 4.29862946e-01
3.11936233e-02 8.77988458e-01 -1.14486851e-01 -9.54756588e-02
1.96045503e-01 4.14732158e-01 2.93030024e-01 -6.02236629e-01
-1.63729712e-01 -4.48196530e-01 -2.21082628e-01 -6.08578503e-01
-1.16910741e-01 -1.05020535e+00 -8.74182999e-01 -4.99378115e-01
-4.90126878e-01 2.18274653e-01 4.07347888e-01 6.95046663e-01
5.47512844e-02 -3.50885183e-01 1.01359785e+00 -3.14564824e-01
-1.39324033e+00 -7.44765580e-01 -1.36681437e+00 3.48955750e-01
4.54303026e-01 -4.15386200e-01 -1.04282081e+00 9.52647701e-02] | [8.293940544128418, 4.240356922149658] |
d678feb0-36fc-41c8-8a47-bcf31f0ef798 | solitary-pulmonary-nodules-prediction-for | 2305.10466 | null | https://arxiv.org/abs/2305.10466v1 | https://arxiv.org/pdf/2305.10466v1.pdf | Solitary pulmonary nodules prediction for lung cancer patients using nomogram and machine learning | Lung cancer(LC) is a type of malignant neoplasm that originates in the bronchial mucosa or glands.As a clinically common nodule,solitary pulmonary nodules(SPNs) have a significantly higher probability of malignancy when they are larger than 8 mm in diameter.But there is also a risk of lung cancer when the diameter is less than 8mm,the purpose of this study was to create a nomogram for estimating the likelihood of lung cancer in patients with SPNs of 8 mm or smaller using computed tomography(CT) scans and biomarker information.Use CT scans and various biomarkers as input to build predictive models for the likelihood of lung cancer in patients with SPNs of 8 mm or less.The age,precursor gastrin-releasing peptide (ProGRP),gender,Carcinoembryonic Antigen(CEA),and stress corrosion cracking(SCC) were independent key tumor markers and were entered into the nomogram.The developed nomogram demonstrated strong accuracy in predicting lung cancer risk,with an internal validation area under the receiver operating characteristics curve(ROC) of 0.8474.The calibration curves plotted showed that the nomogram predicted the probability of lung cancer with good agreement with the actual probability.In this study,we finally succeeded in constructing a suitable nomogram that could predict the risk of lung cancer in patients with SPNs<=8 mm in diameter.The model has a high level of accuracy and is able to accurately distinguish between different patients,allowing clinicians to develop personalized treatment plans for individuals with SPNs. | ['Gongjin Song', 'Hailan Zhang'] | 2023-05-17 | null | null | null | null | ['computed-tomography-ct'] | ['methodology'] | [-2.82677621e-01 3.04918528e-01 -6.76199913e-01 2.66989201e-01
-3.88289958e-01 -1.40415519e-01 2.28446558e-01 2.62187183e-01
-2.45210141e-01 4.87021983e-01 8.10802355e-02 -8.23065996e-01
-1.64962515e-01 -9.87293303e-01 -1.20980456e-01 -6.66274071e-01
-3.31658348e-02 9.85927284e-01 6.08307779e-01 2.30920091e-01
-2.16982275e-01 8.20567548e-01 -1.02826464e+00 -4.95081469e-02
9.62222755e-01 8.76919091e-01 7.32686698e-01 5.94048381e-01
-1.02968514e-01 4.67480630e-01 8.91659483e-02 7.87436813e-02
2.23464563e-01 -6.36621535e-01 -3.83556724e-01 -1.59835517e-01
-4.30809520e-02 -1.62896231e-01 9.45630819e-02 5.20623863e-01
7.98943713e-02 -6.06030464e-01 1.06531107e+00 -9.21591043e-01
1.11397080e-01 5.92354715e-01 -2.41741136e-01 -4.77203727e-02
-4.10449728e-02 -7.62002468e-02 6.57204807e-01 -8.23283911e-01
2.46320277e-01 6.87969148e-01 7.69816637e-01 7.22022474e-01
-6.42757773e-01 -5.14549315e-01 -4.70548421e-01 -4.81992885e-02
-1.21197832e+00 5.05918860e-01 -1.81729823e-01 -5.57965398e-01
5.40138066e-01 7.71928132e-01 1.22694397e+00 3.50224584e-01
9.11969304e-01 6.31095022e-02 9.00477052e-01 -4.45728928e-01
4.14067619e-02 5.23210466e-01 -2.39826471e-01 8.54275584e-01
1.11352980e+00 2.11696133e-01 1.58994853e-01 -3.34802985e-01
8.51526678e-01 5.83337128e-01 -1.04860865e-01 -3.05332392e-01
-1.20748591e+00 8.03605199e-01 4.83750433e-01 1.07369709e+00
-2.60893822e-01 2.00576097e-01 1.48327220e-02 -1.77527830e-01
-2.45193303e-01 1.03548646e-01 -4.30437744e-01 3.31765234e-01
-5.91949582e-01 -1.65351704e-01 1.12219739e+00 3.81157696e-01
1.04690976e-01 -3.76400501e-01 -6.02917038e-02 6.76770926e-01
6.10943377e-01 7.07075179e-01 1.16491604e+00 -1.90281853e-01
3.90638635e-02 8.59710991e-01 -1.21499710e-01 -1.58858746e-01
-5.52969635e-01 -4.11888003e-01 -9.07899380e-01 -1.30639181e-01
5.47266722e-01 2.68655755e-02 -8.85319948e-01 1.23652351e+00
-2.58576777e-02 -1.41084880e-01 2.57764477e-02 5.85757613e-01
4.75433946e-01 2.63684034e-01 5.54683387e-01 -4.04802680e-01
1.76052630e+00 -7.91350424e-01 -2.46150777e-01 -8.20043311e-02
9.15112495e-01 -3.61026794e-01 7.27928400e-01 4.02544811e-02
-8.37937593e-01 -2.30778933e-01 -6.89124763e-01 5.01369476e-01
-6.27217889e-02 5.02277792e-01 5.79761028e-01 1.06868887e+00
-8.62146080e-01 4.90110308e-01 -9.58883107e-01 -9.50044453e-01
-2.24647492e-01 5.12401581e-01 -2.56805360e-01 2.35342279e-01
-1.25246620e+00 1.12165082e+00 5.09931326e-01 -3.76155227e-01
-5.63240051e-01 -6.11123621e-01 -2.36320317e-01 2.78106451e-01
1.89294890e-01 -1.23857713e+00 1.14082789e+00 -3.97216439e-01
-1.00856233e+00 7.28804410e-01 -9.85829681e-02 -4.28643554e-01
3.87583137e-01 3.56556177e-01 -4.14141059e-01 1.84970960e-01
4.84078750e-02 1.34144828e-01 4.72610712e-01 -6.23753905e-01
-9.57044840e-01 -4.64921802e-01 -7.56515205e-01 2.28466243e-01
-2.87610054e-01 -3.77367705e-01 -1.45675659e-01 -2.78004080e-01
4.40027565e-01 -1.31103885e+00 -5.64893842e-01 -5.50580993e-02
-6.65461272e-02 -5.23836017e-01 6.55907035e-01 -4.42771226e-01
1.25928068e+00 -1.80342317e+00 -4.15191114e-01 6.09032214e-01
3.57123643e-01 4.11526710e-02 5.72695971e-01 3.04791421e-01
1.11007988e-01 7.85321474e-01 -2.54389718e-02 8.77996504e-01
-1.22582056e-01 1.63150653e-01 5.10503471e-01 1.57986268e-01
-1.48312256e-01 1.02038109e+00 -7.23640978e-01 -8.13326180e-01
2.30763227e-01 1.15722604e-01 -3.60653438e-02 8.12690333e-02
-1.63058326e-01 -4.63586077e-02 -7.69234180e-01 5.79584360e-01
3.74263823e-01 -4.58765090e-01 5.46247423e-01 2.11885497e-01
-9.81304646e-02 -2.06882223e-01 -6.34201527e-01 5.60758471e-01
-3.84354681e-01 8.61108899e-02 -1.56122133e-01 -7.92958885e-02
1.03361416e+00 7.87751973e-01 7.42940843e-01 -2.16666594e-01
-2.91877426e-02 7.44580865e-01 4.14771825e-01 -6.60531700e-01
-3.73469204e-01 -8.30216765e-01 2.34349087e-01 2.50265479e-01
-5.76073110e-01 -5.91843009e-01 3.30094278e-01 -2.55552791e-02
1.19462967e+00 -7.69560039e-01 1.10456586e+00 -3.17380220e-01
9.48292315e-01 -5.52142784e-02 1.33647576e-01 4.47876692e-01
1.11172497e-01 1.51445135e-01 8.34528327e-01 -1.30648747e-01
-8.00810754e-01 -1.10467136e+00 -7.45488286e-01 1.90928847e-01
-1.31224245e-01 4.67329137e-02 -3.08888614e-01 -7.41844237e-01
3.96244645e-01 6.71747327e-01 -5.25497198e-01 3.16894650e-01
-2.93657690e-01 -9.10445988e-01 3.22320342e-01 4.98590201e-01
4.36267287e-01 -6.28023565e-01 -3.16632330e-01 2.79595703e-01
1.43809691e-01 -4.71817225e-01 -4.35837060e-02 4.04163092e-01
-1.47896981e+00 -1.53640258e+00 -8.55302274e-01 -7.80388713e-01
7.98529625e-01 2.37294436e-01 8.03559124e-01 6.49245918e-01
-1.78353876e-01 4.52499539e-01 6.86072707e-02 -6.56632066e-01
-8.29311609e-01 2.14131236e-01 -2.08401620e-01 -5.41473866e-01
2.48795658e-01 -2.60379493e-01 -6.92765236e-01 8.04741561e-01
-7.55938411e-01 -4.60345149e-02 1.35150945e+00 5.45244336e-01
7.65177846e-01 3.94014753e-02 6.27397895e-01 -9.81736481e-01
4.94883716e-01 -9.69824493e-01 -2.76828051e-01 1.14944398e-01
-1.14495325e+00 -2.01568633e-01 7.20112562e-01 -1.67835295e-01
-7.96129882e-01 -3.42741632e-03 -1.98385231e-02 1.79371119e-01
-2.08888888e-01 5.78150094e-01 1.85025275e-01 -4.97350022e-02
6.41649187e-01 -7.63727631e-03 2.03238398e-01 -4.22540829e-02
-4.61960703e-01 4.11809415e-01 1.09135225e-01 4.71073650e-02
7.69135356e-01 1.92361176e-01 6.99371815e-01 -9.00514781e-01
-5.67420125e-01 -9.57195759e-01 -5.16453087e-01 -1.59749389e-01
9.00844097e-01 -7.92655170e-01 -5.26231050e-01 1.78507090e-01
-1.57734111e-01 -1.95748150e-01 -1.46665841e-01 1.26345420e+00
-3.19358110e-01 4.75212187e-01 -4.56283271e-01 -4.66831058e-01
-5.71095049e-01 -4.70618188e-01 2.62004822e-01 3.72244149e-01
-3.15760911e-01 -1.31386983e+00 2.61437953e-01 2.21160933e-01
6.10594630e-01 2.27611408e-01 1.24471593e+00 -1.22990906e+00
-5.89780986e-01 -8.25308800e-01 -1.82839438e-01 9.56660230e-03
1.19036362e-02 8.65397975e-02 -2.84951121e-01 -4.54461202e-02
3.68640423e-01 1.10987276e-01 6.44556463e-01 4.30401921e-01
8.82528365e-01 -2.77345896e-01 -1.02242768e+00 3.52272332e-01
1.71687412e+00 5.78495324e-01 5.21984100e-01 1.92755654e-01
3.21957439e-01 3.11737657e-01 6.24253571e-01 2.15701789e-01
-6.84863329e-02 1.63532302e-01 6.92374945e-01 1.62017122e-01
-1.62886165e-03 -3.38110566e-01 3.92196141e-02 5.47816813e-01
-3.96026760e-01 -4.72253829e-01 -1.11344707e+00 2.90348351e-01
-1.03183293e+00 -9.56291616e-01 -9.25348818e-01 2.24769926e+00
3.99660766e-01 1.91280663e-01 -2.67569739e-02 5.62335923e-03
6.51780248e-01 -3.10695052e-01 -3.89828086e-01 -2.54195333e-01
4.33445215e-01 -9.32473615e-02 8.06878090e-01 1.92006826e-01
-4.78068769e-01 1.15218893e-01 6.69099951e+00 6.20800078e-01
-1.28725863e+00 -2.35376522e-01 4.88273233e-01 3.46240997e-01
-4.01609212e-01 2.54583240e-01 -8.96564603e-01 4.00044590e-01
1.01538324e+00 -5.20300508e-01 -3.10051769e-01 9.35534239e-01
9.15429294e-02 -7.49604881e-01 -9.12606239e-01 3.91068041e-01
-3.17497812e-02 -9.32221770e-01 -2.46697754e-01 5.43893814e-01
6.16926670e-01 9.64264199e-02 -1.69442430e-01 3.01733613e-01
-2.86155015e-01 -1.01293051e+00 -6.13921136e-02 5.41795492e-01
1.00852668e+00 -1.30398408e-01 1.33847547e+00 7.28453696e-01
-1.13958919e+00 -5.42310253e-02 -2.70245552e-01 1.42556816e-01
-9.83707458e-02 6.22765422e-01 -2.07630491e+00 2.61423469e-01
1.99319974e-01 3.68877240e-02 -8.13265443e-01 1.24000132e+00
-1.03002191e-01 7.40039885e-01 -6.98767066e-01 -6.80796087e-01
-5.44757918e-02 -7.61379525e-02 1.38183713e-01 7.43981242e-01
1.01488388e+00 3.94339383e-01 -3.44388068e-01 6.78843319e-01
2.55790442e-01 8.09597433e-01 -3.13483268e-01 -1.56771824e-01
5.91720283e-01 1.41101193e+00 -1.13335419e+00 -2.90056735e-01
-5.83289146e-01 2.03693092e-01 -4.44108099e-01 -3.39842558e-01
-6.53513908e-01 1.69826776e-01 -4.08277065e-01 9.27148819e-01
1.11388184e-01 5.23844719e-01 -3.60391885e-01 -6.19925201e-01
-4.13446665e-01 -2.09592476e-01 6.41418457e-01 -4.72269863e-01
-8.30519199e-01 3.94464552e-01 5.49805872e-02 -1.29109275e+00
-4.39778417e-01 -6.97768986e-01 -1.12771297e+00 6.76698327e-01
-8.53775203e-01 -1.03052974e+00 -7.08039999e-01 -2.29705814e-02
1.05413971e-02 -1.34987667e-01 9.14801002e-01 -4.50483769e-01
-2.13063493e-01 1.27478048e-01 1.87148467e-01 -3.10538203e-01
4.93486434e-01 -1.57706761e+00 -5.24776876e-01 -2.73276307e-03
-6.07873559e-01 3.47798079e-01 2.67372847e-01 -1.09942997e+00
-9.05395329e-01 -9.09389615e-01 1.16161990e+00 -2.41957247e-01
6.30990863e-01 4.44439143e-01 -5.35949647e-01 4.00760651e-01
-3.55383277e-01 -2.40551338e-01 1.03713536e+00 -3.77540231e-01
2.70660520e-01 -7.77312415e-03 -1.21627820e+00 4.18702602e-01
2.64485300e-01 1.44606857e-02 -3.25227946e-01 4.11595583e-01
-2.70450432e-02 -2.13894397e-01 -1.48101437e+00 6.00563645e-01
8.76036167e-01 -1.25690138e+00 9.45699096e-01 2.25819319e-01
-8.80225736e-04 -1.39530957e-01 4.50598985e-01 -6.70479536e-01
-6.33520424e-01 1.49408937e-01 3.75473410e-01 5.02340019e-01
8.02394032e-01 -8.07088971e-01 1.09309697e+00 8.75987589e-01
1.11337908e-01 -1.11665618e+00 -6.98215425e-01 -6.91849887e-01
1.41472816e-01 1.59014180e-01 3.49817574e-01 4.41884637e-01
2.47874781e-01 -1.47575382e-02 4.00732636e-01 -1.69960797e-01
1.25328958e-01 -2.21806362e-01 4.42605048e-01 -1.54234385e+00
-1.86290264e-01 -4.66601163e-01 -3.71171892e-01 -4.08122480e-01
-4.19470221e-01 -1.05840945e+00 -6.38786316e-01 -2.06541038e+00
5.92444777e-01 -5.82243741e-01 -1.71497688e-01 3.25265199e-01
-8.81319419e-02 -2.48771086e-01 2.41386086e-01 6.26334190e-01
1.84555948e-01 -1.73834473e-01 1.49795318e+00 3.07693124e-01
-2.49189258e-01 9.07118320e-01 -4.49094743e-01 9.03221667e-01
1.06203687e+00 -6.80265844e-01 -4.31467324e-01 7.29955435e-01
1.24306403e-01 9.01354551e-01 2.30807871e-01 -1.06312871e+00
1.38346327e-03 -5.96034348e-01 5.78670502e-01 -1.05263376e+00
-1.19907469e-01 -1.32339644e+00 8.92787039e-01 1.62576044e+00
1.18979685e-01 -3.01890224e-01 -1.68527320e-01 5.29073775e-01
4.97460142e-02 -1.11606026e+00 7.90392697e-01 -3.99905056e-01
-3.02824527e-01 2.87604928e-01 -8.61070275e-01 -4.25054133e-01
1.62901342e+00 -6.08168960e-01 -1.55086011e-01 -3.80486518e-01
-8.03663433e-01 -6.30540401e-03 7.40923047e-01 -2.69595534e-01
3.18986714e-01 -1.09051979e+00 -5.53608477e-01 1.08302295e-01
7.17888698e-02 5.30680791e-02 7.13647529e-02 1.42609107e+00
-1.14565134e+00 1.01842642e+00 -1.09489918e-01 -3.51723492e-01
-1.37587774e+00 4.04724568e-01 7.38068044e-01 -7.97442079e-01
-1.55020639e-01 6.30239964e-01 2.57733792e-01 -2.18677577e-02
-9.04948190e-02 -7.45292902e-01 -4.40854222e-01 -1.50130704e-01
-1.42536551e-01 6.99442208e-01 -2.52160758e-01 -1.09033607e-01
-2.07718119e-01 5.48344076e-01 9.35641825e-02 1.78784698e-01
7.98282683e-01 8.63136053e-02 -5.20206869e-01 3.99771243e-01
7.57193923e-01 6.29731834e-01 -3.57103080e-01 2.81455755e-01
1.96984053e-01 -1.33675352e-01 -3.13096613e-01 -9.55000281e-01
-6.62077725e-01 1.44627422e-01 6.90318346e-01 4.23452199e-01
8.67546320e-01 4.03304815e-01 7.66973734e-01 1.71910062e-01
5.91782443e-02 -6.61423087e-01 1.20327231e-02 1.61324352e-01
7.20867932e-01 -1.08770907e+00 2.74022192e-01 -6.92582667e-01
-6.57366872e-01 1.54501379e+00 6.90985620e-01 -5.83206862e-02
8.54447842e-01 1.43617794e-01 6.76508695e-02 -5.64014129e-02
-9.85234082e-01 -5.38197644e-02 3.46334696e-01 3.60531390e-01
8.04440558e-01 4.43188578e-01 -1.21021366e+00 9.21580434e-01
-1.89643979e-01 2.81020164e-01 4.59349871e-01 5.07344127e-01
-1.20210803e+00 -1.24526572e+00 -6.13411069e-01 1.15229547e+00
-6.75941527e-01 1.67802274e-01 -4.50868547e-01 1.47962737e+00
2.22872019e-01 3.50358814e-01 -1.07538551e-01 -1.01724910e-02
8.38412791e-02 2.36695543e-01 2.56168872e-01 -5.37541628e-01
-6.38719320e-01 2.99509853e-01 7.37944543e-02 1.04169443e-01
-1.51121616e-01 -6.62068129e-01 -1.66642904e+00 -5.13029061e-02
-6.70190930e-01 5.59128881e-01 7.57878065e-01 5.91377139e-01
-4.58673090e-01 1.24124937e-01 5.50717294e-01 8.95315036e-03
-6.32431567e-01 -1.11758828e+00 -1.16046703e+00 -1.08797243e-02
-1.44178644e-01 -2.07602307e-01 -8.83220851e-01 -2.13888451e-01] | [15.376104354858398, -2.3576602935791016] |
3063c7d4-6a65-4bf9-a3f4-979ad7a4f307 | leaps-end-to-end-one-step-person-search-with | 2303.11859 | null | https://arxiv.org/abs/2303.11859v1 | https://arxiv.org/pdf/2303.11859v1.pdf | LEAPS: End-to-End One-Step Person Search With Learnable Proposals | We propose an end-to-end one-step person search approach with learnable proposals, named LEAPS. Given a set of sparse and learnable proposals, LEAPS employs a dynamic person search head to directly perform person detection and corresponding re-id feature generation without non-maximum suppression post-processing. The dynamic person search head comprises a detection head and a novel flexible re-id head. Our flexible re-id head first employs a dynamic region-of-interest (RoI) operation to extract discriminative RoI features of the proposals. Then, it generates re-id features using a plain and a hierarchical interaction re-id module. To better guide discriminative re-id feature learning, we introduce a diverse re-id sample matching strategy, instead of bipartite matching in detection head. Comprehensive experiments reveal the benefit of the proposed LEAPS, achieving a favorable performance on two public person search benchmarks: CUHK-SYSU and PRW. When using the same ResNet50 backbone, our LEAPS obtains a mAP score of 55.0%, outperforming the best reported results in literature by 1.7%, while achieving around a two-fold speedup on the challenging PRW dataset. Our source code and models will be released. | ['Yanwei Pang', 'Fahad Khan', 'Jin Xie', 'Rao Muhammad Anwer', 'Jiale Cao', 'Zhiqiang Dong'] | 2023-03-21 | null | null | null | null | ['person-search', 'human-detection'] | ['computer-vision', 'computer-vision'] | [-1.05034411e-01 -4.61989902e-02 -1.33032754e-01 -4.75755990e-01
-9.29412127e-01 -3.55853558e-01 6.49674535e-01 -8.56310204e-02
-9.17763948e-01 5.93651772e-01 5.90672374e-01 3.33795816e-01
-3.00062299e-01 -7.40201652e-01 -5.28657913e-01 -6.02140367e-01
-1.23834282e-01 1.09504032e+00 4.50768262e-01 -2.70363957e-01
-1.83565155e-01 1.07064739e-01 -1.49684453e+00 1.13363221e-01
8.08898747e-01 8.74237239e-01 1.64196476e-01 3.96913499e-01
4.31319952e-01 3.13841313e-01 -5.04910707e-01 -6.59479618e-01
6.75556481e-01 -4.68061380e-02 -6.51820064e-01 -1.34291410e-01
5.86559832e-01 -3.66613597e-01 -8.34546149e-01 7.02353060e-01
1.21959662e+00 4.85023528e-01 3.99666190e-01 -1.08984458e+00
-5.75773001e-01 5.69220603e-01 -7.18645453e-01 5.49777091e-01
8.53221536e-01 4.21459734e-01 1.25325751e+00 -1.12967801e+00
5.83551884e-01 1.36604834e+00 7.05192924e-01 7.33605325e-01
-1.36280012e+00 -1.04126346e+00 2.46280760e-01 2.93561995e-01
-1.83607173e+00 -3.76683891e-01 4.27835584e-01 -1.58666223e-01
8.31553221e-01 4.08365428e-01 7.80815661e-01 1.00058579e+00
-2.28432551e-01 1.06002569e+00 8.28952730e-01 -4.51804660e-02
-5.45533299e-02 -8.81353691e-02 4.77002740e-01 8.11094284e-01
2.38466784e-01 2.25297719e-01 -7.73757577e-01 -5.79820931e-01
4.83689845e-01 2.44174272e-01 -3.10675800e-01 -2.88912296e-01
-1.22476172e+00 6.94995224e-01 8.39819729e-01 2.73023220e-03
-4.85169142e-01 -1.74515024e-02 3.37734878e-01 3.16741131e-02
1.12608306e-01 8.64851698e-02 -8.95919949e-02 9.53248888e-02
-1.11790669e+00 9.37415898e-01 6.46697223e-01 9.47931886e-01
5.62003076e-01 -5.02389431e-01 -1.10508394e+00 9.40653682e-01
1.58354133e-01 8.14981282e-01 4.93804723e-01 -3.84047657e-01
6.83812857e-01 7.16758966e-01 1.51067853e-01 -1.02470767e+00
-5.96720517e-01 -8.50153267e-01 -9.66972232e-01 -5.07343650e-01
2.45070398e-01 1.11873277e-01 -1.15566111e+00 1.82664502e+00
5.10965645e-01 2.48504296e-01 -2.84407049e-01 1.34801078e+00
1.21341479e+00 4.76353198e-01 1.37030035e-01 2.91447669e-01
1.82440627e+00 -1.35833573e+00 -3.63895260e-02 -3.76450628e-01
2.50913233e-01 -2.18149930e-01 8.17974389e-01 8.06509256e-02
-1.16708982e+00 -7.04296768e-01 -7.58747935e-01 -1.45721391e-01
-5.19186035e-02 3.59396517e-01 4.00349408e-01 5.51808119e-01
-1.03504837e+00 1.19392306e-01 -3.66003722e-01 -4.05102968e-01
5.34286082e-01 7.81159461e-01 -5.93312025e-01 -2.57521421e-01
-1.19668341e+00 3.97999138e-01 2.88352907e-01 -7.65388412e-03
-8.88244450e-01 -9.15679574e-01 -7.11275876e-01 2.82513738e-01
5.36230206e-01 -1.11696434e+00 8.84715378e-01 -3.55447799e-01
-1.05871797e+00 1.01095295e+00 -4.33462530e-01 -5.91111422e-01
8.32535267e-01 -3.15173388e-01 -3.07972342e-01 9.48424116e-02
4.12659585e-01 6.02267563e-01 8.23200464e-01 -8.51914644e-01
-9.30565536e-01 -5.52405536e-01 -1.50426328e-01 3.15614492e-01
-3.73501778e-01 2.19799101e-01 -1.28310919e+00 -7.18462169e-01
1.47423623e-02 -1.12902617e+00 -4.83484864e-01 -2.70776987e-01
-5.75738370e-01 -6.52846277e-01 1.42928034e-01 -7.56195903e-01
1.30304122e+00 -1.84994006e+00 3.06561202e-01 6.35357559e-01
7.15170622e-01 2.31893808e-01 -2.87974626e-01 2.02202037e-01
1.81059539e-01 -4.58606333e-01 5.48514314e-02 -6.36833489e-01
1.82257816e-01 -3.20057750e-01 -1.06801447e-02 6.28673851e-01
-3.27723116e-01 1.24968910e+00 -7.22703874e-01 -4.43159729e-01
1.10653844e-02 3.38418096e-01 -7.90315688e-01 1.75081968e-01
3.51968855e-01 1.46597385e-01 -6.06513083e-01 7.58127868e-01
6.77822530e-01 -3.96924436e-01 -2.10703555e-02 -1.97477818e-01
9.19282660e-02 8.51150900e-02 -1.18946195e+00 1.58609021e+00
-6.38887985e-03 1.48293823e-01 -7.35112205e-02 -8.02843511e-01
8.85032117e-01 -2.53269672e-01 4.42614079e-01 -9.09163296e-01
-6.30791709e-02 1.04108844e-02 -1.38440758e-01 1.11338738e-02
6.44192517e-01 5.73767424e-01 -4.73077506e-01 4.64390218e-01
7.60024115e-02 9.84207034e-01 2.39934847e-01 4.75222051e-01
1.29566979e+00 -2.22651362e-01 1.48841769e-01 -5.47699451e-01
9.72646534e-01 -2.93143213e-01 8.20895493e-01 1.27457833e+00
-3.79646748e-01 4.97989327e-01 5.50421365e-02 -7.52014041e-01
-8.74135375e-01 -1.13969338e+00 1.65284142e-01 1.59665775e+00
4.57948774e-01 -7.06225514e-01 -6.26692832e-01 -8.02892208e-01
3.54532897e-01 2.53477935e-02 -6.47382081e-01 -7.79639557e-02
-8.92118037e-01 -7.73487866e-01 7.56174922e-01 4.36729014e-01
8.90049756e-01 -9.40181017e-01 -2.54937857e-01 7.07689300e-02
-3.72906178e-01 -1.02983570e+00 -1.17961740e+00 -3.00882578e-01
-1.30541906e-01 -9.22259450e-01 -1.29492354e+00 -9.15959358e-01
7.81034529e-01 4.59651470e-01 1.10039961e+00 2.04937086e-01
-5.76357543e-01 3.13861638e-01 -1.55428529e-01 1.66671053e-01
3.32686335e-01 5.52933037e-01 2.75641888e-01 -8.93065054e-03
5.89144170e-01 -3.32411140e-01 -1.30320752e+00 7.50035167e-01
-2.31369391e-01 -1.10191658e-01 7.24082768e-01 1.01196361e+00
6.39218748e-01 -3.93343568e-01 4.13270503e-01 -2.94542491e-01
5.02827525e-01 -4.59262997e-01 -4.48784053e-01 3.49116564e-01
-5.48146009e-01 4.29042354e-02 2.99911499e-01 -4.64437574e-01
-8.75688076e-01 1.82222500e-01 -1.76971778e-01 -1.17917582e-01
1.55980706e-01 5.72969504e-02 -1.71757504e-01 -1.13466881e-01
7.06022263e-01 6.76315963e-01 -2.44566590e-01 -4.32300061e-01
1.85693935e-01 5.78303695e-01 7.89798498e-01 -6.27006710e-01
1.12418389e+00 5.63097894e-01 -3.96029115e-01 -5.10720074e-01
-6.77927077e-01 -1.02835655e+00 -2.84876078e-01 -8.22207853e-02
6.71361148e-01 -1.34272921e+00 -1.12341404e+00 4.79089677e-01
-8.14225674e-01 -6.72084689e-02 -2.07875088e-01 1.69964254e-01
-3.65101397e-02 3.52269113e-01 -5.28972864e-01 -5.01751184e-01
-1.03715599e+00 -1.06552565e+00 1.35209012e+00 4.72902149e-01
-1.41489103e-01 -3.42315555e-01 1.36775911e-01 8.06536138e-01
4.10890996e-01 -1.28733948e-01 2.10207120e-01 -1.03840768e+00
-7.39452779e-01 -3.41466308e-01 -4.52005506e-01 -2.84696341e-01
-2.14842439e-01 -1.06422198e+00 -7.21090496e-01 -8.06832016e-01
-8.74050796e-01 -3.09158444e-01 1.32843900e+00 7.16176778e-02
1.00998557e+00 -2.23144442e-01 -8.72475445e-01 8.04701388e-01
1.10248458e+00 -3.61477554e-01 4.46359068e-01 3.28794688e-01
5.73320448e-01 3.85820091e-01 4.85591382e-01 6.51090860e-01
9.12616551e-01 1.22560179e+00 -1.22919932e-01 -1.51618838e-01
-2.38291502e-01 -5.10549903e-01 2.31646061e-01 9.04400647e-02
-4.37750667e-01 -1.92011207e-01 -8.25799525e-01 6.47464037e-01
-2.15598178e+00 -1.25677156e+00 3.33193213e-01 2.28016353e+00
7.25374937e-01 4.29318883e-02 8.55798781e-01 -2.84336150e-01
9.16309893e-01 1.85395703e-01 -6.58376038e-01 5.90699315e-01
-2.82804482e-02 2.24495545e-01 5.58888912e-01 2.98037171e-01
-1.25706840e+00 1.07347333e+00 5.07324219e+00 1.09478176e+00
-5.37721574e-01 2.87556827e-01 3.95990014e-01 -3.50374609e-01
-4.28720005e-02 -3.02713513e-01 -1.55565786e+00 6.54336989e-01
3.32733989e-01 -2.87198186e-01 3.96063715e-01 8.70950341e-01
1.09640703e-01 4.58310172e-02 -9.56905127e-01 1.39344823e+00
3.33015442e-01 -1.19609249e+00 9.42933634e-02 2.40510970e-01
3.59860837e-01 -3.25643234e-02 -1.23620862e-02 6.39401674e-01
2.70956844e-01 -8.50773692e-01 6.20726585e-01 5.61617553e-01
5.26491702e-01 -9.29450929e-01 5.57073295e-01 3.49535495e-01
-1.75853646e+00 -4.83461916e-01 -3.87746155e-01 3.26428533e-01
2.73470312e-01 4.52745497e-01 -5.85030913e-01 6.02994740e-01
1.09768558e+00 4.27916735e-01 -7.97827482e-01 1.30176592e+00
4.11117226e-02 3.16426635e-01 -5.86753309e-01 -8.75204653e-02
1.48860477e-02 9.99973416e-02 8.03251743e-01 1.41662300e+00
8.10524747e-02 1.16971955e-01 5.26120782e-01 6.87105656e-01
-2.46618196e-01 1.72969997e-01 5.84789030e-02 6.44917488e-01
6.73718393e-01 1.30601966e+00 -5.53603530e-01 -3.96887958e-01
-2.14759707e-01 1.39752638e+00 5.09130120e-01 2.52025157e-01
-1.03027856e+00 -3.03362489e-01 8.09714973e-01 3.49323362e-01
4.55552787e-01 8.20229426e-02 4.09012914e-01 -1.19187212e+00
3.19752872e-01 -8.44155967e-01 9.35082674e-01 -2.38530919e-01
-1.36810720e+00 6.65802598e-01 1.61932036e-01 -9.33177590e-01
-1.59866631e-01 -1.17753290e-01 -5.71915805e-01 9.53793228e-01
-1.28084636e+00 -1.61615109e+00 -6.41494155e-01 9.15042460e-01
6.34093106e-01 -5.76798081e-01 4.98322487e-01 5.15685916e-01
-6.96323097e-01 1.42549169e+00 -3.21329385e-01 3.47076595e-01
7.17009664e-01 -9.74425912e-01 7.30321884e-01 8.28065693e-01
-5.22993095e-02 7.80619740e-01 4.44834232e-01 -8.04057956e-01
-1.37232792e+00 -1.02947640e+00 9.20923769e-01 -3.42342764e-01
1.95913211e-01 -8.08524549e-01 -4.72786754e-01 5.30949235e-01
-1.26909211e-01 1.19477637e-01 5.70129573e-01 2.97692537e-01
-4.00224417e-01 -2.59318471e-01 -1.19954383e+00 5.87345779e-01
1.75525951e+00 -7.03627765e-02 -6.08422220e-01 4.40133184e-01
5.73079169e-01 -4.80356067e-01 -5.74352384e-01 3.53221208e-01
7.84725368e-01 -7.97947764e-01 1.66332376e+00 -4.92035896e-01
-1.05472520e-01 -3.23417544e-01 1.21750414e-01 -8.82032335e-01
-1.03210032e+00 -7.73676991e-01 -1.87028721e-01 1.09317768e+00
3.04803073e-01 -7.76548207e-01 1.06513238e+00 6.75278425e-01
3.79358947e-01 -8.44656587e-01 -1.04317546e+00 -8.77784133e-01
-4.10694033e-01 -3.80915925e-02 6.28873706e-01 2.86969453e-01
-1.93377256e-01 2.02827170e-01 -7.16508448e-01 2.43724331e-01
1.09321511e+00 1.94411442e-01 1.14857221e+00 -1.25598133e+00
-6.31280184e-01 -2.39531487e-01 -5.00253797e-01 -1.48677659e+00
1.21445477e-01 -1.10487425e+00 -5.37640825e-02 -1.36421716e+00
9.66605723e-01 -5.78337133e-01 -3.22390079e-01 5.75173974e-01
-5.00958800e-01 3.81231397e-01 3.92704606e-01 4.79774565e-01
-1.01108682e+00 7.19579101e-01 7.40707874e-01 -3.95577431e-01
-4.43644196e-01 3.46862346e-01 -7.73886323e-01 3.95809650e-01
5.45259655e-01 -3.50688547e-01 -1.46139920e-01 -1.49221525e-01
-5.09588234e-02 -3.77469867e-01 7.13440061e-01 -1.08493567e+00
7.43567288e-01 4.24863547e-01 7.09521949e-01 -8.21593642e-01
4.58567381e-01 -2.91928619e-01 -2.45098379e-02 6.31296873e-01
-3.25169444e-01 1.08922176e-01 -1.06874518e-01 7.39712298e-01
1.79039851e-01 1.60501629e-01 5.70932984e-01 -4.68975678e-02
-8.72782588e-01 7.35421658e-01 1.36245653e-01 1.63714528e-01
1.09620845e+00 -1.49168745e-01 -2.74690479e-01 -3.02899122e-01
-6.49743557e-01 7.90952504e-01 1.70259804e-01 5.47088265e-01
7.78524518e-01 -1.27374327e+00 -1.01971781e+00 3.06290746e-01
2.79584497e-01 -2.17581153e-01 5.57233751e-01 7.90610254e-01
-3.53468917e-02 4.72004443e-01 1.51598498e-01 -5.74252725e-01
-1.53337383e+00 4.60809827e-01 2.79675961e-01 -6.79387391e-01
-9.64280069e-01 1.05272126e+00 3.42991620e-01 -5.20217896e-01
3.72065037e-01 4.44588870e-01 -6.04235232e-02 -1.56931698e-01
9.31854248e-01 5.67224979e-01 -2.21432894e-01 -7.93228328e-01
-6.92303121e-01 4.95621055e-01 -4.45231885e-01 -4.36472483e-02
1.12607181e+00 -2.67579090e-02 1.88817292e-01 -6.23008728e-01
8.31555307e-01 9.85801667e-02 -9.58551526e-01 -6.58387184e-01
-1.37425691e-01 -5.73528826e-01 -2.69742221e-01 -7.06987083e-01
-1.08307672e+00 -1.42439250e-02 7.76564181e-01 -3.55929494e-01
1.06229055e+00 5.12034059e-01 1.23263383e+00 5.37047148e-01
6.60820305e-01 -1.07137275e+00 -4.44253441e-03 2.70324618e-01
9.36778665e-01 -1.08122265e+00 2.17714101e-01 -4.06672060e-01
-5.34025252e-01 4.30056661e-01 7.32604027e-01 -1.73175871e-01
5.18067598e-01 -6.54547140e-02 -5.62934875e-01 -1.90594539e-01
-5.01455486e-01 -5.62765479e-01 7.28218496e-01 4.11817908e-01
-2.08348632e-01 4.53035533e-02 -4.59130257e-01 1.01442194e+00
-3.07565302e-01 -1.84923168e-02 -4.31231737e-01 5.49724519e-01
-3.85198206e-01 -9.37226951e-01 -3.77486020e-01 5.55282474e-01
-1.86431214e-01 -1.56278580e-01 -2.53377944e-01 6.54353142e-01
2.17498273e-01 6.79602504e-01 -6.94612861e-02 -6.45487428e-01
5.97250938e-01 -1.76486745e-01 2.89822310e-01 -2.39583895e-01
-8.79276156e-01 -2.85447925e-01 -1.65443972e-03 -9.61790264e-01
-1.61344483e-01 -7.23412216e-01 -1.01326418e+00 -4.16059613e-01
-7.31982142e-02 1.01530142e-01 2.64071003e-02 7.85818160e-01
6.50854349e-01 6.50823116e-02 4.53911126e-01 -8.89156818e-01
-5.46200514e-01 -8.97822618e-01 -1.60932213e-01 5.00532508e-01
4.55882773e-02 -7.84360349e-01 -1.00258708e-01 -4.06562567e-01] | [14.871386528015137, 0.7978830337524414] |
d461097a-6776-42e9-8b9b-38a56b46f6fd | casa-nlu-context-aware-self-attentive-natural | 1909.08705 | null | https://arxiv.org/abs/1909.08705v1 | https://arxiv.org/pdf/1909.08705v1.pdf | CASA-NLU: Context-Aware Self-Attentive Natural Language Understanding for Task-Oriented Chatbots | Natural Language Understanding (NLU) is a core component of dialog systems. It typically involves two tasks - intent classification (IC) and slot labeling (SL), which are then followed by a dialogue management (DM) component. Such NLU systems cater to utterances in isolation, thus pushing the problem of context management to DM. However, contextual information is critical to the correct prediction of intents and slots in a conversation. Prior work on contextual NLU has been limited in terms of the types of contextual signals used and the understanding of their impact on the model. In this work, we propose a context-aware self-attentive NLU (CASA-NLU) model that uses multiple signals, such as previous intents, slots, dialog acts and utterances over a variable context window, in addition to the current user utterance. CASA-NLU outperforms a recurrent contextual NLU baseline on two conversational datasets, yielding a gain of up to 7% on the IC task for one of the datasets. Moreover, a non-contextual variant of CASA-NLU achieves state-of-the-art performance for IC task on standard public datasets - Snips and ATIS. | ['Mona Diab', 'Garima Lalwani', 'Peng Zhang', 'Arshit Gupta'] | 2019-09-18 | casa-nlu-context-aware-self-attentive-natural-1 | https://aclanthology.org/D19-1127 | https://aclanthology.org/D19-1127.pdf | ijcnlp-2019-11 | ['dialogue-management'] | ['natural-language-processing'] | [ 4.61105496e-01 3.42538327e-01 -3.22989464e-01 -6.24362886e-01
-6.20587170e-01 -7.27093458e-01 1.02637148e+00 3.13812256e-01
-3.63393366e-01 7.68974483e-01 9.00650263e-01 -5.19714117e-01
4.41187024e-01 -4.42146182e-01 -1.75331607e-01 -1.27705336e-01
1.07978016e-01 8.52672338e-01 3.70768845e-01 -6.58055186e-01
2.06886068e-01 -1.84949070e-01 -1.16041100e+00 7.78669953e-01
5.30616105e-01 9.67170417e-01 2.99535424e-01 9.61169720e-01
-7.39860654e-01 1.23811781e+00 -6.44162357e-01 9.51859206e-02
-2.73034275e-01 -5.91747582e-01 -1.50219274e+00 8.29084367e-02
-9.56267342e-02 -4.16887045e-01 -4.63137105e-02 2.95276642e-01
2.98742414e-01 4.55777764e-01 5.13062119e-01 -9.70474243e-01
8.79383236e-02 8.88044417e-01 7.42678493e-02 5.39139844e-02
9.38450575e-01 -8.81202926e-04 1.34355307e+00 -6.04512095e-01
6.66018605e-01 1.65315235e+00 4.02310818e-01 9.35892522e-01
-1.34353936e+00 -1.84821293e-01 5.01734138e-01 -6.98048174e-02
-6.71421528e-01 -7.15846658e-01 5.92893124e-01 -3.53182584e-01
1.43098021e+00 3.53387505e-01 9.26956609e-02 1.25286233e+00
-1.51919082e-01 1.14145672e+00 9.81852591e-01 -7.87041903e-01
4.23386633e-01 1.64028317e-01 8.60489011e-01 1.43337056e-01
-9.11345005e-01 -2.90176004e-01 -6.26465201e-01 -3.38259548e-01
1.53544441e-01 -1.29281491e-01 -3.55002098e-02 2.39203483e-01
-9.06260192e-01 1.04911530e+00 6.11926429e-02 4.10368681e-01
-2.77301073e-01 -3.35455537e-01 6.96294606e-01 4.44865853e-01
6.21996164e-01 5.79410791e-01 -7.65328288e-01 -6.44529760e-01
-1.08977355e-01 2.95559406e-01 1.37777948e+00 8.56657743e-01
7.61609316e-01 -3.89352590e-01 -5.90057313e-01 1.32050407e+00
1.84682935e-01 -1.15743643e-02 5.01552522e-01 -8.40690613e-01
5.66843152e-01 8.84566963e-01 2.17450172e-01 -3.91272306e-01
-7.35489488e-01 3.39124590e-01 -1.24595106e-01 -4.34603602e-01
5.30219674e-01 -5.84037483e-01 -6.37056351e-01 2.00948310e+00
4.02424037e-01 2.03008547e-01 4.01203364e-01 7.38222539e-01
9.59995151e-01 8.63599122e-01 4.02014762e-01 -2.57262677e-01
1.67450607e+00 -1.00244391e+00 -1.01152122e+00 -8.91416192e-01
8.11255157e-01 -8.43682408e-01 1.28655982e+00 -2.31622308e-02
-8.04271877e-01 -3.19652975e-01 -8.29316199e-01 -2.10522801e-01
-2.71220058e-01 -1.94065288e-01 7.09612310e-01 4.27464008e-01
-8.64056170e-01 1.19381249e-01 -4.63623941e-01 -6.25282228e-01
-3.46666187e-01 2.77375370e-01 -7.33695701e-02 8.56484473e-02
-1.65923142e+00 9.10291553e-01 2.85173655e-01 -2.03161567e-01
-4.82592046e-01 -5.31184852e-01 -1.01308823e+00 8.49703848e-02
8.70926797e-01 -1.38987929e-01 2.05828643e+00 -5.65723598e-01
-1.82867587e+00 5.58330655e-01 -6.75358355e-01 -7.02420175e-01
1.86730087e-01 -4.32076007e-01 -2.21539482e-01 -9.12376568e-02
2.95829568e-02 5.96961915e-01 4.37824786e-01 -1.16116083e+00
-8.30063343e-01 -1.67359814e-01 4.40937966e-01 4.88698483e-01
5.52933663e-02 2.50248462e-02 -6.16876781e-01 -1.10242374e-01
-1.11034013e-01 -1.09903681e+00 -9.89862010e-02 -8.50694776e-01
-4.56045538e-01 -7.54352510e-01 1.05424941e+00 -5.06063342e-01
1.54365754e+00 -2.07905769e+00 1.14816763e-01 -2.12777987e-01
-6.31763116e-02 2.71875739e-01 3.24446671e-02 7.99906194e-01
1.62050232e-01 -7.22933263e-02 -8.26412663e-02 -8.46624076e-01
1.92521423e-01 4.59301770e-01 -5.89676201e-01 -3.40012550e-01
9.07410681e-02 8.13535333e-01 -8.02259445e-01 -3.10165435e-01
5.35996497e-01 1.34120723e-02 -6.04100883e-01 7.66679406e-01
-9.45972919e-01 5.88204503e-01 -4.36924577e-01 4.20043558e-01
3.83374952e-02 -6.31114244e-02 6.49311841e-01 1.42433926e-01
-1.35659844e-01 1.17960906e+00 -7.32722640e-01 1.51700306e+00
-9.78706062e-01 5.93955517e-01 2.58974642e-01 -5.55270612e-01
8.06836486e-01 6.61497176e-01 2.58047014e-01 -6.21651292e-01
1.66994333e-01 -8.99275467e-02 7.91193619e-02 -2.43382439e-01
6.88104272e-01 -6.77508190e-02 -5.88458359e-01 8.09074879e-01
1.29507348e-01 8.14016163e-02 1.83003604e-01 6.49963498e-01
1.19076896e+00 -3.61861765e-01 5.93085647e-01 -9.03439894e-02
6.97679996e-01 1.03604183e-01 4.82237667e-01 9.58565056e-01
-3.81039649e-01 3.11324328e-01 8.70721817e-01 -2.67754763e-01
-5.79330564e-01 -5.90395153e-01 -9.20713097e-02 1.71476781e+00
6.05828390e-02 -5.94289243e-01 -9.79443371e-01 -8.71821702e-01
-1.45605281e-01 1.19801521e+00 -6.13648593e-01 -4.55336049e-02
-6.13822997e-01 -4.58583593e-01 2.36593574e-01 3.52900594e-01
4.22037929e-01 -1.50920975e+00 -2.85946131e-01 5.52312970e-01
-5.59221625e-01 -1.56672657e+00 -7.13618517e-01 4.36529607e-01
-4.37667608e-01 -9.94299054e-01 1.00806549e-01 -4.20141250e-01
1.25462323e-01 1.39797524e-01 1.29266262e+00 -6.31120056e-02
4.48073559e-02 5.33685684e-01 -6.29678667e-01 -3.53960007e-01
-9.17518258e-01 2.11877897e-01 -4.50209305e-02 6.34522140e-02
7.32519209e-01 -2.79586583e-01 -3.62957716e-01 6.35875344e-01
-6.45539343e-01 2.68858045e-01 2.99760312e-01 9.46918309e-01
-9.35945213e-02 -4.88429815e-01 8.72296512e-01 -1.35678780e+00
9.41183269e-01 -5.23446977e-01 -1.16959155e-01 2.57391185e-01
-3.50944579e-01 -4.21671662e-03 4.84317243e-01 -3.86567801e-01
-1.51397502e+00 -1.83079652e-02 -2.85532802e-01 1.66138709e-01
-5.06268859e-01 4.31459874e-01 -2.58827418e-01 6.05313301e-01
5.93157291e-01 1.72201823e-02 -3.45033519e-02 -4.64217067e-01
3.43082666e-01 1.15422130e+00 4.05354828e-01 -7.25133777e-01
-8.99791420e-02 -4.60799187e-02 -6.59290433e-01 -1.19881451e+00
-1.15338767e+00 -9.25280273e-01 -6.79173946e-01 -2.45837286e-01
9.40183878e-01 -7.54422784e-01 -8.81071508e-01 3.49018872e-01
-1.24915516e+00 -7.64838219e-01 6.05864376e-02 1.11139596e-01
-4.01071519e-01 2.58937836e-01 -9.55533743e-01 -1.22031832e+00
-3.80041808e-01 -1.23998451e+00 9.67825532e-01 2.17568606e-01
-8.92597556e-01 -1.17354655e+00 -1.02163509e-01 8.03071558e-01
3.79525840e-01 -2.86749095e-01 1.09911752e+00 -1.38911915e+00
-1.97363541e-01 -1.26478419e-01 -7.40398560e-03 1.01002946e-01
2.88413644e-01 -7.05353975e-01 -1.40150058e+00 5.26335575e-02
1.20353021e-01 -6.91569269e-01 5.31332970e-01 2.52472788e-01
4.75025028e-01 -4.43588197e-01 -3.40422958e-01 -3.03992778e-01
7.37598598e-01 8.26724589e-01 3.02684009e-01 1.11439649e-03
2.60422885e-01 9.01055098e-01 8.42729449e-01 4.16426390e-01
6.22917116e-01 9.47248042e-01 1.02049150e-01 1.48810416e-01
1.81924567e-01 -2.38998622e-01 3.69725764e-01 5.45337558e-01
5.42743802e-01 -3.21019500e-01 -1.04004681e+00 4.98106778e-01
-2.19553804e+00 -6.43653810e-01 1.40090123e-01 1.95876980e+00
1.10591948e+00 3.67455661e-01 1.57916412e-01 -2.12164596e-01
6.09379649e-01 5.65781772e-01 -5.36419272e-01 -7.90312171e-01
2.55666524e-01 -2.42715642e-01 -1.06617674e-01 1.18792474e+00
-1.24843645e+00 1.37970006e+00 5.83172274e+00 4.19323683e-01
-7.73836136e-01 1.00581929e-01 6.56567931e-01 1.58477694e-01
-2.03107577e-02 1.16835423e-01 -1.06551838e+00 3.40018779e-01
1.15632713e+00 1.61867365e-01 5.11761725e-01 9.99202311e-01
2.37246677e-01 -6.20096862e-01 -1.45565736e+00 4.21438485e-01
-1.02344267e-01 -1.12596428e+00 -1.54159591e-01 -2.51476854e-01
3.08304280e-01 -1.06904335e-01 -4.26056743e-01 8.64457190e-01
5.66887617e-01 -7.83884466e-01 1.64801985e-01 3.02648284e-02
3.65545422e-01 -5.81178963e-01 7.79791355e-01 4.91316795e-01
-9.61842418e-01 3.76354600e-03 -1.21645145e-01 -2.21253827e-01
3.10560942e-01 1.34186268e-01 -1.54674137e+00 1.89223275e-01
2.03035578e-01 3.72546941e-01 -7.44122416e-02 -7.64119485e-03
-2.24299952e-01 8.76684010e-01 -2.31761202e-01 -1.64736420e-01
4.84872520e-01 5.39611466e-02 6.71195805e-01 1.46349216e+00
-5.73503315e-01 5.56496382e-01 7.00420797e-01 4.55291063e-01
-1.26409069e-01 1.04938351e-01 -4.47907567e-01 -6.10804819e-02
7.55099058e-01 9.74179566e-01 -4.10183161e-01 -3.85033220e-01
-5.23858905e-01 8.75444055e-01 2.38016784e-01 1.91151172e-01
-3.22933972e-01 1.65506974e-02 9.57288742e-01 -1.88523412e-01
-8.46326500e-02 -3.71251404e-02 -2.86814213e-01 -1.01089633e+00
-2.62626886e-01 -8.47150147e-01 5.92750967e-01 -4.53560442e-01
-9.94034469e-01 7.31210828e-01 3.23485285e-02 -7.49496520e-01
-9.53017950e-01 -3.97302985e-01 -9.03379261e-01 9.70295787e-01
-1.28001904e+00 -9.02240574e-01 -5.26923053e-02 3.89280260e-01
1.32168603e+00 -1.20185599e-01 1.20604849e+00 -9.59687456e-02
-5.70705116e-01 2.76343673e-01 -4.12566513e-01 2.62363911e-01
8.90471399e-01 -1.50029337e+00 5.31730473e-01 4.81489778e-01
-2.85590738e-01 6.36465549e-01 9.08702493e-01 -7.14592934e-01
-1.21051168e+00 -7.98209071e-01 1.18891966e+00 -6.77909911e-01
6.29016101e-01 -7.09829569e-01 -1.00867617e+00 8.92750263e-01
4.80461836e-01 -5.78842282e-01 8.44605684e-01 7.98681796e-01
-5.17670400e-02 4.36761647e-01 -8.91251266e-01 7.79477298e-01
7.19763994e-01 -7.57771969e-01 -9.81208861e-01 2.92838365e-01
1.22644043e+00 -6.09187841e-01 -6.16807044e-01 2.63595939e-01
3.94522637e-01 -9.42430079e-01 8.56899023e-01 -7.46561587e-01
4.08449352e-01 1.99936986e-01 -3.00135970e-01 -1.18865478e+00
2.54149854e-01 -8.49971950e-01 -1.68456838e-01 1.35321045e+00
6.50894761e-01 -4.31523293e-01 5.15663147e-01 1.09426868e+00
-2.07899526e-01 -6.53251350e-01 -7.20479786e-01 -1.31338879e-01
-3.65606338e-01 -8.45867932e-01 3.86761367e-01 8.71790588e-01
6.80799007e-01 1.26582360e+00 -6.03561640e-01 -6.04939498e-02
-5.20790927e-04 1.64649338e-01 9.46683168e-01 -1.16076732e+00
-1.23810984e-01 -2.05248058e-01 2.61782765e-01 -1.39248180e+00
4.54519719e-01 -4.73595500e-01 3.38341266e-01 -1.40522456e+00
-1.18828289e-01 -3.50403398e-01 1.88765541e-01 5.02956271e-01
-4.94375050e-01 -4.87548053e-01 2.45709926e-01 3.00944924e-01
-7.56042421e-01 5.50797164e-01 6.36837125e-01 -4.21821326e-02
-8.87455285e-01 3.20415646e-01 -5.44372439e-01 6.85342848e-01
6.82227910e-01 1.34880664e-02 -5.25792122e-01 1.10956104e-02
-5.04027128e-01 5.78540385e-01 -2.76167393e-01 -6.49095595e-01
3.78689110e-01 -1.62293077e-01 -1.62768722e-01 -7.13033795e-01
7.79689312e-01 -5.73214948e-01 -5.04621089e-01 8.15865248e-02
-9.02763307e-01 -3.77362281e-01 2.84447938e-01 5.10226965e-01
-2.36340702e-01 -2.58736461e-01 5.86106896e-01 -1.99648961e-01
-9.58035171e-01 -2.24123150e-01 -7.99106061e-01 2.51360804e-01
7.99994826e-01 1.87924951e-01 -5.29116571e-01 -8.26276183e-01
-9.73483026e-01 6.59542859e-01 -2.55753510e-02 6.82093918e-01
2.33332202e-01 -7.11363971e-01 -2.53005266e-01 1.24386430e-01
3.08660150e-01 1.35703191e-01 4.22335386e-01 5.46183705e-01
3.87979150e-02 9.95518684e-01 3.14692408e-01 -6.12027109e-01
-1.40459144e+00 1.62638292e-01 3.18343103e-01 -6.49436235e-01
-4.79268879e-01 6.75433397e-01 4.67358232e-01 -9.19505954e-01
6.63633764e-01 -2.83844620e-01 -5.66905320e-01 3.58064264e-01
7.80725896e-01 -4.25012149e-02 -2.26241406e-02 -4.87009615e-01
-2.88527191e-01 -2.00263232e-01 -5.90586483e-01 -4.14495021e-01
8.75017762e-01 -5.26115894e-01 3.61052379e-02 9.20939505e-01
9.19749081e-01 -1.82667017e-01 -1.17823291e+00 -9.15639341e-01
4.89766985e-01 -9.48117524e-02 -1.57793373e-01 -1.26148617e+00
-6.55027479e-02 6.56842291e-01 4.94606830e-02 5.59279740e-01
7.24047959e-01 2.25258142e-01 9.29232776e-01 6.59617662e-01
2.64700025e-01 -1.23696423e+00 1.80457845e-01 1.23399746e+00
8.95893931e-01 -1.53127134e+00 -4.27946568e-01 -5.07140458e-01
-1.24775887e+00 8.17005038e-01 9.50092256e-01 5.07611990e-01
5.21827340e-01 2.01327756e-01 6.54017448e-01 -1.42955678e-02
-1.40686786e+00 -3.36979955e-01 8.34432542e-02 1.46161586e-01
6.73284769e-01 1.43226698e-01 -1.81044698e-01 7.92422950e-01
-5.71992807e-02 -4.30031151e-01 3.62810820e-01 1.24104106e+00
-5.43715775e-01 -1.24445510e+00 -2.89661467e-01 3.57131779e-01
-3.49820167e-01 -1.41571149e-01 -9.08579409e-01 5.57443261e-01
-5.48814774e-01 1.58971596e+00 1.50624484e-01 -5.27411044e-01
4.30266649e-01 8.06669116e-01 -3.80908817e-01 -1.13499844e+00
-9.55044389e-01 2.46651798e-01 8.22921872e-01 -5.56593060e-01
-2.45861650e-01 -8.10341954e-01 -1.34183240e+00 -1.20389894e-01
-2.55064070e-01 4.73510504e-01 5.85230052e-01 1.44824779e+00
2.39765823e-01 5.80288827e-01 8.10306013e-01 -6.95388138e-01
-2.35683635e-01 -1.50140858e+00 -1.63964555e-01 3.69712740e-01
3.46078306e-01 -6.30665421e-01 -3.35617542e-01 -2.59704798e-01] | [12.707925796508789, 7.745953559875488] |
cfa95833-c870-4b79-ac9e-dc1fc29bf0d3 | self-supervised-multi-task-procedure-learning | null | null | https://www.ecva.net/papers/eccv_2020/papers_ECCV/html/2830_ECCV_2020_paper.php | https://www.ecva.net/papers/eccv_2020/papers_ECCV/papers/123620545.pdf | Self-Supervised Multi-Task Procedure Learning from Instructional Videos | We address the problem of unsupervised procedure learning from instructional videos of multiple tasks using Deep Neural Networks (DNNs). Unlike existing works, we assume that training videos come from multiple tasks without key-step annotations or grammars, and the goals are to classify a test video to the underlying task and to localize its key-steps. Our DNN learns task-dependent attention features from informative regions of each frame without ground-truth bounding boxes and learns to discover and localize key-steps without key-step annotations by using an unsupervised subset selection module as a teacher. It also learns to classify an input video using the discovered key-steps using a learnable key-step feature pooling mechanism that extracts and learns to combine key-step based features for task recognition. By experiments on two instructional video datasets, we show the effectiveness of our method for unsupervised localization of procedure steps and video classification. | ['Ehsan Elhamifar', 'Dat Huynh'] | null | null | null | null | eccv-2020-8 | ['procedure-learning'] | ['computer-vision'] | [ 6.92004144e-01 2.86421161e-02 -5.85803270e-01 -5.27531564e-01
-1.00280845e+00 -6.47214472e-01 2.99270719e-01 2.90428042e-01
-6.11976087e-01 5.83116353e-01 1.74673721e-01 -4.38595470e-03
-2.49628812e-01 -6.17265880e-01 -1.38427377e+00 -8.13623905e-01
-1.79372221e-01 1.13740817e-01 2.85621762e-01 5.13976574e-01
3.91931146e-01 3.96492660e-01 -1.86928034e+00 8.91628563e-01
6.52686477e-01 1.04852998e+00 4.88895565e-01 1.05763793e+00
3.75326648e-02 1.64829528e+00 -7.61234581e-01 2.70421177e-01
3.67022425e-01 -5.84498405e-01 -1.01216424e+00 4.19737518e-01
1.10065627e+00 -5.93717873e-01 -5.50987601e-01 1.11021376e+00
3.77300829e-02 7.63811111e-01 6.58154488e-01 -1.15649760e+00
-5.88049531e-01 5.56081712e-01 -6.54352680e-02 4.62676585e-01
6.50335491e-01 4.19024192e-02 7.84664869e-01 -7.16038346e-01
8.57460439e-01 9.77576673e-01 3.95797461e-01 8.26887310e-01
-1.11089265e+00 -4.53782976e-01 3.60672802e-01 2.10936502e-01
-1.10569441e+00 -4.72761542e-01 5.75171471e-01 -5.50042450e-01
8.38013709e-01 -1.34350866e-01 6.73718274e-01 1.29009938e+00
2.75584370e-01 1.27675974e+00 7.15074599e-01 -3.29235047e-01
2.90366739e-01 -1.35256559e-01 2.26664230e-01 1.31162357e+00
1.22929662e-01 -3.31491768e-01 -8.80394042e-01 2.34020039e-01
1.06735706e+00 6.07074976e-01 -5.63019514e-01 -6.31658852e-01
-1.50254726e+00 4.75007147e-01 1.14790767e-01 1.89130917e-01
-3.32155615e-01 2.70989627e-01 4.20805752e-01 5.60636103e-01
2.36031175e-01 6.48517311e-01 -7.78010249e-01 -1.37870580e-01
-9.17094111e-01 1.41029686e-01 7.79868960e-01 1.40508068e+00
9.89334345e-01 -7.86341354e-02 -5.02586126e-01 3.58191758e-01
-3.10012847e-01 1.79116964e-01 6.80680633e-01 -1.35570943e+00
5.74503124e-01 8.07673037e-01 -1.59959227e-01 -4.43570495e-01
-2.27071732e-01 1.65039659e-01 -3.13921094e-01 2.24508688e-01
6.31365061e-01 -1.19087346e-01 -1.28415644e+00 1.38194859e+00
3.02495845e-02 4.27882910e-01 8.64467993e-02 7.05320954e-01
1.11590993e+00 4.85706747e-01 1.81255296e-01 -2.16846511e-01
1.12168562e+00 -1.21192539e+00 -7.53675580e-01 -1.68055668e-01
8.52056563e-01 -2.83634394e-01 8.19962084e-01 3.69784176e-01
-1.37266898e+00 -9.46531475e-01 -7.79164314e-01 -2.98986256e-01
-4.34826791e-01 3.71288657e-01 4.66298968e-01 3.49695794e-02
-1.01502585e+00 5.75047016e-01 -7.83923745e-01 -3.03074926e-01
7.98972964e-01 5.45549095e-01 -8.39022875e-01 -1.94965750e-01
-4.70318735e-01 3.34611356e-01 5.67862630e-01 -1.82825774e-01
-1.71372890e+00 -9.17968929e-01 -1.40400338e+00 4.10419136e-01
9.16770637e-01 -4.94930595e-01 1.34926319e+00 -1.57159734e+00
-1.50625968e+00 9.17870283e-01 -2.14473665e-01 -3.41301829e-01
-5.71568608e-02 -5.00182688e-01 2.31970102e-01 8.14138293e-01
2.38388151e-01 7.37152338e-01 1.06857240e+00 -1.01333773e+00
-9.99378979e-01 -2.29800731e-01 4.95124608e-01 2.17433125e-01
-2.17505679e-01 2.32882742e-02 -3.66110146e-01 -4.63490546e-01
1.62441880e-01 -7.36014307e-01 -2.98074111e-02 -1.20487578e-01
-2.47975066e-01 -3.11561942e-01 8.44756722e-01 -6.57004893e-01
9.60637331e-01 -2.10871887e+00 5.11154532e-01 -1.85825363e-01
4.68899906e-01 -5.78836985e-02 -3.41997623e-01 -1.17842592e-01
-1.57824010e-01 -1.18561156e-01 6.87048957e-03 -1.26784548e-01
-3.73666108e-01 1.84345350e-01 -2.68924106e-02 5.73050559e-01
3.15491378e-01 8.02674830e-01 -1.09369051e+00 -5.38164556e-01
2.11222053e-01 1.43700451e-01 -5.99881470e-01 6.98006094e-01
-3.72007638e-01 2.38858238e-01 -5.92533946e-01 9.66612816e-01
7.64146596e-02 -1.42684489e-01 2.64079809e-01 1.14567643e-02
7.93079957e-02 2.14682117e-01 -1.02007377e+00 2.07433176e+00
-2.29206756e-01 8.60610902e-01 1.68465331e-01 -1.60214651e+00
4.82264072e-01 5.47212481e-01 5.80668032e-01 -4.42060322e-01
1.06484033e-01 -1.13318793e-01 -1.95574790e-01 -1.03987491e+00
1.78193137e-01 5.52952588e-01 -1.62469715e-01 3.00218791e-01
1.24700284e+00 1.03951171e-01 4.05903310e-01 1.59415632e-01
1.49119484e+00 5.86460888e-01 4.57850605e-01 -2.75209665e-01
5.99386275e-01 -4.99977246e-02 6.24476314e-01 9.68841791e-01
-3.44286948e-01 5.76678574e-01 7.11100698e-01 -8.31580043e-01
-6.53201103e-01 -9.00276661e-01 5.26805878e-01 1.83167207e+00
-7.97855034e-02 -2.27136493e-01 -8.10195208e-01 -1.46838915e+00
-1.69328958e-01 2.62325048e-01 -9.76840615e-01 2.58521978e-02
-8.09314609e-01 2.82455653e-01 7.94672966e-02 9.45598006e-01
3.29225153e-01 -1.40628505e+00 -7.50912964e-01 6.54124692e-02
-4.78128567e-02 -1.36039412e+00 -6.10916674e-01 8.05808723e-01
-8.24445903e-01 -1.60500360e+00 -6.19902253e-01 -1.50024617e+00
1.38311350e+00 5.49097955e-01 1.12909377e+00 8.85250568e-02
-2.24262476e-01 8.73455763e-01 -1.22062750e-01 -1.37997776e-01
-6.67297989e-02 -8.07467178e-02 -1.17246531e-01 2.19404371e-03
4.81850147e-01 -2.40426913e-01 -3.11189383e-01 1.31525144e-01
-8.13421130e-01 7.02651218e-02 6.20565712e-01 6.94987297e-01
7.34113693e-01 -1.96135859e-03 1.35318294e-01 -9.47172344e-01
2.70909399e-01 -2.37553835e-01 -7.06099153e-01 4.46266741e-01
2.35302851e-01 5.03895208e-02 7.39024699e-01 -6.72856629e-01
-1.01770175e+00 5.23377001e-01 6.16804123e-01 -8.90295506e-01
-9.02999282e-01 2.56487608e-01 -2.59331256e-01 3.81277082e-03
4.05978888e-01 5.51046073e-01 -5.23790777e-01 -1.99925005e-01
5.10833487e-02 1.74862340e-01 5.84755540e-01 -8.88489962e-01
5.47381759e-01 3.01836699e-01 -1.16971120e-01 -7.10775197e-01
-1.21275806e+00 -6.56532705e-01 -1.12550747e+00 -4.22994912e-01
1.43842530e+00 -1.13593972e+00 -1.08834386e+00 2.63126642e-01
-1.00847816e+00 -7.66703784e-01 -4.53518838e-01 8.30186725e-01
-9.30959225e-01 -3.20921056e-02 -8.29415560e-01 -5.25985360e-01
2.07821295e-01 -1.37317276e+00 1.13862348e+00 5.70964038e-01
-4.91670631e-02 -9.57931161e-01 -2.95509636e-01 4.94970411e-01
-1.36809215e-01 5.20694582e-03 7.80352592e-01 -1.23344946e+00
-1.04123998e+00 -2.09094375e-01 6.53953031e-02 4.13376540e-01
2.02399090e-01 -7.69512132e-02 -9.13990080e-01 -1.22940928e-01
7.66085386e-02 -8.54639411e-01 9.37136352e-01 6.81244135e-01
1.87147617e+00 -2.86281049e-01 -3.09826076e-01 9.05300617e-01
1.19104803e+00 5.76134622e-01 2.89446592e-01 3.05414677e-01
9.59210753e-01 7.29028881e-01 4.28841591e-01 -1.03789821e-01
1.23235583e-01 -4.83477972e-02 2.50897348e-01 2.62718707e-01
1.41897071e-02 -1.92863718e-02 7.07038343e-01 6.95332825e-01
-4.38786477e-01 -2.30135024e-01 -7.63015687e-01 6.16378725e-01
-1.87733901e+00 -1.30981493e+00 5.26194870e-01 2.10272622e+00
6.62583351e-01 1.38935000e-01 -4.67068180e-02 -1.91500396e-01
6.95938110e-01 1.77371845e-01 -3.83699328e-01 -1.87785760e-01
2.62858599e-01 3.96472484e-01 3.88249815e-01 1.34689391e-01
-1.66048884e+00 9.47157323e-01 6.09237194e+00 4.86449122e-01
-8.13646674e-01 -1.71778873e-01 6.32130086e-01 -1.99588045e-01
3.74800771e-01 -2.70003200e-01 -8.75894845e-01 4.09042947e-02
7.73719072e-01 1.10746920e-01 3.59965414e-01 1.13927448e+00
-1.52650513e-02 -2.32182384e-01 -1.69641590e+00 6.78421795e-01
5.61597407e-01 -1.50661695e+00 3.14576104e-02 -2.68405408e-01
1.10905969e+00 -3.08682293e-01 -1.37954429e-01 6.37624979e-01
5.53800762e-01 -9.09800410e-01 4.31045055e-01 6.26333058e-01
5.91776252e-01 -5.58234155e-01 5.11735380e-01 2.56319851e-01
-1.27741337e+00 -4.35241878e-01 -7.29642510e-02 -8.30894932e-02
-5.59645474e-01 -2.62065619e-01 -7.67625511e-01 2.36670107e-01
7.57425606e-01 1.04364669e+00 -3.80890399e-01 1.06070375e+00
-4.94294405e-01 7.13577867e-01 1.15184568e-01 8.70304629e-02
3.63253266e-01 -7.58141875e-02 2.65936941e-01 1.25135386e+00
2.56620139e-01 4.63260487e-02 7.42049217e-01 5.91382027e-01
-5.51483393e-01 -1.69858813e-01 -8.24575841e-01 -3.10097903e-01
2.59905398e-01 1.20505786e+00 -1.02188683e+00 -8.05347502e-01
-8.43434155e-01 1.10414577e+00 4.42321658e-01 8.38017166e-01
-6.09612226e-01 -5.16029418e-01 6.62457585e-01 -2.20193490e-02
6.20981693e-01 -2.56607354e-01 3.20461065e-01 -1.27342427e+00
-6.08450137e-02 -8.40927720e-01 4.48066056e-01 -9.35096323e-01
-7.95833766e-01 1.62703693e-01 2.58316454e-02 -1.31555879e+00
-1.68717220e-01 -9.23928499e-01 -6.78359210e-01 3.13992918e-01
-1.27478468e+00 -1.04615402e+00 -3.59167635e-01 8.44770432e-01
1.31345999e+00 -4.92413610e-01 6.96528614e-01 -5.37551604e-02
-4.95486498e-01 1.86875254e-01 -9.60058197e-02 7.82966912e-01
6.95919156e-01 -1.56973648e+00 -2.56020814e-01 7.73518026e-01
2.35439003e-01 7.24516213e-01 2.49050140e-01 -5.39500535e-01
-1.60958946e+00 -1.24885857e+00 5.37931919e-01 -5.71058691e-01
5.56235850e-01 -6.78910911e-01 -8.22662711e-01 1.32520497e+00
3.74657422e-01 5.33406138e-01 8.49934995e-01 -1.35253310e-01
-1.79423466e-01 -2.17454880e-02 -8.14457536e-01 3.63279253e-01
1.17554688e+00 -7.77721703e-01 -9.07941401e-01 6.64132476e-01
7.15507388e-01 -6.28612518e-01 -7.22912908e-01 1.47952825e-01
4.52488571e-01 -7.05720842e-01 8.14568639e-01 -1.05152786e+00
7.86848843e-01 -1.17658399e-01 -5.39991409e-02 -1.15264690e+00
-2.57994473e-01 -4.69118536e-01 -2.96291709e-01 8.09820890e-01
1.38165012e-01 8.19500387e-02 9.53282416e-01 4.07121211e-01
-6.04984999e-01 -4.25370395e-01 -5.97656310e-01 -3.37997586e-01
-5.06917298e-01 -1.23028040e-01 7.26134628e-02 1.17489433e+00
-2.18246549e-01 1.82260260e-01 -1.89493433e-01 1.72193751e-01
4.28603530e-01 1.98960260e-01 8.15740824e-01 -1.08578360e+00
-1.31578937e-01 -1.38516620e-01 -6.04093373e-01 -1.19040298e+00
7.10067511e-01 -7.51981914e-01 4.81791824e-01 -1.48124278e+00
4.49418664e-01 4.01163965e-01 -5.71349502e-01 6.63016617e-01
-1.95725128e-01 -1.24510467e-01 1.93357840e-02 -1.69371367e-01
-1.53260052e+00 3.76259387e-02 1.35949063e+00 -3.82247716e-01
-3.06654215e-01 4.67416458e-02 -3.10754776e-01 1.02441514e+00
4.14083183e-01 -6.83496773e-01 -5.17606199e-01 -5.54722846e-01
-1.55364782e-01 4.07309324e-01 2.82291055e-01 -1.23126686e+00
4.26974893e-01 -3.66303295e-01 9.51710939e-01 -3.32845360e-01
3.65493894e-02 -9.99639034e-01 -6.91074669e-01 2.41810679e-01
-9.18969452e-01 -1.09526403e-01 1.54359609e-01 6.57354951e-01
-3.44223589e-01 -4.26697135e-01 3.22718918e-01 -5.40233374e-01
-1.26908851e+00 3.65827054e-01 -9.94087636e-01 -6.94109946e-02
1.33655965e+00 -3.86245459e-01 -3.27697784e-01 -1.16214052e-01
-1.09788454e+00 2.20278040e-01 2.57420778e-01 3.72381985e-01
8.21605265e-01 -1.11090851e+00 -3.46965373e-01 4.48599815e-01
2.16342002e-01 3.40239823e-01 2.22146630e-01 5.18498838e-01
-5.82913339e-01 4.14165437e-01 -3.65065753e-01 -7.60309696e-01
-1.50242209e+00 7.51060009e-01 2.24773303e-01 -3.40825468e-02
-5.11594713e-01 1.18018043e+00 6.04565084e-01 -3.94414365e-01
6.91974699e-01 -9.61623549e-01 -3.73189449e-01 1.05720818e-01
6.60511315e-01 -1.99215319e-02 -1.79251149e-01 -2.70675182e-01
-6.06935211e-02 3.69958401e-01 -1.96674332e-01 5.13262570e-01
1.48976660e+00 2.67485827e-01 3.79928313e-02 5.07996798e-01
1.33798814e+00 -2.10046321e-01 -1.59685230e+00 -2.04481959e-01
1.03701659e-01 -2.58216023e-01 -5.00834957e-02 -5.07672668e-01
-1.12861454e+00 7.86507130e-01 2.31135413e-01 -9.67699066e-02
1.10667300e+00 -1.73533857e-02 3.02414358e-01 1.13294256e+00
4.75372262e-02 -1.20991564e+00 5.56277692e-01 6.35639966e-01
4.27273840e-01 -1.40897357e+00 -6.72889650e-02 -1.61811978e-01
-4.57669199e-01 1.57634389e+00 1.17843699e+00 -3.74261200e-01
3.84417087e-01 3.86345983e-01 -9.67441723e-02 -2.60574937e-01
-8.79804432e-01 -2.52600819e-01 4.18782473e-01 5.45693040e-01
4.37397510e-01 -2.27912992e-01 4.26104188e-01 5.12689352e-01
5.92626929e-01 2.44782224e-01 4.51905578e-01 1.59204209e+00
-7.82122076e-01 -5.48132658e-01 -2.07876280e-01 6.96408272e-01
-5.88712215e-01 -1.68318488e-02 -2.32948393e-01 8.49719703e-01
3.80467832e-01 4.32398468e-01 3.90584230e-01 -2.42701918e-01
1.61765099e-01 4.43220764e-01 7.46330321e-01 -1.17417121e+00
-8.18863869e-01 1.53168455e-01 -1.82463214e-01 -9.13998187e-01
-8.13917875e-01 -5.35435200e-01 -1.28146255e+00 2.65738755e-01
-6.60517663e-02 1.99004725e-01 2.85812080e-01 1.02881813e+00
1.00149259e-01 9.92509544e-01 2.84623891e-01 -1.01736140e+00
-1.15231931e-01 -7.78455317e-01 -4.34317470e-01 6.08191133e-01
6.28420830e-01 -5.22710323e-01 -1.99485287e-01 9.39963579e-01] | [8.714776039123535, 0.7132071256637573] |
2c6e50e1-398e-468d-9064-f25fe3612607 | two-branch-multi-scale-deep-neural-network | 2211.16786 | null | https://arxiv.org/abs/2211.16786v1 | https://arxiv.org/pdf/2211.16786v1.pdf | Two-branch Multi-scale Deep Neural Network for Generalized Document Recapture Attack Detection | The image recapture attack is an effective image manipulation method to erase certain forensic traces, and when targeting on personal document images, it poses a great threat to the security of e-commerce and other web applications. Considering the current learning-based methods suffer from serious overfitting problem, in this paper, we propose a novel two-branch deep neural network by mining better generalized recapture artifacts with a designed frequency filter bank and multi-scale cross-attention fusion module. In the extensive experiment, we show that our method can achieve better generalization capability compared with state-of-the-art techniques on different scenarios. | ['Haoliang Li', 'Shiqi Wang', 'Chenqi Kong', 'Jiaxing Li'] | 2022-11-30 | null | null | null | null | ['image-manipulation'] | ['computer-vision'] | [ 2.38891497e-01 -7.50987172e-01 -1.26249745e-01 1.00052021e-01
-8.51496577e-01 -3.79094601e-01 4.03345346e-01 -1.64431408e-01
-3.22046846e-01 4.79455978e-01 -5.82204610e-02 -5.82158506e-01
-2.35005602e-01 -7.18444109e-01 -8.70139301e-01 -6.21654391e-01
-6.30289689e-02 -3.78083944e-01 2.30600774e-01 -2.06007436e-01
5.75447083e-01 5.73229373e-01 -9.23917234e-01 6.93551302e-01
6.95020974e-01 9.91149306e-01 -3.10800448e-02 3.21050286e-01
-5.88727109e-02 7.69427359e-01 -9.91960585e-01 -1.01366448e+00
3.08233857e-01 -8.94391462e-02 -2.81673402e-01 -5.72834350e-02
4.70751643e-01 -5.98202050e-01 -8.79318416e-01 1.60038185e+00
6.18651330e-01 2.13786066e-02 3.77321690e-01 -1.06995511e+00
-1.26798558e+00 5.94588518e-01 -1.17478442e+00 6.48243189e-01
2.11452320e-02 3.10383171e-01 1.72876716e-01 -6.81571960e-01
2.39550859e-01 1.28397703e+00 1.09532428e+00 3.80574554e-01
-8.20913672e-01 -1.15000606e+00 1.70375183e-01 4.84751940e-01
-1.25932705e+00 -1.00768410e-01 1.07210577e+00 -4.92725074e-02
5.57366073e-01 -5.46908304e-02 1.80194452e-01 1.60060942e+00
5.22376239e-01 1.04831660e+00 1.12121558e+00 -2.25838229e-01
-2.61660457e-01 1.66388765e-01 2.67720848e-01 6.81795835e-01
3.73323828e-01 1.63140833e-01 -3.52575451e-01 -2.47041434e-01
8.86943340e-01 5.41948497e-01 -4.66141433e-01 2.01815680e-01
-7.33056784e-01 9.72754478e-01 6.09743536e-01 3.46998036e-01
-2.42540836e-01 1.92440957e-01 5.02649963e-01 1.81378961e-01
1.94952458e-01 1.12918459e-01 -5.07724397e-02 2.63402611e-01
-1.13609111e+00 2.31334165e-01 1.91487551e-01 5.19625247e-01
2.98204303e-01 3.05761755e-01 -2.68491451e-03 6.21700943e-01
7.40456954e-02 3.91885549e-01 8.34266782e-01 -2.14148164e-01
7.54705191e-01 3.52087229e-01 -1.27573326e-01 -1.28516817e+00
-1.98129207e-01 -6.63336813e-01 -1.00020063e+00 9.47728604e-02
1.35614365e-01 -1.35202944e-01 -9.28709209e-01 1.35349107e+00
5.48873581e-02 5.15366077e-01 7.21664652e-02 7.61139095e-01
5.93879819e-01 8.55192721e-01 1.96677029e-01 -1.86987177e-01
1.38765633e+00 -8.97009611e-01 -9.66231942e-01 -1.82640553e-01
-9.42011550e-02 -6.78399563e-01 9.78278577e-01 8.76742721e-01
-8.87683988e-01 -9.64184105e-01 -1.40352929e+00 1.35910198e-01
-4.10028875e-01 5.60084917e-02 8.69786978e-01 1.07739198e+00
-2.89917350e-01 7.80448794e-01 -4.41800773e-01 4.17913273e-02
9.15132642e-01 1.13337487e-01 -5.01202941e-01 -1.92475215e-01
-1.37519670e+00 6.50951087e-01 4.60036159e-01 2.48396173e-01
-7.55022824e-01 -5.21295309e-01 -4.80125517e-01 2.78980285e-01
4.31466043e-01 -1.16135135e-01 8.29360902e-01 -7.43891537e-01
-1.06763172e+00 5.97475588e-01 5.23324966e-01 -8.34854126e-01
4.93912220e-01 -4.62886691e-01 -9.65820193e-01 2.93792874e-01
-3.18271130e-01 8.53123069e-02 1.33666062e+00 -1.08748198e+00
-4.35237437e-01 -6.83407247e-01 -9.76321921e-02 -4.09197509e-01
-8.72161627e-01 1.02158725e-01 -5.75552583e-01 -1.08924460e+00
-2.93471187e-01 -5.71507454e-01 -9.96112451e-02 -1.20793797e-01
-2.49919102e-01 9.78920609e-02 1.25042665e+00 -1.09894824e+00
1.41195190e+00 -2.30492783e+00 -2.22848237e-01 -1.12264985e-02
-6.65618526e-03 9.22743917e-01 -1.20615095e-01 2.03853056e-01
-1.68873206e-01 2.31278673e-01 -2.74937838e-01 3.57019268e-02
-3.34597766e-01 -4.58469540e-01 -7.17176139e-01 5.85262775e-01
1.47320703e-01 7.76487052e-01 -4.39068049e-01 -4.38979864e-01
2.80002326e-01 7.33108521e-01 -1.85710698e-01 -1.85833629e-02
-6.26234047e-04 7.39943013e-02 -5.80612540e-01 6.98525250e-01
1.17060733e+00 -1.70925781e-01 -2.53740419e-02 -3.40503156e-01
4.08033073e-01 -3.12833488e-01 -1.05810320e+00 1.84731448e+00
-3.38435769e-01 5.96088231e-01 -1.85682431e-01 -1.12292898e+00
9.72031236e-01 3.96117680e-02 1.61846042e-01 -7.16805160e-01
4.69820201e-01 -4.71832529e-02 -2.09923029e-01 -7.83024251e-01
3.51674020e-01 -1.86903775e-01 -1.91196442e-01 3.89315993e-01
7.94248208e-02 6.69602394e-01 -2.46899530e-01 1.89443320e-01
1.02684355e+00 5.20099699e-02 -1.12933569e-01 1.14165708e-01
6.48720026e-01 -3.71323109e-01 4.32385445e-01 8.60529006e-01
-3.08846503e-01 4.45168465e-01 2.07397670e-01 -6.12073183e-01
-9.20179188e-01 -1.06604993e+00 -1.65918976e-01 5.00557065e-01
3.31508905e-01 -6.23615496e-02 -9.20393288e-01 -1.12946928e+00
2.42053866e-01 6.96203232e-01 -4.08620119e-01 -6.27955556e-01
-8.07999015e-01 -9.40867066e-01 9.58277106e-01 5.96758723e-01
1.17337394e+00 -9.90486979e-01 -3.12323064e-01 2.64585823e-01
-5.46859503e-02 -1.19125342e+00 -5.66921055e-01 -2.13757858e-01
-7.82204628e-01 -1.13188910e+00 -8.54097247e-01 -6.86924100e-01
1.47686586e-01 5.54781377e-01 4.81978208e-01 3.16461682e-01
-4.90418702e-01 -9.62401330e-02 -4.40846264e-01 -4.11648184e-01
-1.36214629e-01 -6.75698966e-02 -1.50665984e-01 3.15167993e-01
7.33293235e-01 -5.22071779e-01 -4.86316800e-01 -1.17366411e-01
-1.37271214e+00 -3.90269220e-01 7.67415881e-01 9.55604017e-01
1.00871205e-01 7.38358557e-01 7.52650976e-01 -9.95467544e-01
1.03650093e+00 -5.73394179e-01 -5.61825097e-01 4.41822380e-01
-4.92345750e-01 2.68445304e-03 8.02704930e-01 -7.86126852e-01
-1.21838653e+00 -1.36545286e-01 -6.57239780e-02 -1.03335547e+00
1.57685176e-01 3.20842326e-01 -4.10284489e-01 -2.60191798e-01
5.22261083e-01 5.25804698e-01 -1.76459566e-01 -7.80960262e-01
2.27433935e-01 9.15877223e-01 9.96007562e-01 -3.49972457e-01
9.26051319e-01 4.14365172e-01 -1.32434055e-01 -6.49184227e-01
-5.96385598e-01 -1.01278216e-01 -2.54533440e-01 -2.52681017e-01
6.53700590e-01 -7.67907619e-01 -8.64228368e-01 8.19749892e-01
-1.24391866e+00 4.09184724e-01 5.40834427e-01 3.88277262e-01
-1.12471245e-01 9.15679157e-01 -7.79127657e-01 -9.68201280e-01
-6.12956166e-01 -1.02203214e+00 8.31787884e-01 5.35517037e-01
5.34668922e-01 -6.43072963e-01 -2.97950506e-01 5.22153676e-01
2.69808948e-01 3.07906121e-01 8.36737633e-01 -8.10968101e-01
-5.86265683e-01 -4.97956723e-01 -3.83200735e-01 5.66995263e-01
-2.14773923e-01 -2.42057860e-01 -8.63438785e-01 -5.16838014e-01
4.58856910e-01 -1.61712229e-01 1.29472458e+00 1.04096174e-01
1.75428033e+00 -3.00411910e-01 -4.96137083e-01 8.50915670e-01
1.59117258e+00 4.52121615e-01 1.00800562e+00 5.24661601e-01
5.26112556e-01 3.14338028e-01 3.78206670e-01 3.65192771e-01
-3.46919119e-01 3.69775265e-01 4.87289011e-01 1.14463240e-01
5.94471619e-02 -4.66850340e-01 2.48944223e-01 2.74657965e-01
2.64126927e-01 -5.47187030e-01 -5.54387093e-01 3.90973747e-01
-1.78338230e+00 -1.30804515e+00 4.47023772e-02 2.10120511e+00
4.40076053e-01 4.74829137e-01 -5.12760319e-02 2.92060494e-01
1.26082397e+00 3.69368196e-01 -5.66256225e-01 -3.13353896e-01
-1.18495665e-01 4.62160558e-01 7.77784586e-01 -7.27241347e-03
-1.35161865e+00 8.93534899e-01 6.04661417e+00 1.38406074e+00
-1.15161633e+00 4.09791023e-01 5.06744385e-01 -1.19992353e-01
-4.04053256e-02 -2.74522543e-01 -8.20150316e-01 1.03213203e+00
8.32469642e-01 2.65650481e-01 5.09641409e-01 7.43001699e-01
-3.03399056e-01 4.43713784e-01 -4.89420414e-01 1.27600300e+00
5.70495009e-01 -1.45229053e+00 1.82101086e-01 -3.58197698e-03
2.97441036e-01 -4.76083845e-01 4.69165415e-01 3.06426615e-01
1.57341864e-02 -1.01447976e+00 5.09840727e-01 4.50178474e-01
7.10185409e-01 -1.25289869e+00 6.68494165e-01 3.49187165e-01
-7.02370286e-01 -5.12265503e-01 -4.83655632e-01 3.75979066e-01
2.85756618e-01 5.95803022e-01 -6.76773936e-02 6.10823989e-01
6.91569924e-01 3.10825080e-01 -4.57439363e-01 9.67706919e-01
-2.07162082e-01 5.79312503e-01 -2.28397567e-02 1.93988923e-02
4.53023940e-01 5.15592359e-02 4.37999994e-01 1.16932333e+00
3.59363049e-01 -5.54149300e-02 -1.05024524e-01 7.40741551e-01
-4.37044680e-01 -1.54002070e-01 -6.68647707e-01 -2.32877731e-01
4.62673098e-01 1.14842451e+00 -5.60614407e-01 -3.00985217e-01
-4.50898916e-01 1.18062842e+00 1.84029266e-01 1.90050408e-01
-1.27487838e+00 -7.85527468e-01 2.55550534e-01 1.25917569e-01
6.07058346e-01 -1.96641058e-01 5.60191087e-02 -1.32177150e+00
1.19566575e-01 -1.01716912e+00 5.72622180e-01 -7.96823800e-01
-1.44866776e+00 5.83142400e-01 -2.98598647e-01 -1.13875008e+00
2.12964222e-01 -9.29246485e-01 -6.89003527e-01 6.50275052e-01
-1.53707504e+00 -1.39622879e+00 -1.57536745e-01 8.20308506e-01
5.72176635e-01 -6.93601847e-01 3.74327749e-01 6.31724179e-01
-7.05354512e-01 9.67893660e-01 3.57911468e-01 3.85746121e-01
7.06965387e-01 -3.92509639e-01 5.25777698e-01 1.19162762e+00
3.90145838e-01 8.72490227e-01 3.82890075e-01 -8.44926953e-01
-1.48585379e+00 -1.03292239e+00 2.26744283e-02 -4.46798950e-02
5.06597102e-01 -5.51313423e-02 -1.13227034e+00 7.02087402e-01
3.34861636e-01 5.68450987e-02 5.93979955e-01 -2.91138381e-01
-7.48324215e-01 -2.02272058e-01 -1.50083995e+00 3.84211063e-01
6.00503802e-01 -5.03914297e-01 -6.21322870e-01 2.39927664e-01
7.05363631e-01 -1.00053728e-01 -5.21542013e-01 2.89794028e-01
6.61317110e-01 -9.21058357e-01 1.26902342e+00 -8.21107388e-01
4.15715307e-01 -1.51300877e-01 -7.80384839e-02 -8.59636486e-01
-5.39299488e-01 -6.73533738e-01 -4.10575986e-01 1.37346220e+00
-9.43826139e-02 -4.10568744e-01 8.71589899e-01 -9.34566185e-03
2.49076501e-01 -3.16113949e-01 -8.64793181e-01 -1.01570523e+00
1.92459226e-01 -2.43929803e-01 7.23097920e-01 9.58731592e-01
-3.08325022e-01 2.29409277e-01 -1.02488816e+00 3.54025811e-01
9.82240558e-01 -9.98892486e-02 6.17076039e-01 -8.49241138e-01
-5.01945853e-01 -3.06147993e-01 -4.45415020e-01 -8.10107529e-01
1.62033960e-01 -5.76670527e-01 -5.59822738e-01 -8.54375184e-01
4.67620581e-01 9.38764513e-02 -7.03551769e-01 1.35093749e-01
-4.57080007e-01 3.37823719e-01 3.79800886e-01 2.72411048e-01
-3.57591242e-01 3.62010568e-01 9.44492400e-01 -5.07121146e-01
2.54986167e-01 -8.81778300e-02 -8.41748476e-01 6.23389900e-01
7.67539322e-01 -6.67784572e-01 -3.02123964e-01 -5.17693579e-01
-1.70434505e-01 1.37117609e-01 4.77271467e-01 -1.21054530e+00
1.69057041e-01 5.89563400e-02 8.63257945e-01 -7.29481578e-01
2.80909359e-01 -8.92569125e-01 1.35491565e-01 6.66369975e-01
-2.20723346e-01 1.17178246e-01 4.58912104e-01 8.99705470e-01
-1.57401428e-01 -6.43778920e-01 1.02177048e+00 -3.66037488e-01
-6.63027823e-01 2.78909743e-01 -1.13907412e-01 -3.55140299e-01
1.03465366e+00 -1.19204074e-01 -5.90260029e-01 -1.80423230e-01
-1.35977447e-01 -6.14388213e-02 1.45523071e-01 6.70036852e-01
8.54962587e-01 -1.46710026e+00 -6.20512128e-01 2.75328904e-01
-2.38257110e-01 -6.51024044e-01 7.35040188e-01 2.45241299e-01
-4.33898628e-01 2.38356099e-01 -5.14151454e-01 -1.07719392e-01
-1.27935076e+00 1.28465509e+00 1.05072871e-01 -4.12213027e-01
-7.49065459e-01 7.16202259e-01 5.40891215e-02 7.86584243e-02
2.08936110e-01 3.90984654e-01 -1.35210603e-01 -2.49169290e-01
9.48189139e-01 3.31776619e-01 -7.18205124e-02 -3.80727172e-01
-1.99375913e-01 4.11352754e-01 -5.70235133e-01 1.70296386e-01
1.35466087e+00 1.51642144e-01 3.12149167e-01 -2.09075406e-01
1.38162076e+00 7.82974530e-03 -1.18465650e+00 -2.58347422e-01
-2.67974496e-01 -8.70585799e-01 3.44649106e-01 -6.73531115e-01
-1.45024657e+00 1.15049291e+00 8.61372292e-01 1.70872435e-01
1.25189817e+00 -5.70970833e-01 1.37896419e+00 3.36051583e-01
2.84514129e-01 -1.04324317e+00 3.91777635e-01 -1.48062810e-01
7.10158229e-01 -1.21089375e+00 3.01362723e-01 -2.91836355e-02
-5.65967143e-01 1.11484742e+00 5.58889866e-01 -5.76355577e-01
6.12220109e-01 2.82740325e-01 -2.03887299e-01 -2.47708932e-02
-2.36741245e-01 3.68420810e-01 -1.57040715e-01 7.33010769e-01
-1.11722596e-01 -4.04494166e-01 -4.03588504e-01 1.01084459e+00
1.79077759e-01 9.57431048e-02 3.88612449e-01 9.17280138e-01
-3.90153557e-01 -9.73650753e-01 -7.08817244e-01 3.70376557e-01
-1.07501733e+00 -1.86191857e-01 -1.53841794e-01 9.09680545e-01
9.03143212e-02 7.94547617e-01 1.26845036e-02 -6.64889634e-01
1.55551583e-01 -6.88487962e-02 6.89170659e-01 1.22641616e-01
-7.58635044e-01 3.79740968e-02 -2.88046867e-01 -3.20416391e-01
-2.08024591e-01 -4.45224673e-01 -8.44762504e-01 -5.63429952e-01
-7.55215347e-01 -8.18455815e-02 7.49797702e-01 6.87512219e-01
3.56489211e-01 5.55351794e-01 6.95645154e-01 -3.88335943e-01
-9.74537134e-01 -1.02156889e+00 -7.45676994e-01 7.00018227e-01
1.30622894e-01 -6.84393108e-01 -1.85302854e-01 4.33291309e-02] | [12.392779350280762, 0.986628532409668] |
5b5adb8d-afb0-4a15-87a6-2927cc75fe69 | micro-stripes-analyses-for-iris-presentation | 2010.14850 | null | https://arxiv.org/abs/2010.14850v2 | https://arxiv.org/pdf/2010.14850v2.pdf | Micro Stripes Analyses for Iris Presentation Attack Detection | Iris recognition systems are vulnerable to the presentation attacks, such as textured contact lenses or printed images. In this paper, we propose a lightweight framework to detect iris presentation attacks by extracting multiple micro-stripes of expanded normalized iris textures. In this procedure, a standard iris segmentation is modified. For our presentation attack detection network to better model the classification problem, the segmented area is processed to provide lower dimensional input segments and a higher number of learning samples. Our proposed Micro Stripes Analyses (MSA) solution samples the segmented areas as individual stripes. Then, the majority vote makes the final classification decision of those micro-stripes. Experiments are demonstrated on five databases, where two databases (IIITD-WVU and Notre Dame) are from the LivDet-2017 Iris competition. An in-depth experimental evaluation of this framework reveals a superior performance compared with state-of-the-art algorithms. Moreover, our solution minimizes the confusion between textured (attack) and soft (bona fide) contact lens presentations. | ['Arjan Kuijper', 'Florian Kirchbuchner', 'Naser Damer', 'Meiling Fang'] | 2020-10-28 | null | null | null | null | ['iris-segmentation'] | ['medical'] | [ 6.77760899e-01 -4.28217277e-02 -4.05945629e-01 -3.07835877e-01
-4.23861623e-01 -4.90277231e-01 5.14667153e-01 -1.23827443e-01
-1.88374758e-01 1.62164748e-01 1.18605420e-02 -4.95833397e-01
-2.12377161e-01 -3.66548210e-01 -4.28422749e-01 -7.59882569e-01
3.68747413e-02 1.14066459e-01 3.20261456e-02 1.45801440e-01
7.40845144e-01 5.20203948e-01 -1.78218687e+00 7.88195789e-01
1.01532817e+00 1.15621698e+00 -8.93156528e-01 7.55646408e-01
1.27271101e-01 3.47269565e-01 -6.88566566e-01 -7.42205918e-01
6.23975098e-01 -4.83247310e-01 -7.43363202e-01 5.65547466e-01
1.19981539e+00 -4.53046173e-01 1.07750401e-01 1.38959122e+00
5.35914600e-01 -3.63437086e-01 6.17887378e-01 -7.28370607e-01
-3.02062213e-01 2.88985282e-01 -1.20993721e+00 2.54047513e-01
4.49686378e-01 4.70730096e-01 3.36695701e-01 -6.37404144e-01
5.94245970e-01 1.25285053e+00 2.56595403e-01 6.46455467e-01
-1.15198004e+00 -7.07294524e-01 -2.64064491e-01 -9.56325699e-03
-1.18771446e+00 -3.50661576e-01 4.88880277e-01 -6.02468133e-01
3.99292856e-01 9.57790673e-01 6.20058060e-01 9.53246057e-01
1.91803411e-01 9.86843944e-01 1.84933603e+00 -6.34300828e-01
-7.65977725e-02 3.72059137e-01 3.90201032e-01 6.33776963e-01
3.93741488e-01 3.19524199e-01 -3.76208395e-01 -3.26358616e-01
5.24858773e-01 -1.20835051e-01 -1.94020867e-01 -4.12524454e-02
-7.91943550e-01 4.23886031e-01 2.96105836e-02 1.61105469e-01
-2.37843003e-02 -7.11468101e-01 1.75432682e-01 2.79118419e-01
4.38966960e-01 3.27585757e-01 1.54670641e-01 4.07806272e-03
-9.51927543e-01 -9.73316878e-02 6.61638737e-01 4.84837085e-01
2.75413841e-01 -4.17680889e-01 -4.17218506e-01 7.94763327e-01
3.39071691e-01 5.42779028e-01 2.59831429e-01 -1.64689824e-01
4.98194367e-01 1.05446064e+00 -1.05326235e-01 -8.94894123e-01
-9.00130421e-02 -4.53261524e-01 -8.68416071e-01 5.80599189e-01
6.59171939e-01 -1.19789012e-01 -1.38668454e+00 6.08884394e-01
6.00383997e-01 3.96814227e-01 6.25804886e-02 1.18309343e+00
8.53667140e-01 8.61033574e-02 -3.96174878e-01 6.76080510e-02
1.70714271e+00 -7.18187571e-01 -4.56010163e-01 1.55911118e-01
4.32469815e-01 -1.38731337e+00 8.74308407e-01 8.67242992e-01
-9.44987476e-01 -3.60315412e-01 -9.99614239e-01 1.47891507e-01
-1.89535111e-01 6.19724751e-01 2.14155287e-01 1.38594925e+00
-7.19996750e-01 2.54514635e-01 -5.62175572e-01 -3.85830611e-01
6.15088046e-01 6.93371177e-01 -1.86485767e-01 1.89865828e-01
-5.60596108e-01 4.96739656e-01 -2.72828955e-02 1.57322899e-01
-2.00130880e-01 -3.90272230e-01 -6.33627474e-01 -4.84768271e-01
1.88004628e-01 -8.72561261e-02 8.75276029e-01 -1.03289795e+00
-1.82992268e+00 1.46729124e+00 -2.18819067e-01 -3.88324052e-01
6.31274939e-01 -9.19359550e-02 -5.88681519e-01 3.09750706e-01
-5.48883379e-01 1.31456241e-01 1.35477185e+00 -1.10516608e+00
-8.75176370e-01 -5.50011396e-01 -6.86923265e-02 -7.42170203e-04
-3.94046336e-01 6.10802054e-01 -6.25279248e-01 -5.39841115e-01
2.80033946e-01 -1.07377291e+00 1.34556904e-01 -2.22894460e-01
-1.08571827e+00 -1.65999785e-01 5.15987813e-01 -6.74529314e-01
1.49057579e+00 -2.41241336e+00 -1.44074246e-01 8.51601660e-01
3.81383359e-01 8.55388582e-01 5.01035750e-02 -6.90891445e-02
-1.55253336e-01 2.60143429e-01 3.13821211e-02 -4.36406344e-01
-4.40899916e-02 -2.30918244e-01 -5.45189261e-01 7.03976691e-01
2.02564880e-01 3.77185583e-01 -2.84266353e-01 -4.53771561e-01
3.64356816e-01 2.34994844e-01 -2.83518106e-01 -2.47047283e-02
1.61956385e-01 4.10065800e-01 -3.87462914e-01 1.14348936e+00
1.10724020e+00 -9.38538164e-02 5.55306859e-03 -1.96195424e-01
-1.64216653e-01 6.34527905e-03 -1.39360166e+00 1.03271365e+00
2.94916540e-01 5.20402789e-01 -2.83363521e-01 -3.40565950e-01
8.87172222e-01 8.26015323e-02 -4.58388813e-02 -7.30880380e-01
4.10288125e-01 2.57250458e-01 2.29142293e-01 -8.01139891e-01
4.22309041e-01 4.55772668e-01 4.14816976e-01 5.18728137e-01
-5.30655086e-01 2.78962195e-01 2.14964300e-01 -2.40306705e-01
4.28389251e-01 -6.64085150e-02 -1.00147082e-02 -1.19724974e-01
8.13872278e-01 -1.79534286e-01 3.43013942e-01 6.53031766e-01
-3.45191270e-01 6.18434608e-01 8.68849993e-01 -8.11023593e-01
-7.67767072e-01 -5.66846132e-01 -7.97077835e-01 3.51023138e-01
3.89973193e-01 -3.45168471e-01 -1.26294589e+00 -9.05465424e-01
3.46592180e-02 5.36737032e-02 -8.49496961e-01 2.16517091e-01
-3.23334217e-01 -1.04589450e+00 7.22384453e-01 -1.31127313e-01
4.53407168e-01 -7.55770504e-01 -6.50117993e-01 -3.90415281e-01
2.10812166e-01 -7.23508239e-01 -5.82965612e-01 -7.55228281e-01
-5.06120741e-01 -1.47792840e+00 -7.07083941e-01 -7.00459778e-01
1.15379214e+00 -3.29215964e-03 3.73007536e-01 2.89335579e-01
-8.98829997e-01 -1.30981058e-01 4.45410376e-03 -5.81964850e-01
-2.14632019e-01 -2.59686053e-01 1.97956353e-01 1.05682027e+00
1.01611304e+00 1.88778564e-01 -7.93317974e-01 5.78890920e-01
-7.75858879e-01 -1.37595877e-01 7.32307017e-01 7.07955241e-01
7.50242412e-01 -1.69012770e-01 -3.16920072e-01 -1.06712306e+00
5.86567700e-01 -6.66004494e-02 -9.14828777e-01 3.67597818e-01
-5.93287408e-01 -3.43995750e-01 -3.49711217e-02 -8.21642876e-01
-8.91723692e-01 -1.07590407e-01 3.73053044e-01 -4.69498992e-01
-4.44151670e-01 1.81418821e-01 3.18325311e-01 -6.16958678e-01
8.86644602e-01 1.06888786e-01 2.64327168e-01 -6.75533295e-01
-1.25029147e-01 1.28115559e+00 4.21015054e-01 -3.94959748e-01
7.68846989e-01 4.59031135e-01 -1.19037442e-02 -8.31798673e-01
-3.11768025e-01 -4.71881598e-01 -4.48527545e-01 -7.55951777e-02
6.98912382e-01 -3.96354049e-01 -1.08992517e+00 1.18259776e+00
-7.33868480e-01 1.43796233e-02 1.49773911e-01 6.36700273e-01
5.28217107e-02 6.29126906e-01 -7.25853682e-01 -9.46545601e-01
-3.46651405e-01 -1.47433257e+00 1.02342856e+00 9.01994228e-01
-2.23042831e-01 -4.56321597e-01 1.57455057e-02 7.85558522e-01
1.12386584e-01 5.59919417e-01 7.88042426e-01 -7.08048284e-01
-6.00478411e-01 -4.39680398e-01 -3.01670343e-01 2.60164648e-01
1.43486485e-01 6.20970488e-01 -1.26282573e+00 -4.25370932e-01
-5.17904274e-02 -1.31925613e-01 7.92875350e-01 2.87211001e-01
1.19580758e+00 -4.15758848e-01 -3.09395462e-01 1.08977175e+00
1.26198578e+00 4.07791197e-01 9.96416152e-01 4.60210592e-01
2.99106628e-01 7.86483705e-01 5.26365638e-01 2.72603005e-01
-9.63151753e-02 5.29552162e-01 2.54654288e-01 -7.34579921e-01
-1.10897481e-01 1.62608683e-01 3.30855995e-02 5.35233803e-02
-3.57835203e-01 -2.73830146e-02 -1.02221560e+00 2.87046194e-01
-1.31944180e+00 -6.84007406e-01 -2.43549407e-01 2.49633837e+00
9.89400387e-01 1.84118688e-01 2.34324574e-01 1.57445565e-01
8.67057085e-01 -6.40355721e-02 -5.79648614e-01 -6.29568994e-01
-2.47607589e-01 5.63876569e-01 4.45924938e-01 3.85963112e-01
-1.40421355e+00 8.12555492e-01 5.76903248e+00 9.75405991e-01
-1.55386472e+00 -4.73071456e-01 9.34040189e-01 -3.76116067e-01
2.53160685e-01 -3.42479408e-01 -9.70879674e-01 5.57143271e-01
5.74040711e-01 3.13331276e-01 3.04124266e-01 3.46327275e-01
-2.03538299e-01 -2.97456086e-01 -8.03953767e-01 1.26815999e+00
2.96998143e-01 -1.19664955e+00 6.52248636e-02 5.31583071e-01
8.12226653e-01 -2.40928069e-01 8.83445859e-01 -3.83260250e-01
-2.12454826e-01 -1.17119598e+00 1.72426030e-01 7.51450956e-01
1.30157483e+00 -6.54758096e-01 8.07669580e-01 -1.46849394e-01
-5.99770665e-01 -6.38663396e-02 -2.44455963e-01 3.10636163e-01
-6.02023840e-01 3.93401027e-01 -8.41946661e-01 5.94155371e-01
7.15285718e-01 4.85830069e-01 -9.05515909e-01 1.39235878e+00
-6.23174235e-02 7.76223779e-01 -3.95609558e-01 7.70441294e-02
1.95862707e-02 -3.96361053e-01 8.15723598e-01 1.03929937e+00
5.59939258e-02 1.71452314e-01 -1.46799400e-01 6.63861454e-01
1.51446879e-01 4.08008158e-01 -2.91182280e-01 -8.85435939e-02
1.37471156e-02 1.20959532e+00 -6.46342099e-01 -3.11805338e-01
-4.70817536e-01 6.32811069e-01 -2.45562702e-01 4.46702659e-01
-2.35436723e-01 -5.82053185e-01 7.44134188e-01 1.67126253e-01
2.66667753e-02 4.19375122e-01 -6.47171676e-01 -1.16108012e+00
1.72381163e-01 -1.55460072e+00 4.00434226e-01 -3.18040550e-01
-1.14985371e+00 7.54431605e-01 -4.54031974e-01 -1.75039780e+00
-6.19960111e-03 -7.25849807e-01 -5.97764075e-01 1.33380949e+00
-1.32863140e+00 -1.27323365e+00 -2.38989070e-01 6.13723040e-01
1.25650048e-01 -6.38826489e-01 6.05006635e-01 8.68827403e-02
-1.09505582e+00 1.21987283e+00 -1.00929499e-01 4.74911183e-01
9.83201981e-01 -1.13769507e+00 4.08709377e-01 1.17791867e+00
-3.19793858e-02 9.71127570e-01 2.91421771e-01 -7.44425297e-01
-1.28801560e+00 -6.30461276e-01 6.92010224e-01 -4.34755981e-01
3.56100857e-01 -5.56471162e-02 -9.11519051e-01 3.22245538e-01
1.99816734e-01 -1.25481248e-01 1.01201415e+00 7.45880604e-02
-4.50764418e-01 3.79150058e-03 -1.27608800e+00 6.79394782e-01
4.27422851e-01 -4.58864421e-01 -5.25971234e-01 2.15625569e-01
-2.11115718e-01 -1.05719960e+00 -9.29904819e-01 6.21491730e-01
8.56672525e-01 -1.15576029e+00 6.37415826e-01 -7.71153212e-01
4.11945134e-01 -5.94572067e-01 3.61623824e-01 -6.55105472e-01
1.79170236e-01 -1.18462420e+00 7.31802220e-03 8.90190005e-01
3.40707004e-01 -8.55967760e-01 7.77567208e-01 7.11021364e-01
2.65570998e-01 -9.66223478e-01 -6.71589851e-01 -3.29988539e-01
-3.71705025e-01 2.47748598e-01 7.94334769e-01 1.06303334e+00
9.29264277e-02 -4.20485109e-01 -3.25575143e-01 5.16105533e-01
9.62218523e-01 3.03922802e-01 1.13372231e+00 -1.25113416e+00
-9.71330926e-02 -6.28600717e-01 -7.82658041e-01 -7.04175889e-01
-5.26623607e-01 -5.95236182e-01 -6.21166050e-01 -6.86460972e-01
5.82789518e-02 -1.86474383e-01 -2.99710602e-01 6.86045647e-01
-2.47521490e-01 7.63311923e-01 1.49274059e-02 3.23918909e-01
-3.53884026e-02 -4.52390492e-01 1.40457487e+00 -2.60806292e-01
-5.12696147e-01 4.54499900e-01 -5.91369390e-01 5.17732143e-01
6.56663358e-01 -1.43145040e-01 -8.32244009e-02 -3.48672494e-02
-9.32412222e-02 -2.57677913e-01 2.51734436e-01 -7.34750926e-01
3.32299471e-01 -5.91859445e-02 5.57377398e-01 -5.64835787e-01
-1.22893222e-01 -3.63036335e-01 -1.39177009e-01 5.38375080e-01
-3.61027032e-01 -6.50745153e-01 4.21343118e-01 1.32695109e-01
-1.24674529e-01 5.51925451e-02 1.03857732e+00 4.43159342e-01
-1.02567188e-01 3.74606848e-01 -1.36998355e-01 -2.81400800e-01
1.03327203e+00 -7.85142720e-01 -8.58713090e-01 3.10048014e-01
-7.39316940e-01 1.10757127e-01 6.35689199e-01 4.88920212e-01
6.06521666e-01 -7.92436838e-01 -8.38964522e-01 1.18952000e+00
2.99058229e-01 -2.42713287e-01 3.07545573e-01 1.13476717e+00
-6.89775288e-01 1.77009836e-01 -2.26279587e-01 -8.62076581e-01
-2.00839472e+00 3.25738639e-01 5.53346634e-01 9.31410193e-02
-6.55030012e-01 8.71625304e-01 2.08881050e-02 9.98064652e-02
4.17417049e-01 -5.66093385e-01 -5.30962586e-01 8.53274986e-02
1.17175138e+00 3.06618094e-01 1.55925974e-01 -5.09005487e-01
-5.11846952e-02 9.92601454e-01 -6.77634180e-01 1.33496970e-01
7.09109366e-01 9.22674909e-02 -5.95643818e-01 -1.38946339e-01
7.05734372e-01 4.15139079e-01 -9.73758399e-01 -4.34444457e-01
-2.65268907e-02 -1.06350279e+00 -2.16956288e-01 -9.80584562e-01
-9.74633634e-01 8.18307817e-01 1.18603313e+00 2.25369036e-01
1.19241548e+00 -4.99049723e-01 6.45274043e-01 -5.04946448e-02
2.52859313e-02 -8.86345029e-01 -3.00895542e-01 -3.58161852e-02
5.61204553e-01 -9.78338718e-01 -9.25317407e-02 -6.33184791e-01
-5.72034538e-01 1.21305084e+00 5.89576960e-01 2.28560530e-02
4.06788230e-01 1.41178578e-01 6.29014373e-01 -2.14124694e-01
-2.25636914e-01 -1.59616932e-01 1.16552222e+00 4.65281665e-01
5.41734919e-02 7.78718516e-02 -4.31562215e-01 2.22269967e-01
-2.26835147e-01 -7.31487945e-02 5.35678625e-01 6.74287736e-01
-1.61901176e-01 -1.21579719e+00 -6.84783936e-01 7.44248688e-01
-7.98745215e-01 -7.57655129e-02 -7.79643714e-01 4.64133918e-01
2.02140093e-01 9.37308729e-01 9.77387130e-02 -5.88137090e-01
3.26180011e-01 2.17620260e-03 2.96908170e-01 -4.40505624e-01
-1.01170063e+00 5.73755145e-01 -8.88218135e-02 -5.84708452e-01
-3.36875290e-01 -5.83501399e-01 -6.58447027e-01 -3.45433950e-01
-1.86927214e-01 -1.53935522e-01 6.50248766e-01 7.60076284e-01
3.71472955e-01 1.23210631e-01 7.96624601e-01 -2.52462566e-01
-4.97329921e-01 -9.52224314e-01 -8.15712392e-01 5.86182356e-01
7.52812564e-01 -2.62411088e-01 -5.59637249e-01 7.41793960e-02] | [3.7397847175598145, -3.634079933166504] |
a54ce186-e0ec-4710-964d-9387022b7a85 | ranking-based-siamese-visual-tracking | 2205.11761 | null | https://arxiv.org/abs/2205.11761v1 | https://arxiv.org/pdf/2205.11761v1.pdf | Ranking-Based Siamese Visual Tracking | Current Siamese-based trackers mainly formulate the visual tracking into two independent subtasks, including classification and localization. They learn the classification subnetwork by processing each sample separately and neglect the relationship among positive and negative samples. Moreover, such tracking paradigm takes only the classification confidence of proposals for the final prediction, which may yield the misalignment between classification and localization. To resolve these issues, this paper proposes a ranking-based optimization algorithm to explore the relationship among different proposals. To this end, we introduce two ranking losses, including the classification one and the IoU-guided one, as optimization constraints. The classification ranking loss can ensure that positive samples rank higher than hard negative ones, i.e., distractors, so that the trackers can select the foreground samples successfully without being fooled by the distractors. The IoU-guided ranking loss aims to align classification confidence scores with the Intersection over Union(IoU) of the corresponding localization prediction for positive samples, enabling the well-localized prediction to be represented by high classification confidence. Specifically, the proposed two ranking losses are compatible with most Siamese trackers and incur no additional computation for inference. Extensive experiments on seven tracking benchmarks, including OTB100, UAV123, TC128, VOT2016, NFS30, GOT-10k and LaSOT, demonstrate the effectiveness of the proposed ranking-based optimization algorithm. The code and raw results are available at https://github.com/sansanfree/RBO. | ['Qiang Ling', 'Feng Tang'] | 2022-05-24 | null | http://openaccess.thecvf.com//content/CVPR2022/html/Tang_Ranking-Based_Siamese_Visual_Tracking_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Tang_Ranking-Based_Siamese_Visual_Tracking_CVPR_2022_paper.pdf | cvpr-2022-1 | ['visual-tracking'] | ['computer-vision'] | [-3.26503068e-01 -2.18615666e-01 -4.70342129e-01 -2.51239568e-01
-5.50916255e-01 -5.19692659e-01 5.48824549e-01 -2.00476591e-02
-4.38083977e-01 7.29992330e-01 -3.08422089e-01 1.50067639e-02
-7.77499080e-02 -4.41831708e-01 -6.94974184e-01 -9.36619282e-01
-7.65419006e-02 3.33096415e-01 8.34029078e-01 2.83509731e-01
7.44017437e-02 2.50555187e-01 -1.47799134e+00 -2.27486178e-01
9.48585868e-01 1.45858216e+00 1.76093906e-01 2.09663495e-01
2.52861559e-01 4.80247736e-01 -6.80990815e-01 -3.32225591e-01
3.87507945e-01 -5.81811555e-02 3.23492475e-02 -2.01594889e-01
5.87202191e-01 -9.02136788e-02 -2.10785165e-01 1.51650286e+00
3.67471099e-01 6.20251000e-02 4.79037374e-01 -1.75645494e+00
-2.45301887e-01 3.07491928e-01 -8.94151032e-01 3.44169527e-01
-8.68059099e-02 3.93174917e-01 9.52965081e-01 -9.39908564e-01
3.68797451e-01 1.18041134e+00 5.05739331e-01 4.17579174e-01
-9.08404112e-01 -1.04781461e+00 5.39314926e-01 2.79127032e-01
-1.50901997e+00 -3.45075935e-01 6.24962807e-01 -6.79092050e-01
1.11721337e-01 3.27889711e-01 7.22916365e-01 9.39936996e-01
1.80382222e-01 7.95767188e-01 9.26464856e-01 2.19176441e-01
2.72799190e-02 3.21227938e-01 3.91847342e-01 6.28263950e-01
7.29908407e-01 4.29387033e-01 -4.50816423e-01 -7.72774145e-02
4.21494544e-01 -9.79023278e-02 -4.15162295e-01 -6.63625419e-01
-1.37393177e+00 5.11235893e-01 5.49519420e-01 -1.82296455e-01
-8.14087316e-02 3.33531685e-02 3.79726917e-01 -8.05852041e-02
3.25619698e-01 -1.99492753e-01 -3.27316135e-01 1.51168704e-01
-8.94775510e-01 7.22783431e-02 3.10248464e-01 1.13500297e+00
7.44994640e-01 1.45205319e-01 -5.48354328e-01 5.07710218e-01
8.40981185e-01 8.92852664e-01 2.17172548e-01 -5.74806154e-01
6.01388574e-01 4.37063098e-01 5.39735675e-01 -1.23976576e+00
-1.37268692e-01 -8.76048803e-01 -6.03022814e-01 3.25982392e-01
5.28116703e-01 -2.61467487e-01 -6.52196586e-01 1.82921600e+00
8.10977817e-01 4.70867366e-01 -1.61574602e-01 1.38267016e+00
6.11203432e-01 5.47415555e-01 1.20681472e-01 -3.01394612e-01
1.41248584e+00 -1.01699424e+00 -6.17710590e-01 -2.63534397e-01
2.69289196e-01 -7.92834580e-01 7.66126633e-01 2.90699184e-01
-6.09104037e-01 -7.28387535e-01 -1.21510673e+00 5.47390580e-01
-1.19747028e-01 8.25447142e-01 3.83590162e-01 4.16384161e-01
-4.96695101e-01 2.67397881e-01 -8.69117737e-01 -1.84866965e-01
5.09743690e-01 2.90542603e-01 3.58448289e-02 4.14537728e-01
-1.01717806e+00 5.67241073e-01 3.92755985e-01 4.04304236e-01
-1.04913354e+00 -5.55578947e-01 -4.52568740e-01 -5.69063909e-02
7.07555890e-01 -2.74710536e-01 1.01006079e+00 -8.32948506e-01
-1.22960830e+00 5.52390397e-01 -5.99600002e-02 -4.56305534e-01
7.22276568e-01 -2.99626827e-01 -5.32562554e-01 -3.53330731e-01
3.73478413e-01 5.75850010e-01 9.59957957e-01 -1.30442119e+00
-1.14100826e+00 -3.32748622e-01 -6.82193190e-02 2.45655373e-01
-2.12910935e-01 -1.08937599e-01 -6.43606484e-01 -5.69151819e-01
3.19942906e-02 -8.92848909e-01 1.38381153e-01 3.46282244e-01
-4.66205120e-01 -4.51226711e-01 9.92098451e-01 -2.50428557e-01
1.12967110e+00 -2.31130147e+00 -9.01018828e-02 2.39491224e-01
2.13675529e-01 4.09366459e-01 3.35980090e-03 -2.38826960e-01
2.06159353e-01 -2.33669698e-01 3.32611591e-01 -1.93269894e-01
6.42020404e-02 -7.85462186e-02 -3.54988337e-01 8.49714220e-01
2.33208850e-01 5.76320350e-01 -1.12880564e+00 -7.44953632e-01
3.29156190e-01 2.54331648e-01 -2.42396548e-01 1.83516085e-01
-2.53354281e-01 3.75867426e-01 -6.78074539e-01 7.08480358e-01
8.96200716e-01 -1.76226661e-01 -5.91834784e-02 -5.56730688e-01
-3.85000527e-01 8.95985961e-02 -1.44972849e+00 1.04626310e+00
1.02992326e-01 4.53261942e-01 2.22219542e-01 -6.72011256e-01
9.22526181e-01 -1.66607648e-01 4.17064428e-01 -5.51123142e-01
1.46511942e-01 8.74534249e-02 5.81306145e-02 -9.53290835e-02
3.91009003e-01 2.04687297e-01 8.20932612e-02 -1.32497162e-01
-7.96077028e-02 6.21731520e-01 2.39452556e-01 1.29637927e-01
5.72898090e-01 4.03611243e-01 1.82206817e-02 -3.31360072e-01
7.96139479e-01 -1.23537779e-02 1.31553161e+00 6.87357724e-01
-7.42234945e-01 2.29469106e-01 4.32804734e-01 -1.37379155e-01
-3.85161817e-01 -1.20824134e+00 -8.86761621e-02 9.02553082e-01
1.03230143e+00 -2.01717719e-01 -4.51375157e-01 -9.10171032e-01
9.61713418e-02 3.77774686e-01 -4.50976819e-01 -2.71347493e-01
-4.20733690e-01 -6.00436985e-01 4.47305143e-01 3.83296520e-01
3.61206919e-01 -8.36064994e-01 -6.48042202e-01 -9.04524401e-02
-6.82683140e-02 -9.38391030e-01 -7.88133979e-01 3.88816521e-02
-6.18074298e-01 -1.29697895e+00 -4.52234089e-01 -5.80135047e-01
6.98692620e-01 5.07995129e-01 7.29933441e-01 6.19682819e-02
-6.46660700e-02 1.32685781e-01 -1.84013501e-01 -3.04077893e-01
1.36129573e-01 -1.27206862e-01 3.81106079e-01 3.02695096e-01
2.00119182e-01 -6.50680717e-03 -6.56799972e-01 8.40622425e-01
-2.33667046e-01 -9.53749865e-02 5.46966732e-01 7.51614511e-01
9.30446565e-01 -4.73668538e-02 4.05758828e-01 -2.70455599e-01
4.35317121e-02 -5.99050641e-01 -1.33171809e+00 3.67652506e-01
-4.76506382e-01 -8.31771865e-02 6.58033848e-01 -7.73511827e-01
-8.07648003e-01 1.23785712e-01 3.09067786e-01 -8.71954560e-01
1.39614448e-01 7.26646632e-02 -4.28155124e-01 -2.97879547e-01
3.13538313e-01 2.91258126e-01 -1.72358572e-01 -2.62517273e-01
-3.95672210e-02 2.52233028e-01 5.17273128e-01 -4.71645027e-01
1.24923921e+00 3.97494048e-01 -6.32791221e-02 -6.69169068e-01
-8.44322145e-01 -5.54085195e-01 -1.62152305e-01 -7.50229836e-01
7.72097409e-01 -9.80626822e-01 -8.60904694e-01 3.64384651e-01
-9.61826026e-01 7.62332454e-02 -1.57571644e-01 7.54273415e-01
-4.16751914e-02 4.75281715e-01 -2.61547510e-02 -1.09854734e+00
-2.18249217e-01 -1.23825979e+00 1.07391584e+00 8.27417910e-01
2.88101137e-01 -5.37739635e-01 -1.64464578e-01 7.44571090e-02
1.43095046e-01 1.23195298e-01 1.51408777e-01 -6.34199381e-01
-1.01650357e+00 -1.91303119e-01 -4.10319507e-01 1.56420544e-01
-1.23119794e-01 2.71136612e-01 -7.89699316e-01 -5.39924026e-01
-2.79582828e-01 -2.28159770e-01 8.13860357e-01 4.32633638e-01
8.83410215e-01 -9.91051644e-02 -6.90710425e-01 5.88867068e-01
1.19285142e+00 2.57103503e-01 3.07892233e-01 1.31171003e-01
6.29483938e-01 1.55777559e-01 1.57509851e+00 4.67996061e-01
1.92353442e-01 9.69251692e-01 9.24313009e-01 1.68182805e-01
-5.68237193e-02 -2.85195172e-01 6.21653259e-01 5.93252540e-01
9.98431221e-02 -2.59806335e-01 -4.62770402e-01 3.43242615e-01
-2.00952411e+00 -8.46331894e-01 -3.42910618e-01 2.63165736e+00
4.33203429e-01 5.77007949e-01 2.39582971e-01 -2.84298390e-01
1.05253255e+00 3.67529243e-01 -7.32796550e-01 5.83273232e-01
2.64243055e-02 -4.20818180e-01 7.22020030e-01 4.36731011e-01
-1.45830417e+00 8.51514876e-01 4.28534746e+00 1.19749701e+00
-1.06864858e+00 1.63842380e-01 2.84525275e-01 -2.66109765e-01
1.95440322e-01 1.22376740e-01 -1.36906254e+00 1.00455570e+00
4.15971339e-01 -1.09337553e-01 1.16552785e-01 1.00950122e+00
3.56965005e-01 -1.87542677e-01 -7.85058618e-01 8.57324839e-01
-3.01458031e-01 -9.67342377e-01 -2.23489136e-01 -6.62852451e-02
2.00605482e-01 1.36868671e-01 2.24894494e-01 3.91575873e-01
1.11072727e-01 -4.01683152e-01 1.19511175e+00 6.38322830e-01
3.94464552e-01 -5.77507973e-01 6.88150346e-01 2.76173234e-01
-1.74038291e+00 2.14293133e-02 -5.69848835e-01 3.45726341e-01
7.55837262e-02 6.28995419e-01 -3.37191582e-01 8.23915303e-01
8.94357562e-01 9.78705823e-01 -7.29581058e-01 1.46474981e+00
-2.55065650e-01 5.48852503e-01 -5.65095067e-01 -1.11321732e-01
7.63242990e-02 -2.92076111e-01 9.96206939e-01 8.15165341e-01
2.62246698e-01 -2.97459215e-01 6.57036662e-01 8.10742915e-01
1.67033359e-01 -5.60451075e-02 -1.19076326e-01 2.22799182e-01
8.03528607e-01 1.49034035e+00 -7.47811377e-01 -2.54348844e-01
-2.95220256e-01 4.80327457e-01 2.88214922e-01 2.87332743e-01
-1.61524034e+00 -2.62081116e-01 7.94215858e-01 -8.28982610e-03
4.74656016e-01 -2.90811155e-02 -8.32569152e-02 -1.28800213e+00
3.08766097e-01 -6.32128060e-01 3.27919632e-01 -4.25956547e-01
-1.18262577e+00 4.38048512e-01 -2.85486057e-02 -1.80594862e+00
4.19259936e-01 -4.61753130e-01 -7.36460805e-01 6.93398297e-01
-1.44399512e+00 -1.00292611e+00 -5.05654156e-01 3.54877263e-01
3.39326143e-01 -8.71152803e-02 -1.68293893e-01 6.92420840e-01
-9.97561872e-01 7.63237298e-01 -1.87208831e-01 1.10846162e-01
8.67792845e-01 -9.48592961e-01 -1.67431161e-01 1.06503654e+00
2.18667388e-02 4.93921101e-01 7.28740573e-01 -8.53469908e-01
-1.19237196e+00 -1.26713705e+00 4.85890567e-01 -2.06162199e-01
7.46273160e-01 -2.15493456e-01 -7.35297263e-01 4.18758273e-01
-2.01542377e-01 4.23364222e-01 1.65062711e-01 -1.29470780e-01
-1.57175079e-01 -5.40472984e-01 -8.48116636e-01 5.04753828e-01
1.00845909e+00 4.44389321e-02 -2.83127278e-01 3.23415548e-01
5.09159803e-01 -4.39536184e-01 -7.37072110e-01 6.00721538e-01
6.43246531e-01 -9.09531713e-01 1.06102371e+00 -1.93652093e-01
-2.03454509e-01 -1.12202346e+00 7.62570873e-02 -8.04940820e-01
-2.24892139e-01 -1.95776120e-01 -3.67056787e-01 1.51496315e+00
2.55232573e-01 -7.35242903e-01 7.38454580e-01 -2.91371141e-02
-1.02901921e-01 -8.15773606e-01 -1.11316657e+00 -1.01902807e+00
-4.69322830e-01 -4.32167128e-02 3.30288947e-01 5.84361315e-01
-5.86776257e-01 1.49907678e-01 -6.14261091e-01 6.52211249e-01
1.11252999e+00 2.62769043e-01 8.94931674e-01 -1.42635715e+00
-2.53923863e-01 -4.18636024e-01 -4.83563185e-01 -1.35916591e+00
-3.12775299e-02 -6.99345052e-01 2.33957365e-01 -1.00859690e+00
1.15072884e-01 -8.25153768e-01 -5.05446911e-01 4.61628586e-01
-3.86114866e-01 1.06329061e-01 3.28941733e-01 4.84817803e-01
-1.16653919e+00 8.32985044e-01 1.04316509e+00 -2.33279198e-01
-2.42143217e-02 2.92284995e-01 -3.61267686e-01 6.66710556e-01
7.05147266e-01 -7.68952131e-01 -2.75006801e-01 -1.91157341e-01
-4.18027714e-02 -1.33054271e-01 6.46592736e-01 -1.35953784e+00
4.91717219e-01 -2.40363851e-01 4.86088425e-01 -1.01576543e+00
2.60099977e-01 -9.61629212e-01 1.92184001e-01 6.72609985e-01
-8.82003158e-02 -2.78484195e-01 1.70832351e-01 9.25728500e-01
-1.59783483e-01 -4.03961586e-03 1.09450209e+00 4.61537957e-01
-8.21252048e-01 5.28858244e-01 5.62371910e-02 2.45075166e-01
1.32553947e+00 -2.22915769e-01 -5.07869661e-01 -3.77229340e-02
-4.46189851e-01 8.85989428e-01 3.88466835e-01 6.12173200e-01
3.37365985e-01 -1.50328124e+00 -5.31399310e-01 1.32113069e-01
1.18312925e-01 -1.21936351e-01 1.05157092e-01 1.25020170e+00
-8.20359401e-03 3.45316887e-01 -4.51259278e-02 -9.78246689e-01
-1.40258312e+00 4.62704003e-01 4.46389824e-01 -3.75435293e-01
-3.26609224e-01 7.37549365e-01 4.57628697e-01 -2.11137518e-01
5.59716046e-01 -3.04277301e-01 -2.48763636e-01 7.69391060e-02
3.34438205e-01 4.04416919e-01 -4.04361874e-01 -7.76895702e-01
-8.51347148e-01 6.05275452e-01 5.55513650e-02 2.47060910e-01
6.52745008e-01 -3.67923677e-02 2.06553996e-01 2.15831771e-01
7.09127545e-01 1.07807763e-01 -1.64863300e+00 -1.37420297e-01
1.16425097e-01 -5.74678421e-01 -1.67088151e-01 -5.56402504e-01
-1.37357807e+00 6.24191642e-01 8.27021718e-01 4.35131565e-02
8.23063970e-01 -2.52643973e-02 5.19283891e-01 4.87711653e-02
4.53130305e-01 -9.98068273e-01 -1.03846630e-02 3.80844206e-01
4.55851376e-01 -1.32188618e+00 9.02387574e-02 -5.08967817e-01
-5.36188960e-01 6.64137542e-01 1.28363132e+00 -1.48712635e-01
4.39398348e-01 5.66029772e-02 -1.68885991e-01 1.94131788e-02
-6.31087959e-01 -2.72528529e-01 5.99473715e-01 3.59485179e-01
9.23299789e-02 7.53590316e-02 -4.72333372e-01 6.63630366e-01
2.18477070e-01 -1.97023392e-01 -7.87957478e-03 6.64739609e-01
-5.35771728e-01 -6.87788963e-01 -5.91856182e-01 4.36687469e-01
-2.32547104e-01 2.35981300e-01 -2.15179741e-01 7.59242117e-01
4.35276955e-01 8.71816576e-01 -2.35984903e-02 -4.71603334e-01
3.55988473e-01 -3.23763728e-01 2.03629360e-01 -2.60495007e-01
-3.93032253e-01 2.97528982e-01 -4.06702980e-02 -7.26225376e-01
-3.95096302e-01 -7.16512322e-01 -1.18423986e+00 -3.81986499e-02
-8.24804306e-01 3.35981786e-01 2.39154905e-01 6.45752013e-01
2.64428407e-01 5.70367336e-01 5.43037355e-01 -8.35038185e-01
-8.19276690e-01 -6.43591464e-01 -4.49989080e-01 -1.67400762e-02
3.45540136e-01 -1.26837242e+00 -4.92298454e-01 -3.63587618e-01] | [6.317736625671387, -2.134039878845215] |
881b0e0f-3d6c-4801-b7f1-16e62d4c2d95 | unsupervised-video-object-segmentation-with-2 | null | null | https://www.ecva.net/papers/eccv_2020/papers_ECCV/html/2189_ECCV_2020_paper.php | https://www.ecva.net/papers/eccv_2020/papers_ECCV/papers/123590477.pdf | Unsupervised Video Object Segmentation with Joint Hotspot Tracking | Object tracking is a well-studied problem in computer vision while identifying salient spots of objects in a video is a less explored direction in the literature. Video eye gaze estimation methods aim to tackle a related task but salient spots in those methods are not bounded by objects and tend to produce very scattered, unstable predictions due to the noisy ground truth data. We reformulate the problem of detecting and tracking of salient object spots as a new task called object hotspot tracking. In this paper, we propose to tackle this task jointly with unsupervised video object segmentation, in real-time, with a unified framework to exploit the synergy between the two. Specifically, we propose a Weighted Correlation Siamese Network (WCS-Net) which employs a Weighted Correlation Block (WCB) for encoding the pixel-wise correspondence between a template frame and the search frame. In addition, WCB takes the initial mask / hotspot as guidance to enhance the influence of salient regions for robust tracking. Our system can operate online during inference and jointly produce the object mask and hotspot track-lets at 33 FPS. Experimental results validate the effectiveness of our network design, and show the benefits of jointly solving the hotspot tracking and object segmentation problems. In particular, our method performs favorably against state-of-the-art video eye gaze models in object hotspot tracking, and outperforms existing methods on three benchmark datasets for unsupervised video object segmentation. | ['Radomír Měch', 'You He', 'Zhe Lin', 'Jianming Zhang', 'Huchuan Lu', 'Lu Zhang'] | null | null | null | null | eccv-2020-8 | ['unsupervised-video-object-segmentation'] | ['computer-vision'] | [ 3.70325238e-01 9.46971588e-03 -4.93899703e-01 -1.43371135e-01
-2.07839489e-01 -2.24321127e-01 1.56631276e-01 -3.35781902e-01
-5.16758263e-01 3.95846665e-01 -1.36021003e-01 6.21830299e-02
-1.26751676e-01 -6.53394908e-02 -9.03700173e-01 -8.32352400e-01
8.15738142e-02 1.77414417e-01 8.42288077e-01 1.78218111e-01
3.98629367e-01 2.62678385e-01 -1.78623617e+00 -1.76003009e-01
1.02301490e+00 1.22518682e+00 6.43328667e-01 5.03355086e-01
1.03716001e-01 8.43594849e-01 -3.07565659e-01 -9.83232334e-02
3.49252790e-01 -3.27441394e-01 -7.16284275e-01 3.75584930e-01
7.37932801e-01 -2.32401237e-01 -3.57191339e-02 1.36694098e+00
3.09169084e-01 3.07539284e-01 3.55352432e-01 -1.66270995e+00
-3.23876679e-01 3.31141412e-01 -1.32038188e+00 8.44885051e-01
3.15578170e-02 3.68686289e-01 9.91043210e-01 -9.05186713e-01
6.32352054e-01 1.06494439e+00 4.31507528e-01 5.70224166e-01
-1.06253636e+00 -7.54387498e-01 5.64369977e-01 4.10912305e-01
-1.41056740e+00 -5.08084536e-01 8.90299261e-01 -4.95274067e-01
4.67368007e-01 2.46318176e-01 7.65916288e-01 5.68406105e-01
1.26081286e-03 1.42660248e+00 7.78830349e-01 -1.42081410e-01
1.63676813e-01 4.41073552e-02 1.80980206e-01 7.87385702e-01
2.40850642e-01 1.01607427e-01 -8.11071038e-01 1.96228221e-01
6.86887681e-01 3.17547023e-01 -5.52323401e-01 -5.44427097e-01
-1.14458406e+00 5.84273100e-01 6.24908507e-01 2.27888569e-01
-4.35518295e-01 9.11698192e-02 -4.11420241e-02 -3.02157879e-01
6.02094591e-01 1.76878855e-01 -3.15947205e-01 2.91400049e-02
-1.44974816e+00 2.59489328e-01 3.88489813e-01 1.02202439e+00
7.49318957e-01 -1.16125695e-01 -4.97102916e-01 5.12159109e-01
6.13484800e-01 3.45097005e-01 2.49817491e-01 -9.61707771e-01
3.09150845e-01 7.36011803e-01 1.87809959e-01 -1.12382829e+00
-4.29683000e-01 -3.87082338e-01 -4.26229805e-01 1.38732359e-01
6.73198581e-01 -2.09632054e-01 -9.80449319e-01 1.60779297e+00
7.28204727e-01 7.16175616e-01 -4.22325730e-01 1.50126672e+00
8.10002565e-01 6.19739950e-01 1.82556346e-01 -4.43325728e-01
1.36867309e+00 -1.30570829e+00 -7.27392793e-01 -5.31667948e-01
1.76168516e-01 -5.86553931e-01 7.45500684e-01 7.15131462e-02
-1.28533959e+00 -5.66002548e-01 -6.48710430e-01 -2.44704727e-03
-3.73334326e-02 4.58782390e-02 4.47472632e-01 4.09543812e-01
-1.08940363e+00 1.35146901e-01 -7.43573368e-01 -2.11528957e-01
8.25011015e-01 5.37489772e-01 1.08618423e-01 1.62903488e-01
-7.53901005e-01 7.01790631e-01 5.60373664e-01 3.36518288e-01
-7.55371630e-01 -9.74809229e-01 -8.65183830e-01 2.15523407e-01
9.93636668e-01 -4.46407795e-01 1.14737689e+00 -1.31877935e+00
-1.26140285e+00 7.97624230e-01 -5.37608385e-01 -4.58731741e-01
3.74533713e-01 -2.17621446e-01 -1.88086331e-01 3.36213380e-01
1.81237713e-01 1.18485332e+00 1.22285259e+00 -1.18834651e+00
-1.15359485e+00 -3.37116867e-01 -1.45566761e-01 2.79676080e-01
-1.53311282e-01 5.08217096e-01 -1.05044281e+00 -5.05187452e-01
-8.97101779e-03 -9.13687289e-01 -1.58364281e-01 3.13240647e-01
-4.72575128e-01 -4.82182771e-01 1.19291520e+00 -4.33697164e-01
1.43799925e+00 -2.14721894e+00 2.55286247e-01 8.63440558e-02
6.34266555e-01 4.70082343e-01 4.72858846e-02 -5.42451143e-01
9.93812308e-02 -2.08868325e-01 -2.26575255e-01 -4.15808350e-01
-1.92292258e-01 -6.47037253e-02 -1.55243322e-01 7.38974690e-01
4.02322263e-01 1.11808157e+00 -1.02126598e+00 -1.02308166e+00
2.13517562e-01 3.04756820e-01 -4.33404744e-01 9.28125903e-02
-4.40853119e-01 3.54064345e-01 -5.26970625e-01 9.73744571e-01
6.22744262e-01 -7.11133122e-01 -2.87910461e-01 -3.62013914e-02
-3.56267035e-01 -3.38394433e-01 -1.14766872e+00 1.41621125e+00
2.91785926e-01 9.26957905e-01 1.19594604e-01 -6.77057624e-01
5.50641954e-01 -8.26694146e-02 6.80371106e-01 -5.01919329e-01
4.41530734e-01 -9.58153978e-02 1.63423002e-01 -6.38562500e-01
7.03973830e-01 3.21544111e-01 4.74184752e-01 4.87799942e-01
5.65187968e-02 3.82472456e-01 4.14399773e-01 5.07059693e-02
4.75471169e-01 2.19510451e-01 1.59928389e-02 -2.77429134e-01
5.12684703e-01 -6.27660677e-02 8.29751849e-01 6.33175671e-01
-7.52504706e-01 7.62897670e-01 4.12070841e-01 -1.29807800e-01
-7.08487451e-01 -5.50348759e-01 -3.03607993e-02 1.40228415e+00
8.68605077e-01 -7.67809823e-02 -1.04406106e+00 -7.05407023e-01
-4.66061458e-02 3.25303853e-01 -7.26352453e-01 7.01381341e-02
-4.94835556e-01 -6.27104759e-01 7.46202320e-02 4.77188110e-01
4.27558482e-01 -1.35190487e+00 -1.07521117e+00 -1.77589148e-01
-1.88749745e-01 -1.16282523e+00 -1.02165902e+00 -1.70221105e-01
-6.20391607e-01 -1.26863420e+00 -1.13319802e+00 -9.06821251e-01
8.35355759e-01 7.63787091e-01 9.52120066e-01 3.16389591e-01
-2.93777049e-01 2.23446637e-01 -3.13969180e-02 -4.45059568e-01
2.31508657e-01 4.92190756e-02 -4.41646241e-02 5.74981272e-01
5.45925975e-01 -3.70083898e-02 -8.72306406e-01 7.06270158e-01
-9.10914898e-01 1.51999548e-01 5.11576474e-01 5.92287719e-01
6.24913275e-01 -1.07678510e-01 1.42946318e-01 -6.41733885e-01
2.11211860e-01 -5.81731021e-01 -1.01880753e+00 3.12170535e-01
-4.82850134e-01 -2.61317939e-01 -6.64552907e-03 -5.63264787e-01
-9.62750912e-01 2.40678042e-01 6.01752996e-01 -8.68818581e-01
-5.86248040e-02 2.58785337e-01 -7.02819694e-03 -4.09450740e-01
2.37987414e-01 2.29129285e-01 1.92897469e-01 -1.46463051e-01
1.78139374e-01 4.60919827e-01 3.88521492e-01 -1.20118126e-01
7.14185894e-01 6.97859347e-01 -1.84746757e-01 -6.87388778e-01
-1.07897735e+00 -9.75197792e-01 -5.97357690e-01 -7.91007936e-01
1.01541054e+00 -8.23519349e-01 -1.00773573e+00 3.52421880e-01
-1.04805958e+00 -2.64142275e-01 -8.85623619e-02 3.83528113e-01
-4.92813230e-01 1.93992659e-01 -8.48656669e-02 -9.77401435e-01
-2.83426315e-01 -1.21202862e+00 1.34473193e+00 9.44080651e-01
-3.73586006e-02 -8.47460449e-01 -2.32643187e-01 3.99195284e-01
1.82637393e-01 4.87112477e-02 1.64529979e-01 -3.60078394e-01
-1.12486899e+00 1.03246234e-01 -6.68579936e-01 -3.01066916e-02
-8.84780213e-02 1.80755079e-01 -8.97988141e-01 -1.82464287e-01
-2.07308024e-01 3.41831222e-02 1.01045442e+00 9.18383539e-01
1.09427106e+00 -4.37277593e-02 -7.27503181e-01 7.68369377e-01
1.10202765e+00 8.63451809e-02 3.23883384e-01 2.69830257e-01
8.45999360e-01 7.44900703e-01 9.47756290e-01 2.83681184e-01
2.68065810e-01 7.82725990e-01 5.82826853e-01 -2.73546070e-01
-5.98963611e-02 8.45964104e-02 1.60633758e-01 2.31597558e-01
-1.66477099e-01 -2.11103380e-01 -8.89090776e-01 8.31564009e-01
-2.17321467e+00 -1.04308689e+00 -1.30837783e-01 1.97983301e+00
8.17024648e-01 9.46304798e-02 5.13830423e-01 -2.71307468e-01
1.07003248e+00 3.25848341e-01 -8.01681817e-01 4.19562876e-01
-1.05447643e-01 -2.84249604e-01 6.25226259e-01 2.78388381e-01
-1.19156682e+00 1.07630324e+00 5.59502411e+00 6.53490603e-01
-1.15154791e+00 1.79320559e-01 6.74910665e-01 -4.28667575e-01
2.90699691e-01 -1.52424246e-01 -1.12396133e+00 9.23426867e-01
3.98575425e-01 9.92137939e-03 2.67037928e-01 7.78143942e-01
3.79948080e-01 -4.99417633e-01 -9.00513947e-01 1.07636225e+00
3.90137136e-01 -1.30676115e+00 -4.64869529e-01 -8.73472318e-02
9.69516575e-01 1.42185062e-01 4.49333340e-01 -1.95960626e-01
-1.09730132e-01 -7.01402664e-01 9.62962627e-01 7.49105990e-01
3.29807341e-01 -4.19224650e-01 4.52503204e-01 3.01806271e-01
-1.20819318e+00 -2.83940166e-01 -1.76139697e-01 2.16045737e-01
2.95358509e-01 1.97150394e-01 -6.48630083e-01 9.84004512e-03
1.10686457e+00 1.00197899e+00 -6.79635286e-01 1.73775482e+00
-1.87159926e-02 5.40722251e-01 -3.90664011e-01 -2.64924854e-01
4.10234898e-01 -1.44636750e-01 8.31304014e-01 8.67070138e-01
-1.20001577e-01 -6.07304159e-04 1.48169100e-01 1.24724591e+00
1.09511800e-01 -2.16586992e-01 -7.55002722e-02 1.67700350e-01
2.11010769e-01 1.26329124e+00 -1.10919726e+00 -2.59887427e-01
-3.90209347e-01 5.72717667e-01 1.68640167e-01 6.13967359e-01
-1.06142330e+00 -1.44043341e-01 5.97941995e-01 2.23576903e-01
7.81228065e-01 8.96977335e-02 -2.29818046e-01 -9.84605253e-01
1.43205926e-01 -4.76391226e-01 2.39637613e-01 -9.65902507e-01
-8.93992126e-01 3.26379091e-01 5.21772318e-02 -1.31924868e+00
4.74029370e-02 -4.63461816e-01 -7.55053997e-01 7.28384972e-01
-1.90693963e+00 -9.31160271e-01 -5.55460989e-01 7.36644149e-01
6.95055366e-01 -1.30078122e-02 -2.54540503e-01 2.74521202e-01
-9.68629420e-01 4.40869510e-01 -1.10930137e-01 1.04169793e-01
4.38233703e-01 -9.99763370e-01 1.68204576e-01 1.17386639e+00
1.97125643e-01 5.12867630e-01 6.92485809e-01 -7.11551726e-01
-1.12279844e+00 -1.23949289e+00 6.23678863e-01 -4.38430130e-01
5.66595972e-01 -2.02182144e-01 -1.02668822e+00 4.86925542e-01
2.43306115e-01 3.94387811e-01 2.22409204e-01 -1.59338266e-01
1.41513258e-01 -2.52780486e-02 -7.50797749e-01 6.12637043e-01
1.01664305e+00 -1.74401075e-01 -4.00221229e-01 2.99859583e-01
7.47690380e-01 -6.84488654e-01 -2.45828494e-01 2.56759167e-01
3.09071273e-01 -8.96769702e-01 9.19742405e-01 -4.89191949e-01
2.07798511e-01 -7.54294932e-01 5.19597709e-01 -7.33748972e-01
-1.97677448e-01 -9.44065571e-01 -5.12546420e-01 1.01931655e+00
2.99638748e-01 -2.83643961e-01 1.04352558e+00 7.65577972e-01
2.32384782e-02 -7.85327017e-01 -7.77009130e-01 -4.14425164e-01
-7.94363558e-01 -2.68092334e-01 3.94492358e-01 6.48756444e-01
-3.25445086e-01 9.27663296e-02 -3.24062765e-01 2.48629645e-01
8.26155424e-01 2.60753572e-01 7.28852153e-01 -1.32088459e+00
3.17858644e-02 -8.27526927e-01 -3.91151369e-01 -1.29712284e+00
1.70694336e-01 -4.73260254e-01 4.86869961e-01 -1.08633161e+00
3.74148488e-01 -3.68129194e-01 -4.39628333e-01 3.06323797e-01
-7.01608777e-01 4.86556321e-01 4.58831668e-01 3.58010501e-01
-1.27926910e+00 4.12833214e-01 1.33768535e+00 -2.04722047e-01
-4.83501047e-01 2.81267613e-01 -6.14678323e-01 8.59612465e-01
3.73383939e-01 -4.91453677e-01 -3.16316247e-01 -3.13834071e-01
7.71183446e-02 -1.78580761e-01 6.09108508e-01 -8.82053435e-01
8.44965041e-01 -2.76088834e-01 3.30164224e-01 -8.15029442e-01
2.65563548e-01 -9.34955418e-01 -2.57755816e-01 1.21074095e-01
-1.56605840e-01 -2.53905416e-01 1.51302621e-01 8.45125556e-01
-1.72785178e-01 -2.39652127e-01 8.02501619e-01 2.72473127e-01
-1.07329226e+00 6.16538286e-01 6.15680106e-02 2.62628347e-01
1.36555231e+00 -7.08008647e-01 -2.21810356e-01 -4.71727690e-03
-4.99012798e-01 7.04951406e-01 4.01921332e-01 7.34723687e-01
6.40009105e-01 -8.58174860e-01 -3.73357773e-01 3.00017029e-01
1.93442907e-02 2.43439227e-01 3.62707019e-01 1.40109682e+00
-2.88502932e-01 4.12274480e-01 -1.07784599e-01 -1.15865552e+00
-1.44244003e+00 7.20031321e-01 3.63917023e-01 2.71627367e-01
-4.82349724e-01 1.13975012e+00 5.30137539e-01 2.82632530e-01
6.40763283e-01 -4.95406032e-01 -5.81285179e-01 1.75771371e-01
6.98069811e-01 3.42818290e-01 -3.41541231e-01 -1.01739526e+00
-2.96817392e-01 7.56371915e-01 -2.10172996e-01 3.99426460e-01
1.19552994e+00 -4.13763702e-01 -1.61666706e-01 2.88790643e-01
8.54376674e-01 -4.60145772e-01 -1.89575565e+00 -6.40293956e-01
1.64156035e-01 -6.60788536e-01 3.27203035e-01 -4.64707285e-01
-1.58926952e+00 5.19780457e-01 5.90244591e-01 2.16529682e-01
1.17151606e+00 2.75294036e-01 7.57360756e-01 -8.44893083e-02
-6.62296712e-02 -1.16816521e+00 4.86162938e-02 2.59683430e-01
4.61949944e-01 -1.56767893e+00 5.07029938e-03 -5.08539677e-01
-8.39262545e-01 7.29695201e-01 1.00120044e+00 -5.14216013e-02
6.50686204e-01 -9.89595354e-02 -5.88951372e-02 -3.92566353e-01
-5.65331876e-01 -6.16890788e-01 8.92403424e-01 3.02671611e-01
-1.00775443e-01 -3.16447526e-01 6.86324984e-02 3.32975894e-01
3.09145004e-01 1.36258736e-01 1.43427595e-01 8.49727809e-01
-6.89092994e-01 -3.36905479e-01 -5.41381776e-01 4.77764606e-01
-5.32126546e-01 -1.14001326e-01 -1.21307671e-01 6.75306022e-01
1.70982897e-01 8.04362535e-01 4.00372624e-01 -9.02929623e-03
-5.29027842e-02 -2.27469102e-01 1.48994103e-01 -3.59737933e-01
-5.36246955e-01 2.63502270e-01 -6.13922775e-01 -8.00667584e-01
-9.57768381e-01 -8.14532161e-01 -1.09955513e+00 6.16543181e-02
-7.56742656e-01 1.67020913e-02 3.92100424e-01 1.02854943e+00
4.58711982e-01 5.66544235e-01 4.31676447e-01 -1.21523285e+00
2.45086346e-02 -7.55669177e-01 -6.04262233e-01 2.35052139e-01
6.93314493e-01 -1.07034647e+00 -1.62482277e-01 2.70181119e-01] | [9.298672676086426, -0.22327828407287598] |
ffde47c2-ffd0-4b7a-ae9b-1483912bc78f | split-kalmannet-a-robust-model-based-deep | 2210.09636 | null | https://arxiv.org/abs/2210.09636v1 | https://arxiv.org/pdf/2210.09636v1.pdf | Split-KalmanNet: A Robust Model-Based Deep Learning Approach for SLAM | Simultaneous localization and mapping (SLAM) is a method that constructs a map of an unknown environment and localizes the position of a moving agent on the map simultaneously. Extended Kalman filter (EKF) has been widely adopted as a low complexity solution for online SLAM, which relies on a motion and measurement model of the moving agent. In practice, however, acquiring precise information about these models is very challenging, and the model mismatch effect causes severe performance loss in SLAM. In this paper, inspired by the recently proposed KalmanNet, we present a robust EKF algorithm using the power of deep learning for online SLAM, referred to as Split-KalmanNet. The key idea of Split-KalmanNet is to compute the Kalman gain using the Jacobian matrix of a measurement function and two recurrent neural networks (RNNs). The two RNNs independently learn the covariance matrices for a prior state estimate and the innovation from data. The proposed split structure in the computation of the Kalman gain allows to compensate for state and measurement model mismatch effects independently. Numerical simulation results verify that Split-KalmanNet outperforms the traditional EKF and the state-of-the-art KalmanNet algorithm in various model mismatch scenarios. | ['Namyoon Lee', 'Yonina C. Eldar', 'Nir Shlezinger', 'Jeonghun Park', 'Geon Choi'] | 2022-10-18 | null | null | null | null | ['simultaneous-localization-and-mapping'] | ['computer-vision'] | [-2.78940946e-01 -4.26092029e-01 5.77516295e-02 -1.53765216e-01
-4.24870133e-01 -2.85940766e-01 6.94171131e-01 -1.49894714e-01
-7.68323302e-01 7.47901618e-01 -1.92609485e-02 -2.08374068e-01
-3.70877922e-01 -5.09522080e-01 -8.77203584e-01 -7.49499977e-01
-3.60296339e-01 4.11911070e-01 7.69782588e-02 -2.38165453e-01
2.81721830e-01 5.24812877e-01 -1.18218207e+00 -8.62968385e-01
9.36101735e-01 8.15485299e-01 8.51265371e-01 5.72543800e-01
7.52581805e-02 7.07951069e-01 -4.06118989e-01 3.36504161e-01
4.19134647e-01 -1.89137802e-01 -2.24427104e-01 -3.41935098e-01
3.40514094e-01 -4.58029002e-01 -9.27904248e-01 1.29663563e+00
5.36401570e-01 4.90581810e-01 1.25197947e-01 -1.16420674e+00
3.67175266e-02 3.88797760e-01 -1.22249862e-02 6.75640851e-02
2.53760397e-01 5.44355018e-03 4.80763704e-01 -8.68686497e-01
4.86922890e-01 1.11664927e+00 1.08835065e+00 7.60103464e-02
-8.21761370e-01 -8.60366404e-01 6.60234839e-02 2.25175634e-01
-1.79686892e+00 -3.74032378e-01 4.35195833e-01 -4.48712051e-01
9.23242748e-01 -3.32459331e-01 8.01561475e-01 7.18692362e-01
1.01225781e+00 4.14542824e-01 9.57236767e-01 -6.11849204e-02
2.60642707e-01 -1.97280988e-01 -7.49768168e-02 8.31111610e-01
3.20253223e-01 7.71375060e-01 -7.44156599e-01 -5.11333235e-02
1.02461398e+00 3.33340108e-01 -3.84740353e-01 -8.00380111e-01
-1.43710530e+00 7.55756617e-01 8.80779684e-01 -7.24147353e-03
-6.57577634e-01 6.70938015e-01 1.68156996e-01 4.67274845e-01
1.18526667e-01 5.45273304e-01 -2.00610310e-01 -3.18612397e-01
-9.61973965e-01 1.74031436e-01 8.37710619e-01 8.27592850e-01
1.23485589e+00 4.53941792e-01 5.21146834e-01 1.42949402e-01
6.59850955e-01 1.20003450e+00 6.51007175e-01 -8.00971389e-01
3.46725911e-01 3.09508413e-01 6.59021020e-01 -1.06295657e+00
-9.81896400e-01 -8.15091789e-01 -1.06647623e+00 1.75940916e-01
1.39470473e-01 -5.25781691e-01 -9.25923407e-01 1.76828754e+00
3.12635124e-01 9.68738914e-01 2.99706459e-01 9.29581463e-01
3.05006117e-01 8.44749868e-01 -5.23672223e-01 -8.62162113e-02
6.85479045e-01 -9.61739540e-01 -1.20245838e+00 -8.88864100e-01
5.54814577e-01 -6.52521610e-01 4.75690290e-02 7.03072175e-02
-4.01394367e-01 -7.11229086e-01 -1.39870083e+00 3.08324546e-01
-4.91450429e-01 2.80222803e-01 5.36288917e-01 1.98374972e-01
-1.19101512e+00 8.30006242e-01 -1.25267041e+00 -4.84636664e-01
-4.96727377e-01 6.06092691e-01 -5.30808091e-01 2.31827945e-01
-1.42573464e+00 1.63553047e+00 7.09175408e-01 7.77102888e-01
-1.13412750e+00 -2.86901742e-01 -1.06613231e+00 -2.12876976e-01
3.78500134e-01 -7.55403817e-01 1.15685344e+00 -3.72331440e-01
-2.06276393e+00 -3.37420292e-02 -3.66449833e-01 -1.03597736e+00
2.80808568e-01 -8.32303941e-01 -3.99763256e-01 -4.24691975e-01
1.26360282e-01 3.19404185e-01 9.12385464e-01 -8.36977303e-01
-7.79269755e-01 -1.94096163e-01 -2.44970992e-01 3.77214640e-01
6.37457550e-01 -6.62967026e-01 -1.44049525e-01 -7.65461773e-02
9.50650752e-01 -1.27556264e+00 -5.62381446e-01 -4.65478748e-01
8.10332820e-02 2.78928488e-01 6.16730332e-01 -7.12364674e-01
1.05895483e+00 -2.18504715e+00 3.51281613e-02 3.55257124e-01
6.68042451e-02 9.30638239e-03 -1.39345437e-01 7.49330699e-01
3.33956808e-01 -6.51692390e-01 2.89042920e-01 -4.00235236e-01
2.49178149e-02 3.85990739e-01 -4.89492804e-01 1.07816160e+00
-4.48911101e-01 8.92721534e-01 -1.21150577e+00 2.26509050e-01
5.72188616e-01 5.48928559e-01 1.30748947e-03 2.40314141e-01
1.83790445e-01 6.43379450e-01 -2.67452300e-01 3.26464698e-03
6.12751782e-01 -1.58168808e-01 1.13048494e-01 -2.17579618e-01
-6.40881360e-01 5.77340066e-01 -1.65856004e+00 2.33089519e+00
-6.63657904e-01 7.15752125e-01 3.69879037e-01 -5.93106091e-01
1.16496420e+00 2.47767359e-01 2.51474321e-01 -6.24470532e-01
3.11855465e-01 5.76344788e-01 -1.41274417e-02 -1.08755752e-02
8.73223126e-01 -2.25064456e-02 -1.60167664e-01 8.26879442e-02
2.33733222e-01 -2.52570480e-01 -1.41379938e-01 1.10673055e-01
1.07112086e+00 2.80251443e-01 5.32201290e-01 -2.55256027e-01
5.87891757e-01 -2.65974313e-01 9.21511173e-01 1.20484304e+00
-1.08291976e-01 -1.47191584e-01 -5.02455115e-01 -6.17869318e-01
-7.26734996e-01 -1.10010064e+00 2.44726717e-01 5.33875704e-01
7.64430821e-01 -4.04861927e-01 4.05087210e-02 -9.48377401e-02
2.48375729e-01 3.44649196e-01 -3.01665008e-01 -3.06357890e-01
-6.09788656e-01 -2.96256244e-01 2.42537156e-01 2.75643349e-01
8.88211966e-01 -5.17986715e-01 -7.77140081e-01 6.36327624e-01
-9.31708589e-02 -9.63461936e-01 -2.14712694e-01 3.41321260e-01
-7.87099063e-01 -8.45458269e-01 -2.40202248e-01 -6.12070858e-01
4.97850984e-01 5.49055159e-01 4.04567868e-01 -3.17374587e-01
4.98179138e-01 2.27607638e-01 -1.65130243e-01 -2.93427467e-01
-2.81877041e-01 -9.30873584e-03 6.77117825e-01 1.12690944e-02
-1.96018696e-01 -5.69600105e-01 -3.04810792e-01 4.26586300e-01
-3.26183736e-01 2.35747918e-01 8.27862978e-01 1.01433456e+00
3.69583577e-01 6.10544309e-02 2.50186622e-01 -3.08146328e-01
4.23008204e-01 -1.72335640e-01 -1.39180815e+00 -5.88608533e-02
-5.52172899e-01 2.69519210e-01 4.17621970e-01 -2.49928400e-01
-9.32024598e-01 4.44023758e-01 -1.17868289e-01 -3.57940286e-01
2.64290452e-01 9.00000036e-01 1.28203630e-01 -5.72313964e-01
2.57550150e-01 5.68075716e-01 2.73646802e-01 -4.44494873e-01
4.10109133e-01 3.17289710e-01 1.00619674e+00 1.00376487e-01
1.12986112e+00 6.17524564e-01 3.34651351e-01 -6.25877678e-01
-6.33156836e-01 -8.37368250e-01 -7.18305349e-01 -2.17452109e-01
5.18815756e-01 -1.41719425e+00 -8.27042937e-01 7.50735939e-01
-1.27414930e+00 -3.59677702e-01 2.72831600e-02 1.24522281e+00
-5.83490431e-01 2.70652652e-01 -4.59093451e-01 -8.58604610e-01
-2.56794035e-01 -1.09918499e+00 7.81853676e-01 3.27283919e-01
9.73238125e-02 -1.00157773e+00 4.51728940e-01 -5.81068277e-01
6.27506375e-01 3.78038250e-02 -1.38917625e-01 -3.51511568e-01
-9.17626441e-01 -6.16796672e-01 5.94208436e-03 1.12690084e-01
2.58568134e-02 -6.64547026e-01 -4.67656165e-01 -8.81549299e-01
2.28112966e-01 1.82935879e-01 7.68897414e-01 4.17098522e-01
-1.64588571e-01 -1.27442122e-01 -6.03171170e-01 9.90357101e-01
1.70015430e+00 3.98858994e-01 3.04965228e-01 5.48784673e-01
7.10721254e-01 -1.66081801e-01 7.26286948e-01 4.22788531e-01
4.23148960e-01 4.77954328e-01 6.09761059e-01 3.39805156e-01
3.57089162e-01 -5.76894820e-01 6.15038872e-01 1.18876934e+00
2.55512834e-01 3.45000736e-02 -7.70313740e-01 3.47218990e-01
-2.26824665e+00 -5.89099824e-01 7.53309131e-02 2.37538266e+00
8.46569240e-02 6.93145394e-02 -7.26746440e-01 -3.21310997e-01
6.19973302e-01 4.08626169e-01 -5.40071726e-01 1.49352685e-01
3.70651819e-02 -2.75569856e-01 1.22559106e+00 1.04432178e+00
-1.23579168e+00 1.02580285e+00 5.75132656e+00 4.28227693e-01
-1.33615792e+00 -6.24925364e-03 -8.56138885e-01 3.96543831e-01
4.14133489e-01 5.64983904e-01 -1.09323049e+00 4.46965188e-01
1.28544211e+00 -2.63914704e-01 6.02630675e-01 8.28371048e-01
3.16007942e-01 -3.70118260e-01 -9.09712434e-01 1.25536859e+00
-5.15756682e-02 -1.41094363e+00 -4.86368418e-01 9.32093188e-02
7.81633496e-01 8.23800445e-01 -2.83157110e-01 4.60484296e-01
5.56794405e-01 -4.24151838e-01 7.55642235e-01 8.89192343e-01
2.25437641e-01 -5.07644653e-01 1.31938076e+00 7.52395153e-01
-1.43548942e+00 -2.35670477e-01 -4.49080348e-01 -6.64976060e-01
7.09892273e-01 5.57202756e-01 -1.12911916e+00 7.80651987e-01
3.39234352e-01 1.06689370e+00 -2.38893241e-01 1.54312623e+00
-5.76111734e-01 1.65945664e-01 -7.30867922e-01 -1.08022176e-01
5.75905919e-01 -3.51385623e-01 7.87262321e-01 6.53310180e-01
6.24119759e-01 -2.34332874e-01 6.19321883e-01 5.20129740e-01
3.28068942e-01 -4.01173234e-01 -6.46235704e-01 2.75103748e-01
7.23767757e-01 9.36728239e-01 -2.90871203e-01 -4.33173746e-01
-1.78571850e-01 7.90876389e-01 2.26194113e-01 3.45948368e-01
-4.83427614e-01 -4.15867060e-01 6.78775251e-01 -6.40684366e-01
3.04085106e-01 -9.67457592e-01 4.17564005e-01 -1.15493166e+00
-2.33346120e-01 -3.10045630e-01 -1.27147466e-01 -7.76201129e-01
-8.57592881e-01 4.71766889e-01 -2.58231342e-01 -1.54057169e+00
-6.32335424e-01 -3.39931518e-01 -3.17452997e-01 1.09470475e+00
-1.50249541e+00 -8.80787432e-01 -4.19534475e-01 2.48717070e-01
1.59331679e-01 -1.12381361e-01 6.56491995e-01 1.07371546e-01
-2.95073211e-01 -1.84772432e-01 8.14566553e-01 7.82822073e-02
7.56260872e-01 -1.11944818e+00 8.20597053e-01 1.14838076e+00
1.29446805e-01 9.80506003e-01 1.09664643e+00 -1.03762996e+00
-1.81641340e+00 -1.01530528e+00 9.49529052e-01 -1.59024015e-01
9.14809525e-01 -3.77193332e-01 -5.86875200e-01 1.06078172e+00
-3.60743180e-02 -8.21917653e-02 -2.16162845e-01 -8.82269591e-02
1.51302099e-01 -3.22416186e-01 -4.73431677e-01 1.40465215e-01
6.32946193e-01 -7.44562387e-01 -6.24083221e-01 1.00343890e-01
7.05036521e-01 -1.01715732e+00 -6.21118844e-01 5.20272255e-01
7.20440209e-01 -8.77308011e-01 7.16078043e-01 2.25670207e-02
-8.37058365e-01 -9.22500372e-01 -1.20705120e-01 -1.77934170e+00
-5.29535949e-01 -7.80290127e-01 -1.85564622e-01 4.95124757e-01
-2.03201342e-02 -1.25419402e+00 7.34442711e-01 -5.05544730e-02
-3.22646022e-01 -2.09859625e-01 -1.28779030e+00 -1.17297220e+00
-5.53791761e-01 -3.80846590e-01 6.26676440e-01 5.19132733e-01
-3.69891912e-01 2.85296708e-01 -9.02002275e-01 1.00741410e+00
6.55189812e-01 -5.26208803e-02 1.20934248e+00 -1.24943948e+00
-3.93984020e-02 3.18852179e-02 -7.35302031e-01 -1.74686301e+00
2.41253525e-01 -4.01852876e-01 6.08719945e-01 -1.68796110e+00
-5.56761205e-01 -2.15821043e-01 -2.71829724e-01 -5.63928522e-02
1.65224716e-01 -3.34371209e-01 3.46948057e-01 3.60415876e-01
-7.07120180e-01 8.03000689e-01 8.31247270e-01 6.55172206e-03
-3.28033775e-01 2.67446965e-01 2.49597326e-01 9.15655851e-01
4.70189899e-01 -4.16835874e-01 -2.84572959e-01 -6.99917495e-01
4.07637924e-01 4.20170307e-01 4.09443438e-01 -1.69922054e+00
1.03798175e+00 1.42424569e-01 1.69880345e-01 -1.14091766e+00
6.83979690e-01 -9.98573244e-01 5.57290196e-01 8.78982663e-01
1.88768655e-01 3.28483909e-01 8.64812657e-02 9.46405470e-01
-3.97158295e-01 -3.02730426e-02 4.02860045e-01 -6.52708858e-02
-1.47343802e+00 3.24824423e-01 -4.12092298e-01 -5.21257401e-01
3.73001724e-01 2.61080675e-02 -1.95856988e-01 -7.24946022e-01
-6.23094678e-01 5.31999350e-01 3.28619093e-01 3.80091459e-01
7.51733184e-01 -1.31618357e+00 -3.59391004e-01 5.28462648e-01
-2.37774074e-01 1.76123604e-02 4.93644834e-01 1.00581288e+00
-5.94349384e-01 7.34669745e-01 -1.22722030e-01 -6.17588699e-01
-5.90691924e-01 3.53904545e-01 6.83730781e-01 -2.01941162e-01
-5.01238883e-01 6.35243356e-01 -5.93643785e-02 -8.26002121e-01
2.88095623e-01 -4.24078494e-01 2.52796531e-01 -3.10867369e-01
5.89666307e-01 4.60535496e-01 -6.08201958e-02 -8.15419376e-01
-5.00453651e-01 6.88673019e-01 2.38609254e-01 -2.97787726e-01
1.06284308e+00 -7.38037109e-01 -2.00109825e-01 7.40845978e-01
1.15088248e+00 -1.84075341e-01 -1.39846528e+00 -6.09058917e-01
2.00376026e-02 -3.67705673e-01 2.87778616e-01 -4.14561749e-01
-5.17629147e-01 6.11131608e-01 8.30346644e-01 -2.89594799e-01
6.22211576e-01 -5.93588054e-01 7.48349786e-01 9.06440079e-01
1.11544847e+00 -9.94720459e-01 -4.04979587e-01 1.31094635e+00
5.21401644e-01 -1.15463257e+00 1.98764786e-01 3.57334204e-02
-2.19416711e-02 9.96815383e-01 4.72740412e-01 -6.02716386e-01
7.89858878e-01 2.71454435e-02 3.96546394e-01 9.07964930e-02
-5.58197498e-01 -1.65682361e-01 1.00437947e-01 3.02269995e-01
-2.62201190e-01 3.12171113e-02 4.76099364e-02 5.50332107e-02
-3.95268470e-01 -8.80335420e-02 4.50603724e-01 1.10634065e+00
-7.68420577e-01 -6.28270149e-01 -4.83436823e-01 1.36169881e-01
2.58708268e-01 9.21942368e-02 2.57604152e-01 8.27856660e-01
-9.97255649e-03 6.48329020e-01 7.54841864e-02 -6.41024709e-01
1.59021169e-01 -1.23839386e-01 2.87917346e-01 -1.92296714e-01
-1.09943546e-01 5.75499348e-02 -1.41594201e-01 -9.12135839e-01
-2.84624189e-01 -4.60136473e-01 -1.37422431e+00 -1.61861762e-01
-7.73214161e-01 5.15238404e-01 1.15090013e+00 1.17787099e+00
4.53219891e-01 3.66894424e-01 4.90043253e-01 -1.22653806e+00
-8.80420327e-01 -1.05780315e+00 -9.24543977e-01 -1.95545003e-01
9.01389837e-01 -9.71400857e-01 -4.39491600e-01 -6.44903362e-01] | [7.49202823638916, -2.0712409019470215] |
4625e940-ac91-404e-b392-4bda34324174 | recurrent-convolutional-neural-networks-for-2 | null | null | https://www.aaai.org/ocs/index.php/AAAI/AAAI15/paper/download/9745/9552 | https://www.aaai.org/ocs/index.php/AAAI/AAAI15/paper/download/9745/9552 | Recurrent Convolutional Neural Networks for Text Classification | Text classification is a foundational task in many NLP applications. Traditional text classifiers often rely on many human-designed features, such as dictionaries, knowledge bases and special tree kernels. In contrast to traditional methods, we introduce a recurrent convolutional neural network for text classification without human-designed features. In our model, we apply a recurrent structure to capture contextual information as far as possible when learning word representations, which may introduce considerably less noise compared to traditional window-based neural networks. We also
employ a max-pooling layer that automatically judges which words play key roles in text classification to capture the key components in texts. We conduct experiments on four commonly used datasets. The experimental results show that the proposed method outperforms the state-of-the-art methods on several datasets, particularly on document-level datasets. | ['Jun Zhao', 'Kang Liu', 'Liheng Xu', 'Siwei Lai'] | 2015-01-01 | null | null | null | proceedings-of-the-twenty-ninth-aaai | ['emotion-recognition-in-conversation'] | ['natural-language-processing'] | [ 1.58832088e-01 -4.48624879e-01 -5.35597742e-01 -3.97765547e-01
-4.50530499e-01 -2.23423302e-01 8.33230615e-01 6.71356022e-01
-8.17023933e-01 5.11221230e-01 3.51019859e-01 -5.98295629e-01
9.33514759e-02 -9.66588676e-01 -8.82000253e-02 -5.21264017e-01
1.69603720e-01 1.42794494e-02 1.74093917e-01 -3.22131157e-01
7.24038303e-01 1.24372303e-01 -1.39646304e+00 5.41368902e-01
7.82644987e-01 1.15153217e+00 1.60743162e-01 6.20876372e-01
-8.69560838e-01 1.00616741e+00 -7.33398199e-01 -3.93882692e-01
-1.33775443e-01 -2.60005891e-01 -8.89726341e-01 -7.72967413e-02
2.10503135e-02 -3.77523862e-02 -4.86508340e-01 9.22624648e-01
4.52222109e-01 4.82568860e-01 5.55275142e-01 -6.38557196e-01
-9.33539629e-01 1.08065689e+00 -5.02123296e-01 4.24946636e-01
4.59717274e-01 -4.34005350e-01 1.36934566e+00 -1.15124190e+00
3.40468377e-01 1.13490975e+00 8.74051869e-01 1.29403219e-01
-1.01218510e+00 -6.34130716e-01 5.72937548e-01 3.10725212e-01
-1.20360029e+00 -3.34754944e-01 8.32288861e-01 -2.41027832e-01
1.13809073e+00 1.45402804e-01 3.48649114e-01 1.22533536e+00
5.65270901e-01 1.11882901e+00 7.28262603e-01 -8.07215095e-01
7.91512206e-02 7.64926076e-02 9.07980442e-01 6.18191600e-01
2.05755308e-01 -5.27156234e-01 -5.40157557e-01 -3.37543339e-01
3.58140737e-01 5.68961799e-01 -3.00242990e-01 1.96567014e-01
-1.10248113e+00 1.06075358e+00 2.93796182e-01 6.28084600e-01
-4.10250425e-01 -9.21781063e-02 9.03058171e-01 3.04771990e-01
7.74070024e-01 2.27520734e-01 -6.09162390e-01 -9.79909822e-02
-9.62768674e-01 7.36000016e-02 7.84453392e-01 7.79731929e-01
5.52274525e-01 -6.74564242e-02 -4.99186993e-01 1.11326098e+00
2.05739304e-01 -7.17522353e-02 1.24390888e+00 2.03010693e-01
7.69275546e-01 9.89093602e-01 -1.65307790e-01 -1.05818653e+00
-5.27948678e-01 -4.44716096e-01 -1.08369291e+00 -3.07036281e-01
4.04585376e-02 -2.14937389e-01 -1.10591459e+00 1.08429742e+00
3.90713103e-02 -3.11957672e-02 1.59047186e-01 4.38652188e-01
1.20022202e+00 9.54159021e-01 1.53365567e-01 -2.50315636e-01
1.59805167e+00 -9.16115999e-01 -9.85459149e-01 -2.12527543e-01
8.80361080e-01 -9.79473114e-01 1.20883465e+00 3.51818353e-01
-4.83197123e-01 -4.51717257e-01 -9.00359154e-01 -2.02213705e-01
-8.46040130e-01 5.01257420e-01 6.48000658e-01 7.26896226e-01
-3.82749021e-01 5.05886495e-01 -5.23420274e-01 -4.08914030e-01
3.37047935e-01 -1.03354543e-01 -1.49899676e-01 7.95696974e-02
-1.53851128e+00 8.89170289e-01 6.07348025e-01 3.21020782e-01
-2.61803925e-01 -4.16670948e-01 -8.96645784e-01 1.84060976e-01
3.17834169e-01 -3.08442920e-01 1.34302223e+00 -8.60447049e-01
-1.63449943e+00 6.94577277e-01 -3.67060393e-01 -4.80078757e-01
2.99727917e-01 -4.18906659e-01 -4.78146315e-01 -2.51533296e-02
-1.15219094e-01 -7.49308541e-02 8.49856973e-01 -6.67841136e-01
-7.67683506e-01 -1.54502913e-01 -1.72019657e-02 1.85269624e-01
-8.90303552e-01 3.59290689e-01 -2.43788779e-01 -1.14112079e+00
4.27056104e-02 -5.00581682e-01 -3.01819354e-01 -2.02302009e-01
-5.25685608e-01 -8.56160939e-01 1.07328773e+00 -3.80287409e-01
1.60478067e+00 -1.85907435e+00 -2.24907458e-01 1.00254603e-02
8.79140720e-02 3.92939001e-01 1.25894323e-01 7.22744405e-01
-9.32458863e-02 9.53405499e-02 7.95834064e-02 -4.36796874e-01
5.32341115e-02 8.00220594e-02 -6.56914413e-01 3.31106544e-01
-1.11934301e-02 8.17632198e-01 -6.51084423e-01 -5.09963036e-01
2.09488988e-01 4.56512213e-01 -2.20541488e-02 1.26664951e-01
-2.77079165e-01 -3.73823196e-01 -8.42150807e-01 4.06786084e-01
3.85506392e-01 -3.69963020e-01 8.47946927e-02 1.31148621e-01
-1.44543737e-01 8.44129801e-01 -1.15978372e+00 1.56538033e+00
-7.20253170e-01 7.77811289e-01 -3.45153481e-01 -1.46492851e+00
1.02664125e+00 5.61783671e-01 8.38245749e-02 -3.04764450e-01
5.28236032e-01 -9.24542248e-02 -2.03118443e-01 -4.37137067e-01
7.60668039e-01 5.22477217e-02 -1.02022514e-01 6.53389573e-01
1.20346926e-01 2.92460293e-01 2.02356890e-01 2.26824775e-01
9.72569406e-01 -3.93141657e-01 6.48086727e-01 -3.24986219e-01
7.15590775e-01 -2.06555203e-01 5.34570515e-01 8.16198826e-01
-6.39440417e-02 4.92419273e-01 5.63890040e-01 -8.00641000e-01
-7.37232506e-01 -2.62877911e-01 -4.66366291e-01 1.62194169e+00
-1.47294685e-01 -7.66694963e-01 -3.11985135e-01 -8.59151661e-01
-4.52342890e-02 6.61428154e-01 -8.93579423e-01 -3.98314297e-02
-4.00408745e-01 -1.04636264e+00 5.64958215e-01 6.60192788e-01
5.87615430e-01 -1.21065474e+00 -3.82209033e-01 3.80368084e-01
-1.52929602e-02 -1.04160655e+00 -5.51079273e-01 7.02310264e-01
-8.57229888e-01 -9.35965121e-01 -6.75382912e-01 -1.02545750e+00
5.31124890e-01 5.46301603e-01 9.96830285e-01 1.94438800e-01
-3.14192474e-01 -2.06676409e-01 -9.32765067e-01 -5.44967830e-01
-7.76953176e-02 6.65531576e-01 -2.51326859e-01 8.05514306e-02
9.18750107e-01 -2.20702603e-01 -5.08012950e-01 5.21917529e-02
-8.74434471e-01 -5.24154380e-02 5.52306294e-01 1.16320777e+00
1.90053388e-01 3.50451201e-01 6.58654630e-01 -1.25119936e+00
1.23238111e+00 -4.60114092e-01 -3.56768370e-01 3.53181899e-01
-6.49624765e-01 8.13091844e-02 1.04260015e+00 -7.63653040e-01
-1.02380693e+00 -1.10007569e-01 -1.39388189e-01 2.35591717e-02
-1.69745803e-01 9.84691620e-01 2.66057849e-01 2.99903333e-01
8.17969084e-01 5.50883651e-01 -5.17574966e-01 -5.26364088e-01
2.76931256e-01 1.10500598e+00 -1.13274671e-01 -6.02209389e-01
4.67549413e-01 1.76283762e-01 -4.23105806e-01 -9.92799640e-01
-1.17169940e+00 -7.34517336e-01 -6.70709550e-01 2.14981332e-01
5.01332641e-01 -7.39901960e-01 -6.37537956e-01 3.88386309e-01
-1.23709965e+00 7.71870241e-02 2.23140083e-02 5.43953240e-01
-7.18085095e-03 3.36913675e-01 -8.70698869e-01 -9.16431427e-01
-8.93646061e-01 -8.52139831e-01 8.06529999e-01 3.03195477e-01
-2.33672723e-01 -1.13586819e+00 -4.13417220e-02 1.68048348e-02
4.21870142e-01 -2.05839470e-01 9.49145734e-01 -1.10761178e+00
1.74847096e-01 -5.42281806e-01 -2.45300382e-01 2.03535974e-01
3.43758881e-01 2.85272207e-03 -1.02471149e+00 -1.71594754e-01
-6.29465654e-02 -4.58217412e-01 1.31309795e+00 2.64603823e-01
1.40161109e+00 -2.72181183e-01 -4.34008449e-01 2.99460441e-01
1.05486155e+00 2.06933752e-01 4.49659675e-01 4.39654946e-01
6.31183624e-01 4.16883588e-01 3.75203848e-01 6.83682442e-01
2.06853762e-01 2.81911492e-01 -1.90127552e-01 -4.79439050e-02
4.56359029e-01 -1.97474211e-01 2.64740586e-01 1.04781425e+00
1.43064648e-01 -4.52561170e-01 -1.02789855e+00 5.35257041e-01
-2.08041406e+00 -8.09193730e-01 -1.69051230e-01 1.80794668e+00
1.11931980e+00 4.85988587e-01 -1.51572943e-01 4.51008707e-01
8.46775353e-01 4.63240981e-01 -2.86641538e-01 -5.46555519e-01
1.22165447e-02 4.41805780e-01 2.08452910e-01 1.78873032e-01
-1.48695934e+00 1.11288393e+00 6.74249315e+00 9.67064440e-01
-1.14059484e+00 -4.71895896e-02 4.04341847e-01 5.24786450e-02
1.00064613e-01 -1.78805828e-01 -1.15258145e+00 4.11753923e-01
8.77001882e-01 -3.64795446e-01 -1.96416825e-01 1.07221973e+00
5.73792085e-02 8.75507668e-02 -9.94156718e-01 8.41122329e-01
2.68189222e-01 -1.40403247e+00 2.28338107e-01 -3.65863234e-01
5.39921820e-01 -1.91010252e-01 -1.48555087e-02 4.91423249e-01
5.39260685e-01 -1.11529422e+00 3.70834470e-01 3.16952378e-01
4.66011465e-01 -8.91841292e-01 1.00765312e+00 4.10454392e-01
-1.42751527e+00 -7.02944919e-02 -6.61662936e-01 -2.32474715e-01
-2.70321041e-01 8.83056462e-01 -6.29094183e-01 5.19970715e-01
6.13753319e-01 9.82291460e-01 -4.27212059e-01 7.54927635e-01
-3.55743259e-01 8.36903572e-01 -1.47567749e-01 -6.88186646e-01
5.12420535e-01 1.15895346e-01 8.45121220e-02 1.66752517e+00
-6.62640408e-02 1.06222071e-01 3.81077468e-01 4.11331236e-01
-4.58777577e-01 7.19129682e-01 -5.70144773e-01 -1.67136624e-01
3.19544494e-01 1.32283854e+00 -9.01631176e-01 -5.51383674e-01
-6.62996650e-01 8.70412707e-01 3.84129524e-01 4.08388019e-01
-3.13547522e-01 -1.25124896e+00 4.95282650e-01 -3.23394448e-01
5.05114913e-01 -1.56085312e-01 -2.77926445e-01 -1.53642905e+00
1.64232329e-01 -7.93278039e-01 5.68029583e-01 -3.39710087e-01
-1.53166986e+00 7.64349818e-01 -3.69931668e-01 -1.09114349e+00
-2.39636421e-01 -8.67828012e-01 -8.73684943e-01 9.59759474e-01
-1.62004077e+00 -1.05044580e+00 -9.84229073e-02 6.35442793e-01
9.85587835e-01 -4.01071459e-01 1.07985437e+00 9.49503854e-02
-8.83380532e-01 7.50036955e-01 4.45566565e-01 7.30828345e-01
7.36406386e-01 -1.08301854e+00 5.41223049e-01 7.69982636e-01
3.23395938e-01 9.85819697e-01 3.99536312e-01 -4.16091323e-01
-1.26998842e+00 -9.13056254e-01 1.09727240e+00 -1.53611913e-01
9.32840943e-01 -6.00337982e-01 -1.22949052e+00 7.35565305e-01
3.25766027e-01 2.38139153e-01 9.40797389e-01 6.15428030e-01
-5.98196149e-01 -6.57745078e-02 -6.15961611e-01 4.91867214e-01
6.08259082e-01 -7.20398188e-01 -1.17293727e+00 3.53922546e-01
7.25726664e-01 -1.26006827e-01 -5.66792846e-01 3.04127503e-02
7.26276815e-01 -4.57944632e-01 6.83341742e-01 -8.88084710e-01
4.59885567e-01 2.01121643e-02 2.88011581e-02 -1.46854913e+00
-2.67223001e-01 -4.75705475e-01 -9.05706659e-02 1.24319410e+00
4.79044437e-01 -7.72536993e-01 8.40236902e-01 1.73274860e-01
1.43579423e-01 -8.42025876e-01 -7.46480763e-01 -5.21047115e-01
3.66811931e-01 -4.90652531e-01 2.91461200e-01 1.35971272e+00
4.45635945e-01 7.04391003e-01 -2.11607486e-01 -2.39085138e-01
1.73831478e-01 3.67903292e-01 4.43617374e-01 -1.53307116e+00
2.04606816e-01 -7.30771840e-01 -4.34642643e-01 -1.18001819e+00
5.36814213e-01 -9.25889909e-01 1.94515884e-02 -1.48807526e+00
1.98560417e-01 -3.15294981e-01 -6.59818113e-01 5.48643053e-01
-5.19789696e-01 -1.74055576e-01 -2.61187822e-01 1.40968338e-01
-5.01878500e-01 6.94529653e-01 9.21364903e-01 -3.12364399e-01
-2.05360219e-01 5.39903454e-02 -8.18351030e-01 7.67396212e-01
8.81327987e-01 -5.86655974e-01 -2.76705772e-01 -3.69091183e-01
3.23115200e-01 -3.74755621e-01 -2.78216690e-01 -5.43628633e-01
6.11266792e-01 -2.55479306e-01 7.14056492e-01 -8.15162957e-01
-6.90168515e-02 -5.81055284e-01 -6.94502831e-01 2.99007118e-01
-7.50432312e-01 -1.03481794e-02 6.05013333e-02 5.03822863e-01
-4.04676646e-01 -5.38622856e-01 6.19876504e-01 -1.08206928e-01
-3.81411314e-01 1.65541589e-01 -7.27266371e-01 -6.87686950e-02
6.75180614e-01 7.90489018e-02 -2.46449143e-01 -1.74613342e-01
-2.34401211e-01 3.01888078e-01 -2.10674986e-01 5.92770457e-01
6.65183306e-01 -1.23074877e+00 -7.38039255e-01 2.09392965e-01
3.70327234e-01 -1.11635521e-01 -1.11651815e-01 1.62475869e-01
-3.63596678e-01 6.57483757e-01 2.40261331e-01 -2.31194809e-01
-1.22611475e+00 5.36243856e-01 2.10639551e-01 -5.35795331e-01
-9.48286474e-01 7.21318841e-01 -7.46892840e-02 -4.29954618e-01
5.46045601e-01 -6.56713843e-01 -6.43691897e-01 4.60866809e-01
8.47615421e-01 -6.33639023e-02 2.99543560e-01 -3.52026582e-01
-2.81651199e-01 4.88654196e-01 -7.83491313e-01 2.70524204e-01
1.29275846e+00 4.63420935e-02 -1.26195261e-02 8.59921575e-01
1.15715802e+00 -2.06102431e-01 -5.08574605e-01 -8.92229140e-01
2.55317301e-01 -2.58061826e-01 4.84484106e-01 -4.31914717e-01
-7.42490053e-01 1.12878525e+00 2.56672442e-01 6.34873331e-01
8.67523551e-01 -3.84178549e-01 8.80733013e-01 1.04511356e+00
-1.20143212e-01 -1.46048427e+00 -3.86660136e-02 1.04356241e+00
6.66976631e-01 -1.33315825e+00 3.04952383e-01 -1.46288335e-01
-4.72846597e-01 1.58063793e+00 3.86261523e-01 -2.52175122e-01
1.09750974e+00 1.99806824e-01 2.14083344e-01 1.43460492e-02
-1.15641594e+00 -2.70702571e-01 3.12491536e-01 5.41770197e-02
1.03479373e+00 -1.50948584e-01 -5.71492434e-01 7.86390245e-01
-2.05194622e-01 -1.02899566e-01 3.25868547e-01 1.32043874e+00
-7.21067369e-01 -1.18827820e+00 -3.36525619e-01 7.19207466e-01
-8.25668812e-01 -3.97154719e-01 -3.67473572e-01 4.99979347e-01
-4.05608743e-01 8.79682481e-01 1.33402228e-01 -3.24172825e-01
3.22841048e-01 5.16514421e-01 -9.23283845e-02 -8.95420551e-01
-9.20463443e-01 -2.08578244e-01 -2.15005055e-02 2.62152553e-02
-2.66221315e-01 -3.92570227e-01 -1.13445425e+00 -4.15615141e-01
-7.16171622e-01 3.19578111e-01 5.78838110e-01 1.08783746e+00
1.81946576e-01 6.67424560e-01 7.01471984e-01 -5.96543789e-01
-5.40425420e-01 -1.51791441e+00 -4.19389546e-01 2.42448539e-01
3.31320882e-01 -6.05384350e-01 -3.54269058e-01 1.28349304e-01] | [10.705069541931152, 7.670263290405273] |
c155334f-a793-4fff-b6ad-475e8fae4f83 | belittling-the-source-trustworthiness | 1809.00494 | null | http://arxiv.org/abs/1809.00494v1 | http://arxiv.org/pdf/1809.00494v1.pdf | Belittling the Source: Trustworthiness Indicators to Obfuscate Fake News on the Web | With the growth of the internet, the number of fake-news online has been
proliferating every year. The consequences of such phenomena are manifold,
ranging from lousy decision-making process to bullying and violence episodes.
Therefore, fact-checking algorithms became a valuable asset. To this aim, an
important step to detect fake-news is to have access to a credibility score for
a given information source. However, most of the widely used Web indicators
have either been shut-down to the public (e.g., Google PageRank) or are not
free for use (Alexa Rank). Further existing databases are short-manually
curated lists of online sources, which do not scale. Finally, most of the
research on the topic is theoretical-based or explore confidential data in a
restricted simulation environment. In this paper we explore current research,
highlight the challenges and propose solutions to tackle the problem of
classifying websites into a credibility scale. The proposed model automatically
extracts source reputation cues and computes a credibility factor, providing
valuable insights which can help in belittling dubious and confirming trustful
unknown websites. Experimental results outperform state of the art in the
2-classes and 5-classes setting. | ['Jens Lehmann', 'Piyush Chawla', 'Diego Esteves', 'Aniketh Janardhan Reddy'] | 2018-09-03 | belittling-the-source-trustworthiness-1 | https://aclanthology.org/W18-5508 | https://aclanthology.org/W18-5508.pdf | ws-2018-11 | ['web-credibility', 'subjectivity-analysis'] | ['methodology', 'natural-language-processing'] | [-4.05828267e-01 4.49842364e-02 -5.79348087e-01 -2.00260088e-01
-9.60096657e-01 -8.46504927e-01 7.24033952e-01 7.33650446e-01
-2.42693573e-01 9.17229712e-01 -4.71881106e-02 -2.70210296e-01
8.05452242e-02 -8.10083032e-01 -6.76075995e-01 -4.86476779e-01
2.59891581e-02 2.62409329e-01 7.11750805e-01 -3.15509975e-01
7.67477691e-01 5.81801981e-02 -1.20033336e+00 5.44696562e-02
1.10656762e+00 1.11308861e+00 -4.37458038e-01 2.09077090e-01
-4.07615490e-02 6.91644490e-01 -8.05855989e-01 -1.34739637e+00
2.53650576e-01 -3.56253833e-01 -5.75188935e-01 -2.19062462e-01
5.63555062e-02 -4.63436931e-01 -2.83634812e-01 1.72933006e+00
1.49470076e-01 -3.43518049e-01 3.13641042e-01 -1.51594067e+00
-6.93362057e-01 7.32619703e-01 -6.49395287e-01 3.78382206e-01
4.59012598e-01 -3.23823661e-01 1.05780458e+00 -6.33047760e-01
6.94149196e-01 9.25444186e-01 4.47606355e-01 -7.51099959e-02
-5.80725610e-01 -6.56052649e-01 -1.70971230e-01 5.43646634e-01
-1.13357520e+00 -2.32874423e-01 7.84792423e-01 -3.96296382e-01
-1.06080398e-01 3.13981116e-01 5.27630925e-01 1.41283309e+00
7.28390962e-02 6.52354419e-01 1.79879069e+00 -2.40169629e-01
2.10061803e-01 8.61896098e-01 2.35856652e-01 5.87416589e-01
1.10564816e+00 -2.09835723e-01 -7.28514254e-01 -6.48748815e-01
4.63600010e-01 -5.61114848e-02 -3.96675199e-01 -2.28057578e-02
-8.19987297e-01 1.07897747e+00 4.12168145e-01 3.73934507e-01
-3.30759257e-01 -4.27654773e-01 4.73143965e-01 4.86500353e-01
6.40394509e-01 4.12763000e-01 -1.04583099e-01 -3.94231647e-01
-9.36260045e-01 8.55071247e-02 1.14562166e+00 6.46514118e-01
4.29452091e-01 -3.49154651e-01 2.93509334e-01 3.90881687e-01
2.66474009e-01 4.16710109e-01 5.86152971e-01 -6.39334381e-01
5.06729722e-01 6.74247622e-01 5.79725325e-01 -1.92905402e+00
1.03970230e-01 -6.48467124e-01 -6.67294621e-01 -7.00182188e-03
8.53831649e-01 2.23107427e-01 -1.36246607e-01 1.08983552e+00
4.53414053e-01 1.37217030e-01 -3.49508703e-01 1.25861323e+00
3.95046383e-01 3.28537583e-01 -2.65261710e-01 -2.63054252e-01
1.52172148e+00 -7.36374497e-01 -7.81527281e-01 1.05493933e-01
3.19098532e-01 -7.90772617e-01 9.86151636e-01 7.95802295e-01
-6.90224648e-01 1.29890248e-01 -1.06181407e+00 1.37002066e-01
-6.70919836e-01 -1.16943836e-01 7.09264159e-01 9.93280649e-01
-5.25801599e-01 7.79247820e-01 -6.05212986e-01 -2.55071908e-01
6.19022906e-01 -3.94333810e-01 -2.78609246e-01 -2.41046250e-01
-1.67235029e+00 9.38510716e-01 2.49265686e-01 -1.84564903e-01
-8.62062514e-01 5.34059759e-03 -1.54137000e-01 -1.07660450e-01
7.98210382e-01 -3.93317975e-02 9.04748142e-01 -1.03808010e+00
-1.12625003e+00 7.80944824e-01 4.62374777e-01 -6.33437335e-01
1.06951940e+00 -1.53888568e-01 -6.10402346e-01 4.03201818e-01
4.19723004e-01 -3.88784379e-01 9.39059138e-01 -1.38210881e+00
-5.97615957e-01 -4.20460016e-01 2.83638597e-01 -2.17830807e-01
-6.39896929e-01 1.34806544e-01 -2.02777833e-01 -6.90229952e-01
1.71520501e-01 -7.43374765e-01 1.15367211e-01 -1.46249002e-02
-7.13648260e-01 -1.39739439e-02 6.23352647e-01 -1.02505946e+00
1.34698963e+00 -1.76585877e+00 -5.14047265e-01 2.65648752e-01
3.83881301e-01 3.97128612e-01 5.04749417e-01 7.28545368e-01
4.82928783e-01 5.95935464e-01 1.48337036e-01 2.25723498e-02
-1.20863974e-01 -1.92807972e-01 -3.80669713e-01 8.81325006e-01
-2.43692607e-01 3.92646372e-01 -1.23123384e+00 -7.53339112e-01
-4.43469256e-01 1.48809925e-01 -1.97100371e-01 -1.26364427e-02
1.26283849e-02 3.13392431e-01 -8.87210369e-01 1.01094425e+00
7.42244184e-01 -5.78014314e-01 6.31818846e-02 -2.20011780e-03
-1.06161214e-01 4.54584479e-01 -9.40443516e-01 1.09844530e+00
-1.29658550e-01 5.83927095e-01 1.10027067e-01 -6.66110277e-01
7.81425238e-01 2.36098051e-01 1.61076352e-01 -2.71179348e-01
3.90880674e-01 5.72754920e-01 -3.62362206e-01 -4.46939677e-01
6.05145454e-01 1.87699229e-01 -2.98570525e-02 6.69330657e-01
-4.18276489e-01 3.37810367e-01 -3.81775424e-02 6.46214008e-01
1.19586933e+00 -1.98223680e-01 3.61319631e-01 8.28548744e-02
4.03316468e-01 3.07064474e-01 3.05909276e-01 7.28395283e-01
-5.66190839e-01 3.10423881e-01 8.99521887e-01 -7.15408325e-02
-8.46580088e-01 -7.11774588e-01 -8.50223154e-02 6.40856266e-01
4.69727606e-01 -3.30579191e-01 -7.62578011e-01 -1.00820088e+00
5.64569384e-02 6.17248416e-01 -3.73007983e-01 -9.30113569e-02
4.39061336e-02 -5.25968075e-01 6.54729068e-01 -2.02256486e-01
9.19145703e-01 -5.26264429e-01 -4.02593464e-01 1.51316971e-01
-6.96074843e-01 -1.06734729e+00 -1.62157342e-01 -4.23691899e-01
-6.36269331e-01 -1.51526582e+00 -5.07457435e-01 -1.62535116e-01
6.40010118e-01 7.73735821e-01 8.57051849e-01 3.58030707e-01
1.48101375e-01 9.76426341e-03 -8.04598212e-01 -2.59225488e-01
-5.58737397e-01 -1.25602990e-01 1.13268912e-01 3.45986545e-01
2.90352166e-01 -4.51898098e-01 -6.35452449e-01 6.75488710e-01
-8.75289440e-01 -4.85572845e-01 6.78490698e-01 7.89469123e-01
2.07000718e-01 3.86702269e-01 8.72876704e-01 -1.17638409e+00
9.22717452e-01 -1.02005780e+00 -5.47011793e-01 1.67102531e-01
-9.51881826e-01 -3.31567049e-01 6.30449653e-01 -4.34473842e-01
-7.97855914e-01 -6.10405982e-01 3.23195487e-01 -1.69297621e-01
1.38307720e-01 7.90185690e-01 1.94053762e-02 -1.04751118e-01
7.44358361e-01 4.91377935e-02 -1.46644384e-01 -4.66474950e-01
2.13372916e-01 1.05017412e+00 2.52751410e-01 -3.93690944e-01
1.26307178e+00 6.67569637e-01 -3.91894042e-01 -6.02931440e-01
-1.23670053e+00 -6.86922133e-01 -2.95406282e-01 -3.86378795e-01
1.53733969e-01 -6.99508846e-01 -7.19191253e-01 5.49682736e-01
-1.14604867e+00 5.78173637e-01 3.67891967e-01 3.31195444e-01
1.44397944e-01 9.10810351e-01 -6.66505754e-01 -9.73514020e-01
-1.17877692e-01 -7.27556705e-01 4.58985150e-01 1.95340604e-01
8.74521658e-02 -6.96411610e-01 1.92654673e-02 8.44391465e-01
5.65434694e-01 5.15655160e-01 2.51611501e-01 -1.05556881e+00
-7.19461501e-01 -1.00710654e+00 -4.78480160e-01 3.54254246e-01
-8.48029703e-02 -9.60447937e-02 -8.28212619e-01 -1.99745059e-01
1.82853073e-01 -4.24319327e-01 5.97558558e-01 -3.57563645e-01
7.28381217e-01 -9.79841888e-01 -2.93400675e-01 -1.67567074e-01
1.26165068e+00 -3.87101710e-01 4.46783811e-01 7.71665454e-01
2.17390627e-01 6.63644493e-01 9.81123269e-01 8.21069121e-01
5.79343438e-01 4.20086652e-01 8.36068809e-01 5.31200409e-01
3.67478549e-01 -5.58246315e-01 2.63251007e-01 9.20502961e-01
-1.56564116e-01 -1.95915535e-01 -7.89124846e-01 4.19761240e-01
-1.69325233e+00 -1.09475827e+00 -4.74708855e-01 2.32208991e+00
7.93349802e-01 5.98746717e-01 1.68906823e-01 4.09514576e-01
1.06351638e+00 5.49926423e-02 -3.26742321e-01 1.01093508e-01
-5.27461395e-02 -5.35246670e-01 7.08104074e-01 1.12392612e-01
-9.46927190e-01 6.86298609e-01 4.68349886e+00 8.23680043e-01
-8.64047706e-01 5.04995227e-01 8.29543829e-01 4.95909572e-01
-1.59573749e-01 2.55105942e-01 -5.39389670e-01 8.85028601e-01
8.50448191e-01 -3.85178059e-01 3.53459686e-01 1.09493542e+00
2.58746177e-01 -5.51343679e-01 -4.67076540e-01 7.92312503e-01
3.80088925e-01 -9.26502526e-01 -2.13076487e-01 1.53938055e-01
5.31347454e-01 -6.08244576e-02 2.51796916e-02 1.19234905e-01
3.32079232e-01 -4.18876588e-01 8.06230485e-01 4.39003855e-01
4.86213923e-01 -5.45628428e-01 1.08274019e+00 7.57814109e-01
-4.71888930e-01 7.68163130e-02 -5.13117135e-01 1.42353192e-01
1.53400645e-01 1.10141838e+00 -7.13732481e-01 5.34186482e-01
8.33097756e-01 5.72066247e-01 -8.89490485e-01 1.32674766e+00
-5.23931324e-01 9.57207859e-01 -2.45021284e-01 -4.72509116e-01
7.71779791e-02 -1.63792163e-01 7.09544063e-01 6.72830284e-01
3.14651191e-01 3.45103280e-03 2.45400108e-02 5.98540246e-01
-3.96189749e-01 1.11003965e-01 -4.81635153e-01 -3.45002949e-01
6.83505416e-01 1.50229537e+00 -9.32806909e-01 -2.79535890e-01
-2.52535313e-01 9.62530375e-01 2.99971730e-01 -7.79317543e-02
-9.85659540e-01 -3.77540290e-01 1.87437251e-01 6.77415967e-01
9.18669801e-04 -2.62575001e-01 -1.04743063e-01 -1.45019054e+00
1.30055994e-01 -8.71296644e-01 3.76591027e-01 -6.71356976e-01
-1.59897375e+00 4.73270297e-01 -2.03354508e-01 -1.44463134e+00
2.50036329e-01 -2.60008305e-01 -4.30446804e-01 3.80274177e-01
-1.82410598e+00 -8.53434443e-01 -4.38776046e-01 5.17526865e-01
3.24502327e-02 -4.47664149e-02 3.48694146e-01 4.20530856e-01
-4.44490373e-01 5.52496135e-01 4.25399989e-01 3.24423552e-01
1.03879142e+00 -9.08647001e-01 -2.15466823e-02 8.34238291e-01
1.85376778e-01 6.70376539e-01 9.00404155e-01 -8.56619656e-01
-1.29939187e+00 -7.02779293e-01 9.26834345e-01 -6.14416957e-01
1.44594967e+00 -1.94718346e-01 -9.46488380e-01 2.06607878e-01
-1.27179667e-01 1.16496153e-01 5.48691392e-01 -1.05332032e-01
-7.71836758e-01 -5.93537278e-02 -1.45298421e+00 3.10427368e-01
6.41149938e-01 -4.78847533e-01 -3.31821322e-01 5.79874635e-01
3.86270642e-01 -1.42709702e-01 -6.56551778e-01 -4.49631333e-01
4.67852294e-01 -1.28433216e+00 4.69653845e-01 -5.80737233e-01
6.17058933e-01 -9.88642201e-02 2.21369803e-01 -1.33016467e+00
6.15180545e-02 -5.40735841e-01 -2.06449807e-01 1.39315498e+00
2.41761908e-01 -9.47260082e-01 7.60809839e-01 4.43085462e-01
2.94695258e-01 -2.65562922e-01 -1.09376895e+00 -6.28670633e-01
-3.58272791e-01 -1.12289786e-01 2.07028836e-01 1.36658287e+00
2.02751160e-01 -1.00983351e-01 -4.09936994e-01 3.83657217e-01
8.15842927e-01 1.21524177e-01 6.34265482e-01 -1.37229443e+00
-2.40184829e-01 -2.96229064e-01 -4.70119476e-01 -5.91070592e-01
-2.68786281e-01 -6.57604456e-01 -4.40757453e-01 -1.17954898e+00
2.67483622e-01 -4.82696384e-01 -1.73731908e-01 2.46160120e-01
-1.28933057e-01 2.13848129e-01 -6.90786317e-02 7.44425178e-01
-9.05018926e-01 2.75543332e-01 1.09779334e+00 -3.49499807e-02
2.44356796e-01 2.56436169e-01 -8.02126169e-01 7.05843270e-01
9.18666542e-01 -1.00488317e+00 -2.11270172e-02 2.36133516e-01
8.16622615e-01 2.80032098e-01 6.41151547e-01 -8.34478676e-01
3.54109913e-01 -1.22278556e-01 -1.68790072e-01 -3.28338563e-01
5.20969331e-02 -9.19780314e-01 -1.61115304e-01 3.05902660e-01
-1.50078833e-01 2.15427168e-02 -6.27316535e-01 1.36402690e+00
-4.00546283e-01 -4.20199543e-01 6.81727529e-01 -2.91004539e-01
1.37402024e-02 5.61463535e-02 -1.89297318e-01 2.51537204e-01
1.03250730e+00 -1.88883021e-02 -9.35529828e-01 -8.20511997e-01
-2.89730877e-01 -6.38837665e-02 5.19420981e-01 2.97421634e-01
3.58047664e-01 -1.08357716e+00 -7.65005708e-01 -4.35673535e-01
1.72531798e-01 -5.99515498e-01 -8.49706605e-02 1.03461194e+00
-5.13164103e-01 2.30644301e-01 -1.82327181e-01 -6.36043549e-02
-9.03650165e-01 5.96835017e-01 -1.93771452e-01 -3.80255520e-01
-1.10162377e-01 4.39816445e-01 -8.00078988e-01 7.27269799e-02
9.18784738e-02 2.48953477e-01 -4.31681663e-01 4.06860352e-01
6.04409277e-01 7.11745858e-01 1.79644585e-01 -7.85922289e-01
-1.74558520e-01 -1.90004572e-01 -1.71085581e-01 -2.21207608e-02
1.21439517e+00 -2.67545313e-01 -3.75268161e-01 4.37000513e-01
9.26445186e-01 4.26799893e-01 -7.88615108e-01 -3.18184584e-01
3.38305682e-01 -1.04900813e+00 1.10353298e-01 -9.87612128e-01
-1.01135027e+00 6.16008639e-01 1.79216973e-02 9.52295721e-01
5.04712641e-01 -1.05677977e-01 7.76571393e-01 4.05062318e-01
1.00693274e+00 -1.19684422e+00 4.60103810e-01 3.13848145e-02
6.87380314e-01 -1.58714819e+00 2.15092301e-01 -6.17486596e-01
-7.39341021e-01 9.11971748e-01 3.19772989e-01 9.07014869e-03
7.86881506e-01 -3.55073273e-01 -6.17356338e-02 -2.26925313e-01
-3.32119226e-01 1.58564210e-01 -4.53231968e-02 2.79173613e-01
2.49485433e-01 -4.09427620e-02 -8.51899803e-01 9.76824880e-01
-4.72213142e-02 -1.87286407e-01 1.15987575e+00 7.78966010e-01
-6.16141021e-01 -8.92639399e-01 -5.79405844e-01 6.44183278e-01
-1.01452923e+00 9.15468019e-03 -6.08430803e-01 7.53788948e-01
-3.19082856e-01 1.26199055e+00 -8.15125525e-01 -3.10142130e-01
4.72607985e-02 -1.67692065e-01 -2.90912449e-01 -2.41687134e-01
-5.77501893e-01 -3.64376813e-01 3.97285342e-01 -4.63258296e-01
-4.20684367e-01 -6.85569882e-01 -7.11626887e-01 -7.49867678e-01
-1.02723157e+00 5.26933968e-01 9.95431542e-01 5.37974775e-01
2.63022989e-01 -2.50975192e-01 9.22273397e-01 -2.83661574e-01
-1.04026628e+00 -8.50781441e-01 -6.85325623e-01 4.82971907e-01
-2.37183701e-02 -9.50737357e-01 -8.91741157e-01 -2.74162740e-01] | [8.128194808959961, 10.215356826782227] |
79e2e95f-9299-47c6-9261-562056adedcf | representing-focus-in-ltag | null | null | https://aclanthology.org/W12-4609 | https://aclanthology.org/W12-4609.pdf | Representing Focus in LTAG | null | ['Kata Balogh'] | 2012-09-01 | representing-focus-in-ltag-1 | https://aclanthology.org/W12-4609 | https://aclanthology.org/W12-4609.pdf | ws-2012-9 | ['dialogue-management'] | ['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] | [-7.336844444274902, 3.683255195617676] |
723f9b98-0ce4-494d-bdc9-06cef388eefb | emulating-the-dynamics-of-complex-systems | 2306.16335 | null | https://arxiv.org/abs/2306.16335v1 | https://arxiv.org/pdf/2306.16335v1.pdf | Emulating the dynamics of complex systems using autoregressive models on manifolds (mNARX) | In this study, we propose a novel surrogate modelling approach to efficiently and accurately approximate the response of complex dynamical systems driven by time-varying exogenous excitations over extended time periods. Our approach, that we name \emph{manifold nonlinear autoregressive modelling with exogenous input} (mNARX), involves constructing a problem-specific exogenous input manifold that is optimal for constructing autoregressive surrogates. The manifold, which forms the core of mNARX, is constructed incrementally by incorporating the physics of the system, as well as prior expert- and domain- knowledge. Because mNARX decomposes the full problem into a series of smaller sub-problems, each with a lower complexity than the original, it scales well with the complexity of the problem, both in terms of training and evaluation costs of the final surrogate. Furthermore, mNARX synergizes well with traditional dimensionality reduction techniques, making it highly suitable for modelling dynamical systems with high-dimensional exogenous inputs, a class of problems that is typically challenging to solve.Since domain knowledge is particularly abundant in physical systems, such as those found in engineering applications, mNARX is well suited for these applications. We demonstrate that mNARX outperforms traditional autoregressive surrogates in predicting the response of a classical coupled spring-mass system excited by a one-dimensional random excitation. Additionally, we show that mNARX is well suited for emulating very high-dimensional time- and state-dependent systems, even when affected by active controllers, by surrogating the dynamics of a realistic aero-servo-elastic onshore wind turbine simulator. In general, our results demonstrate that mNARX offers promising prospects for modelling complex dynamical systems, in terms of accuracy and efficiency. | ['Bruno Sudret', 'Stefano Marelli', 'Styfen Schär'] | 2023-06-28 | null | null | null | null | ['dimensionality-reduction'] | ['methodology'] | [-2.78190672e-02 1.44537417e-02 3.54275733e-01 5.13391793e-01
-4.18717772e-01 -6.70602262e-01 5.50244868e-01 -1.44155219e-01
-7.37856477e-02 6.73971891e-01 -1.60706446e-01 -3.85512471e-01
-7.10760236e-01 -4.79097456e-01 -8.52252126e-01 -9.60941136e-01
-1.35002777e-01 5.67326069e-01 -2.54914910e-01 -5.56807399e-01
-3.30520511e-01 5.91604114e-01 -1.49532115e+00 -5.31221807e-01
8.41180682e-01 7.62912989e-01 2.02955648e-01 7.81304121e-01
6.17528260e-01 5.19050956e-01 -3.63273263e-01 4.75122422e-01
3.57370615e-01 -3.74446958e-01 -5.87928772e-01 1.60597160e-01
-2.30824485e-01 -1.76736824e-02 -5.39613128e-01 6.55802488e-01
4.58313972e-01 8.88439417e-01 7.94208229e-01 -9.08594131e-01
-3.15809131e-01 1.67911038e-01 1.01397745e-02 -5.19796945e-02
-2.84570921e-02 4.78274137e-01 5.78235447e-01 -1.03260815e+00
2.55028367e-01 1.14058471e+00 8.17664206e-01 4.60049838e-01
-1.62653244e+00 -1.21281922e-01 3.15669179e-01 -4.73150462e-01
-1.32351637e+00 -2.19607323e-01 8.98650050e-01 -9.20746267e-01
8.17055166e-01 4.14202690e-01 5.25811851e-01 1.08072686e+00
2.60353565e-01 3.04520458e-01 8.89086723e-01 -1.49510786e-01
4.60438639e-01 -1.08694639e-02 1.53568149e-01 2.22381368e-01
2.25896034e-02 5.27564883e-01 2.31462140e-02 -4.59862322e-01
1.03912735e+00 5.72485551e-02 -3.65231574e-01 -5.21657526e-01
-1.20140684e+00 6.24211133e-01 2.13538423e-01 1.89624026e-01
-8.19509387e-01 3.19492459e-01 1.46932721e-01 4.77289885e-01
5.26731789e-01 9.33222651e-01 -5.36185801e-01 -2.94601947e-01
-4.33186144e-01 3.96137327e-01 6.59193575e-01 8.43220592e-01
2.77196735e-01 9.11689043e-01 4.01995659e-01 8.19656253e-01
2.01336741e-01 7.55532861e-01 4.79973167e-01 -1.09571803e+00
1.62598431e-01 4.46621567e-01 6.35792136e-01 -8.21314991e-01
-4.23688233e-01 -6.21527970e-01 -1.19326663e+00 4.21439439e-01
2.49409646e-01 -5.69882572e-01 -6.82767153e-01 1.77213168e+00
5.40651262e-01 4.07217801e-01 2.92491913e-01 1.05529499e+00
8.45207125e-02 1.07984126e+00 -2.09769458e-01 -4.97159004e-01
1.02059519e+00 -5.83397031e-01 -6.57896459e-01 -2.25901678e-01
5.14563620e-01 -4.67121631e-01 1.09124839e+00 4.20902938e-01
-1.32731771e+00 -7.91282594e-01 -7.86746621e-01 3.54392141e-01
-3.44519407e-01 1.69434175e-01 2.33149782e-01 1.60464309e-02
-1.02279997e+00 9.62852418e-01 -1.14373779e+00 -7.75171863e-03
-4.94329214e-01 4.43100721e-01 -1.61021665e-01 4.84784693e-01
-1.36800909e+00 1.03323591e+00 1.59401521e-01 5.57556212e-01
-1.03489554e+00 -1.24033093e+00 -8.03498268e-01 1.13790341e-01
5.35449922e-01 -9.21889842e-01 1.19563305e+00 -4.47658271e-01
-1.99029648e+00 2.99170185e-02 3.68873961e-02 -3.51949960e-01
3.62654805e-01 -2.48073414e-01 -4.59574997e-01 -1.69581801e-01
-4.33369517e-01 -1.74971461e-01 1.20801318e+00 -1.02828968e+00
4.02977467e-02 -1.19981140e-01 4.09691669e-02 1.20536827e-01
-2.18073711e-01 -2.78525442e-01 8.48848596e-02 -7.51408577e-01
-1.25045493e-01 -1.33199883e+00 -8.16040635e-01 -2.82716453e-01
-5.02909832e-02 -4.09226343e-02 1.01329458e+00 -4.78038311e-01
1.36914968e+00 -2.19961715e+00 8.91073823e-01 3.40806991e-01
1.98698401e-01 6.25805557e-01 6.26175478e-02 9.51581717e-01
-5.77116728e-01 1.93153732e-02 -3.25442404e-01 -1.15972213e-01
1.18085213e-01 2.08672032e-01 -8.47831011e-01 3.85201156e-01
5.08632958e-01 7.61134028e-01 -9.28823173e-01 1.68708175e-01
5.13948023e-01 6.65977657e-01 -2.09212556e-01 3.16730708e-01
-2.40749300e-01 7.90030062e-01 -7.00150311e-01 -6.50775805e-02
1.66254714e-01 -3.88838559e-01 -1.57252148e-01 1.34426773e-01
-2.54460394e-01 -2.21232191e-01 -1.45819902e+00 1.11501575e+00
-9.54738021e-01 3.97648633e-01 5.75348973e-01 -1.38307023e+00
1.02080762e+00 6.45291865e-01 7.33050644e-01 -1.06882825e-02
1.58339322e-01 2.35760435e-01 -1.30246043e-01 -3.10686857e-01
3.05247396e-01 -2.24533573e-01 -1.48795888e-01 2.49508888e-01
-1.00421764e-01 -6.67949200e-01 -9.46600512e-02 3.50790732e-02
1.04251003e+00 1.92248002e-01 3.04678708e-01 -5.52978456e-01
7.05655754e-01 2.03853086e-01 4.32750046e-01 2.89010346e-01
3.70704681e-01 1.76278755e-01 3.97738904e-01 -3.84419799e-01
-1.15761924e+00 -8.83461952e-01 -1.89439908e-01 5.92933357e-01
-4.31407057e-02 -1.59608990e-01 -6.49358273e-01 1.80914775e-01
1.31781831e-01 6.58141792e-01 -5.87642848e-01 -4.59658235e-01
-7.49727905e-01 -5.79623163e-01 -5.51897585e-02 5.61944783e-01
-2.02288553e-01 -6.75860226e-01 -5.02377748e-01 6.95360243e-01
3.98122430e-01 -1.04310644e+00 -3.55693847e-01 2.00548828e-01
-1.05173147e+00 -1.00722373e+00 -7.35993564e-01 -4.65153754e-01
6.79037631e-01 -6.30007014e-02 9.02144969e-01 -2.15281561e-01
-4.58595604e-01 8.79335821e-01 1.16891272e-01 -3.64147097e-01
-7.12070823e-01 -2.24254444e-01 6.51102722e-01 1.44412935e-01
-4.90503162e-01 -7.57926226e-01 -2.74882495e-01 4.79625612e-01
-8.79245758e-01 -2.11829588e-01 3.24945241e-01 1.14314222e+00
7.46120036e-01 1.98715031e-01 6.54258072e-01 -7.38771856e-01
8.88405442e-01 -5.91943026e-01 -1.15182292e+00 -1.34890586e-01
-3.80385250e-01 -2.40725260e-02 1.41216540e+00 -9.47809458e-01
-8.16761851e-01 2.77983725e-01 3.13332200e-01 -1.04559743e+00
1.26981094e-01 8.04976463e-01 1.93085551e-01 -3.31681743e-02
7.58553386e-01 2.23021314e-01 4.11283761e-01 -5.86715996e-01
2.92352498e-01 3.82725418e-01 6.60475373e-01 -7.06836581e-01
1.19822001e+00 1.61875016e-03 6.18094385e-01 -1.17025888e+00
-4.34463888e-01 -4.71123070e-01 -6.09268308e-01 -1.80700645e-01
4.13143784e-01 -9.16524768e-01 -9.74708259e-01 5.24808824e-01
-7.75634408e-01 -5.95063210e-01 -8.30250323e-01 7.02297151e-01
-9.79377091e-01 -1.70845259e-02 -5.41185439e-01 -1.19875813e+00
-1.27748594e-01 -9.91884470e-01 1.04252923e+00 -2.23105848e-01
-2.38044322e-01 -1.44715488e+00 2.42358640e-01 -2.66399384e-01
5.08742154e-01 7.58281231e-01 8.71860862e-01 -3.13085109e-01
-2.23516822e-01 -4.23030764e-01 6.04184508e-01 5.33604801e-01
2.27485262e-02 1.65191412e-01 -6.26242578e-01 -5.77942252e-01
4.85125870e-01 3.12069058e-02 2.01710403e-01 4.40253317e-01
7.50137269e-01 -3.50707918e-01 -2.88064182e-01 2.89729685e-01
1.29712355e+00 -8.85750167e-03 1.14096470e-01 -2.71970630e-01
6.72806263e-01 5.83215535e-01 7.33807027e-01 4.96659130e-01
-1.16739191e-01 7.67893672e-01 3.91048759e-01 -2.11458817e-01
3.89568388e-01 -6.22284561e-02 5.10317922e-01 1.25693178e+00
-1.94086313e-01 7.04691233e-03 -1.02410448e+00 6.00787640e-01
-1.87324607e+00 -6.64576590e-01 -3.06598932e-01 2.32999587e+00
5.00919759e-01 -3.27869803e-01 1.92219496e-01 2.68361181e-01
6.80133104e-01 -1.67670757e-01 -9.37766671e-01 -4.68524784e-01
1.65380593e-02 2.96905786e-01 1.23678349e-01 6.38191640e-01
-1.03484404e+00 4.87593770e-01 6.64763069e+00 5.33792794e-01
-1.12955117e+00 -2.69141555e-01 2.21022457e-01 4.38184366e-02
1.99191466e-01 1.41577581e-02 -6.42298818e-01 5.39369345e-01
1.63141847e+00 -7.19121635e-01 7.92297602e-01 9.85595703e-01
7.55563855e-01 3.82921219e-01 -1.07283247e+00 6.13184094e-01
-5.35233676e-01 -1.03434110e+00 -3.28813493e-01 1.30485132e-01
8.20751607e-01 -9.38747898e-02 2.91678011e-01 4.65893209e-01
4.10763949e-01 -8.48734736e-01 3.02677155e-01 8.08615148e-01
5.13216138e-01 -4.51601475e-01 4.65956062e-01 7.11205125e-01
-1.29058647e+00 -2.74869114e-01 -1.78747118e-01 -1.84322074e-01
3.23382556e-01 4.24782097e-01 -3.57641220e-01 5.94615102e-01
2.07316384e-01 8.85835886e-01 4.30984050e-02 7.11543858e-01
2.14655533e-01 6.60826683e-01 -7.22621679e-01 1.45595610e-01
2.32365191e-01 -6.51768327e-01 9.41415966e-01 5.11122227e-01
5.14615715e-01 3.82385969e-01 3.51867050e-01 9.34960425e-01
3.96840185e-01 -1.11511044e-01 -9.13891435e-01 -2.30915502e-01
2.82861292e-01 1.09280598e+00 5.44952936e-02 -3.22038323e-01
1.70473922e-02 2.18281195e-01 -6.01341911e-02 7.20285773e-01
-9.03877497e-01 -3.22980314e-01 9.22874928e-01 3.03390026e-01
2.36088127e-01 -6.65492296e-01 2.37258166e-01 -1.01167762e+00
1.58371836e-01 -9.04786408e-01 9.70683061e-03 -8.06048036e-01
-1.19701207e+00 6.78272605e-01 2.24827439e-01 -1.92458498e+00
-1.03282666e+00 -6.38874352e-01 -4.99919027e-01 1.11818409e+00
-1.16168344e+00 -7.97933817e-01 -1.46077365e-01 4.73197609e-01
4.54803735e-01 9.75788087e-02 1.01564407e+00 4.88937050e-02
-8.75167012e-01 -7.29711577e-02 7.01435626e-01 -3.95305872e-01
3.00895810e-01 -1.29129255e+00 3.60354245e-01 8.04676950e-01
-4.24676478e-01 7.70733714e-01 1.08814430e+00 -3.32987040e-01
-2.02859116e+00 -1.43435204e+00 1.42492145e-01 -4.94550258e-01
1.13898206e+00 -1.32061169e-01 -1.36236620e+00 5.74737787e-01
-2.86405027e-01 3.33906710e-01 2.75162190e-01 -2.87292808e-01
2.48199955e-01 -5.59381358e-02 -6.53417110e-01 4.78502095e-01
6.93395436e-01 -4.60796148e-01 -6.06662929e-01 5.35052419e-01
7.76662827e-01 -7.00507283e-01 -1.50368547e+00 5.12620389e-01
9.47959796e-02 3.57958712e-02 1.14873922e+00 -9.63041008e-01
1.36433810e-01 -5.61173618e-01 2.38980412e-01 -1.86308730e+00
-3.70960534e-01 -1.35302114e+00 -7.08382845e-01 9.45235431e-01
3.23112071e-01 -1.01174033e+00 2.76840925e-01 8.45568120e-01
-3.92397642e-01 -7.83425510e-01 -7.78594553e-01 -1.22970569e+00
3.92588079e-01 -1.71447620e-01 5.54658055e-01 8.33814442e-01
-7.89555758e-02 3.29197109e-01 -5.87024748e-01 6.75433338e-01
3.40045810e-01 1.28883040e-02 9.62071359e-01 -1.33141923e+00
-5.16051829e-01 -2.13118970e-01 -1.83707580e-01 -1.02287984e+00
3.01221400e-01 -3.41882050e-01 1.58742175e-01 -1.05617547e+00
-7.76088536e-01 -3.35468471e-01 -1.20957695e-01 3.09062880e-02
-2.35926658e-01 -3.70933950e-01 2.12469567e-02 3.30999047e-01
7.78739378e-02 9.01270688e-01 1.16152036e+00 1.01187475e-01
-5.73904216e-01 4.32183653e-01 -1.71338901e-01 6.98603392e-01
6.23709142e-01 -1.31733432e-01 -7.38638341e-01 -2.88136274e-01
-7.08078891e-02 5.95223248e-01 3.94068033e-01 -1.08525324e+00
3.07175487e-01 -3.79744381e-01 -1.03561930e-01 -3.86661530e-01
8.70370686e-01 -1.11400890e+00 8.45057189e-01 3.71737480e-01
-2.02795327e-01 3.91049206e-01 4.67427522e-01 8.85100901e-01
-5.53665578e-01 9.22236815e-02 7.99165666e-01 1.27379403e-01
-3.67915124e-01 3.03602725e-01 -6.00596070e-01 1.07275680e-01
1.11145830e+00 -5.13566546e-02 -1.73972294e-01 -3.71405840e-01
-1.13631070e+00 4.27325487e-01 3.59508872e-01 2.55723983e-01
3.34806949e-01 -1.06793404e+00 -5.28940678e-01 3.21712673e-01
-2.18411162e-01 2.17990085e-01 5.27071834e-01 9.13647711e-01
-8.53702724e-02 4.96987730e-01 1.09843820e-01 -6.87318623e-01
-8.50882649e-01 8.97524297e-01 5.02147853e-01 -3.08110684e-01
-7.22190440e-01 2.04897493e-01 3.11783761e-01 -4.02928233e-01
-3.05001289e-01 -5.46267748e-01 -6.78159073e-02 -3.04651439e-01
2.24549800e-01 5.17529011e-01 7.64809772e-02 -6.25379384e-01
3.97280641e-02 8.15431178e-01 5.63950360e-01 1.47002026e-01
1.49892187e+00 2.77888924e-02 2.44169822e-03 7.58073807e-01
8.60614002e-01 -3.76560807e-01 -1.43587065e+00 -3.17561269e-01
-2.22426176e-01 -1.39718791e-02 1.91582277e-01 -2.62764633e-01
-8.11205447e-01 8.27753723e-01 1.30466729e-01 4.86288220e-01
1.17347467e+00 -4.44082499e-01 4.10426795e-01 5.32325387e-01
2.27468252e-01 -1.02884710e+00 -9.51448828e-02 7.48609126e-01
1.23674071e+00 -6.55451953e-01 1.60044841e-02 -3.70481998e-01
-5.35409331e-01 1.04206264e+00 4.73311126e-01 -6.78987861e-01
8.90105844e-01 3.21768433e-01 -1.40706539e-01 6.48111552e-02
-1.18562722e+00 2.59624988e-01 5.18726826e-01 3.90531719e-01
-4.95294034e-02 -1.33170141e-02 9.44943875e-02 5.63024163e-01
-1.86885577e-02 -1.37202308e-01 5.94139099e-01 8.56802702e-01
-4.25172485e-02 -1.04691911e+00 -5.47697842e-01 2.09882125e-01
-8.69007632e-02 2.70424426e-01 -7.92371556e-02 1.07597387e+00
-6.28934443e-01 7.03431606e-01 -8.32920372e-02 -2.35607132e-01
7.63646781e-01 -4.79883291e-02 -1.99045315e-01 -6.46085858e-01
-6.76741958e-01 2.01064095e-01 -8.95703062e-02 -4.54793155e-01
-9.25207213e-02 -5.96571267e-01 -9.43811715e-01 -1.28340423e-01
-4.40909743e-01 5.01721025e-01 5.95349729e-01 6.35681331e-01
7.90213525e-01 8.30436349e-01 8.97689223e-01 -1.08164835e+00
-1.11007714e+00 -8.58817458e-01 -5.63935995e-01 1.55212894e-01
7.00874805e-01 -1.02747011e+00 -6.79448187e-01 1.56866863e-01] | [6.479795455932617, 3.454958915710449] |
9863cf2a-2db1-4e8a-9ea9-960f927fee3f | self-supervised-learning-for-cardiac-mr-image | 1907.02757 | null | https://arxiv.org/abs/1907.02757v1 | https://arxiv.org/pdf/1907.02757v1.pdf | Self-Supervised Learning for Cardiac MR Image Segmentation by Anatomical Position Prediction | In the recent years, convolutional neural networks have transformed the field of medical image analysis due to their capacity to learn discriminative image features for a variety of classification and regression tasks. However, successfully learning these features requires a large amount of manually annotated data, which is expensive to acquire and limited by the available resources of expert image analysts. Therefore, unsupervised, weakly-supervised and self-supervised feature learning techniques receive a lot of attention, which aim to utilise the vast amount of available data, while at the same time avoid or substantially reduce the effort of manual annotation. In this paper, we propose a novel way for training a cardiac MR image segmentation network, in which features are learnt in a self-supervised manner by predicting anatomical positions. The anatomical positions serve as a supervisory signal and do not require extra manual annotation. We demonstrate that this seemingly simple task provides a strong signal for feature learning and with self-supervised learning, we achieve a high segmentation accuracy that is better than or comparable to a U-net trained from scratch, especially at a small data setting. When only five annotated subjects are available, the proposed method improves the mean Dice metric from 0.811 to 0.852 for short-axis image segmentation, compared to the baseline U-net. | ['Florian Guitton', 'Giacomo Tarroni', 'Yike Guo', 'Wenjia Bai', 'Steffen E. Petersen', 'Jinming Duan', 'Daniel Rueckert', 'Chen Chen', 'Paul M. Matthews'] | 2019-07-05 | null | null | null | null | ['small-data'] | ['computer-vision'] | [ 4.36735690e-01 3.07106674e-01 -2.68876344e-01 -7.90427983e-01
-6.59737885e-01 -4.73589987e-01 2.97140867e-01 5.61919570e-01
-8.05100381e-01 6.94687068e-01 -1.37885004e-01 -3.14283408e-02
-1.36848658e-01 -5.74961662e-01 -5.18805027e-01 -7.99701869e-01
-8.76523703e-02 4.01480466e-01 3.12500030e-01 1.28116950e-01
-5.29689230e-02 4.48623121e-01 -1.23075402e+00 -2.06725448e-01
9.74834859e-01 1.30912817e+00 3.45948815e-01 3.63842070e-01
1.10015839e-01 6.54813588e-01 -2.85878688e-01 -5.32558560e-02
1.56607807e-01 -5.12124479e-01 -8.77335370e-01 3.12445641e-01
1.86861441e-01 -1.93936393e-01 -1.48389235e-01 8.48727584e-01
6.06938243e-01 1.05188623e-01 6.62933350e-01 -6.87449396e-01
-2.27216810e-01 4.24995035e-01 -4.54025030e-01 3.55064601e-01
-1.94827989e-01 4.92384806e-02 9.45727527e-01 -4.47829008e-01
6.03665590e-01 4.09860402e-01 6.58193648e-01 4.44206774e-01
-1.32232523e+00 -3.81809533e-01 -3.31573218e-01 6.09894842e-02
-1.23899865e+00 -2.38340363e-01 9.38609004e-01 -6.43049121e-01
4.81321543e-01 1.49466127e-01 5.53699434e-01 7.62421608e-01
1.79349616e-01 8.02116692e-01 1.09725821e+00 -2.85573006e-01
2.53023803e-01 1.05682708e-01 1.82892188e-01 8.31383049e-01
-1.90942008e-02 -4.74211611e-02 -1.35253042e-01 1.96027905e-02
9.06476915e-01 2.95828938e-01 -2.92767644e-01 -7.66371250e-01
-1.38731670e+00 9.06210184e-01 8.11253667e-01 6.83263421e-01
-4.54268724e-01 -1.68931425e-01 5.83192289e-01 1.63957819e-01
5.77175736e-01 5.76171935e-01 -4.10816193e-01 -1.35747105e-01
-1.15107107e+00 -3.02894473e-01 4.32279289e-01 4.94253069e-01
7.63603568e-01 -1.07814126e-01 -5.90361357e-02 8.87843966e-01
5.54262809e-02 3.51274729e-01 8.30955744e-01 -6.11423373e-01
8.55816230e-02 6.68267846e-01 -2.45469123e-01 -9.98316526e-01
-8.10145497e-01 -6.71163380e-01 -1.32196283e+00 2.10334808e-01
6.28292501e-01 -3.30152631e-01 -8.98813426e-01 1.48625839e+00
3.46574277e-01 -2.49475151e-01 -2.36512214e-01 9.69272494e-01
5.47357678e-01 2.60532886e-01 -1.05235435e-01 -4.08397913e-01
9.27941799e-01 -9.01468694e-01 -6.64141834e-01 -1.16591245e-01
8.12830091e-01 -5.71521819e-01 1.01731575e+00 3.53411496e-01
-8.74330282e-01 -5.34704566e-01 -1.00732589e+00 2.44171068e-01
-1.64292127e-01 2.07091331e-01 7.96484828e-01 5.74675262e-01
-9.21061754e-01 9.36586201e-01 -1.13342810e+00 -1.19138651e-01
9.91045654e-01 6.12792671e-01 -6.13603234e-01 1.57478638e-02
-9.31786776e-01 8.01479936e-01 4.08795238e-01 2.54890501e-01
-6.01787269e-01 -5.09588242e-01 -9.10697401e-01 5.90552911e-02
5.06344616e-01 -4.24744904e-01 1.02804816e+00 -1.09008694e+00
-1.40485835e+00 9.24395502e-01 1.67993680e-01 -5.46962619e-01
7.44698286e-01 -1.03498034e-01 3.99790034e-02 4.45185244e-01
1.38479680e-01 6.72408164e-01 8.05121660e-01 -1.03579021e+00
-3.27661306e-01 -5.51922262e-01 -1.98877752e-01 -2.92101335e-02
-4.32931095e-01 -2.34510690e-01 -1.80748463e-01 -7.11374342e-01
2.26766914e-01 -8.67842972e-01 -5.34016252e-01 1.91699728e-01
-2.42352262e-01 -7.66816065e-02 6.94920301e-01 -5.45157015e-01
9.08036888e-01 -2.13507342e+00 1.09642982e-01 4.07998532e-01
4.96390492e-01 3.69123757e-01 3.43096644e-01 -7.96502158e-02
-6.34521246e-02 -7.74812177e-02 -7.12783873e-01 -7.47413039e-02
-4.18731958e-01 3.08156908e-01 3.07675481e-01 6.74148738e-01
1.72854826e-01 9.87910211e-01 -1.03799284e+00 -7.40259409e-01
4.28128421e-01 3.12904507e-01 -4.68264222e-01 3.24052185e-01
8.13368261e-02 1.08948350e+00 -4.85592782e-01 2.40355879e-01
2.60258824e-01 -3.88596177e-01 2.12993518e-01 -1.00383908e-01
1.50525421e-01 -2.00720310e-01 -8.35519731e-01 1.90281987e+00
-5.98994255e-01 5.77611089e-01 -1.13981657e-01 -1.61479902e+00
1.04345262e+00 4.02144104e-01 9.96079206e-01 -6.89641237e-01
4.01301891e-01 3.16454142e-01 3.04462075e-01 -5.06433725e-01
-1.90086320e-01 -3.75424981e-01 2.57756636e-02 5.60236096e-01
3.25276405e-01 -7.76314214e-02 1.96206212e-01 -7.52014816e-02
1.12740970e+00 -1.89176798e-01 4.56634879e-01 -3.52126002e-01
6.54843211e-01 -2.06240878e-01 5.25915921e-01 5.45489728e-01
-3.63010615e-01 6.96108222e-01 3.68325979e-01 -6.63618386e-01
-9.89451826e-01 -8.10856998e-01 -4.31429952e-01 6.64193213e-01
4.43227924e-02 -1.09718785e-01 -9.24306571e-01 -9.88286138e-01
-2.68157095e-01 1.06062502e-01 -7.62930334e-01 -2.43715972e-01
-5.73036969e-01 -7.12999403e-01 2.90558934e-01 8.38963628e-01
5.29889882e-01 -1.21968949e+00 -9.33784246e-01 3.32169473e-01
7.36721382e-02 -1.12000096e+00 -4.21484739e-01 4.82179403e-01
-1.20085573e+00 -1.08746231e+00 -1.08084333e+00 -8.99824560e-01
1.08497036e+00 -7.90847540e-02 9.35711741e-01 2.35528201e-01
-5.67651093e-01 1.82918936e-01 -3.76257479e-01 -1.49061352e-01
-2.07579389e-01 1.55232504e-01 -2.63730697e-02 2.38719672e-01
1.61515735e-02 -7.63157547e-01 -8.09610903e-01 3.36437225e-01
-9.88264978e-01 5.47187701e-02 8.25435877e-01 1.22603500e+00
8.30721498e-01 -1.03279963e-01 7.52367198e-01 -1.07968390e+00
2.33585954e-01 -1.60204902e-01 -3.74747306e-01 3.78979817e-02
-6.40686214e-01 8.58777836e-02 9.29778457e-01 -3.04138958e-01
-7.48568773e-01 4.77523446e-01 -2.31691286e-01 -2.57061392e-01
-3.51749539e-01 5.75337708e-01 6.25867471e-02 -3.20885062e-01
7.50782788e-01 2.90646732e-01 4.31834668e-01 -2.87289470e-01
2.71313906e-01 6.17972851e-01 5.08969188e-01 -3.11661631e-01
6.34252727e-01 5.00296056e-01 7.87543952e-02 -6.96434021e-01
-9.92399573e-01 -6.05040312e-01 -1.39698875e+00 -2.51126468e-01
9.44184601e-01 -4.19149518e-01 -4.70724136e-01 3.97569329e-01
-7.25777149e-01 -4.81866717e-01 -4.87498730e-01 6.90359294e-01
-6.78973377e-01 6.22589946e-01 -3.91528130e-01 -3.39419723e-01
-5.22993863e-01 -1.19613850e+00 7.68725336e-01 1.83310285e-01
-1.90401569e-01 -1.12118757e+00 -1.20736025e-01 3.70011896e-01
4.67984319e-01 5.64354599e-01 8.36695075e-01 -7.66535580e-01
-1.51642680e-01 -4.35803980e-01 -2.54494995e-01 5.97168624e-01
5.50281823e-01 -5.80705702e-01 -7.40624070e-01 -2.88559645e-01
8.77949297e-02 -5.01514375e-01 6.71055913e-01 6.23554409e-01
1.54235637e+00 5.83715625e-02 -9.61260200e-02 5.78183472e-01
1.11248279e+00 1.61954880e-01 3.16432089e-01 1.14607625e-01
7.81600058e-01 4.74474460e-01 4.25509453e-01 3.08515728e-01
1.22942068e-01 5.42644680e-01 3.39911520e-01 -5.86552680e-01
8.27012658e-02 1.78822026e-01 -3.84045631e-01 9.34794843e-01
-3.34341735e-01 4.56951678e-01 -1.03481305e+00 5.43151081e-01
-1.90953553e+00 -5.71347237e-01 1.07134588e-01 2.43347263e+00
1.14479339e+00 2.42081076e-01 2.08203599e-01 3.54125470e-01
5.43024898e-01 -7.39841610e-02 -5.72065294e-01 3.29517946e-02
1.95512459e-01 3.62474561e-01 3.99346679e-01 2.25697562e-01
-1.33447683e+00 6.74825191e-01 5.94026613e+00 5.74107468e-01
-1.41792035e+00 2.47748643e-01 9.10901010e-01 2.66766906e-01
2.55858630e-01 -3.93580616e-01 -7.39229992e-02 4.31590796e-01
7.72746801e-01 1.23352356e-01 1.51413351e-01 8.85897875e-01
2.33639434e-01 -1.28386557e-01 -1.13964927e+00 1.06023121e+00
1.26533210e-01 -1.24518824e+00 -3.90853286e-01 1.28267454e-02
7.30775177e-01 -5.13939038e-02 -1.36274934e-01 -4.23916019e-02
-2.25112393e-01 -1.08162606e+00 2.13124052e-01 3.74540180e-01
7.43870258e-01 -7.78475344e-01 9.91567314e-01 4.53870237e-01
-7.94945836e-01 3.87250781e-02 -3.53381157e-01 5.05368523e-02
1.83603257e-01 8.11554670e-01 -7.52514660e-01 4.32793289e-01
5.74649990e-01 6.94421530e-01 -4.89921838e-01 1.15671933e+00
-2.11859852e-01 6.82848632e-01 -1.82124451e-01 1.67361513e-01
4.65783089e-01 -5.62806800e-02 1.51068747e-01 1.02602518e+00
-1.28956214e-02 9.44739282e-02 3.94770235e-01 6.48983002e-01
-6.83751777e-02 4.93030965e-01 -4.87933159e-01 4.46104072e-02
-2.28061572e-01 1.43350959e+00 -1.27804542e+00 -3.04340363e-01
-2.57450491e-01 9.96731997e-01 3.41767132e-01 -6.79978030e-03
-6.53590202e-01 -4.76965219e-01 -8.72977450e-02 1.53821185e-01
2.36939698e-01 -2.98080921e-01 -4.21380252e-01 -1.05892384e+00
3.41485552e-02 -5.22171259e-01 3.04207772e-01 -2.07963228e-01
-1.23156071e+00 7.16901839e-01 -3.20289820e-01 -1.22591078e+00
-4.04965222e-01 -5.53025007e-01 -5.54850399e-01 4.40921307e-01
-1.50343251e+00 -8.64085674e-01 -4.61869627e-01 5.94452143e-01
4.14653629e-01 -1.38236225e-01 9.00672317e-01 3.96861494e-01
-2.88807392e-01 5.25520980e-01 2.58126706e-01 5.15157759e-01
6.95427299e-01 -1.48906755e+00 -1.58844993e-01 5.34057617e-01
3.10776204e-01 4.95643407e-01 3.99155557e-01 -3.10642689e-01
-9.62912619e-01 -8.79904091e-01 6.10813558e-01 -4.98574004e-02
5.41251421e-01 -2.46290997e-01 -8.76038790e-01 4.02975351e-01
-8.99951458e-02 7.40250289e-01 9.45817173e-01 5.13968058e-02
6.37100041e-02 -2.22325489e-01 -1.24543774e+00 2.51027465e-01
7.07327306e-01 -2.77154118e-01 -5.39702654e-01 4.82919067e-01
2.78150529e-01 -3.77786934e-01 -1.26206553e+00 3.68917733e-01
4.00752366e-01 -8.00615788e-01 8.61822486e-01 -4.24728334e-01
5.12280285e-01 -1.07650861e-01 2.59482086e-01 -1.36527550e+00
-1.42113239e-01 -3.95297170e-01 1.45315185e-01 7.90787876e-01
4.23656374e-01 -5.40196359e-01 7.83143401e-01 5.48816144e-01
-1.58938065e-01 -1.08837748e+00 -1.00111544e+00 -6.13969505e-01
1.65094897e-01 -1.90253377e-01 9.31765325e-03 1.11252630e+00
1.07893817e-01 3.77773911e-01 -3.22829515e-01 -2.39771619e-01
7.22725809e-01 1.84561923e-01 4.12844062e-01 -1.40785992e+00
-3.10672462e-01 -3.15653533e-01 -8.12995017e-01 -7.94077873e-01
1.68357626e-01 -1.13111877e+00 1.47130325e-01 -1.45855331e+00
2.67854273e-01 -6.55800402e-01 -4.69542533e-01 6.51808798e-01
-1.40796870e-01 6.05373442e-01 -7.91376233e-02 2.69272745e-01
-6.13256574e-01 5.21138787e-01 1.55643213e+00 -2.68505841e-01
-1.75163522e-01 2.14169994e-01 -5.64491272e-01 7.77130961e-01
8.08039308e-01 -4.39079136e-01 -4.02058452e-01 -2.09135920e-01
-2.20369816e-01 1.50029436e-01 2.53852934e-01 -1.02962649e+00
1.07080519e-01 2.42206648e-01 7.27131426e-01 -1.94593996e-01
-4.64236699e-02 -1.04230380e+00 -3.85779768e-01 5.55661082e-01
-4.46298718e-01 -3.52343321e-01 -8.55238959e-02 4.05037582e-01
-4.50240076e-01 -3.35453242e-01 9.44522977e-01 -2.81933635e-01
-4.86060590e-01 5.37850916e-01 -1.46540150e-01 -4.06041071e-02
1.12311912e+00 -2.19570741e-01 2.61379898e-01 -3.15285951e-01
-1.16299403e+00 6.93238080e-02 1.07033856e-01 1.15495086e-01
4.39133883e-01 -1.14366460e+00 -4.48002428e-01 1.58136591e-01
1.24416426e-02 3.01862299e-01 3.44205916e-01 1.25768983e+00
-5.08730233e-01 3.77419770e-01 -4.33360130e-01 -1.01599908e+00
-1.01700628e+00 4.57952648e-01 1.87617391e-01 -4.79301304e-01
-1.05630636e+00 4.86117125e-01 1.29459873e-01 -4.72766906e-01
3.13343257e-02 -2.41034225e-01 -3.77945840e-01 -1.42807909e-03
4.45592284e-01 -2.91697718e-02 2.31964320e-01 -6.98327959e-01
-2.09455833e-01 5.76646805e-01 -1.84917197e-01 3.34414303e-01
1.64564395e+00 3.93644385e-02 4.87306826e-02 3.72672319e-01
1.35371399e+00 -3.80076975e-01 -1.35111392e+00 -4.26330954e-01
1.75752401e-01 -4.03708667e-01 2.78546989e-01 -7.48005092e-01
-1.45116603e+00 1.12157726e+00 8.36005390e-01 6.36164770e-02
1.11755908e+00 6.40957430e-02 8.57028604e-01 4.56133246e-01
4.07739431e-01 -1.02068245e+00 2.16715351e-01 1.67594552e-01
5.57693064e-01 -1.68361700e+00 -9.79333445e-02 -3.37424964e-01
-8.04060757e-01 1.25915384e+00 3.88714135e-01 -2.75640130e-01
6.36943519e-01 2.13177338e-01 2.46225566e-01 -2.01484054e-01
-1.68237798e-02 -2.99245626e-01 4.01334316e-01 6.25911236e-01
6.00996971e-01 -7.49355853e-02 -4.94381726e-01 6.18856370e-01
-5.63207008e-02 1.10281847e-01 3.02837610e-01 8.94805312e-01
-4.20670509e-01 -1.02933776e+00 3.00368387e-02 7.49049962e-01
-7.31358767e-01 1.67251721e-01 -8.16172063e-02 6.81356728e-01
-1.34625182e-01 5.98985791e-01 -2.86678784e-02 -7.96660930e-02
7.89466873e-02 -3.07818204e-02 4.53695327e-01 -6.49725795e-01
-5.23046374e-01 1.47689760e-01 -2.85847396e-01 -5.58966935e-01
-5.89443564e-01 -5.13157248e-01 -1.35583353e+00 3.50261092e-01
-4.80704159e-01 1.50387511e-01 4.83885705e-01 1.24166846e+00
8.51208419e-02 5.97577333e-01 9.40456629e-01 -9.08743560e-01
-5.32705247e-01 -9.66396034e-01 -6.32839203e-01 6.50504231e-01
1.61778435e-01 -7.90408731e-01 -6.55238703e-02 1.78143218e-01] | [14.714964866638184, -2.2536938190460205] |
811e57ba-e8dc-402b-8939-cf493589e4c6 | delving-globally-into-texture-and-structure | 2209.08217 | null | https://arxiv.org/abs/2209.08217v1 | https://arxiv.org/pdf/2209.08217v1.pdf | Delving Globally into Texture and Structure for Image Inpainting | Image inpainting has achieved remarkable progress and inspired abundant methods, where the critical bottleneck is identified as how to fulfill the high-frequency structure and low-frequency texture information on the masked regions with semantics. To this end, deep models exhibit powerful superiority to capture them, yet constrained on the local spatial regions. In this paper, we delve globally into texture and structure information to well capture the semantics for image inpainting. As opposed to the existing arts trapped on the independent local patches, the texture information of each patch is reconstructed from all other patches across the whole image, to match the coarsely filled information, specially the structure information over the masked regions. Unlike the current decoder-only transformer within the pixel level for image inpainting, our model adopts the transformer pipeline paired with both encoder and decoder. On one hand, the encoder captures the texture semantic correlations of all patches across image via self-attention module. On the other hand, an adaptive patch vocabulary is dynamically established in the decoder for the filled patches over the masked regions. Building on this, a structure-texture matching attention module anchored on the known regions comes up to marry the best of these two worlds for progressive inpainting via a probabilistic diffusion process. Our model is orthogonal to the fashionable arts, such as Convolutional Neural Networks (CNNs), Attention and Transformer model, from the perspective of texture and structure information for image inpainting. The extensive experiments over the benchmarks validate its superiority. Our code is available at https://github.com/htyjers/DGTS-Inpainting. | ['Yong Rui', 'Meng Wang', 'Yang Wang', 'Haipeng Liu'] | 2022-09-17 | null | null | null | null | ['image-inpainting'] | ['computer-vision'] | [ 1.32393926e-01 2.17283860e-01 -2.39693463e-01 -1.67435750e-01
-8.30406845e-01 -1.22630484e-01 3.03140461e-01 -1.61289185e-01
2.34550610e-01 5.48012197e-01 4.24187124e-01 4.39582437e-01
-5.14147896e-03 -1.03977132e+00 -1.10094476e+00 -8.76872659e-01
3.26754838e-01 2.41002262e-01 2.89627701e-01 -4.91185129e-01
1.96624219e-01 2.18795091e-01 -1.55299282e+00 6.88953817e-01
7.72392631e-01 1.40423203e+00 7.44681180e-01 1.75496638e-01
-3.67264777e-01 1.05375731e+00 6.76423386e-02 -3.01792890e-01
1.72855139e-01 -4.03232187e-01 -6.28883779e-01 2.47742400e-01
6.01363838e-01 -4.59446847e-01 -6.92012429e-01 1.23244715e+00
1.78927928e-01 -1.48477480e-01 2.25051686e-01 -8.20942938e-01
-1.20878088e+00 5.44204831e-01 -8.50471139e-01 -6.34828284e-02
2.22547010e-01 3.64410460e-01 9.71638620e-01 -1.09299970e+00
6.94607198e-01 1.14097774e+00 6.08023882e-01 2.35661775e-01
-8.90575826e-01 -3.51788938e-01 3.30884248e-01 2.33392224e-01
-1.42084360e+00 -3.08250308e-01 1.21001303e+00 -2.21020356e-01
6.18617415e-01 2.39904329e-01 7.86879897e-01 1.08013332e+00
5.23251474e-01 9.17947471e-01 1.10673630e+00 -6.40554950e-02
5.56179043e-03 -1.99568272e-01 -2.94214994e-01 6.62486911e-01
-3.31270933e-01 7.75410831e-02 -7.64331460e-01 2.52327770e-01
1.22540700e+00 4.48200375e-01 -3.47469807e-01 -1.12026498e-01
-1.06992459e+00 5.08272171e-01 6.87840283e-01 3.64094675e-01
-6.49914145e-01 2.58203000e-01 6.47578686e-02 2.56307900e-01
8.38356376e-01 1.93473816e-01 -5.79982877e-01 1.43846825e-01
-1.14358556e+00 1.27630949e-01 3.97427678e-01 1.19848764e+00
1.33620739e+00 -5.83229819e-03 -4.43642318e-01 8.23074222e-01
2.94204324e-01 3.57135147e-01 4.20273542e-01 -9.09169137e-01
4.64991331e-01 7.54578650e-01 1.26156106e-03 -1.32224905e+00
1.53001010e-01 -6.09480798e-01 -1.07240593e+00 8.20816495e-03
1.41006149e-02 4.67846632e-01 -9.35443640e-01 1.70607388e+00
3.48617852e-01 4.64309961e-01 -4.18432057e-01 1.00879455e+00
7.66978502e-01 7.91656315e-01 -7.58427009e-02 -1.86613686e-02
1.53284562e+00 -1.31011426e+00 -7.48808444e-01 -3.88526082e-01
-4.58443724e-02 -1.03458178e+00 1.18578660e+00 2.06432104e-01
-1.30403674e+00 -8.84621322e-01 -9.33192849e-01 -6.04490697e-01
-1.26268595e-01 -9.33005363e-02 5.13776124e-01 3.98988500e-02
-1.22506130e+00 6.73068285e-01 -7.63390124e-01 -1.80372387e-01
7.51548946e-01 4.79678772e-02 -3.30136955e-01 -4.57136095e-01
-1.14036536e+00 6.01180911e-01 8.64635259e-02 4.24812138e-01
-1.07430124e+00 -8.26858521e-01 -8.46827745e-01 7.64101669e-02
2.25482941e-01 -8.42107892e-01 7.33023107e-01 -1.54983032e+00
-1.60960221e+00 8.04869652e-01 -2.22412139e-01 -2.53726542e-01
6.05712533e-01 -3.41987997e-01 -1.55664399e-01 1.69211209e-01
4.26054686e-01 8.26423347e-01 1.06554973e+00 -1.31527877e+00
-4.76170897e-01 -2.62531608e-01 -9.35072452e-02 2.59198397e-01
-2.07617894e-01 -2.07153112e-01 -9.56866801e-01 -1.11480618e+00
2.26489827e-01 -5.19996643e-01 -8.24482888e-02 3.78586262e-01
-2.70963848e-01 1.80487573e-01 9.92560685e-01 -1.01104701e+00
1.15799510e+00 -2.32983208e+00 4.82990831e-01 -1.47883385e-01
4.19451863e-01 -1.62084475e-01 -2.94376314e-01 5.29121161e-01
-7.18997642e-02 -2.14419380e-01 -5.56792021e-01 -5.20112872e-01
-7.11891726e-02 3.96695375e-01 -6.35225058e-01 4.12308455e-01
5.55135250e-01 1.30028510e+00 -8.24046075e-01 -3.85615796e-01
2.55792707e-01 5.92837691e-01 -6.68516755e-01 3.95379484e-01
-5.34955561e-01 6.65501356e-01 -6.75933599e-01 8.10398459e-01
1.01675487e+00 -3.07526261e-01 -1.70137361e-01 -5.22139728e-01
-1.71012446e-01 1.11271732e-01 -8.89057398e-01 2.36948276e+00
-4.14764017e-01 3.50979090e-01 3.31876040e-01 -1.03356969e+00
9.06338751e-01 1.95284113e-01 6.29305303e-01 -9.55885351e-01
3.63603272e-02 1.93501383e-01 -5.74428618e-01 -5.35546541e-01
4.92339909e-01 -9.82462689e-02 6.57233745e-02 1.93619236e-01
8.55093151e-02 -1.19300440e-01 -3.81033480e-01 1.04486912e-01
8.57634008e-01 6.00079358e-01 -1.29123658e-01 -2.08960220e-01
4.16781664e-01 -1.48282111e-01 4.69085783e-01 4.12956297e-01
1.20453708e-01 1.22100353e+00 2.21261576e-01 -7.62231469e-01
-1.19953239e+00 -1.10658622e+00 1.18531857e-03 9.03772593e-01
5.52488327e-01 -3.50329369e-01 -9.05696929e-01 -3.20481539e-01
-7.87063241e-02 1.51654974e-01 -1.04726362e+00 -2.20153883e-01
-5.36779702e-01 -4.30351645e-01 7.46356323e-02 3.43529016e-01
1.00692499e+00 -1.20568252e+00 -9.08192098e-02 4.12006378e-01
-5.09737492e-01 -9.16638672e-01 -7.88008034e-01 1.50902607e-02
-6.45928264e-01 -8.42912078e-01 -7.49951124e-01 -8.89458358e-01
5.26919127e-01 2.85649568e-01 1.14537823e+00 2.15396225e-01
-2.78328568e-01 2.63762027e-02 -4.78625417e-01 5.01560271e-02
-1.00964464e-01 -4.41760011e-02 -5.17489135e-01 6.08842254e-01
-2.40590926e-02 -1.02337480e+00 -1.07898140e+00 4.00614411e-01
-1.20771539e+00 4.24946815e-01 9.34741735e-01 8.52457702e-01
1.03820252e+00 1.59201965e-01 1.97670668e-01 -7.71035373e-01
4.60602269e-02 -8.13720882e-01 -1.73744336e-01 2.24983126e-01
-2.56794780e-01 -4.29787412e-02 5.67286909e-01 -4.55349863e-01
-1.09046578e+00 4.75388393e-02 -3.26228380e-01 -7.89895296e-01
-3.39823775e-02 4.52262700e-01 -5.56814313e-01 -5.55219427e-02
2.64669538e-01 7.50348687e-01 4.47745174e-02 -8.39065373e-01
3.42929453e-01 2.96168178e-01 4.81951952e-01 -7.52717972e-01
7.01115191e-01 7.97253668e-01 -2.93620586e-01 -3.63854706e-01
-9.93367672e-01 -1.88484939e-03 -5.59963763e-01 -6.62531778e-02
9.26032901e-01 -1.08743572e+00 -1.97540849e-01 6.52871013e-01
-1.06432211e+00 -4.11621094e-01 -6.18932307e-01 -4.56485674e-02
-7.48599112e-01 4.40016180e-01 -9.02510166e-01 -2.33504593e-01
-2.64193773e-01 -1.22604430e+00 1.53327799e+00 1.31396890e-01
1.68080442e-02 -7.80049145e-01 -1.09353952e-01 4.39664871e-01
6.82412982e-01 1.72331616e-01 8.32931995e-01 1.96928576e-01
-1.17891169e+00 7.25321993e-02 -4.75616962e-01 3.23315710e-01
2.68385202e-01 -3.55027378e-01 -1.05760920e+00 -1.57060578e-01
4.11042005e-01 -1.92643836e-01 1.06861687e+00 4.79054391e-01
1.43813896e+00 -4.24209028e-01 -1.41650468e-01 8.80908728e-01
1.58233774e+00 -2.09229410e-01 1.03882229e+00 2.26915404e-01
1.08933687e+00 5.95514655e-01 4.40398753e-01 3.23443472e-01
4.87258732e-01 6.85820639e-01 5.89769840e-01 -3.98451954e-01
-6.95669055e-01 -7.48577654e-01 2.78264374e-01 8.18765879e-01
7.58784488e-02 -2.10724652e-01 -4.96857643e-01 4.17114437e-01
-1.98361361e+00 -9.19899106e-01 1.16816223e-01 1.84626484e+00
1.06808424e+00 -1.36670351e-01 -3.71291727e-01 -2.09667966e-01
6.85070932e-01 4.70082611e-01 -5.16954958e-01 1.25557706e-01
-4.51120257e-01 3.86643291e-01 3.28889012e-01 6.39916599e-01
-8.75617325e-01 1.12993622e+00 5.22153378e+00 1.46004558e+00
-1.12570679e+00 3.83215338e-01 1.13295722e+00 1.38152819e-02
-6.13676965e-01 2.27139011e-01 -4.45741773e-01 7.76711941e-01
3.83207738e-01 4.76758093e-01 7.22170055e-01 4.88348603e-01
1.81508049e-01 -1.05891533e-01 -8.72411489e-01 9.61596847e-01
-5.60640506e-02 -1.56311989e+00 3.39301318e-01 -1.14602651e-02
9.50447202e-01 6.18242286e-02 3.88881743e-01 -6.51222887e-03
1.33944988e-01 -1.14253914e+00 1.18154216e+00 8.63659322e-01
1.01726639e+00 -5.35770178e-01 5.52202284e-01 1.66497707e-01
-1.32975018e+00 -3.22821247e-03 -4.57249343e-01 -5.41221239e-02
-5.84482811e-02 8.37919831e-01 -7.45397713e-03 8.91906202e-01
8.73021364e-01 1.22350872e+00 -3.78991872e-01 5.19026518e-01
-2.27691963e-01 4.10370857e-01 -1.72391206e-01 6.66052639e-01
3.12071502e-01 -3.38424414e-01 2.55927414e-01 8.48469853e-01
3.71316373e-01 -6.74231648e-02 1.52229279e-01 1.35377669e+00
-3.08265369e-02 -1.69379357e-02 -4.14883554e-01 3.18649799e-01
2.55100638e-01 1.28449631e+00 -5.78290045e-01 -3.62903833e-01
-4.32157904e-01 1.32505512e+00 5.65620542e-01 3.84496063e-01
-9.62225258e-01 8.82545412e-02 5.42451859e-01 5.05559325e-01
4.56441373e-01 -7.53696337e-02 -5.01383126e-01 -1.19303644e+00
3.15369964e-01 -8.73603344e-01 3.82530913e-02 -9.94128168e-01
-1.56105483e+00 6.41448975e-01 -3.78578663e-01 -1.49089909e+00
5.01593590e-01 -2.40224868e-01 -5.47396839e-01 1.13838041e+00
-1.74155831e+00 -1.76326501e+00 -3.68428141e-01 8.04248214e-01
8.89615178e-01 1.22233883e-01 5.59620142e-01 5.19864857e-01
-5.20624280e-01 3.14978302e-01 -3.05274911e-02 -3.16722170e-02
6.33158863e-01 -8.57339919e-01 3.03803325e-01 7.09157705e-01
-2.59159263e-02 4.01981413e-01 4.05630797e-01 -8.64813626e-01
-1.66007376e+00 -1.28766310e+00 6.02054775e-01 -3.30472171e-01
6.65152371e-01 -3.61581713e-01 -1.05242419e+00 5.97862005e-01
4.41714823e-01 2.59257644e-01 1.41260698e-02 -3.84524256e-01
-5.35245061e-01 -3.27902079e-01 -8.68358254e-01 5.23773074e-01
1.22029018e+00 -6.93771064e-01 -2.29015544e-01 3.22392464e-01
8.40978384e-01 -5.57055235e-01 -9.11675334e-01 3.06313276e-01
3.72106045e-01 -1.15506959e+00 1.01296675e+00 -7.11613372e-02
1.03206432e+00 -5.80750167e-01 -3.87773216e-01 -9.68756795e-01
-5.96615374e-01 -7.67152488e-01 9.31440964e-02 1.39815569e+00
1.27164930e-01 -4.74510282e-01 7.27869809e-01 3.56240958e-01
-4.90203291e-01 -1.01486540e+00 -9.51550901e-01 -2.47597516e-01
-1.33735901e-02 -4.14465338e-01 8.37680221e-01 9.45762157e-01
-5.26374817e-01 5.50534874e-02 -6.79182827e-01 1.67895734e-01
4.94443625e-01 5.08126915e-01 4.48486060e-01 -6.62913740e-01
-5.60753167e-01 -4.61436152e-01 -1.55107662e-01 -1.40615439e+00
3.61618362e-02 -8.52412820e-01 1.27112046e-02 -1.38608432e+00
4.04942840e-01 -4.93048877e-01 -3.84662271e-01 5.07035971e-01
-2.73721159e-01 5.56886196e-01 -7.24737644e-02 5.46980083e-01
-5.54723263e-01 9.36152220e-01 1.65978849e+00 -3.57039273e-01
1.49441540e-01 -4.05749083e-01 -9.34683502e-01 5.38330615e-01
6.24460876e-01 -4.50270981e-01 -3.66586149e-01 -8.17667723e-01
2.73530424e-01 1.68698162e-01 7.22041190e-01 -8.68538439e-01
3.35631698e-01 -2.45321184e-01 6.02164745e-01 -4.24789906e-01
3.70695174e-01 -9.55353439e-01 5.72257459e-01 9.19910371e-02
-7.00303912e-02 -6.97152764e-02 1.24227382e-01 7.51385987e-01
-6.82878673e-01 3.16549033e-01 9.39593375e-01 -3.25962722e-01
-7.47330010e-01 9.49342132e-01 8.28121826e-02 5.56706861e-02
8.43420506e-01 -3.02230060e-01 -1.97991729e-01 -2.28438437e-01
-5.87022424e-01 -3.73764038e-02 7.77656257e-01 5.16500950e-01
7.82887518e-01 -1.34617937e+00 -6.70745671e-01 6.07872546e-01
5.55911884e-02 3.68443966e-01 8.66232574e-01 9.17189419e-01
-5.39006591e-01 -8.82268623e-02 -4.74719673e-01 -5.65374613e-01
-5.81447959e-01 7.76142716e-01 3.68724406e-01 -3.00191641e-01
-7.20815659e-01 9.28829789e-01 8.84888828e-01 4.88360450e-02
-5.67571148e-02 -3.84001791e-01 1.71237677e-01 -1.52583212e-01
5.10950267e-01 -2.16946051e-01 -4.14713323e-02 -6.46957099e-01
-3.79836224e-02 1.00244701e+00 -1.86231539e-01 1.65197834e-01
1.55222797e+00 -4.37154144e-01 -6.57314181e-01 2.27742895e-01
1.22683859e+00 1.32334515e-01 -1.77652538e+00 -4.19418514e-01
-3.80457968e-01 -5.67913890e-01 1.82267830e-01 -6.53273463e-01
-1.50702488e+00 7.50768602e-01 3.30083221e-01 -7.48658478e-02
1.41675735e+00 8.80460888e-02 1.07904112e+00 -4.87889796e-01
4.81331706e-01 -8.08560431e-01 1.71137393e-01 2.34576762e-01
1.13749623e+00 -1.10207677e+00 -2.03658357e-01 -5.84805131e-01
-5.07142246e-01 9.39359307e-01 5.89396000e-01 -6.23865485e-01
8.09236884e-01 3.82828802e-01 -1.83895230e-01 -3.75748277e-01
-7.84022748e-01 6.69220611e-02 2.61085004e-01 4.54970688e-01
3.12874973e-01 2.71137618e-03 1.15150675e-01 6.54925406e-01
7.65738040e-02 2.02279259e-02 -7.27994740e-02 6.21678650e-01
-3.49154472e-01 -1.12336147e+00 -3.02006185e-01 1.80067539e-01
-2.96412826e-01 -4.84185874e-01 -2.68367350e-01 5.31138599e-01
4.86355633e-01 6.48961484e-01 1.40564337e-01 -3.50699246e-01
1.99660450e-01 -2.65322596e-01 3.62655371e-01 -5.37508547e-01
-5.43612182e-01 2.83129275e-01 -4.06291068e-01 -9.46969092e-01
-3.25618029e-01 -3.56264114e-01 -8.94667625e-01 -3.37573498e-01
1.85767189e-01 -1.28472656e-01 8.76584873e-02 7.32402384e-01
6.20407701e-01 5.42060852e-01 7.20631421e-01 -9.89263475e-01
-6.09014407e-02 -9.44025099e-01 -8.09983790e-01 4.61294770e-01
3.72091740e-01 -4.98242617e-01 -2.15246513e-01 9.96287093e-02] | [11.260812759399414, -1.2752388715744019] |
0b95804f-5532-4b04-a169-6fc48d681f28 | uav-trajectory-planning-for-data-collection | null | null | https://ieeexplore.ieee.org/document/8842600 | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=8842600 | UAV Trajectory Planning for Data Collection from Time-Constrained IoT Devices | The global evolution of wireless technologies and intelligent sensing devices are transforming the realization of smart cities. Among the myriad of use cases, there is a need to support applications whereby low-resource IoT devices need to upload their sensor data to a remote control centre by target hard deadlines; otherwise, the data becomes outdated and loses its value, for example, in emergency or industrial control scenarios. In addition, the IoT devices can be either located in remote areas with limited wireless coverage or in dense areas with relatively low quality of service. This motivates the utilization of UAVs to offload traffic from existing wireless networks by collecting data from time-constrained IoT devices with performance guarantees. To this end, we jointly optimize the trajectory of a UAV and the radio resource allocation to maximize the number of served IoT devices, where each device has its own target data upload deadline. The formulated optimization problem is shown to be mixed integer non-convex and generally NP-hard. To solve it, we first propose the high-complexity branch, reduce and bound (BRB) algorithm to find the global optimal solution for relatively small scale scenarios. Then, we develop an effective sub-optimal algorithm based on successive convex approximation in order to obtain results for larger networks. Next, we propose an extension algorithm to further minimize the UAV's flight distance for cases where the initial and final UAV locations are known a priori. We demonstrate the favourable characteristics of the algorithms via extensive simulations and analysis as a function of various system parameters, with benchmarking against two greedy algorithms based on distance and deadline metrics. | ['Ali Ghrayeb', 'Tri Minh Nguyen', 'Chadi M. Assi', 'Sanaa Sharafeddine', 'Moataz Samir'] | 2019-09-17 | null | null | null | ieee-transactions-on-wireless-communications-1 | ['trajectory-planning'] | ['robots'] | [ 2.83965886e-01 2.47338280e-01 -1.64300308e-01 2.31570508e-02
-3.10080796e-01 -8.22420239e-01 -1.37923220e-02 1.45408258e-01
-2.63796002e-01 9.94752526e-01 -5.32005250e-01 -4.94664252e-01
-9.31520879e-01 -1.07438290e+00 -3.99567693e-01 -1.05528462e+00
-3.23730826e-01 5.41458070e-01 2.34594211e-01 -4.84678000e-02
-3.43746781e-01 6.21600032e-01 -1.22717261e+00 -8.06200743e-01
9.08034742e-01 1.44215989e+00 6.02061689e-01 5.41634858e-01
4.62937891e-01 -1.23291895e-01 -7.20598578e-01 2.38813367e-02
6.29851937e-01 9.85119194e-02 -4.67454970e-01 2.98180133e-01
-6.64572299e-01 -3.95282984e-01 2.50536621e-01 8.78020763e-01
5.31200826e-01 1.28974468e-01 3.64527345e-01 -1.71069610e+00
-6.84557557e-02 2.67607242e-01 -6.45604193e-01 6.26143217e-02
8.75900015e-02 -4.36641909e-02 5.69723845e-01 -2.03421749e-02
3.85033637e-01 7.46692300e-01 3.64418149e-01 2.55851567e-01
-8.31538737e-01 -4.99852985e-01 3.62412661e-01 -6.50050044e-02
-1.15286052e+00 -4.30557013e-01 5.70138514e-01 -8.14141426e-03
5.53789973e-01 4.77472007e-01 6.77616656e-01 3.47582638e-01
-3.50049064e-02 1.29450783e-01 5.35898030e-01 -2.19936833e-01
4.69905913e-01 -1.28952086e-01 -4.65247303e-01 2.98283815e-01
7.57573009e-01 -3.51978578e-02 3.71723622e-01 -7.81170577e-02
2.91012257e-01 2.14989647e-01 -3.07606876e-01 -2.41721600e-01
-1.22593141e+00 6.04759336e-01 5.52162409e-01 3.04983407e-01
-9.67757285e-01 1.04935460e-01 -3.07711083e-02 3.02411050e-01
4.37328547e-01 1.26185253e-01 -7.21970797e-01 7.73271024e-02
-6.98643804e-01 -1.44347861e-01 8.43478024e-01 1.22036386e+00
3.73573601e-01 -1.53777108e-01 1.40680090e-01 3.09661925e-01
3.91932994e-01 9.91667092e-01 -1.61558956e-01 -8.66663337e-01
6.29285574e-01 3.97894591e-01 7.83407867e-01 -1.03296018e+00
-7.10700989e-01 -4.16496962e-01 -1.13687110e+00 -1.17698215e-01
3.13305736e-01 -6.93902254e-01 -4.10927027e-01 1.59332907e+00
8.38305056e-01 2.16392521e-02 1.75155059e-01 9.03210461e-01
2.81947218e-02 9.40981686e-01 -7.16757774e-02 -1.08554804e+00
1.08115304e+00 -3.91682595e-01 -4.84147608e-01 -6.17656037e-02
4.20616597e-01 -5.84645092e-01 4.74843591e-01 1.70886993e-01
-1.05183601e+00 -1.76832192e-02 -8.46231282e-01 6.14508569e-01
-4.30300593e-01 3.16389114e-01 2.71717519e-01 7.71221101e-01
-8.23789001e-01 1.17354579e-01 -7.49653697e-01 -6.98130906e-01
3.30506146e-01 7.06433594e-01 3.94062191e-01 -1.94965065e-01
-7.58982182e-01 2.27365240e-01 3.72505248e-01 1.30263820e-01
-6.81395888e-01 -6.04137838e-01 -4.11233544e-01 4.67217900e-02
7.20419109e-01 -9.34964418e-01 9.31293666e-01 -4.69790280e-01
-1.39205444e+00 1.31398961e-01 1.26295045e-01 -4.69384313e-01
3.63419235e-01 4.91383970e-01 -3.17240536e-01 1.84818864e-01
2.04012588e-01 4.40795600e-01 7.42328346e-01 -9.63395774e-01
-1.10187340e+00 -3.26710701e-01 5.74771285e-01 2.07281075e-02
-5.42092979e-01 -1.50665283e-01 -8.16586912e-02 -2.48545364e-01
-7.71685243e-02 -1.15068197e+00 -6.39431179e-01 -5.29040731e-02
-3.44999611e-01 -2.86606282e-01 1.29511344e+00 -1.32721558e-01
9.84075189e-01 -2.08706737e+00 2.98451800e-02 2.51723439e-01
-3.52767333e-02 1.28357410e-01 -6.32201433e-02 4.02285814e-01
4.83720273e-01 1.46181479e-01 -4.96827103e-02 -1.40100732e-01
-7.95921236e-02 4.36367810e-01 -1.18140317e-01 4.52397585e-01
-1.96704000e-01 4.43391532e-01 -1.11456978e+00 -1.59491718e-01
2.79410511e-01 3.03771526e-01 -2.65209705e-01 1.79066770e-02
-3.01948100e-01 5.56521952e-01 -1.10014391e+00 7.87133396e-01
8.31182897e-01 -3.86770904e-01 2.14115605e-01 -1.91711575e-01
-4.63555843e-01 -1.86268985e-01 -1.09000397e+00 1.13864183e+00
-6.87521279e-01 2.32752994e-01 8.78342271e-01 -1.32720017e+00
5.23984492e-01 3.61257493e-01 1.31218076e+00 -3.33954245e-01
4.45660323e-01 2.07274184e-01 -3.60272765e-01 -4.51150656e-01
2.73804694e-01 1.26558781e-01 -2.86663592e-01 4.41307962e-01
-8.54544282e-01 -1.02644578e-01 3.97757679e-01 5.40377572e-02
1.31498933e+00 -4.00542527e-01 3.26846123e-01 -2.12486863e-01
4.66465116e-01 8.35601836e-02 4.53620613e-01 4.56802815e-01
-1.95549965e-01 -2.80742943e-01 1.39031466e-02 -2.47016042e-01
-6.94181800e-01 -8.15047383e-01 1.39505357e-01 6.88374102e-01
5.67133367e-01 9.09648389e-02 -5.46841085e-01 -5.35721242e-01
-1.62609354e-01 5.02980113e-01 -7.13528991e-02 1.30465537e-01
-2.07237646e-01 -8.89439702e-01 -2.53037691e-01 -1.12084255e-01
5.64327896e-01 -4.45501953e-01 -1.40879571e+00 4.61757779e-01
-4.26887795e-02 -1.50209153e+00 -2.48903468e-01 1.43989846e-01
-5.95755696e-01 -1.20864546e+00 -6.66315198e-01 -4.91387159e-01
9.99296367e-01 8.92801106e-01 6.51713312e-01 -3.78207490e-02
-9.37958658e-02 7.77593017e-01 -5.48106015e-01 -7.68541276e-01
1.99298728e-02 3.39919746e-01 2.28866801e-01 -3.55239697e-02
-1.22776330e-01 -6.43697023e-01 -7.83891261e-01 6.73339367e-01
-8.54422688e-01 -2.12195128e-01 4.35440511e-01 2.21444502e-01
8.87442708e-01 1.12541199e+00 8.44254553e-01 -1.17013983e-01
6.24061942e-01 -8.72149646e-01 -1.27307951e+00 1.89881280e-01
-3.01011890e-01 -3.66245836e-01 7.81804025e-01 -4.48897004e-01
-5.00669181e-01 4.81408000e-01 4.01318222e-01 -1.97228909e-01
7.32862875e-02 2.88066983e-01 -3.91273290e-01 -2.20702156e-01
-1.69358522e-01 -1.72538161e-01 -2.17573851e-01 -3.25101521e-03
1.47959396e-01 8.31588626e-01 1.08792618e-01 -3.46830398e-01
1.16238105e+00 7.92594135e-01 4.64090526e-01 -9.87358332e-01
-8.41121435e-01 -3.88617873e-01 -1.92917094e-01 -5.51178336e-01
6.54181600e-01 -7.22785115e-01 -1.09110296e+00 3.87265608e-02
-1.04247761e+00 -2.95967430e-01 -3.89032274e-01 4.69600737e-01
-3.77194077e-01 5.57993501e-02 2.33578265e-01 -1.09214723e+00
-2.76502311e-01 -1.16935635e+00 1.07859731e+00 3.43982399e-01
3.49288136e-01 -7.69128919e-01 -3.85474980e-01 1.48396134e-01
5.68243325e-01 6.46231651e-01 4.56622928e-01 -1.52463093e-01
-8.40782762e-01 -2.27180943e-01 -2.93094844e-01 -1.49091706e-01
4.25732732e-01 -2.82306522e-01 -1.77817091e-01 -6.88655734e-01
-2.11612448e-01 1.49978369e-01 6.65672719e-02 5.94659805e-01
1.11267865e+00 -6.95794523e-01 -9.74460542e-01 4.52190161e-01
1.51787233e+00 3.66909027e-01 6.83691800e-02 -2.07469612e-02
2.25194171e-01 6.35230541e-01 1.11894095e+00 9.55809951e-01
5.56396782e-01 6.89524770e-01 1.30594194e+00 1.51201308e-01
4.60192382e-01 2.52598226e-01 2.12330386e-01 5.23663998e-01
-8.64390656e-02 -9.25794661e-01 -5.12205660e-01 7.45597601e-01
-1.89122319e+00 -5.71790576e-01 -6.27756715e-02 2.42789698e+00
1.98211726e-02 -7.12404326e-02 4.02761549e-01 3.33939791e-01
7.46828198e-01 -6.88212216e-02 -8.10257316e-01 -4.84655648e-02
2.47994795e-01 -3.04059446e-01 1.15721345e+00 1.04018517e-01
-9.04083908e-01 5.08724630e-01 4.93831110e+00 6.54144406e-01
-9.83022571e-01 3.74257952e-01 3.72773290e-01 -2.63063878e-01
-6.04659133e-02 -2.02562883e-01 -5.06969154e-01 7.38634050e-01
1.06770670e+00 -3.88981968e-01 7.79367745e-01 7.13459849e-01
9.57673371e-01 -3.14736247e-01 -5.23762584e-01 8.05001259e-01
-6.21778250e-01 -9.70577121e-01 -6.96338475e-01 5.87157547e-01
8.39070320e-01 1.11413755e-01 -1.68744043e-01 -4.18148816e-01
4.50738192e-01 -2.49281392e-01 5.08240700e-01 2.50119090e-01
7.49544621e-01 -9.50770497e-01 4.86291736e-01 6.09557450e-01
-1.66155684e+00 -6.79173410e-01 -5.66283941e-01 -2.13364273e-01
7.24359155e-01 1.14753973e+00 -7.08005667e-01 8.17234337e-01
7.36085713e-01 4.15650964e-01 1.82773173e-01 1.23506129e+00
-7.20914155e-02 3.21141005e-01 -9.81443584e-01 -4.57525313e-01
2.65995234e-01 -2.75302529e-01 9.07979369e-01 3.89255255e-01
9.03855383e-01 6.09168649e-01 5.26950657e-01 4.80299026e-01
-5.98206930e-03 -2.25342289e-02 -7.70236909e-01 2.71732658e-02
9.19836223e-01 1.45470822e+00 -1.04062712e+00 1.37262464e-01
-3.22987467e-01 5.75490773e-01 -3.56820673e-01 4.41046119e-01
-9.61205900e-01 -4.28356707e-01 7.85577357e-01 1.44017428e-01
4.53730583e-01 -5.27925611e-01 -1.04966266e-02 -5.00206470e-01
8.14838186e-02 -6.51413798e-02 2.65019089e-01 -6.38888478e-01
-8.75724256e-01 3.80528390e-01 8.82857442e-02 -1.38084912e+00
-1.80853620e-01 -2.77819961e-01 -5.27695835e-01 1.76690906e-01
-1.67836249e+00 -8.87191653e-01 -4.57625180e-01 7.30101287e-01
3.98932904e-01 1.39014840e-01 4.03223038e-01 5.19971013e-01
-5.99760056e-01 2.38334201e-03 2.88025409e-01 -6.13143563e-01
-7.39470050e-02 -5.94471633e-01 -2.94335991e-01 9.05290186e-01
-4.85978812e-01 -2.69541070e-02 8.37118030e-01 -4.64059055e-01
-1.51897883e+00 -1.41336131e+00 6.00900054e-01 2.07479984e-01
7.07932711e-01 -9.59883407e-02 2.13266224e-01 2.70031005e-01
2.99961250e-02 1.21860176e-01 2.74322182e-01 -6.48106158e-01
7.40926743e-01 -5.53065598e-01 -1.69280100e+00 3.83264124e-01
1.01322913e+00 3.47836792e-01 3.78814518e-01 8.33591759e-01
8.81313503e-01 4.48970571e-02 -8.49989176e-01 5.10955870e-01
2.42673084e-01 -4.74165678e-01 7.90344715e-01 -1.13611855e-01
-5.95958352e-01 -5.95685065e-01 -3.71199787e-01 -1.36278856e+00
-3.64602208e-02 -1.16636527e+00 -1.31323617e-02 1.27898562e+00
3.23980689e-01 -8.00929785e-01 7.01277316e-01 4.26493347e-01
4.95769680e-02 -6.84065640e-01 -1.34105980e+00 -1.03625977e+00
-4.71356928e-01 -4.50460911e-01 9.79027689e-01 4.48627084e-01
-2.34849364e-01 1.77292019e-01 -2.29037642e-01 8.18290234e-01
7.07887113e-01 1.78023785e-01 6.39691591e-01 -1.43454504e+00
2.42799103e-01 3.83961089e-02 -9.12098661e-02 -1.23272109e+00
2.58970559e-02 -2.63716936e-01 9.46468860e-02 -1.73143685e+00
-4.64965731e-01 -1.04875839e+00 9.56488401e-02 4.30883646e-01
6.30826235e-01 2.41947733e-02 2.47316614e-01 7.29234517e-02
-8.98111701e-01 4.15870696e-01 1.11761570e+00 -2.95061082e-01
-3.87214512e-01 8.13825071e-01 -4.46382821e-01 5.67876041e-01
1.04223764e+00 -4.71534431e-01 -9.72537458e-01 -3.80212307e-01
4.71872896e-01 4.08712745e-01 2.99535811e-01 -1.13584054e+00
2.90757775e-01 -4.76906747e-01 -3.36711377e-01 -7.15658069e-01
3.00749421e-01 -2.01784492e+00 3.39455873e-01 6.24977350e-01
3.45046192e-01 1.99929774e-01 -2.29803771e-01 8.48841250e-01
3.89446944e-01 -7.21657202e-02 5.95668197e-01 2.88436949e-01
-2.49258727e-01 6.30565107e-01 -5.49167454e-01 -2.71340102e-01
1.77461815e+00 -3.41528565e-01 -1.34224921e-01 -6.03320003e-01
-5.53877771e-01 7.54285574e-01 3.20692718e-01 1.79904804e-01
3.49857986e-01 -9.99409258e-01 -2.84181803e-01 -1.15163304e-01
-2.52157420e-01 1.45388961e-01 1.31281152e-01 1.18116200e+00
-9.68319178e-02 6.86472595e-01 1.37410909e-01 -4.33388501e-01
-1.13023436e+00 7.76003659e-01 9.96415094e-02 -2.61845261e-01
-6.02720119e-02 3.63411933e-01 -2.44112611e-01 -2.24628951e-02
1.91582009e-01 -6.49679184e-01 4.31958027e-02 6.57171980e-02
2.85725027e-01 7.57262826e-01 -9.19023827e-02 -5.92354178e-01
-3.88494700e-01 6.29620492e-01 6.77434802e-01 3.28988612e-01
1.40107322e+00 -9.24134672e-01 -9.28523764e-02 -4.40953493e-01
8.05439949e-01 -2.54048724e-02 -1.17325354e+00 1.78974777e-01
-2.27890760e-01 -2.34728649e-01 2.87877828e-01 -5.41905701e-01
-1.60409224e+00 1.53925300e-01 6.05811834e-01 1.11808908e+00
1.76808572e+00 1.62253246e-01 1.08415687e+00 3.24801385e-01
1.07041705e+00 -9.59560573e-01 -2.66881198e-01 7.41449073e-02
3.49041969e-01 -8.59505951e-01 -9.66524519e-03 -7.74973989e-01
5.19696064e-02 8.93404424e-01 1.09471165e-01 2.30020612e-01
9.08358514e-01 2.76581109e-01 -4.47832614e-01 -1.58620089e-01
-5.31749725e-01 -6.04640365e-01 -3.67275953e-01 9.07137811e-01
-3.78123373e-01 4.21972424e-01 -5.22567391e-01 4.72515255e-01
-1.03072882e-01 -1.28778871e-02 4.44387823e-01 7.92633414e-01
-7.00300992e-01 -1.09329808e+00 -3.56402755e-01 4.93309289e-01
-3.61999035e-01 3.55335236e-01 2.95748911e-03 7.14636087e-01
3.26373041e-01 1.63482440e+00 -8.35996307e-03 -1.72149882e-01
2.44050667e-01 -8.18643928e-01 1.54009297e-01 -2.11763442e-01
1.25706062e-01 -2.50656903e-03 1.24162942e-01 -4.20192003e-01
-8.44277382e-01 -6.31996274e-01 -1.11998975e+00 -2.03308389e-01
-5.20691633e-01 3.23410332e-01 9.40986156e-01 1.01966107e+00
5.25250494e-01 6.28866434e-01 1.06797838e+00 -8.48022819e-01
-3.40849966e-01 -4.78360534e-01 -5.08466005e-01 -5.29728353e-01
3.42356980e-01 -6.61209464e-01 -3.45742106e-01 -2.03372344e-01] | [5.894326686859131, 1.5370090007781982] |
f71f8964-9740-4ef0-a13f-a4b27da67811 | a-max-affine-spline-perspective-of-recurrent | null | null | https://openreview.net/forum?id=BJej72AqF7 | https://openreview.net/pdf?id=BJej72AqF7 | A MAX-AFFINE SPLINE PERSPECTIVE OF RECURRENT NEURAL NETWORKS | We develop a framework for understanding and improving recurrent neural net-works (RNNs) using max-affine spline operators (MASO). We prove that RNNs using piecewise affine and convex nonlinearities can be written as a simple piecewise affine spline operator. The resulting representation provides several new perspectives for analyzing RNNs, three of which we study in this paper. First, we show that an RNN internally partitions the input space during training and that it builds up the partition through time. Second, we show that the affine parameter of an RNN corresponds to an input-specific template, from which we can interpret an RNN as performing a simple template matching (matched filtering) given the input. Third, by closely examining the MASO RNN formula, we prove that injecting Gaussian noise in the initial hidden state in RNNs corresponds to an explicit L2 regularization on the affine parameters, which links to exploding gradient issues and improves generalization. Extensive experiments on several datasets of various modalities demonstrate and validate each of the above analyses. In particular, using initial hidden states elevates simple RNNs to state-of-the-art performance on these datasets. | ['Richard Baraniuk', 'Zichao Wang', 'Randall Balestriero'] | 2019-05-01 | null | null | null | iclr-2019-5 | ['l2-regularization'] | ['methodology'] | [ 4.13004696e-01 4.67077494e-01 -1.70843899e-01 -2.94715703e-01
-6.65965021e-01 -6.42524600e-01 4.46159959e-01 -5.44967830e-01
-3.52075577e-01 4.53190714e-01 3.91401619e-01 -3.56088668e-01
-1.32573098e-01 -4.29414868e-01 -1.13816249e+00 -1.00212693e+00
1.11676835e-01 3.19558829e-01 -1.82004794e-01 -3.61478299e-01
-1.22871220e-01 6.53338552e-01 -1.12979043e+00 2.64341593e-01
6.79402292e-01 8.49528372e-01 -5.48026077e-02 7.02906251e-01
1.14585303e-01 7.25645304e-01 -4.02763307e-01 -3.88182312e-01
5.01976788e-01 -4.38064665e-01 -8.93313229e-01 -1.78694189e-01
5.55816114e-01 -1.53445974e-02 -6.87352479e-01 1.02495861e+00
2.82543182e-01 4.08088982e-01 6.18855655e-01 -9.79175150e-01
-9.19724166e-01 8.26163590e-01 -1.26417890e-01 1.78436097e-02
-1.71196505e-01 -8.42008889e-02 1.05276275e+00 -9.10379589e-01
6.61071360e-01 1.07492626e+00 1.17215848e+00 9.63160932e-01
-1.52099609e+00 -1.90607756e-01 2.52413630e-01 -3.11207473e-01
-9.54151869e-01 -4.93409485e-01 7.47493863e-01 -4.35089231e-01
6.52448058e-01 5.40214300e-01 4.57369685e-01 1.20290899e+00
-1.47948666e-02 1.11872756e+00 9.98753190e-01 -3.73484790e-01
2.47674324e-02 4.59741987e-02 6.47053421e-01 8.25839996e-01
-4.23590034e-01 1.92172050e-01 -2.75385618e-01 8.32129866e-02
1.04854751e+00 2.45198607e-01 -4.58210647e-01 -2.82573432e-01
-1.08596587e+00 7.43417025e-01 7.10412502e-01 3.94275904e-01
-3.14685076e-01 2.88005829e-01 5.21618366e-01 3.67881030e-01
4.42133337e-01 3.47115874e-01 -3.18584681e-01 -4.18177107e-03
-9.30255175e-01 2.08803802e-03 5.80900967e-01 7.74536371e-01
4.59284693e-01 3.66655797e-01 -2.04057321e-01 9.20165956e-01
8.89720768e-02 4.06539589e-01 4.74690616e-01 -1.03064167e+00
5.17042279e-01 2.62948990e-01 -2.50516206e-01 -4.31625366e-01
-6.22450054e-01 -8.58238757e-01 -1.10097873e+00 1.51718244e-01
7.66041696e-01 -1.52580485e-01 -1.16960752e+00 2.04372120e+00
-3.21863860e-01 2.57672846e-01 3.51219565e-01 7.40393043e-01
6.12794876e-01 8.03008795e-01 -2.19820485e-01 -3.42018098e-01
1.04269528e+00 -9.88760889e-01 -9.02858973e-01 1.65916502e-01
4.44221467e-01 -2.46060312e-01 1.16177583e+00 3.31779838e-01
-1.47801828e+00 -5.86893082e-01 -8.78530145e-01 -3.22679758e-01
-2.96122521e-01 2.56172985e-01 4.43198085e-01 3.82951200e-01
-1.36784208e+00 1.02379739e+00 -1.12085950e+00 -1.48449630e-01
2.28509143e-01 4.11366761e-01 -1.55122653e-01 4.47034717e-01
-1.21050668e+00 1.03395367e+00 1.44320741e-01 5.87721884e-01
-8.25403333e-01 -9.06539679e-01 -6.75557613e-01 -6.81173801e-03
1.57980975e-02 -7.33873963e-01 1.15849233e+00 -1.03455627e+00
-1.60468447e+00 6.95170045e-01 -4.92627442e-01 -9.68154788e-01
6.34255946e-01 -3.50906819e-01 -3.04405421e-01 -4.93158214e-02
-3.34047645e-01 5.24689376e-01 9.93178427e-01 -1.37295163e+00
-5.97126707e-02 -5.15074670e-01 2.89979633e-02 1.70706421e-01
-1.78401589e-01 -1.70828432e-01 -3.85054678e-01 -7.83147335e-01
4.97098416e-01 -1.15812206e+00 -3.63746315e-01 -2.61231333e-01
-5.56895316e-01 -1.52667806e-01 6.11349940e-01 -7.36529410e-01
1.25099862e+00 -2.14177918e+00 6.36595905e-01 4.83085960e-01
3.94657165e-01 1.88210174e-01 4.16142456e-02 4.87429015e-02
-3.00842732e-01 1.89356148e-01 -3.65050882e-01 -5.92214167e-01
-9.23025683e-02 4.45799738e-01 -8.26148093e-01 5.70323884e-01
1.32964194e-01 1.24064267e+00 -6.08360887e-01 4.14611548e-02
-7.46248513e-02 8.83865535e-01 -4.00262564e-01 -1.60456225e-01
-1.87920351e-02 4.22411054e-01 -9.81770381e-02 2.81276464e-01
3.98303479e-01 -1.31015331e-01 -1.39711514e-01 -2.64119118e-01
-2.00257361e-01 2.54777342e-01 -7.73293376e-01 1.49586642e+00
-5.51739275e-01 9.76382494e-01 -1.80348232e-01 -1.12318873e+00
9.28475499e-01 3.78798455e-01 3.43450099e-01 -3.08047533e-01
1.15970924e-01 2.44502097e-01 -1.92803487e-01 -2.36186013e-01
4.09416109e-01 -2.87943125e-01 2.43400782e-01 2.99377263e-01
2.40484804e-01 3.76609802e-01 -1.62358537e-01 -2.71925665e-02
8.44489276e-01 1.13323770e-01 -3.27843904e-01 -3.12182456e-01
4.53373492e-01 -4.27529365e-01 3.92897040e-01 8.48018885e-01
-2.47127842e-02 8.24916959e-01 6.28212810e-01 -4.00105298e-01
-1.26526058e+00 -1.34985662e+00 -3.31680387e-01 9.85082805e-01
-3.66419762e-01 8.66295248e-02 -9.56802607e-01 -4.12024528e-01
-2.53163993e-01 5.60834169e-01 -1.03720820e+00 -1.12881906e-01
-9.03762937e-01 -5.32516003e-01 8.13367784e-01 9.01370764e-01
3.59344393e-01 -1.11573207e+00 -2.97991663e-01 1.30319431e-01
-1.78980395e-01 -8.02194655e-01 -8.20415914e-01 4.98033524e-01
-1.40873194e+00 -5.21169007e-01 -1.17240989e+00 -8.84151816e-01
8.48611474e-01 1.45547790e-02 8.64458919e-01 -2.88513213e-01
2.13913694e-01 3.50968778e-01 3.32542583e-02 -2.28358656e-01
-4.30490613e-01 2.10288465e-01 2.87963539e-01 2.05928326e-01
-3.05369925e-02 -6.51648998e-01 -2.43453562e-01 2.63551176e-01
-7.71161437e-01 1.46855295e-01 3.69254470e-01 9.81436551e-01
7.69150436e-01 -3.97207379e-01 4.76724595e-01 -9.76898730e-01
7.14042544e-01 -1.86105415e-01 -4.92393225e-01 4.93436933e-01
-3.20176333e-01 6.85809553e-01 7.65636027e-01 -6.59018278e-01
-9.33220685e-01 1.89171791e-01 -2.14745045e-01 -8.19271505e-01
2.59858072e-01 3.57015342e-01 2.71664441e-01 -4.78925146e-02
8.38004470e-01 4.16618526e-01 2.35562772e-01 -5.00801146e-01
6.23602629e-01 1.79603875e-01 9.03258860e-01 -6.30017757e-01
9.54938352e-01 6.58264577e-01 -1.58297643e-02 -8.06535780e-01
-1.11244369e+00 -2.57671058e-01 -8.22040796e-01 -3.26934725e-01
9.07525301e-01 -5.73941112e-01 -1.14121366e+00 2.56769627e-01
-1.22205126e+00 -4.96369451e-01 -7.49829769e-01 4.69336420e-01
-8.59278321e-01 -2.55670547e-02 -8.82374883e-01 -1.04684103e+00
-4.28508073e-01 -1.19071066e+00 7.07089126e-01 2.49402851e-01
-1.12359971e-01 -1.23558283e+00 8.00762475e-02 1.28000841e-01
3.93071681e-01 1.74518600e-01 7.45058537e-01 -5.62734365e-01
-4.12820995e-01 -3.62004596e-03 1.90405510e-02 6.20370746e-01
-3.37422639e-01 2.39430666e-02 -1.18681026e+00 -1.02296121e-01
4.70834225e-01 -4.39268835e-02 1.14887643e+00 7.12101400e-01
1.13617969e+00 -5.06154001e-01 -4.76346835e-02 9.69404817e-01
1.19745851e+00 2.22099423e-01 7.48783588e-01 2.02955514e-01
8.54291201e-01 5.39312303e-01 -5.42065538e-02 -1.19277418e-01
-4.19950262e-02 5.17480075e-01 1.73757777e-01 -3.26439828e-01
3.17436941e-02 -3.37331980e-01 5.82673430e-01 1.05817509e+00
-6.89953685e-01 4.24917042e-01 -6.24153495e-01 3.82596105e-01
-1.95947552e+00 -1.12962210e+00 -2.50743657e-01 2.17557383e+00
9.17613745e-01 1.75757408e-01 2.36949876e-01 5.47765270e-02
6.49442196e-01 6.89652562e-02 -7.12573588e-01 -4.75140601e-01
-4.44046050e-01 2.65293807e-01 6.74571753e-01 8.13036978e-01
-9.28053260e-01 8.62862408e-01 6.92868757e+00 5.98015368e-01
-1.18456244e+00 9.61238444e-02 5.61502457e-01 -4.36374508e-02
-4.50771272e-01 -2.52917171e-01 -8.26927960e-01 2.11991489e-01
1.00475955e+00 -9.32255667e-03 8.95987689e-01 6.63323998e-01
3.01889688e-01 6.64758086e-01 -1.27650154e+00 7.89663494e-01
-1.05293848e-01 -1.56697309e+00 -3.41921374e-02 8.40373486e-02
8.70247066e-01 2.33757704e-01 5.61586142e-01 3.70712489e-01
1.90353230e-01 -1.12122881e+00 8.14542711e-01 8.65766883e-01
6.08461201e-01 -6.74007952e-01 3.75730395e-01 2.28950128e-01
-1.01847279e+00 5.64857796e-02 -3.66769463e-01 2.24742070e-01
9.84991565e-02 1.74047902e-01 -6.04265153e-01 3.17853868e-01
3.17858428e-01 6.58783793e-01 -2.53536910e-01 7.38175035e-01
-1.30404979e-01 6.14893794e-01 -3.20030153e-01 1.73474088e-01
4.25619662e-01 -5.10434270e-01 5.88080764e-01 1.16461492e+00
-2.57654563e-02 1.48516119e-01 -3.66003484e-01 1.10067940e+00
-3.54206413e-01 -1.76406398e-01 -6.35197341e-01 1.92204088e-01
-9.62515082e-03 9.15980518e-01 -3.76907766e-01 -2.36531392e-01
-2.27817625e-01 8.69344831e-01 6.51362658e-01 8.77424359e-01
-8.61959636e-01 -2.94693619e-01 6.99673355e-01 -1.44613892e-01
1.03708237e-01 -1.88125446e-01 -7.23494411e-01 -1.23039556e+00
1.39658794e-01 -7.31058300e-01 1.92357525e-01 -7.05592334e-01
-1.08550215e+00 8.31628859e-01 -2.53961474e-01 -1.05697680e+00
-7.20738396e-02 -6.39778376e-01 -6.04211092e-01 9.51093197e-01
-1.14123034e+00 -8.88742805e-01 1.13654695e-01 6.08161509e-01
5.46647966e-01 -4.73775016e-03 7.66052425e-01 1.74131095e-01
-7.29599237e-01 8.17482471e-01 3.24440598e-01 3.27882707e-01
9.13689062e-02 -1.24471998e+00 6.42516315e-01 8.27846646e-01
4.24937040e-01 1.24539912e+00 7.65795112e-01 -3.02182496e-01
-1.35868824e+00 -9.69574630e-01 7.72707820e-01 -5.50557077e-01
8.06578398e-01 -3.73663306e-01 -1.38399208e+00 1.16328323e+00
6.70773163e-02 -4.06625941e-02 3.26713175e-01 3.27569634e-01
-3.65018100e-01 2.93882797e-03 -8.13484311e-01 7.95256019e-01
1.11878920e+00 -8.76825869e-01 -7.38524616e-01 3.59988540e-01
8.71183753e-01 -7.02027798e-01 -9.87322688e-01 4.10301328e-01
7.04296112e-01 -7.13974118e-01 9.59190488e-01 -1.22481191e+00
2.75814354e-01 -1.92561969e-01 -2.81053707e-02 -1.31898367e+00
-2.09646061e-01 -1.06137705e+00 -3.26930553e-01 9.15114462e-01
7.73465276e-01 -6.73144102e-01 8.70403409e-01 8.46299291e-01
-3.95118445e-01 -1.16762400e+00 -7.93069005e-01 -8.45504224e-01
2.83239484e-01 -5.81067562e-01 1.94238067e-01 8.71924818e-01
-6.20668791e-02 1.52577460e-01 -4.02966946e-01 2.06877187e-01
4.38408881e-01 -3.46746445e-01 2.91236281e-01 -1.02703357e+00
-3.07534456e-01 -7.09456742e-01 -2.30130851e-01 -1.51142454e+00
2.88372606e-01 -1.13951957e+00 9.53666866e-02 -1.38842142e+00
-8.40245187e-02 -4.79765385e-01 -3.66108179e-01 5.33300519e-01
6.47764280e-02 2.00000301e-01 3.30794692e-01 4.50854808e-01
-1.77464634e-01 3.98793519e-01 1.23653960e+00 -3.86470333e-02
-6.13318205e-01 5.57637036e-01 -4.62458372e-01 9.06174362e-01
7.67325938e-01 -2.52636075e-01 -2.62948394e-01 -4.63771164e-01
1.55834049e-01 1.21291302e-01 4.30364490e-01 -6.91191077e-01
4.02759671e-01 2.47662321e-01 2.39656582e-01 -5.59823930e-01
5.05220652e-01 -7.81882048e-01 8.76874849e-02 5.96548736e-01
-9.17365134e-01 4.12055254e-02 1.64052978e-01 4.03338641e-01
-1.17120342e-02 -3.37557554e-01 9.18705761e-01 1.31650874e-02
-1.64221570e-01 1.76991105e-01 -1.44478619e-01 2.61647463e-01
5.90126932e-01 -2.78265387e-01 -2.87860811e-01 -4.39212799e-01
-1.29751337e+00 1.83775648e-01 -6.18712604e-03 2.27778539e-01
5.67371130e-01 -1.40188694e+00 -4.57154363e-01 3.21117818e-01
-4.33940291e-01 -4.44027632e-02 1.49620369e-01 1.07029355e+00
-2.79319882e-01 4.32423890e-01 3.05375457e-03 -6.21944547e-01
-1.15470088e+00 3.86269301e-01 8.52765501e-01 -1.75430790e-01
-7.68983483e-01 9.56600964e-01 2.28850305e-01 -5.11739552e-01
7.76851237e-01 -7.53060758e-01 -1.63514748e-01 9.30458158e-02
3.26619327e-01 4.82860833e-01 -2.34157834e-02 -6.21018052e-01
-6.79481924e-02 7.29038298e-01 -1.63148746e-01 -3.57138067e-01
1.50823450e+00 1.78857550e-01 -2.26097945e-02 8.80426466e-01
1.55875766e+00 -2.17464775e-01 -1.61184061e+00 -4.94710147e-01
-8.26931000e-02 1.81468099e-01 -1.90859258e-01 -4.79752243e-01
-1.20781577e+00 9.54576135e-01 3.50001663e-01 4.52309519e-01
9.97926176e-01 2.37588845e-02 5.99142909e-01 5.32451987e-01
-1.38283253e-01 -9.17756796e-01 -1.80423677e-01 8.98076355e-01
1.08214450e+00 -7.85083354e-01 -4.36311781e-01 -8.34081843e-02
-8.34798217e-01 1.41509521e+00 1.06953554e-01 -3.89642060e-01
6.21360779e-01 2.21029967e-01 1.02960408e-01 -3.15481052e-02
-5.34773052e-01 -5.05779758e-02 6.52420163e-01 3.02952170e-01
3.30854714e-01 7.29850233e-02 1.65503789e-02 4.19658005e-01
-5.21043479e-01 -2.73301542e-01 3.59494627e-01 3.55374694e-01
-2.58125484e-01 -7.47863114e-01 -2.70363808e-01 4.08520162e-01
-4.36335474e-01 -2.28688091e-01 -1.41446859e-01 6.53843701e-01
-3.14927131e-01 3.03968817e-01 8.13637897e-02 -3.04621875e-01
5.55979788e-01 3.89857888e-01 5.45210719e-01 -2.91860461e-01
-6.88758194e-01 5.57457563e-03 -2.62955397e-01 -3.66602898e-01
-2.70145059e-01 -7.59721160e-01 -1.52477801e+00 -5.78808710e-02
-2.64590681e-01 -2.31259614e-02 8.68512392e-01 1.00099504e+00
4.73444052e-02 8.19283843e-01 5.03068745e-01 -8.73040676e-01
-8.87923539e-01 -7.72917867e-01 -3.64978462e-01 2.54721135e-01
8.58748794e-01 -3.15309137e-01 -5.18276334e-01 2.20178634e-01] | [7.820025444030762, 3.543745756149292] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.